Introduction

Serious infections caused by antibiotic-resistant bacteria have become a major global problem, indicating the need for further innovation in antimicrobial research and development to provide the next generation of antimicrobial drugs. The biosynthesis of secondary metabolites requires multi-step enzymatic pathways starting with intermediates of primary metabolites as precursors [1]. Genes involved in the biosynthesis of antibiotics and other secondary metabolites have been cloned and characterized from a wide variety of organisms in recent decades, revealing some of their complex genetic organization and biosynthetic mechanisms [1,2,3].

Toxins produced by phytopathogenic bacteria increase the severity of disease in plants. Several of the phytotoxins from pseudomonads also possess antibacterial activity [4, 5]. In many cases, the structure of toxins, the nature of intermediates in their biosynthesis and their mode of action in plant diseases are known.

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) [6,7,8,9]. Furthermore, albicidin inhibits the in vitro supercoiling activity of Escherichia coli DNA gyrase, with IC50 (40–50 nM) below most coumarins and quinolones by blocking the religation of the cleaved DNA intermediate during the gyrase catalytic sequence and inhibits the relaxation of supercoiled DNA by gyrase and topoisomerase IV [10]. Recently, a gene cluster spanning more than 50 kb in the genome of X. albilineans associated with albicidin production has been cloned [11,12,13,14,15]. Most importantly the chemical structure for albicidin has been solved and de nova synthesis has been achieved.

In this review, the relationship between albicidin and sugarcane leaf-scald disease will first be described. Structure and biosynthesis of albicidin will be discussed. Mechanisms of albicidin action and resistance will then be summarized.

Albicidin and sugarcane leaf-scald disease

Sugarcane leaf-scald disease

Leaf-scald disease is a major disease of sugarcane (Saccharum spp. hybrids), with the potential to cause severe losses of cane yield and quality in susceptible cultivars. It is a vascular disease caused by the gram-negative bacterium X. albilineans [9, 16]. Leaf-scald disease has been reported in more than 50 countries [16] and continues to spread into new areas [17,18,19]. Symptoms including the emergence of chlorotic leaves, wilting, necrosis and sometimes rapid death of plants, often appear after a prolonged latent period [20]. In its latent period, the pathogen can remain dormant until environmental conditions are favorable for symptom expression. However, when the disease enters its acute phase, entire blocks of apparently healthy cane can die off and become completely unharvestable over a period of only a few weeks. The pathogen is mechanically transmitted during harvesting, and can also spread naturally, for example by wind-blown exudates under cyclonic conditions [16]. X. albilineans infects several grasses other than sugarcane. Control of the disease is based on the use of resistant varieties as this reservoir of the pathogen in weed grasses makes eradication virtually impossible [16].

Electronic microscopic examination of the structure of infected leaves revealed that the pathogen is confined to the xylem vessels or adjacent intercellular spaces during the early stages of disease development [21]. Mature chloroplasts were absent from white leaf tissue and the plastids in these tissues were proplastids, etioplasts, and vesicular forms smaller than chloroplasts [21]. Birch and Patil [7] suggest that the characteristic white pencil lines and chlorosis of emerging leaves are due to the production by X. albilineans of a diffusible phytotoxin which selectively blocks chloroplast differentiation.

Albicidin antibiotics and phytotoxins

Chlorosis inducing isolates of X. albilineans produce a family of potent antibiotics in culture. The major antimicrobial component, called albicidin, is a low molecular weight compound with several aromatic rings [7]. The compound is soluble in polar organic solvents and partially soluble in water. Albicidin is rapidly bactericidal at nanomolar concentrations against a wide range of gram-positive and gram-negative bacteria (Table 1), but it shows no cytotoxicity to cultured mammalian cells at 8 µg ml−1 [22]. Albicidins are therefore of great interest as potential clinical antibacterial drugs. This clinical interest is further heightened by the recent discovery that malaria and toxoplasma parasites contain vestigial plastids that are essential for their survival [21].

Table 1 Activity of albicidin against gram-positive and gram-negative bacteria
Full size table

Albicidin-deficient (Tox−) mutants of X. albilineans fail to cause chlorosis or any other symptoms of leaf-scald disease in inoculated sugarcane [6, 8]. Transgenic sugarcanes that express the albicidin detoxification enzyme (AlbD) block systemic disease [9, 23]. These results indicate that albicidin phytotoxins are responsible for the characteristic chlorotic symptoms in X. albilineans infected sugarcane and that they play an overall role in systemic disease development by weakening host defenses [23].

Effect of albicidin on chloroplasts

It was postulated that albicidin might cause the blocked chloroplast differentiation, in diseased plants, as chloroplasts exhibit prokaryotic-like mechanisms of DNA replication [6]. Furthermore, it has been shown that plant chloroplasts have functional DNA gyrase that shows sensitivity to quinolones and as such, this is the most likely target of albicidin [24, 25]. Albicidin mutants and revertants showed a correlation between albicidin production and ability to cause disease [6]. An albicidin resistant bacterial strain of Pantoea dispersa (SB1403), showed a strong capacity for enzymatic detoxification of albicidin [26]. Susceptible sugarcane plants co-inoculated with P. dispersa showed a 98% reduction in the frequency of white pencil lines, even with a tenfold excess of X. albilineans inoculum. Also, the pathogen could not be reisolated and mutants of P. dispersa that failed to produce the detoxification enzyme were less effective in biocontrol [20, 26].

Albicidin is a potent inhibitor of DNA gyrase

Early studies suggested that albicidins block DNA replication in bacteria and sugarcane proplastids and inhibit replication of bacteriophage T4 and T7 [6, 7, 22]. The primary mode of action was shown to be a rapid and complete block of DNA synthesis. Albicidin also resulted in partial inhibition of RNA and protein synthesis, but this probably reflects decreasing cell viability [22]. Furthermore, albicidin did not appear to bind directly to DNA, as no change in the absorption spectra is observed on mixing albicidin with E. coli DNA [22]. The kinetics of DNA replication inhibition by albicidin closely resemble the effects of inhibition of DNA gyrase by the coumarin and quinolone antibiotics [22, 27, 28]. This selective inhibition of DNA synthesis without binding suggested a specific interaction of albicidin with an essential DNA replication protein [22].

Recent work by Hashimi et al. [10] showed that albicidin is a potent inhibitor of DNA Gyrase with minimum inhibitory concentrations less than most quinolones. Albicidin was shown to inhibit the religation of the cleaved DNA intermediate during the gyrase catalytic sequence, and also inhibit the relaxation of supercoiled DNA by gyrase and topoisomerase IV. Furthermore, E. coli strains harboring quinolone (GyrA S83L) and CcdB (GyrA R462C) resistant mutations showed cross-resistance to albicidin suggesting a similar mechanism of action.

Albicidin resistance mutants of E. coli have not demonstrated cross-resistance to inhibitors of the gyrase subunits, or to other DNA replication inhibitors [22]. This might be hampered by the fact that albicidin resistant E. coli arise at high frequency by mutations in tsx, the gene for an outer membrane protein involved in the active uptake of nucleosides from the surrounding medium [29, 30].

DNA gyrase as a drug target

DNA gyrase is known to be the target of several classes of antimicrobial agents [31]. Quinolones are synthetic compounds that target DNA gyrase A and act by stabilizing the DNA gyrase-DNA cleavage complex, thus inhibiting DNA supercoiling. Shortly after binding the Gyrase-DNA complex, quinolones induce a conformational change in the enzyme. Once the gyrase has cleaved the DNA, the quinolone traps this complex and prevents the religation of the DNA strands. Consequently, the quinolone-gyrase-DNA complex inhibits DNA replication [32].

In contrast, coumarins (coumermycin A1 and novobiocin) are natural products of Streptomyces which inhibit DNA gyrase by competing with ATP for binding to the Gyrase B subunit [27, 31, 33]. Enzymatic analysis of the 43 kDa amino-terminal fragment of GyrB shows that it contains a coumarin sensitive ATPase activity. It has been shown that novobiocin is a noncompetitive inhibitor of the ATPase activity and binds a monomer of the 47 kDa GyrB fragment, whereas coumermycin which resembles a dimer of novobiocin, forms a complex with a dimer of the 47 kDa fragment [31].

Selective site-specific mutagenesis of DNA gyrase has revealed that mutations which confer resistance to quinolones map to both gyrase A and B subunits, while those conferring resistance to coumarins map to gyrase B [31]. Mutations which confer resistance to these drugs are shown in Fig. 1.

Fig. 1
figure 1

Domain organization of E. coli DNA gyrase. The a GyrA (97 kDa) and b GyrB (90 kDa) proteins are represented as linear blocks with proposed domain boundaries indicated. Key amino acids referred to in the text are also shown. Amino acid 122 in GyrA is the active-site tyrosine. Amino acids whose mutation leads to drug resistance are numbered in italics. Quinolone-resistance mutations map to GyrA (86–106) and GyrB (426 and 447). Mutation at GyrB (136) confers coumarin resistance, a mutation at GyrB (751) confers Microcin B17 resistance and mutations at GyrA (214 and 462) confer resistance to the F plasmid protein CcdB. Adapted from [31, 68,69,70]

Full size image

Other agents which target DNA gyrase include microcin B17, CcdB, simocyclinones, cyclothialidine, cinodine and clerocidin [31, 34, 35]. Microcin B17, CcdB, and clerocidin inhibit the supercoiling assay by stabilizing the Gyrase-DNA complex [31, 36]. In contrast, simocyclinones inhibit an early catalytic step of the gyrase by interfering with enzyme-DNA binding [34]. Cyclothialidine inhibits the ATPase activity of gyrase by competing for the ATP binding site [37]. The precise mode of action of these agents with gyrase is incompletely understood, although resistance to some of these compounds has been mapped to a specific unit of gyrase.

Structure and function of the albicidin biosynthetic pathway

Transposon mutagenesis revealed that at least two gene clusters spanning more than 60 kb in the genome X. albilineans are involved in albicidin production [38, 39]. Subsequently, three genes required for albicidin biosynthesis were identified, cloned, and sequenced from two Queensland strains of X. albilineans [14, 40, 41]. A study with X. albilineans strain Xa23R1 from Florida revealed that three gene clusters, containing a total of 22 open reading frames (ORFs), are involved in albicidin biosynthesis [11, 42]. Figure 2 shows the major albicidin biosynthetic cluster containing 19 ORFs.

Fig. 2
figure 2

Physical map and genetic organization of the DNA region containing the albicidin gene cluster involved in albicidin production. The location and direction of the 19 open reading frames (ORFs) identified in the gene cluster are shown by thick arrows. The three polyketide synthase and nonribosomal peptide synthase genes are shown by black and thick arrows. Modified from [11] to show the arrangement of genes involved in albicidin biosynthesis

Full size image

Nonribosomal peptide and polyketide synthases

Polyketide synthases (PKSs) and nonribosomal peptide synthetases (NRPSs) are structurally and mechanistically related to fatty acid synthases, all of which catalyze the synthesis of biopolymers in the absence of a nucleic acid or other templates [43].

In an NRPS the modules can be further subdivided into three functional domains: the adenylation domain (A) activates amino acids, a peptidyl carrier protein (PCP) attaches the growing polypeptide and a condensation domain (C) that catalyzes the formation of the peptide bond [43]. By analogy, a PKS module consists of three domains: the acyltransferase domain activates acyl-CoA; the acyl carrier protein (ACP) domain tethers the growing polyketide and the ketosynthase domain (KS) forms the bond. Each NRPS or PKS system also has a domain for loading a starter unit onto the first PKS/NRPS module and a chain-terminating thioesterase (TE) domain [44].

The NRPS substrate-binding pockets are so highly specific for their substrates that predictive models based upon consensus signature motifs for known substrates have been determined. Substrate specificity is determined at the binding pocket, consisting of a stretch of about 100 amino acid residues between highly conserved motif A4 and A5 [45]. 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 8 or 10 amino acids lining the pocket [46, 47].

Sequence analysis of the X. albilineans biosynthetic region showed that it contains one large multifunctional fused PKS-NRPS gene (designated xabB/AlbI) and two small NRPS genes (designated albIV and albIX) [11, 14]. xabB (syn. albI) encodes an enzyme of 6879 amino acids (aa) with several domains involved in polyketide and peptide synthesis. The PKS region of XabB is divided into three modules (Fig. 3). The module designated PKS-1 contains acyl-CoA ligase and acyl carrier protein (ACP1) domains. The module designated PKS-2 contains ®-ketoacyl synthase (KS1) and ®-ketoacyl reductase (KR) domains followed by two consecutive ACP domains (ACP2 and ACP3). The module designated PKS-3 contains a KS domain (KS2) followed by a PCP domain (PCP1) [11, 14].

Fig. 3
figure 3

Model for the synthesis of albicidin by the three polyketide synthase (PKS) modules and the seven nonribosomal peptide synthase (NRPS) modules identified in albicidin biosynthetic cluster. Boxes with dotted lines around the chemical structure of albicidin indicate the stepwise synthesis by PKSs and NRPSs. Abbreviations: A adenylation, ACP acyl carrier protein, AL acyl-CoA ligase, C condensation, KR ®-ketoacyl reductase, KS ®-ketoacyl synthase, NRPS nonribosomal peptide synthase, PCP peptidyl carrier protein, PKS polyketide synthase, TE thioesterase. The question mark in the NRPS-2 domain indicates that this A domain is incomplete. Modified from [11]

Full size image

The PKS module of XabB is connected to four NRPS modules, by the PCP1 domain. The first three NRPS domains contain the general order of C, A and PCP domains typical of such enzymes [48] and NRPS-4 contains only a single C domain. The NRPS region of XabB contains a duplicated region corresponding to NRPS-1 and NRPS-3. This indicates that two identical amino acids are added to the growing chain by NRPS-1 and NRPS-3 separated by an amino acid that is added by NRPS-2 [11].

AlbIV forms one NRPS which contains a single A domain followed by a PCP domain. NRPS AlbIX contains two NRPS modules, the first module containing an A and a PCP domain while the second module contains a C, A and PCP domain. Interestingly the A and PCP domains of both NRPS modules in AlbIX are identical, implying that they load the same amino acid onto the growing albicidin chain. At the end of the polypeptide is a TE domain responsible for terminating the growing chain [11].

Modifying and regulatory enzymes

Once the polypeptide chains have been released, they frequently undergo further enzymatic modification by ancillary enzymes (e.g., methyltransferases, hydroxylases and glycosyl transferases). This subsequent modification is generally required for the final product to be biologically active [49]. Most tailoring enzymes are dedicated to the biosynthetic pathway itself and are encoded by genes that are clustered with the core PKS and NRPS genes [48, 50].

The major biosynthetic gene cluster of albicidin contains two methyltransferases (xabC and xabD) that utilize S-adenosyl-methionine as a co-substrate for O-methylation of small molecules [51,52,53]. Evidence from insertional mutagenesis and complementation proves that xabC is involved in albicidin biosynthesis in X. albilineans [41]. There are four other ORFs designated xabJ (possible alpha/beta fold hydrolase COG0596, with 39% similarity over 260 aa to GenBank Accession AA054683), xabK (probable benzoyl-CoA oxygenase OG3396, with 54% similarity over 426 aa to AAN39376), xabL (probable dienelactone hydrolase COG4188, with 58% similarity over 278 aa to ZP_00172356) [54] and albV (probable carbamoyl transferase, with 46% similarity over 441 aa to AAG02370 from Streptomyces verticillus [11].

The production of antibiotics is associated with tight regulation of expression in the producing organism. This regulation involves switching on the biosynthetic genes during the correct phase of growth and controlling the export of the antibiotic. The biosynthetic gene cluster of albicidin contains one regulatory gene (designated albVIII) [11]. Analysis of albVIII showed it is most similar to the syringomycin regulatory gene syrP from Pseudomonas syringae. Searches of protein sequence databases demonstrated that syrP was most similar to histidine kinases such as the CheA regulatory protein of E. coli. Site-directed insertional mutagenesis of the syrP gene [55], exhibited an unusual pleiotropic phenotype including a failure to produce syringomycin in liquid media in contrast to the production of elevated levels of the toxin on agar media. Furthermore, the syrP mutant was relieved of the suppression of toxin production that accompanies inorganic phosphate concentrations of >1 mM on agar media [55].

Precursor synthesis genes

Polyketide and polypeptide antibiotic biosynthesis begin with activation of starter units followed by elongation of the antibiotic backbone by the large PKS and NRPS enzymes. The albicidin biosynthetic cluster contains several genes which may have a role in the production of precursor molecules. Genes designated ubiC and pabAB show similarity to PHBA (para-hydroxybenzoate) synthase and PABA (para-aminobenzoate) synthase respectively [56, 57]. Furthermore, these albicidin precursors might be activated by another gene xabE (syn. AlbVII) which showed similarity to benzoate CoA ligase.

Resistance genes

Antibiotic biosynthetic genes are often clustered with one or more genes conferring resistance to the antibiotic in bacteria [1]. The major albicidin biosynthetic cluster contains two genes for self-protection. albF is an ABC transporter involved in the active efflux of albicidin. Expression of albF in E. coli increased albicidin resistance by 30–50 fold [54]. Another gene albG shows similarity to quinolone-resistance gene qnr. Expression of albG in E. coli conferred significant albicidin resistance. Furthermore, the addition of purified AlbG to the in vitro DNA gyrase supercoiling assay conferred albicidin resistance suggesting an interaction with the DNA gyrase subunits [10].

Resistance to albicidin is conferred by genes present in other bacteria. AlbD produced by Pantoea dispersa is an endopeptidase that directly cleaves the peptide bond of albicidin and subsequently rendering it inactive [58]. While, AlbA synthesized by Klebsiella oxytoca binds to albicidin and completely inhibits its activity [59, 60].

X. albilineans DNA gyrase

As albicidin is a potent inhibitor of DNA gyrase, the albicidin producer X. albilineans would require resistance mechanisms for self-protection. In vitro supercoiling assays with purified DNA gyrase subunits of X. albilineans showed a 25-fold resistance to albicidin and ciprofloxacin in comparison to the E. coli DNA gyrase [61]. However, when the X. albilineans DNA gyrase A subunit was substituted with E. coli DNA gyrase A in the assay, this combination was as sensitive as the E. coli DNA gyrase AB suggesting that resistance to albicidin is conferred by the DNA gyrase A subunit.

Structure of albicidin

Structural analysis of albicidin has shown that it consists of p-aminobenzoic acids and cyanoalanine with a chemical formula [M+H]C44H39O12N6 and molecular weight of 843.8260 Da [56]. It consists of the nonproteinogenic α-amino acid β-cyano-l-alanine (Cya-3), the aromatic δ-amino acids p-aminobenzoic acid (pABA-2 and pABA-4) and 4-amino-2-hydroxy-3-methoxybenzoic acid (pMBA-5 and pMBA-6). The main structure is linked with 3-(4-hydroxyphenyl)-2-methyl acrylic acid at the N-terminal (Fig. 3).

Strategies for improving the production of albicidin

Albicidin is a potent antibiotic that with minimum inhibitory concentrations significantly less than many current DNA gyrase inhibitors. However, albicidin production is limited from X. albilineans and other strategies are needed to improve the production of albicidin to realize its clinical potential.

Heterologous production

Heterologous biosynthesis of compounds in bacteria relying on transforming the complete biosynthetic cluster into a host bacterium can be used to increase production. Since the complete genome of X. albilineans has been sequenced and the albicidin biosynthetic cluster has been identified, heterologous expression of albicidin would be feasible. Recently, the complete biosynthetic cluster of albicidin was transferred to Xanthomonas axonopodis pv. Vesicatoria [62]. Albicidin production was increased sixfold compared to X. albilineans suggesting a promising strategy for engineering overproduction. The advantage of the heterologous host offers a fast-growing bacterium which is easily amenable to genetic modification compared to X. albilineans. Further modifications to improve albicidin production includes: (i) addition of constitutive promoters to upregulate transcription (ii) the use of codon optimization to improve translation and (iii) addition of albicidin efflux pumps to improve albicidin secretion.

Total synthesis of albicidin

Another strategy to improve albicidin production is chemical synthesis. In 2015, a convergent total synthesis approach was used to synthesize albicidin [63]. Three different fragments of albicidin were synthesized in this strategy; (i) an N-terminal coumaric acid, (ii) a central tripeptide and (iii) C-terminal dipeptide. The central tripeptide was coupled to the C-terminal peptide by BTC-mediated coupling resulting in a pentapeptide. Finally, quantitative BTC-mediated coupling to coumaric acid and global allyl deprotection resulted in albicidin. Antimicrobial activity of the synthesized albicidin was shown to be in accordance with that of natural albicidin (IC50: 40 nM).

The structure and chemical synthesis of albicidin open the possibility of synthesis of albicidin derivatives with improved biological activity. Albicidins were produced with N-terminal acylation to improve antimicrobial activity [64]. Fourteen derivatives were synthesized with variable cinnamoyl, phenylpropanoyl, and benzoyl N-terminal residues. Derivatives with substitutions in the para-position of benzoyl N-terminal group were shown to be the most significant in terms of activity while short N-acetylated derivatives showed significantly reduced activity.

Modification to the central amino acid of albicidin with various amino acids was performed to determine the effect of charge, chirality, and steric bulk on antimicrobial activity [65]. It was found that charged amino acids reduce albicidin activity while uncharged amino acids retain activity. Threonine was found to be the most promising in increasing albicidin activity.

A recent study used LC–MS/MS bioactivity-guided spectral networking analysis of albicidin extracts from X. albilineans to identify eight different natural albicidins with differential activity against gram-positive and gram-negative bacteria [66].

The albicidin biosynthetic cluster also contains an O-Carbamoyl-Transferase (Alb15) which carbamoylates albicidin at the N-terminal resulting in carbamoyl-albicidin [67]. DNA gyrase supercoiling assays showed that carbamoyl-albicidin was six times more potent in inhibiting bacterial gyrase activity when compared to albicidin. In vivo assays showed that carbamoyl-albicidin was differential in its effects on gram-negative bacteria, while it showed similar activity to albicidin in gram-positive strains [67].

Conclusion

Albicidin is a potent DNA gyrase inhibitor with concentrations significantly less than current DNA gyrase inhibitors used clinically. Recent studies have identified the complete biosynthetic cluster of albicidin, produced albicidin in heterologous hosts, solved the structure, and chemically synthesized albicidin and its derivatives. These recent advancements should help to realize the clinical potential of albicidin as new antibiotics are urgently needed to combat multi-antibiotic-resistant bacteria.