A multifunctional polyketide–peptide synthetase essential for albicidin biosynthesis in Xanthomonas albilineans

Guozhong Huang1, Lianhui Zhang1,2 and Robert G. Birch1

Department of Botany, The University of Queensland, Brisbane 4072, Australia1
Institute of Molecular Agrobiology, The National University of Singapore, Singapore1176042

Author for correspondence: Robert G. Birch. Tel: +61 7 33653347. Fax: +61 7 33651699. e-mail: r.birch{at}botany.uq.edu.au


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Albicidins, a family of potent antibiotics and phytotoxins produced by the sugarcane leaf scald pathogen Xanthomonas albilineans, inhibit DNA replication in bacteria and plastids. A gene located by Tn5-tagging was confirmed by complementation to participate in albicidin biosynthesis. The gene (xabB) encodes a large protein (predicted Mr 525695), with a modular architecture indicative of a multifunctional polyketide synthase (PKS) linked to a non-ribosomal peptide synthetase (NRPS). At 4801 amino acids in length, XabB is the largest reported PKS–NRPS. Twelve catalytic domains in this multifunctional enzyme are arranged in the order N terminus–acyl-CoA ligase (AL)–acyl carrier protein (ACP)–ß-ketoacyl synthase (KS)–ß-ketoacyl reductase (KR)–ACP–ACP–KS–peptidyl carrier protein (PCP)–condensation (C)–adenylation–PCP–C. The modular architecture of XabB indicates likely steps in albicidin biosynthesis and approaches to enhance antibiotic yield. The novel pattern of domains, in comparison with known PKS–NRPS enzymes for antibiotic production, also contributes to the knowledge base for rational design of enzymes producing novel antibiotics.

Keywords: leaf scald phytotoxin, nonribosomal peptide synthetase, polyketide synthase

Abbreviations: A, adenylation; ACP, acyl carrier protein; AL, acyl-CoA ligase; AMT, aminotransferase; AT, acyltransferase; C, condensation; Cy, cyclization; DH, dehydratase; KR, ß-ketoacyl reductase; KS, ß-ketoacyl synthase; MT, methyltransferase; NRPS, nonribosomal peptide synthetase; PCP, peptidyl carrier protein; PKS, polyketide synthase; TE, thioesterase

The GenBank accession number for the sequence determined in this work is AF239749.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Many phytopathogenic bacteria and fungi secrete toxins with phytotoxic activity and a broad spectrum of antimicrobial properties (Guenzi et al., 1998 ). Albicidin phytotoxins produced by Xanthomonas albilineans are key pathogenicity factors in the development of leaf scald, one of the most devastating diseases of sugarcane (Saccharum interspecific hybrids) (Ricaud & Ryan, 1989 ; Zhang et al., 1999 ). Albicidins selectively block prokaryote DNA replication and cause the characteristic chlorotic symptoms of leaf scald disease by blocking chloroplast development (Birch & Patil, 1983 , 1985b , 1987a , b ). Because albicidins are rapidly bactericidal at nanomolar concentrations against a broad range of Gram-positive and Gram-negative bacteria, they are also of interest as potential clinical antibiotics (Birch & Patil, 1985a ).

The major antimicrobial component of the family of albicidins produced in culture by X. albilineans has been partially characterized as a low-Mr compound with several aromatic rings (Birch & Patil, 1985a ). Low yields have slowed studies into the chemical structure of albicidin, its application as a tool to study prokaryote DNA replication and its development as a clinical antibiotic (Zhang et al., 1998 ). Genetic analysis of albicidin biosynthesis is likely to indicate approaches to increase yields, probable structural features, and opportunities for engineering novel antibiotics in this family.

Transposon mutagenesis revealed that at least two gene clusters spanning more than 60 kb in the genome of X. albilineans are involved in albicidin production (Rott et al., 1996 ; Wall & Birch, 1997 ). Recently, a gene (xabA) required for albicidin biosynthesis in X. albilineans has been identified and found to encode a phosphopantetheinyl transferase, which may activate albicidin synthetases in the albicidin biosynthetic pathway (Huang et al., 2000b ). Here, we show that a second gene (xabB) required for albicidin biosynthesis encodes a large, multifunctional enzyme, with unusual polyketide synthase (PKS) modules fused to a non-ribosomal peptide synthetase (NRPS).


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains and plasmids.
The properties of bacteria and plasmids used are listed in Table 1.


View this table:
[in this window]
[in a new window]
 
Table 1. Bacterial strains and plasmids

 
Media, culture conditions and antibiotics.
X. albilineans strains were routinely cultured on sucrose peptone (SP) medium at 28 °C (Birch & Patil, 1985b ). Escherichia coli strains were used as hosts in cloning experiments and were grown on LB medium at 37 °C (Sambrook et al., 1989 ). Broth cultures were aerated by shaking at 200 r.p.m. on an orbital shaker. Modified YEB medium (Van Larebeke et al., 1977 ) was used for patch mating. The following antibiotics were added to media as required: kanamycin, 50 µg ml-1; tetracycline, 15 µg ml-1; ampicillin, 100 µg ml-1; and spectinomycin, 50 µg ml-1.

Routine genetic procedures.
Bacterial genomic and plasmid DNA isolation, gel electrophoresis, DNA restriction digests, ligation reactions, transformation and colony hybridization were performed by routine procedures (Sambrook et al., 1989 ). DNA fragments were excised from agarose gels and residual agarose was removed with the BRESAclean DNA purification kit (GeneWorks).

Construction of a X. albilineans partial genomic library.
Genomic DNA from X. albilineans Xa13 was digested with EcoRI and size-fractionated. DNA fragments of 15–20 kb were ligated to dephosphorylated EcoRI-cleaved pBluescript II SK. The ligated DNA was electroporated into E. coli TOP10. Transformants were selected on LB agar medium containing ampicillin, and stored in LB broth with 15% glycerol at -70 °C.

DNA sequencing and analysis.
Sequencing reactions were performed by dideoxynucleotide chain termination (Sanger et al., 1977 ) using a BigDye Terminator Cycle Sequencing kit and a 373A DNA sequencer (PE Applied Biosystems) through the Australian Genome Research Facility. Oligonucleotide primers were purchased from GeneWorks. Analyses of DNA and protein sequences in the GenBank, EMBL, PIR and SWISS-PROT databases were carried out by using the University of Wisconsin Genetics Computer Group (UWGCG) programs BLASTP, FASTA, PILEUP and BESTFIT through WebANGIS version 2.0.

PCR amplification.
BamHI-digested genomic DNA from X. albilineans LS157 was religated at low concentration (0·5 µg ml-1) to generate circular DNA molecules as templates for inverse PCR. Three primers, one from the IS terminal region of Tn5 (IR2: 5'-CGGGATCCTCACATGGAAGTCAGATCCTG-3'), and two flanking the unique BamHI restriction site of Tn5 (BL: 5'-GGGGACCTTGCACAGATAGC-3' and BR: 5'-CATTCCTGTAGCGGATGGAGATC-3'), were used to amplify the sequences flanking the Tn5 insertion in the genome of LS157. The amplified fragments (1·4 kb and 6·0 kb) were cloned into pZErO-2, yielding pZIL and pZIR (Fig. 1).



View larger version (18K):
[in this window]
[in a new window]
 
Fig. 1. Physical and functional map of part of the albicidin biosynthetic gene cluster. (a) Partial physical map of the Tn5 insertion locus in LS157 genomic DNA. Restriction enzymes used: C, ClaI; E, EcoRI; S, SpeI; N, NotI; and B, BamHI. (b) Probes used to recover clone pXABB: probe 1, 1·4 kb EcoRI–NotI fragment digested from pBC157; and probe 2, 0·9 kb PCR product amplified from Xa13 genomic DNA using primers complementary to sequences flanking the Tn5 insertion in LS157. (c) Clones and subclones used for sequencing and described in Table 1. (d) The transcription directions of three putative ORFs in the 16·5 kb EcoRI fragment are indicated by arrows. (e) Organization of X. albilineans XabB, constructed by comparison with known protein sequences. The white box indicates the PKS region, and the black box indicates the NRPS region. Relative positions of deduced catalytic domains are indicated. Horizontal bars indicate proposed biosynthetic modules.

 
PCR was performed in a volume of 50 µl with 200 ng genomic DNA (or 10 ng plasmid DNA), 0·4 ng of each primer µl-1, 0·2 mM of each dNTP, 1·8 mM Mg2+ and 1 unit of elongase enzyme mix (Life Technologies). A 10 min initial denaturation step at 94 °C was followed by 35 thermal cycles of denaturation at 94 °C for 1 min, annealing at 55 °C for 1 min, and extension at 72 °C for 1 min per 1 kb of expected amplification.

Construction of promoter probes and glucuronidase assay.
Plasmid pRG960sd contains a promoterless ß-glucuronidase gene (uidA) downstream of a multiple cloning site (Van den Edde et al., 1992 ). Sequence upstream of xabB (nucleotide residues 1005–1210 or 521–1210) was amplified from pXABB by PCR. Forward primer P1F1 (5'-ACGCGGATCCCAGCAGGGTGTCATACACG-3') or P1F2 (5'-TCGCGGATCCGCGCGATTGAAGTAGTCC-3') contained a BamHI restriction site (underlined). Reverse primer P1R (5'-TCCCCCGGGCGGCCAGCGTGGTGCTACTAC-3') introduced a XmaI restriction site (underlined). PCR fragments were ligated into BamHI/XmaI-cut pRG960sd, yielding pRG960p1 and pRG960p2. These constructs were mobilized from E. coli DH5{alpha} into X. albilineans LS155 as described below.

Promoter strength was quantified by fluorometric analysis of glucuronidase (GUS) activity (Jefferson, 1987 ; Xiao et al., 1992 ). The protein content in cell lysates was determined by the dye-binding method (Bradford, 1976 ) using a Bio-Rad protein assay kit.

Bacterial conjugation.
DNA transfer between E. coli donor JM109(pLAFR3±insert) or DH5{alpha}(pRG960sd±insert) and X. albilineans recipient (LS157, LS-JP2 or LS155), was accomplished by triparental transconjugation with helper strain pRK2013. Mid-exponential-phase cultures of the recipient were spotted onto agar plates containing YEB medium with no antibiotics (20 µl per spot). After the liquid had been absorbed by the agar, 20 µl of mid-exponential-phase culture of the helper was added to each spot. The liquid was again allowed to absorb, and 20 µl of mid-exponential-phase culture of the donor was added to each spot. After incubation of the mating plates overnight at 28 °C, transconjugants were selected on SP plates supplemented with ampicillin, and tetracycline or spectinomycin.

Assay and quantification of albicidin production.
Albicidin was quantified by a microbial plate bioassay as described previously (Birch & Patil, 1985b ), except that the 10 ml basal layer of LB agar and the 5 ml overlayer of 50% LB with 1% agar were supplemented with tetracycline or spectinomycin, and E. coli DH5{alpha}(pLAFR3) or DH5{alpha}(pRG960sd) was used as the indicator strain. This change avoided interference by tetracycline or spectinomycin, which were added to some cultures to ensure retention of pLAFR3 or pRG960sd derivatives in X. albilineans. Inhibition zone widths in the bioassay were converted to albicidin concentrations by interpolation on a dose-response plot produced under the same assay conditions. The plot fitted the formula log[Alb]=0·3 W-0·92, where [Alb] is units of albicidin per 20 µl sample and W is the width in millimeters of the zone of growth inhibition surrounding each well.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Cloning and sequencing of xabB gene required for albicidin production
X. albilineans Tox- mutant LS157 contains a single Tn5 insertion, in a 4·1 kb ClaI restriction fragment or in a 16·5 kb EcoRI restriction fragment (Fig. 1). Selection for kanamycin resistance, following shotgun cloning of ClaI restriction fragments of LS157 DNA into pBluescript II SK, yielded clone pBC157. Sequences flanking the Tn5 insertion in LS157 DNA were amplified by inverse PCR and cloned into pZErO-2, producing pZIL and pZIR. Plasmid pXABB was screened from an EcoRI genomic library of wild-type X. albilineans Xa13, using probes described in the legend to Fig. 1(b). Subclones pSEBL and pSEBR were derived from pXABB (Fig. 1c; Table 1).

The double-strand sequence of the 16511 bp EcoRI genomic fragment in pXABB was obtained by a primer-walking approach, using subclones pBC157, pZIL, pZIR, pSEBL and pSEBR. The Tn5 insertion in the genome of LS157 is accompanied by a 9 bp perfect repeat sequence (GTCCTGAAG), commencing at bp 2490 in the sequence deposited in GenBank with accession number AF239749.

The only ORF longer than 900 bp within the 16·5 kb fragment is disrupted by the Tn5 insertion. This ORF (designated xabB) encodes a protein of 4801 amino acids (Mr 525695). It commences at bp 1230 in the above GenBank sequence with a TTG codon, 6 bp downstream from a RBS, GAGG. There is an alternative start codon (ATG) a further 15 bp downstream. Of the codons in this ORF, 8·5% are rarely used in E. coli. The closest match (TTGAGC-14x-TATAAC) to the consensus -35 (TTGACA) and -10 (TATAAT) sequences for E. coli {sigma}70 promoters occurs 117 bp upstream of the translation initiation codon (Fig. 2).



View larger version (30K):
[in this window]
[in a new window]
 
Fig. 2. Sequence of the region upstream from xabB. The nucleotide sequence is numbered according to the 16511 bp sequence in GenBank with accession no. AF239749. The putative -35 and -10 promoter sequences of xabB and the divergent gene thp are underlined, as are RBS. The transcriptional directions of xabB and thp are indicated by arrows. Translational start codons are indicated by boldface type. Primers P1F1 and P1R are boxed.

 
Downstream by 35 bp from the TAG stop codon of xabB is a probable RBS (GAGG), separated by 6 bp from the ATG start codon of another ORF (designated xabC) in the same orientation as xabB. There are no sequences resembling transcriptional terminators between these ORFs. Overlapping the xabB promoter region is another probable promoter for a divergent transcript including a putative RBS (TGGAGG) and start codon for a gene designated thp, separated by 233 bp from xabB (Figs 1, 2). Analysis of these flanking genes is described by Huang et al. (2000a ).

Sequence homology analyses
Comparison of XabB with available protein sequence databases reveals an N-terminal region (from Met-1 to Asp-3235) similar to many microbial modular PKSs, and a C-terminal region (from Pro-3236 to Asp-4801) similar to NRPSs.

Recognizable PKS domains commencing at the N terminus of XabB are an acyl-CoA ligase (AL), acyl carrier protein (ACP1), ß-ketoacyl synthase (KS1) and ß-ketoacyl reductase (KR), followed by two consecutive ACPs and one KS (Fig. 1). The motifs characteristic of these domains are aligned with those from other organisms in Fig. 3. The AL domain contains the conserved adenylation core sequence (SGSSG) and the ATPase motif (TGD), and shows 22–30% identity and 50–60% similarity to prokaryotic and eukaryotic aromatic-acid CoA ligases and long-chain fatty-acid CoA ligases (protein symbols and GenBank accession nos: MycA, AF184956; LCFA, P46450; PksJ, P40806; ComL2, P14913; FkbB, AF082099; RifA, X86780; RapA, AF040570). The three ACP domains show up to 39·2% identity and 78·6% similarity to ACPs (MycA, AF184956; HMWP1, Y12527; Ta1, AJ006977; PksX6, U11039; EryA1, M63676; MAS, M95808), and all contain a 4'-phosphopantetheinyl-binding cofactor box GxDS(I/L) (Hopwood & Sherman, 1990 ), except that A replaces G in ACP1 (Fig. 3). The two KS domains show up to 56·1% identity and 80·8% similarity to prokaryotic ß-ketoacyl synthases (KSs) (MycA, AF184956; HMWP1, Y12527; Ta1, AJ006977; PksX6, U11039; PksM, O31781; EryA1, M63676; PpsA, G3261605). Both contain motif GPxxxxxxxCSxSL around the active site Cys, and two His residues downstream of the active site Cys, in motifs characteristic of these enzymes (Donadio et al., 1991 ; Hopwood, 1997 ; Huang et al., 1998 ). The KR domain shows up to 27·9% identity and 61·8% similarity to other ß-ketoacyl reductases (HMWP1, Y12527; Ta1, AJ006977; PksX6, U11039; EryA1, M63676; MAS, M95808), and contains the NAD(P)H-binding site GGxGxLG (Scrutton et al., 1990 ).



View larger version (46K):
[in this window]
[in a new window]
 
Fig. 3. Alignment of X. albilineans XabB enzymic domains with those of PKSs and fatty acid synthases from other organisms. Identical amino acids are indicated by boldface type. Asterisks and overlines identify conserved amino acids at catalytic sites. Xal-XabB, X. albilineans XabB for biosynthesis of albicidin (this study); Bsu-MycA, B. subtilis MycA for biosynthesis of mycosubtilin (AF184956); Hin-LCFA, Haemophilus influenzae long-chain fatty-acid-CoA ligase (P46450); Bsu-PksJ, B. subtilis polyketide synthase J (P40806); Pcr-ComL2, Petroselinum crispum 4-coumarate-CoA ligase 2 (P14913); Sma-FkbB, Streptomyces sp. MA6548 FkbB for biosynthesis of FK506 (AF082099); Ame-RifA, Amycolatopsis mediterranei RifA for biosynthesis of rifamycin B (AF040570); Shy-RapA, Streptomyces hygroscopicus RapA for biosynthesis of rapamycin (X86780); Ype-HMWP1, Y. pestis HMWP1 for biosynthesis of yersiniabactin (Y12527); Mxa-Ta1, M. xanthus Ta1 for biosynthesis of TA (AJ006977); Bsu-PksX6, B. subtilis polyketide synthase PksX6 (U11039); Ser-EryA1, Saccharopolyspora erythraea EryA module for biosynthesis of erythromycin (M63676); Bsu-PksM, B. subtilis PKS for a polyketide synthase (O31781); Mtu-PpsA, Mycobacterium tuberculosis PKS for a polyketide synthase (G3261605); Mtu-MAS, M. tuberculosis MAS for biosynthesis of mycocerosic acid (M95808).

 
At the C-terminus of XabB is an apparent peptide synthetase region linked to the PKS module via a peptidyl carrier protein (PCP) domain (Fig. 1). The peptide synthetase region shows 31–38% identity and 60–63% similarity with members of the peptide synthetase family from Bacillus (TycB, AF004835; FenD, AJ011849; BacC, AF007865; GrsB, X61658; DhbF, AF047828; SrfAA, X70356), Streptomyces (SnbDE, Y11547), Pseudomonas (SyrE, AF047828) and E. coli (EntF, P11454). It displays the ordered condensation (C), adenylation (A) and PCP domains typical of such multienzymes (Marahiel et al., 1997 ), followed by an extra C domain. The conserved sequences, characteristic of the domains commonly found in peptide synthetases, are compared with those from XabB in Table 2.


View this table:
[in this window]
[in a new window]
 
Table 2. Comparison of conserved sequences in peptide synthetases and XabB

 
Complementation of xabB and xabC genes
Plasmids pLXABB1 and pLXABB2 contain the 16·5 kb EcoRI fragment from pXABB (Fig. 1), cloned into broad-host-range cosmid pLAFR3 in forward and reverse orientation to the lac promoter, respectively. Mobilization of pLAFR3, pLXABB1, pLXABB2 or pLXABC by bacterial conjugation into Tox- mutants LS157 (xabB::Tn5) and LS-JP2 (xabC insertional mutant) occurred at a frequency of 1·5x10-2 transconjugants per recipient cell. Albicidin production was undetectable in the Tox- mutant LS157 and pLAFR3 controls. Introduction of the xabC gene on pLXABC restored albicidin production in LS-JP2, but not in LS157. Introduction of pLXABB1 or pLXABB2 expressing the full-length xabB and a truncated but functional xabC further enhanced albicidin production in LS-JP2 and in LS157 relative to the wild-type parental strain LS155 (Fig. 4).



View larger version (19K):
[in this window]
[in a new window]
 
Fig. 4. Albicidin production by wild-type X. albilineans LS155 ({blacktriangleup}), or Tox- mutant strains LS-JP2 or LS157 carrying plasmids pLAFR3 ({blacksquare}), pLXABB1 ({circ}), pLXABB2 ({bullet}) or pLXABC ({triangleup}). Panels show results from separate experiments testing complementation of (a) LS-JP2 and (b) LS157. Albicidin concentrations in culture supernatants (units ml-1) were quantified based on inhibition zone width in a microbial bioassay. Data are means±SE of five replicates.

 
Functional analysis of xabB promoter region
GUS activity was undetectable in LS155 and LS155(pRG960sd) controls. Plasmid pRG960p1 or pRG960p2, with 206 bp or 690 bp from the xabB promoter region upstream of GUS, both conferred GUS activity with no difference in expression level or pattern in X. albilineans LS155 (Fig. 5). GUS activity increased with culture density, whereas albicidin accumulation was delayed until late in the exponential-growth phase.



View larger version (24K):
[in this window]
[in a new window]
 
Fig. 5. Relationship between growth ({blacksquare}), albicidin production ({circ}) and GUS activity ({blacktriangleup}) in X. albilineans LS155(pRG960p1) (a) and in LS155(pRG960p2) (b). Data are relative values [100% growth=OD550 1·43; 100% albicidin production=268·5 units ml-1; 100% GUS activity=119 pmol methylumbelliferone min-1 (mg protein)-1] and are means±SE of two replicates. Locations and sizes of inserts in pRG960p1 and pRG960p2 are indicated in Fig. 2 and Table 1.

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The xabB gene required for albicidin biosynthesis in X. albilineans encodes a large (526 kDa) protein with the sequence characteristics of a mixed-function PKS and NRPS. Tox- Tn5-insertional mutant LS157 was complemented in trans by a 16·5 kb EcoRI fragment including xabB (14·4 kb) and a truncated downstream xabC, but not by xabC alone. A 206 bp region upstream of the xabB start codon is a functional promoter during the phase of albicidin accumulation in X. albilineans (Fig. 5). The xabB gene uses TTG as start codon, which may impose post-transcriptional control on the rate of gene product formation (McCarthy & Gualerzi, 1990 ). We have shown elsewhere that the downstream cistron xabC encodes a methyltransferase (MT) involved in albicidin biosynthesis, and that no other downstream cistron in the operon is required for albicidin production (Huang et al., 2000a ).

The N-terminal region of the predicted XabB protein resembles a multifunctional PKS. Bacterial multifunctional PKSs operate as biochemical assembly lines, composed of a series of catalytic domains involved in sequential assembly and modification of acyl groups on the growing polyketide chain (Cane et al., 1998 ; Keating & Walsh, 1999 ). The catalytic domains are arranged in modules, punctuated by acyl carrier protein (ACP) domains that tether the nascent polyketide while it undergoes the catalytic modifications programmed in the associated module. For each polyketide there is an initiation module, a series of elongation modules that define the length and structure of the polyketide chain, and a termination module to release the product from the final tether. The initiation module typically comprises an acyltransferase (AT) domain that couples the initial acyl group from an acyl-CoA substrate to the phosphopantetheinyl tether of the first ACP domain. Each elongation module typically comprises at least a KS, AT and ACP. The AT selects and loads an acyl group (typically malonate) onto the ACP. The KS removes the growing polyketide unit from the upstream ACP and couples it by decarboxylation to this next tethered acyl group (typically extending the chain by one acetyl group). Other catalytic domains (e.g. KR, or dehydratase, DH) within an elongation module can modify the newly elongated polyketide before it is transferred to the next module. A thioesterase (TE) domain in the termination module accomplishes release of the assembled polyketide from the last ACP (Cane et al., 1998 ; Keating & Walsh, 1999 ).

Biosynthesis of a polyketide can involve the sequential action of several PKS proteins, each with one to six elongation modules (MacNeil et al., 1992 ; Aparicio et al., 1996 ). There are variations on the modular PKS design, including participation by some loading domains across modules or in trans from separate proteins (Keating & Walsh, 1999 ), and several examples of hybrid PKS/NRPS proteins (Albertini et al., 1995 ; Gehring et al., 1998 ; Duitman et al., 1999 ; Paitan et al., 1999 ). Subsequent modification of the polyketide by dedicated tailoring enzymes is generally required to complete the biologically active product (Cane et al., 1998 ).

XabB commences with an AL domain (residues 1–629) followed by an ACP domain (ACP1, residues 630–731). In other PKS systems, an N-terminal AL is involved in activation and incorporation of 3,4-dihydroxycyclohexane carboxylic acid, 3-amino-5-hydroxybenzoic acid (AHBA) or a long-chain fatty acid as a starter (Aparicio et al., 1996 ; Motamedi & Shafiee, 1998 ; Tang et al., 1998 ; Duitman et al., 1999 ). It may be possible to determine the substrate specificity of the AL domain by heterologous expression and purification of the module 1 protein containing AL plus ACP1 domains, and ATP-pyrophosphate exchange assays of the purified fusion protein with potential substrates, as applied to NRPS internal A domains (Guenzi et al., 1998 ; Stachelhaus et al., 1998 , 1999 ).

The second module in XabB contains a KS (residues 732–1165) and a KR (residues 1811–1971) upstream of two ACPs (residues 2457–2522 and 2544–2613), but lacks any discernable AT domain (Fig. 1). Tandem ACPs are unusual within PKS modules. Other instances include WA for a green pigment formation and ST for sterigmatocystin biosynthesis in Aspergillus nidulans (Mayorga & Timberlake, 1992 ; Yu & Leonard, 1995 ), the PKS1 product for melanin biosynthesis in Colletotrichum lagenarium (Takano et al., 1995 ) and the pksX7 product for an unknown polyketide in Bacillus subtilis (Albertini et al., 1995 ). The significance of the tandem ACPs in these systems has not been resolved. ST performs the iterative addition of seven acetate units (from malonyl-CoA donors) to a hexanoate primer, to form an octaketide intermediate that is stabilized by cyclization, yielding a product with three aromatic rings (Yu & Leonard, 1995 ). This is interesting in the context of the apparent aromatic structure of albicidin. Although most aromatic polyketides are synthesized by iterative type I or type II PKSs (Shen, 2000 ), there is an example of synthesis by a non-iterative type I PKS in the case of pyoluteorin in Pseudomonas fluorescens (Nowak-Thompson et al., 1997 ).

Possibilities in the case of XabB include use of either or both of ACP2 and ACP3 in chain elongation. One domain could be non-functional, as observed for modification domains in various PKSs (Aparicio et al., 1996 ; Nowak-Thompson et al., 1997 ; Tang et al., 1998 ). Selection between ACP2 and ACP3 could depend on which has been loaded with an appropriate acyl group by a separate AT enzyme. If they have different loading specificities (e.g. for malonyl-CoA versus methylmalonyl-CoA), this could contribute to the production of multiple, structurally related albicidins by the same synthetase. The evidence that domains typically found within a module can in some instances act in trans (Keating & Walsh, 1999 ), including a trans-acting KS in the deduced type I mycobactin PKS (Quadri et al., 1998 ), supports the possibility that ACP2 and ACP3 could act sequentially in chain elongation. Finally, ACP2 and ACP3 could accomplish the iterative biosynthesis of a longer ketide chain if ACP3 serves as a waiting position allowing polymerization of successive products from ACP2, as for the tandem PCP domains in enniatin synthetase (Zocher & Keller, 1997 ). As discussed below, it will be easier to distinguish between these possibilities when the complete structure of albicidin is elucidated.

The third module in XabB contains a KS (residues 2630–3046) followed by a PCP (residues 3221–3307) at the start of the NRPS region. Four other fused PKS/NRPS systems (Albertini et al., 1995 ; Gehring et al., 1998 ; Duitman et al., 1999 ; Paitan et al., 1999 ) are known, three of which lack recognizable AT domains (Fig. 6). Yersinia pestis HMWP1 contains a typical PKS elongation module (including AT) and an NRPS module with a terminating thioesterase (TE) domain. It is the third protein, following an AL (YbtE) and NRPS (HMWP2), in the biosynthetic apparatus for yersiniabactin (Gehring et al., 1998 ). B. subtilis MycA bears the closest resemblance to XabB, showing PKS initiation and elongation modules linked via an aminotransferase (AMT) domain to the NRPS region. In B. subtilis PksX6 and Myxococcus xanthus Ta1, the NRPS region precedes the PKS region. Separate AT enzymes encoded elsewhere in the genome may operate in trans to load the appropriate acyl groups onto the ACPs in the elongation modules of these PKSs. Candidates are a malonyl-CoA transacylase gene (fenF) located immediately upstream of mycA (Duitman et al., 1999 ) and an AT gene located 20 kb upstream of ta1 (Paitan et al., 1999 ). We have not yet identified an AT gene in the xab gene cluster.



View larger version (22K):
[in this window]
[in a new window]
 
Fig. 6. Organization of five known PKS–NRPS enzymes, aligned at the first KS domain. (a) X. albilineans XabB, encoded by xabB for albicidin biosynthesis (this study); (b) B. subtilis MycA for mycosubtilin biosynthesis (Duitman et al., 1999 ); (c) Y. pestis HMWP1 for yersiniabactin biosynthesis (Gehring et al., 1998 ); (d) M. xanthus partial gene product Ta1 for TA biosynthesis (Paitan et al., 1999 ); (e) B. subtilis PksX6 for unknown function (Albertini et al., 1995 ). White boxes indicate PKS regions, black boxes indicate NRPS regions. Relative positions of deduced catalytic domains are indicated. Vertical bars follow the carrier domains at the end of each biosynthetic module.

 
From the characteristics of albicidin (Birch & Patil, 1985a ) and the architecture of the XabB PKS region (Fig. 1), it is likely that: (i) the AL couples CoA to an aryl or acyl residue in an ATP-dependent reaction and loads the activated residue onto the 4'-phosphopantetheine prosthetic arm of ACP1; (ii) an acyl group is loaded onto ACP2 or/and ACP3 by a separate AT; (iii) the KS1 domain accepts the initiation residue from ACP1 onto a conserved cysteine, then transfers it by decarboxylative condensation onto the acyl group tethered to ACP2; (iv) the tethered chain is modified by KR; (v) the ketide chain may be extended at ACP3 as discussed above; (vi) the assembled polyketide intermediate is translocated via KS2 onto the 4-phosphopantetheine prosthetic arm of PCP1, at the start of the NRPS region.

The step to link polyketide and peptide moieties in the nascent antibiotic is of particular interest. In other keto-peptide synthetases, the first PCP is preceded by a C domain, rather than a KS domain as in XabB. The absence of an A domain upstream of PCP1 may indicate involvement of a trans-acting A domain as in HMWP1 (Gehring et al., 1998 ), or transfer without chain extension as in MycA (Duitman et al., 1999 ).

NRPSs typically show a modular architecture and tethered biosynthetic strategy analogous to PKSs (Cane et al., 1998 ; Keating & Walsh, 1999 ). In NRPSs a C domain removes the growing peptide unit from the upstream PCP domain and couples it to the next amino acid in the chain, which has already been selected and loaded by an A domain onto the PCP in the same module (Marahiel et al., 1997 ; Stachelhaus et al., 1998 ). Other catalytic domains (e.g. epimerase or N-methyltransferase) within an elongation module can modify the newly elongated polypeptide before it is transferred to the next module in the biochemical assembly line (Marahiel et al., 1997 ).

The A domain in the NRPS region of XabB contains ten conserved sequences (A1–A10, Table 2) identified as AMP-binding, ATP-Mg2+-binding, adenine-binding or ATPase sites (Turgay et al., 1992 ; Marahiel et al., 1997 ). In other NRPS systems, A domains select and load a particular amino-, hydroxy- or carboxy-acid unit (Marahiel et al., 1997 ). Substrate specificity is determined at the binding pocket, consisting of a stretch of about 100 amino acid residues between highly conserved motifs A4 and A5 (Conti et al., 1997 ). Based on sequence analysis of known A domains, in relation to the crystal structure of the GrsA substrate-binding pocket, models have been developed to predict substrate specificity from eight or ten amino acids lining the pocket (Stachelhaus et al., 1999 ; Challis et al., 2000 ). Based on alignment of complete sequences between A4 and A5 motifs, the A domain from XabB falls in a diverse cluster involved in loading of aromatic amino acids (mostly Tyr), Leu or His.

The model using eight critical residues (http://jhunix.hcf.jhu.edu/~ravel/nrps/) gives no convincing prediction, because the XabB signature sequence AVKYVAND diverges from the strongly conserved residues D and V/I/L at the start and end of this eight-residue signature. Too few A domains have yet been characterized to allow reliable prediction by the model of nonproteinogenic substrates. However, an initial aspartic acid residue is highly conserved in amino-acid-specific A domains, where it is believed to electrostatically stabilize the amino group of the substrate at the binding pocket. The initial alanine residue in the case of XabB is consistent with a hydroxy-acid substrate, by analogy with the initial glycine in the signature sequence (GALHVVGI) of the hydroxyisovaleric-acid loading domain of enniatin synthetase (Haese et al., 1993 ). Incorporation of a hydroxy acid would involve an ester bond in the product chain, which is consistent with the cleavage and inactivation of albicidin by the AlbD esterase (Zhang & Birch, 1997 ).

Based on the architecture of the XabB NRPS region, we infer that the polyketide intermediate tethered on PCP1 is accepted by C1 and coupled to the amino-, hydroxy- or carboxy-acid unit preloaded by A onto PCP2. There are no evident peptide-modifying domains (e.g. epimerase or N-methyltransferase) in the NRPS module. The final C domain at the carboxyl-terminus of XabB may transfer the intermediate product assembled by XabB to a second synthetase protein required for albicidin production as in the mycosubtilin synthetase (Duitman et al., 1999 ). Alternatively, the final C domain may be involved in peptide-chain termination and cyclization, as in the enniatin, HC-toxin, rapamycin and FK506 systems (Konz & Marahiel, 1999 ), which also lack a TE motif for chain termination. The ORF downstream of XabB encodes a MT involved in albicidin biosynthesis (Huang et al., 2000a ), not a functionally linked NRPS.

The full chemical structure of albicidin has not yet been determined. Albicidin is insensitive to proteases, but it is cleaved and detoxified by the AlbD esterase (Birch & Patil, 1985b ; Zhang & Birch, 1997 ). Despite its relatively large size and lack of an apparent nucleoside motif in spectroscopic data, albicidin is actively accumulated in E. coli by illicit uptake through the nucleoside-specific Tsx outer-membrane pore (Birch et al., 1990 ; Fsihi et al., 1993 ). Partial chemical characterization of the major antimicrobial component of the albicidin family, by methods including proton- and C13-NMR and mass spectroscopy, indicates features including: an Mr of 842000; approximately 38 carbon atoms; three or four aromatic rings; at least one COOH group; two OCH3 groups, including a probable O-methyl tyrosine; a trisubstituted double bond; and a CN linkage (Birch & Patil, 1985a ; and unpublished analysis by Richard Moore at University of Hawaii and Mary Garson at University of Queensland).

The molecular logic of PKS and NRPS systems has generally been interpreted in the light of known product structures. We anticipate that the biosynthetic possibilities indicated by the architecture of XabB will assist in the interpretation of chemical data to solve the structure of albicidin. It is therefore interesting to consider the likely structure of the intermediate in albicidin biosynthesis produced by XabB, based on the domain structure of this unusual synthetase (Fig. 7). The O-methyl tyrosine component in NMR data could arise by action of the downstream XabC MT (Huang et al., 2000a ) on an intermediate like the one shown in Fig. 7(b).



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 7. Hypothetical structures of the intermediate in albicidin biosynthesis produced by XabB. (a) General structure, assuming that PCP1 acts only as a transfer point, and that a hydroxy acid is loaded by A onto PCP2, with n>=1 depending on the number of extender units added at ACP2 and ACP3. (b) Stepwise construction of an alternative possible structure, assuming that acetate extender units are added at ACP2 and ACP3, an amino acid residue is added at PCP1, and tyrosine is loaded by A onto PCP2. Loading of one amino acid and one hydroxy acid in the NRPS (black) region, resulting in an amide bond and an ester bond in the product chain, is also consistent with the chemical properties of albicidin.

 
XabB contains five carrier protein (ACP/PCP) domains, to which the growing polyketide or polypeptide chain could be covalently tethered. Each functional ACP or PCP domain must have a specific serine side-chain phosphopantetheinylated by a dedicated phosphopantetheinyl transferase (PPTase) (Lambalot et al., 1996 ). Recently, a gene (xabA) distant from xabB has been identified and found to encode a PPTase required for albicidin biosynthesis in X. albilineans (Huang et al., 2000b ). It is likely that XabB may be a target of XabA, although the ACP and/or PCP specificity of XabA to XabB remains to be experimentally determined.

The characterization of XabB as a multi-modular hybrid enzyme provides new insights into the mechanism of albicidin biosynthesis and possible approaches to engineer the overproduction of albicidins. For example, the complementation experiments (Fig. 4) indicate that increased copy number of xabB stimulates early production of albicidin, but other factors become limiting during idiophase. It may be possible to increase expression of the albicidin synthetase by modifications to the promoter and TTG start codon, or to improve albicidin yields by supplying candidate substrates. The unusual enzyme organization also contributes to the emerging understanding of how microbes generate structural diversity of antibiotics, which will open possibilities for combinatorial engineering of antibiotics of mixed polyketide/peptide origin.


   ACKNOWLEDGEMENTS
 
The authors would like to thank Allan Downie and Gregory Challis for comments on the manuscript, Jaques Ravel for ideas on A domain specificity, Richard Moore and Mary Garson for interpretations of albicidin structure and Marc Van Montagu for kindly providing plasmid pRG960sd. This work was supported by a University of Queensland Postgraduate Research Scholarship to G. Huang, and by grants from the Australian Research Council and the Sugar Research and Development Corporation.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Albertini, A. M., Caramori, T., Scoffone, F., Scotti, C. & Galizzi, A.(1995). Sequence around the 159° region of the Bacillus subtilis genome: the pksX locus spans 33·6 kb. Microbiology 141, 299-309.[Abstract]

Aparicio, J. F., Molnar, I., Schwecke, T., Konig, A., Haydock, S. F., Khaw, L. E., Staunton, J. & Leadlay, P. F.(1996). Organization of the biosynthetic gene cluster for rapamycin in Streptomyces hygroscopicus: analysis of the enzymatic domains in the modular polyketide synthase. Gene 169, 9-16.[Medline]

Birch, R. G. & Patil, S. S.(1983). The relation of blocked chloroplast differentiation to sugarcane leaf scald disease. Phytopathology 73, 1368-1374.

Birch, R. G. & Patil, S. S. (1985a). Antibiotic and process for the production thereof. US Patent 4525354, 25 June 1985.

Birch, R. G. & Patil, S. S.(1985b). Preliminary characterization of an antibiotic produced by Xanthomonas albilineans which inhibits DNA synthesis in Escherichia coli. J Gen Microbiol 131, 1069-1075.[Medline]

Birch, R. G. & Patil, S. S.(1987a). Correlation between albicidin production and chlorosis induction by Xanthomonas albilineans, the sugarcane leaf scald pathogen. Physiol Mol Plant Pathol 30, 199-206.

Birch, R. G. & Patil, S. S.(1987b). Evidence that an albicidin-like phytotoxin induces chlorosis in sugarcane leaf scald disease by blocking plastid DNA replication. Physiol Mol Plant Pathol 30, 207-214.

Birch, R. G., Pemberton, J. M. & Basnayake, W. V.(1990). Stable albicidin resistance in Escherichia coli involves an altered outer-membrane nucleoside uptake system. J Gen Microbiol 136, 51-58.[Medline]

Bradford, M. M.(1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein–dye binding. Anal Biochem 72, 248-254.[Medline]

Cane, D. E., Walsh, C. T. & Khosla, C.(1998). Harnessing the biosynthetic code: combinations, permutations, and mutations. Science 282, 63-68.[Abstract/Free Full Text]

Challis, G. L., Ravel, J. & Townsend, C. A.(2000). Predictive, structure-based model of amino acid recognition by nonribosomal peptide synthetase adenylation domains. Chem Biol 7, 211-224.[Medline]

Conti, E., Stachelhaus, T., Marahiel, M. A. & Brick, P.(1997). Structural basis for the activation of phenylalanine in the non-ribosomal biosynthesis of gramicidin S. EMBO J 16, 4174-4183.[Abstract/Free Full Text]

Ditta, G., Stanfield, S., Corbin, D. & Helinski, D. R.(1980). Broad host range DNA cloning system for Gram-negative bacteria: construction of a gene bank of Rhizobium meliloti. Proc Natl Acad Sci USA 77, 7347-7351.[Abstract]

Donadio, S., Staver, M. J., McAlpine, J. B., Swanson, S. J. & Katz, L.(1991). Modular organization of genes required for complex polyketide biosynthesis. Science 252, 675-679.[Medline]

Duitman, E. H., Hamoen, L. W., Rembold, M. & 10 other authors (1999). The mycosubtilin synthetase of Bacillus subtilis ATCC6633: a multifunctional hybrid between a peptide synthetase, an amino transferase, and a fatty acid synthase. Proc Natl Acad Sci USA 96, 13294–13299.[Abstract/Free Full Text]

Fsihi, H., Kottwitz, B. & Bremer, E.(1993). Single amino acid substitutions affecting the substrate specificity of the Escherichia coli K-12 nucleoside-specific Tsx channel. J Biol Chem 268, 17495-17503.[Abstract/Free Full Text]

Gehring, A. M., DeMoll, E., Fetherston, J. D., Mori, I., Mayhew, G. F., Blattner, F. R., Walsh, C. T. & Perry, R. D.(1998). Iron acquisition in plague: modular logic in enzymatic biogenesis of yersiniabactin by Yersinia pestis. Chem Biol 5, 573-586.[Medline]

Guenzi, E., Galli, G., Grgurina, I., Gross, D. C. & Grandi, G.(1998). Characterization of the syringomycin synthetase gene cluster: a link between prokaryotic and eukaryotic peptide synthetases. J Biol Chem 273, 32857-32863.[Abstract/Free Full Text]

Haese, A., Schubert, M., Herrmann, M. & Zocher, R.(1993). Molecular characterization of the enniatin synthetase gene encoding a multifunctional enzyme catalysing N-methyldepsipeptide formation in Fusarium scirpi. Mol Microbiol 7, 905-914.[Medline]

Hopwood, D. A.(1997). Genetic contributions to understanding polyketide synthases. Chem Rev 97, 2465-2497.[Medline]

Hopwood, D. A. & Sherman, D. H.(1990). Molecular genetics of polyketides and its comparison to fatty acid biosynthesis. Annu Rev Genet 24, 37-66.[Medline]

Huang, W., Jia, J., Edwards, P., Dehesh, K., Schneider, G. & Lindqvist, Y.(1998). Crystal structure of ß-ketoacyl-acyl carrier protein synthase II from E. coli reveals the molecular architecture of condensing enzymes. EMBO J 17, 1183-1191.[Abstract/Free Full Text]

Huang, G., Zhang, L. & Birch, R. G.(2000a). Analysis of the genes flanking xabB: a methyltransferase gene is involved in albicidin biosynthesis in Xanthomonas albilineans. Gene 255, 327-333.[Medline]

Huang, G., Zhang, L. & Birch, R. G.(2000b). Albicidin antibiotic and phytotoxin biosynthesis in Xanthomonas albilineans requires a phosphopantetheinyl transferase gene. Gene 258, 193-199.[Medline]

Jefferson, R. A.(1987). Assaying chimeric genes in plants: the GUS gene fusion system. Plant Mol Biol Rep 5, 387-405.

Keating, T. A. & Walsh, C. T.(1999). Initiation, elongation, and termination strategies in polyketide and polypeptide antibiotic biosynthesis. Curr Opin Chem Biol 3, 598-606.[Medline]

Konz, D. & Marahiel, M. A.(1999). How do peptide synthetases generate structural diversity? Chem Biol 6, R39-R48.[Medline]

Lambalot, R. H., Gehring, A. M., Flugel, R. S., Zuber, P., LaCelle, M., Marahiel, M. A., Reid, R., Khosla, C. & Walsh, C. T.(1996). A new enzyme superfamily – the phosphopantetheinyl transferases. Chem Biol 3, 923-936.[Medline]

McCarthy, J. E. G. & Gualerzi, C.(1990). Translational control of prokaryotic gene expression. Trends Genet 6, 78-85.[Medline]

MacNeil, D. J., Occi, J. L., Gewain, K. M., MacNeil, T., Gibbons, R. H., Ruby, C. L. & Danis, A. J.(1992). Complex organization of the Streptomyces avermitilis genes encoding the avermectin polyketide synthase. Gene 115, 119-125.[Medline]

Marahiel, M. A., Stachelhaus, T. & Mootz, H. D.(1997). Modular peptide synthetases involved in nonribosomal peptide synthesis. Chem Rev 97, 2651-2673.[Medline]

Mayorga, M. E. & Timberlake, W. E.(1992). The developmentally regulated Aspergillus nidulans wA gene encodes a polypeptide homologous to polyketide and fatty acid synthases. Mol Gen Genet 235, 205-212.[Medline]

Motamedi, H. & Shafiee, A.(1998). The biosynthetic gene cluster for the macrolactone ring of the immunosuppressant FK506. Eur J Biochem 256, 528-534.[Abstract]

Nowak-Thompson, B., Gould, S. J. & Loper, J. E.(1997). Identification and sequence analysis of the genes encoding a polyketide synthase required for pyoluteorin biosynthesis in Pseudomonas fluorescens Pf-5. Gene 204, 17-24.[Medline]

Paitan, Y., Alon, G., Orr, E., Ron, E. Z. & Rosenberg, E.(1999). The first gene in the biosynthesis of the polyketide antibiotic TA of Myxococcus xanthus codes for a unique PKS module coupled to a peptide synthetase. J Mol Biol 286, 465-474.[Medline]

Quadri, L. E. N, Sello, J., Keating, T. A., Weinreb, P. H. & Walsh, C. T.(1998). Identification of a Mycobacterium tuberculosis gene cluster encoding the biosynthetic enzymes for assembly of the virulence-conferring siderophore mycobactin. Chem Biol 5, 631-645.[Medline]

Ricaud, C. & Ryan, C. C. (1989). Leaf scald. In Diseases of Sugar Cane: Major Diseases, pp. 39–53. Edited by C. Ricaud, B. T. Egan, A. G. Gillaspie, Jr & C. G. Hughes. Amsterdam: Elsevier.

Rott, P. C., Costet, L., Davis, M. J., Frutos, R. & Gabriel, D. W.(1996). At least two separate gene clusters are involved in albicidin production by Xanthomonas albilineans. J Bacteriol 178, 4590-4596.[Abstract]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Sanger, F., Nicklen, S. & Coulson, A. R.(1977). DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 74, 5463-5467.[Abstract]

Scrutton, N. S., Berry, A. & Perham, R. N.(1990). Redesign of the coenzyme specificity of a dehydrogenase by protein engineering. Nature 343, 38-43.[Medline]

Shen, B.(2000). Biosynthesis of aromatic polyketides. Top Curr Chem 209, 1-51.

Stachelhaus, T., Mootz, H. D., Bergendahl, V. & Marahiel, M. A.(1998). Peptide bond formation in nonribosomal peptide biosynthesis – catalytic role of the condensation domain. J Biol Chem 273, 22773-22781.[Abstract/Free Full Text]

Stachelhaus, T., Mootz, H. D. & Marahiel, M. A.(1999). The specificity-conferring code of adenylation domains in nonribosomal peptide synthetases. Chem Biol 6, 493-505.[Medline]

Staskawicz, B., Dahlbeck, D., Keen, N. & Napoli, C.(1987). Molecular characterization of cloned avirulence genes from race 0 and race 1 of Pseudomonas syringae pv. glycinea. J Bacteriol 169, 5789-5794.[Medline]

Takano, Y., Kubo, Y., Shimizu, K., Mise, K., Okuno, T. & Furusawa, I.(1995). Structural analysis of PKS1, a polyketide synthase gene involved in melanin biosynthesis in Colletotrichum lagenarium. Mol Gen Genet 249, 162-167.[Medline]

Tang, L., Yoon, Y. J., Choi, C. & Hutchinson, C. R.(1998). Characterization of the enzymatic domains in the modular polyketide synthase involved in rifamycin B biosynthesis by Amycolatopsis mediterranei. Gene 216, 255-265.[Medline]

Turgay, K., Krause, M. & Marahiel, M. A.(1992). Four homologous domains in the primary structure of GrsB are related to domains in a superfamily of adenylate-forming enzymes. Mol Microbiol 6, 529-546.[Medline]

Van den Edde, G., Deblaere, R., Goethals, K., Van Montagu, M. & Hosters, M.(1992). Broad host range and promoter selection vectors for bacteria that interact with plants. Mol Plant–Microbe Interact 5, 228-234.[Medline]

Van Larebeke, N., Genetello, C., Hernalsteens, J. P., Depicker, A., Zaenen, I., Messens, E., Van Montagu, M. & Schell, J.(1977). Transfer of Ti plasmids between Agrobacterium strains by mobilization with the conjugative plasmid RP4. Mol Gen Genet 152, 119-124.

Wall, M. K. & Birch, R. G.(1997). Genes for albicidin biosynthesis and resistance span at least 69 kb in the genome of Xanthomonas albilineans. Lett Appl Microbiol 24, 256-260.[Medline]

Xiao, Y., Lu, Y., Heu, S. & Hutcheson, S. W.(1992). Organization and environmental regulation of the Pseudomonas syringae pv. syringae 61 hrp cluster. J Bacteriol 174, 1734-1741.[Abstract]

Yu, J. H. & Leonard, T. J.(1995). Sterigmatocystin biosynthesis in Aspergillus nidulans requires a novel type I polyketide synthase. J Bacteriol 177, 4792-4800.[Abstract]

Zhang, L. & Birch, R. G.(1997). The gene for albicidin detoxification from Pantoea dispersa encodes an esterase and attenuates pathogenicity of Xanthomonas albilineans to sugarcane. Proc Natl Acad Sci USA 94, 9984-9989.[Abstract/Free Full Text]

Zhang, L., Xu, J. & Birch, R. G.(1998). Factors affecting biosynthesis by Xanthomonas albilineans of albicidin antibiotics and phytotoxins. J Appl Microbiol 85, 1023-1028.[Medline]

Zhang, L., Xu, J. & Birch, R. G.(1999). Engineered detoxification confers resistance against a pathogenic bacterium. Nature Biotechnol 17, 1021-1024.[Medline]

Zocher, R. & Keller, U.(1997). Thiol template peptide synthesis systems in bacteria and fungi. Adv Microb Physiol 38, 85-131.[Medline]

Received 14 June 2000; revised 26 December 2000; accepted 8 November 2000.