SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

The 18-membered polyketide macrolide borrelidin exhibits a number of important biological activities, including potent angiogenesis inhibition. This has prompted two recent total syntheses as well as the cloning of the biosynthetic gene cluster from Streptomyces parvulus Tü4055. Borrelidin possesses some unusual structural characteristics, including a cyclopentane carboxylic acid moiety at C17 and a nitrile moiety at C12 of the macrocyclic ring. Nitrile groups are relatively rare in nature, and little is known of their biosynthesis during secondary metabolism. The nitrile group of borrelidin is shown here to arise from the methyl group of a methylmalonyl-CoA extender unit incorporated during polyketide chain extension. Insertional inactivation of two genes in the borrelidin gene cluster, borI (coding for a cytochrome P450 monooxygenase) and borJ (coding for an aminotransferase), generated borrelidin non-producing mutants. These mutants accumulated different compounds lacking the C12 nitrile moiety, with the product of the borI-minus mutant (12-desnitrile-12-methyl-borrelidin) possessing a methyl group and that of the borJ-minus mutant (12-desnitrile-12-carboxyl-borrelidin) a carboxyl group at C12. The former but not the latter was converted into borrelidin when biotransformed by an S. parvulus mutant that is deficient in the biosynthesis of the borrelidin starter unit. This suggests that 12-desnitrile-12-methyl-borrelidin is a competent biosynthetic intermediate, whereas the carboxylated derivative is a shunt metabolite. Bioconversion of 12-desnitrile-12-methyl-borrelidin into borrelidin was also achieved in a heterologous system co-expressing borI and borJ in Streptomyces albus J1074. This bioconversion was more efficient when borK, which is believed to encode a dehydrogenase, was simultaneously expressed with borI and borJ. On the basis of these findings, a pathway is proposed for the formation of the nitrile moiety during borrelidin biosynthesis.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Borrelidin is an 18-membered macrolide polyketide produced by several streptomycete species. The structure of borrelidin (Fig. 1A) was first elucidated in 1967, refined by detailed nuclear magnetic resonance (NMR) analysis and confirmed by X-ray crystallography (Keller-Schierlein, 1967; Anderson et al., 1989; Kuo et al., 1989). Borrelidin is identical to the antibiotic treponemycin (Maehr and Evans, 1987). Borrelidin was discovered as a result of its antibacterial activity (Berger et al., 1949), an activity that involves selective inhibition of threonyl-tRNA synthetase (Paetz and Nass, 1973). Activity of borrelidin against spirochaetes of the genus Treponema (Singh et al., 1985), viruses (Dickinson et al., 1965) and drug-resistant strains of Plasmodium falciparum (Otoguro et al., 2003) has also been described. Recently, borrelidin has returned to prominence with the report that borrelidin displays antiangiogenesis activity (Wakabayashi et al., 1997). In a rat aorta matrix culture model of angiogenesis, borrelidin exhibited a potent angiogenesis-inhibiting effect causing disruption of formed capillary tubes in a dose-dependent manner by inducing apoptosis of the capillary-forming cells with an IC50 value of 0.8 nM (Wakabayashi et al., 1997). The IC50 value for antiangiogenesis activity is 50-fold lower than that for inhibition of protein synthesis via threonyl-tRNA synthetase inhibition, indicating different molecular targets for the compound (Wakabayashi et al., 1997). Borrelidin also displayed potent inhibition of angiogenesis in vivo in a mouse dorsal air sac model, which examines vascular endothelial growth factor (VEGF)-induced angiogenesis, and was very effective in the inhibition of the formation of spontaneous lung metastases of B16-BL6 melanoma cells at the same dosage that inhibited angiogenesis (Funahashi et al., 1999). Borrelidin has also been identified as an inhibitor of cyclin-dependent kinase Cdc28/Cln2 of Saccharomyces cerevisiae (Tsuchiya et al., 2001). Recently, borrelidin has been shown to be involved in the inhibition of growth of human umbilical vein endothelial cells by a threonine-dependent mechanism (probably via inhibition of threonyl-tRNA synthetase) and to induce apoptosis by activation of caspases 3 and 8 (Kawamura et al., 2003). Taken together, these data suggest that borrelidin has potential as a lead compound for the development of novel antitumour agents.

image

Figure 1. A. Structure of borrelidin. B. Organization of the borrelidin gene cluster showing in detail the region containing the borI, borJ and borK genes. A, AvrII; B, BamHI; Bg, BglII; E, EcoRI; K, KpnI; M, MfeI; Mu, MluI; N, NsiI; S, SphI. The grey triangles indicate the location of the insertion of the aac(3)IV gene in single restriction sites or by causing a deletion (indicated by horizontal brackets). The crossed BglII site was eliminated by Klenow polymerase treatment and religation. For details on the organization of the gene cluster, see Olano et al. (2004).

Download figure to PowerPoint

Borrelidin contains a non-glycosylated, macrocyclic polyketide lactone ring and exhibits structural features not frequently found in natural products: the presence of a nitrile moiety at C12 and a 1,2-trans-substituted cyclopentane carboxylic acid moiety attached to C17 (Fig. 1A). Based on its chemical structure, it was predicted that biosynthesis of the borrelidin polyketide skeleton must be performed by a modular polyketide synthase (PKS) that would incorporate trans-cyclopentane-1,2-dicarboxylic acid as the starter unit and seven extension units (three malonyl-CoA and four methylmalonyl-CoA units). This has been verified recently through the cloning, sequencing and characterization of the borrelidin biosynthetic gene cluster from Streptomyces parvulus Tü4055 (Olano et al., 2003; 2004). Here, we report the identification and characterization of the genes involved in the formation of the nitrile moiety and the generation of novel borrelidin analogues by insertional gene inactivation.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

The origin of the carbon atom of the nitrile moiety

Borrelidin contains a nitrile moiety at C12 of the macrolide ring. Sequence analysis of the AT domain of module 3 of the borrelidin PKS (Olano et al., 2004) indicates that the substrate used for the third round of chain extension in borrelidin biosynthesis is methylmalonyl-CoA. This is based on the presence of the active site motif GXSXG (Haydock et al., 1995) and also the expected motif (YASH) for the selection of methylmalonyl-CoA (Reeves et al., 2001; Del Vecchio et al., 2003). Thus, the carbon atom of the nitrile moiety most probably arises from the methyl group of methylmalonyl-CoA. To investigate this further, stable isotope feeding experiments were carried out using sodium [2,3-13C2]-propionate and sodium [1,2-13C2]-acetate as precursors for the methylmalonyl-CoA and malonyl-CoA extender units respectively (Simpson, 1997). These compounds were fed, separately, to growing cultures of S. parvulus Tü4055 and, after 5 days, the resulting cultures were harvested, and the borrelidin produced was extracted and purified. The resultant borrelidin was analysed by inverse-gated 13C-NMR spectroscopy. Using an inverse-gated pulse sequence, it is possible to integrate the peak areas and thus estimate the incorporation level. These data clearly indicate the intact and specific incorporation of malonyl-CoA extender units during the first, second and eighth rounds (all with the same level of incorporation at one-fifth that of natural abundance) and methylmalonyl-CoA extender units during the third to seventh rounds of polyketide chain extension (all with the same level of incorporation at twofold that of natural abundance). Significantly, within the limits of the NMR spectroscopy methods used, it was clear that the carbon atoms arising from C2 and C3 of the fed propionate units were labelled in equal amounts. Therefore, the label was diluted into and taken from one pool only, i.e. methylmalonyl-CoA. If the nitrile group originated from an unusual malonate unit containing an α-nitrile group, or some such intermediate, then the intact incorporation into C12 and C12a would not be the same as in the other propionate positions. This strongly indicates that methylmalonyl-CoA is the extender unit incorporated by module three of the borrelidin PKS.

Gene candidates required for formation of the nitrile moiety

Taken together with the results described above, an examination of the bor gene cluster (Fig. 1B) led us to the hypothesis that formation of the nitrile moiety during borrelidin biosynthesis involves a biological route unprecedented for microorganisms. Of most interest to us were the products of two genes, borI and borJ (Fig. 1B). The borI gene product shares the highest sequence similarity with the cytochrome P450 monooxygenase TylHI from Streptomyces fradiae (41% identity), which catalyses the hydroxylation of an exocyclic methyl group of a macrolactone before addition of a deoxyhexose moiety during tylosin biosynthesis (Fouces et al., 1999), and with two other cytochrome P450 monooxygenases, SuaC from Streptomyces griseolus (accession number AAA26823; 39% identity) and P-450sca-2 from Streptomyces carbophilus involved in pravastatin biosynthesis (Watanabe et al., 1995; 38% identity). The C-terminus of BorI contains a 10-amino-acid signature (FSHGPHQCLG) that is specific to cytochromes P450, with a conserved cysteine residue involved in binding the haem iron in the fifth co-ordination site (Poulos et al., 1987). Based on these similarities, we propose that BorI is a cytochrome P450 hydroxylase that catalyses oxidation of the C12-methyl group during borrelidin biosynthesis. The borJ gene product resembles numerous pyridoxal-phosphate (PLP)-dependent aminotransferases, with the highest similarities occurring for BioA from Kurthia sp. 538-KA26 (Kiyasu et al., 2001; 39% identity), a class III aminotransferase encoded by gene BA_4800 from Bacillus anthracisA2012 (Read et al., 2002; 38% identity), and the adenosylmethionine-8-amino-7-oxononanoate aminotransferase BioA from Bacillus subtilis (accession number BAC03240; 40% identity). These PLP-dependent aminotransferases exhibit a specific amino acid motif, present in BorJ (DLMTVGKALTGG), around the lysine residue (underlined) of the PLP attachment site (Yonaha et al., 1992). BorJ is therefore a candidate to participate in the formation of the borrelidin nitrile group, at the stage of incorporation of the nitrogen atom.

Insertional inactivation of borI and borJ

To assess the involvement of the borI and borJ gene products in nitrile formation, we individually inactivated both genes by inserting an apramycin resistance cassette into the chromosome by gene replacement. For borI, the aac(3)IV gene (conferring apramycin resistance; AmR) was inserted within the coding region and, for borJ, a 453 bp internal fragment of the coding region was replaced by the aac(3)IV gene (Figs 2A and 3A). After transformation of protoplasts of the wild-type S. parvulus strain and selection of primary transformants, HygS-AmR colonies were obtained as the result of a double cross-over event. This was confirmed by Southern hybridization (Figs 2B and 3B). One mutant from each experiment (designated M19B9I1 for borI and M19B9J1 for borJ ) was chosen for further analysis. Culture supernatants of both mutants were analysed by high-performance liquid chromatography (HPLC) linked to electrospray-ionization mass spectrometry (ESI-MS). Neither mutant produced borrelidin but, when either borI or borJ was reintroduced in trans, using expression vectors pborIH or pborJH, respectively (Fig. 1B), borrelidin production was restored. This confirmed the absence of a polar effect upon the expression of downstream genes.

image

Figure 2. Generation and analysis of mutant M19B9I1. A. Scheme representing the replacement in the chromosome of the wild-type borI gene by the in vitro mutated one. aac(3)IV, apramycin resistance gene; hyg, hygromycin resistance gene. bla, β-lactamase gene. B. Southern hybridization showing the gene replacement event by comparison of BamHI-digested chromosomal DNA from the wild-type strain and mutant M19B9I1. A 2.2 kb BglII–NsiI fragment from pSLK (Fig. 1B) was used as probe. C. Chemical structure of 12-desnitrile-12-methyl-borrelidin (2) isolated from cultures of mutant M19B9I1.

Download figure to PowerPoint

image

Figure 3. Generation and analysis of mutant M19B9J1. A. Scheme representing the replacement in the chromosome of the wild-type borI gene by the in vitro mutated one. aac(3)IV, apramycin resistance gene; hyg, hygromycin resistance gene. bla, β-lactamase gene. B. Southern hybridization showing the gene replacement event by comparison of BamHI-digested chromosomal DNA from the wild-type strain and mutant M19B9J1. A 2.2 kb BglII–NsiI fragment from pSLK (Fig. 1B) was used as probe. C. Chemical structure of 12-desnitrile-12-carboxy-borrelidin (3) isolated from cultures of mutant M19B9J1.

Download figure to PowerPoint

Although both mutants are deficient in borrelidin production, they each accumulated a new compound, as observed by HPLC-ESI-MS (LCMS). In the case of M19B9I1, a more lipophilic compound than borrelidin was present, with a maximum UV absorbance at 240 nm (2). Clearly, the cyanodiene chromophore had been disrupted. LCMS was performed using both positive and negative modes, revealing the molecular mass of the accumulated product to be 478.4 amu (m/z = 477.4 recorded for [M–H] and m/z = 501.5 recorded for [M+Na]+), 11 amu less than borrelidin itself. The mutant M19B9J1 contained a compound less lipophilic than borrelidin, with a maximum UV absorbance at 262 nm (3). ESI-MS indicated that the molecular mass of this compound is 508.3 amu (m/z = 507.3 recorded for [M–H] and m/z = 531.2 recorded for [M+Na]+). Therefore, (3) is 19 mass units heavier than borrelidin.

The mutant strains were then fermented to isolate the accumulated products for structural elucidation. Each strain was fermented in two 7 l bioreactors each containing 5 l of PYDG medium. After 6 days of fermentation, the broths were pooled, and the organic residues extracted from the broth were processed by C18 reversed phase and Sephadex LH20 chromatography, before final purification by preparative C18 reversed phase HPLC. Finally, 28 mg (3.5 mg l−1 isolated yield) of (2) and 17 mg (2.1 mg l−1 isolated yield) of (3) were isolated, and both compounds were analysed by nuclear magnetic resonance (NMR) spectroscopy. The NMR spectra were compared with that of the parent compound, and two-dimensional NMR experiments were conducted (1H-1H COSY, HMBC and HMQC) to obtain full assignments. It was found that the structures of the isolated compounds differed from borrelidin only in the substituents at C12. 12-Desnitrile-12-methyl-borrelidin (2) was isolated from mutant M19B9I1, and 12-desnitrile-12-carboxyl-borrelidin (3) was isolated from mutant M19B9J1 (Figs 2C and 3C).

These experiments indicate that borI and borJ are involved in the formation of the nitrile moiety at the C12 methyl group during borrelidin biosynthesis and also confirm that this event occurs after assembly of the polyketide chain. Furthermore, two compounds were isolated that may be intermediates for nitrile formation during borrelidin biosynthesis.

Biotransformation of 12-desnitrile-12-methyl-borrelidin (2) and 12-desnitrile-12-carboxyl-borrelidin (3)

Biotransformation experiments were performed in order to clarify: (i) the order in which the borI and borJ gene products act; and (ii) whether the accumulated compounds (2) and (3) are genuine intermediates in borrelidin biosynthesis or rather shunt products generated by further conversion of the compounds accumulated by the two mutants.

Mutants M19B9I1 and M19B9J1 were grown in the presence of either (2) or (3) and, after 5 days of incubation, the biotransformation products were analysed by HPLC. Mutant M19B9J1 was found to convert (2) into (3) efficiently. Therefore, BorI must act before BorJ. However, neither mutant was able to modify (3) (Table 1).

Table 1. Bioconversion of 12-desnitrile-12-methyl-borrelidin (2) and 12-desnitrile-12-carboxyl-borrelidin (3) by different S. parvulus mutants defective in the production of borrelidin.
MutantGene inactivatedSubstrate accumulatedSubstrate added for bioconversion
23
M19b9i1 borI 223
M19b9j1 borJ 333
SPME borE NoneBorrelidin, 13

To test whether or not the two compounds were indeed biosynthetic intermediates, they were fed independently to cultures of a mutant deficient only in starter unit biosynthesis (Table 1). This mutant contains a disrupted copy of a gene (borE) that could code for an enzyme required for the cyclization of a putative intermediate involved in the biosynthesis of the starter unit in borrelidin biosynthesis, trans-cyclopentane-1,2-dicarboxylic acid (Olano et al., 2004). During normal growth, this mutant does not produce borrelidin, but it recovers the ability to synthesize borrelidin when the medium is supplemented with trans-cyclopentane-1,2-dicarboxylic acid, demonstrating that all the PKS and post-PKS enzymes are catalytically active in this mutant (Olano et al., 2004). Growth of the borE-minus mutant in the presence of (2) led to the formation of borrelidin, although no biotransformation was observed when the strain was fed with (3) (Table 1). Taken together, these data indicate that (2) is a competent biosynthetic intermediate in borrelidin biosynthesis and that (3) is most probably a shunt product.

Heterologous expression of  borI and  borJ in Streptomyces albus J1074

The in vivo activity of borI and borJ gene products was additionally tested in a heterologous host strain. The genes were expressed in S. albus JI074, either alone or in combination, under the control of the strong constitutive promoter of the erythromycin resistance gene, ermE*p (Bibb et al., 1994). Feeding (2) to S. albus expressing borI caused its conversion into (3), while feeding (2) to S. albus expressing borI and borJ together led to its conversion into borrelidin (Fig. 4). These experiments confirm that (2) is a real biosynthetic intermediate in borrelidin biosynthesis. However, no conversion of (3) was achieved by feeding this compound to S. albus expressing either borJ alone or borI and borJ together, providing further evidence that (3) is a shunt product (data not shown).

image

Figure 4. Recovery of (3; DNCB) (white bars) and borrelidin (grey bars) after biotransformation of (2) by recombinant S. albus strains. Cultures on solid media were fed with 2 µg of (3) and, after biotransformation, (3) and borrelidin production was determined by HPLC analysis. Experiments were run in triplicate.

Download figure to PowerPoint

Although these data show that no other genes are absolutely essential for the formation of the nitrile moiety, it remains possible that such genes are normally required and their function is taken over by genes adventitiously present in S. albus. In particular, we considered the possibility that the putative dehydrogenase BorK, the product of a gene lying immediately downstream of borI and borJ in the biosynthetic gene cluster of S. parvulus, might play a role in nitrile production. Inactivation of borK has been shown greatly to reduce borrelidin biosynthesis (Olano et al., 2004). When the borK gene was expressed in S. albus together with borI and borJ, it was found that conversion of (2) into borrelidin was a more efficient process than when borI and borJ were expressed together in the absence of borK (Fig. 4). In contrast, expression of BorK alone or co-expression of borI did not lead to the bioconversion of (2) (Fig. 4).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Macrolide antibiotics are synthesized in a series of condensation reactions catalysed by modular PKSs. Numerous post-polyketide modifications in macrolides have been reported mainly involving methylations, hydroxylations and epoxidations (Weber et al., 1991; Stassi et al., 1993; Rodríguez et al., 1995; Betlach et al., 1998; Gaisser et al., 2002). In addition, most macrolides are modified further by glycosylation (Trefzer et al., 1999). Borrelidin presents a unique structural feature not described for any other known macrolide, a nitrile moiety present at C12 of the macrolide ring. Nitrile-containing natural products produced by microorganisms are relatively rare, although more than 120 nitrile-containing natural products from different sources have been described (Fleming, 1999). Toyocamycin and β-cyanoglutamic acid are antifungal antibiotics that contain nitrile moieties (Nishimura et al., 1956; Naruse et al., 1993). Among the actinomycetes, several nitrile-containing compounds with different chemical structures have been isolated. Saframycin A, cyanocycline A and Dnacin A1 are heterocyclic quinone compounds produced by Streptomyces lavendulae, Streptomyces lusitanus and Actinosynnema pretiosum respectively (Arai et al., 1980; Hayashi et al., 1983; Gould et al., 1993; Hida et al., 1994). All these compounds exhibit antibiotic and antitumour activities. A structure–activity study of saframycin has demonstrated that analogues lacking the nitrile moiety display a much lower cytotoxic activity (Kishi et al., 1984). In the case of pyrrolnitrin (Arima et al., 1965), an antifungal antibiotic isolated from Pseudomonas pyrrocinia, the replacement of the chloro substituent in the 3-position of the pyrrole ring by a cyano group leads to a remarkable enhancement in photostability (Nevill et al., 1988).

There are few biosynthetic routes to nitriles described in the literature. Broadly, they encompass either the metabolism of an amino acid or the introduction of inorganic cyanide. The best understood biosynthetic route is that of cyanogenic glycosides in plants, which involves the conversion of the amine group of amino acids into nitriles (Kato et al., 2000a; Celenza, 2001). The overall pathway involves the oxidation of the amine, probably via an N,N-dihydroxy intermediate, to yield an aldoxime, which then undergoes enzymatic dehydration to produce a nitrile. By way of example, during the formation of cyanogenic glucosides by Sorghum bicolor (Kahn et al., 1999), the cytochromes P450 CYP79A1 and CYP79E1 have been shown to catalyse the N,N-dihydroxylation of a primary amine and the dehydration of the resultant aldoxime respectively. Alternatively, the aldoxime may, in plants, be converted to a glucoinolate, which will yield a nitrile upon hydrolysis (Fleming, 1999 and references cited therein).

The ability of an actinomycete cytochrome P450 (NocL) to catalyse the two oxidative steps required to convert a tyrosine-derived amine function to an oxime during nocardicin A biosynthesis has been reported (Kelly and Townsend, 2002). The enzyme NocL displays 32% sequence identity to BorI (C. Townsend, personal communication) and was purified from an Escherichia coli strain carrying the nocL gene cloned from the nocardicin producer Nocardia uniformis ssp. tsuyamanesis. Incubation of NocL in conjunction with spinach ferredoxin, spinach ferredoxin-NADP+ reductase, NADPH and the appropriate nocardicin intermediates was sufficient to convert the amine function to an oxime.

The process by which aldoximes, via nitriles, are degraded to carboxylic acids appears to be widespread in bacteria (Kato et al., 2000b). This pathway uses a haem-containing aldoxime dehydratase, examples of which have been isolated from Rhodococcus (Xie et al., 2003) and Bacillus (Kato et al., 2000a) species. In Rhodococcus, these enzymes are found clustered with nitrile dehydratase and amidase, thereby forming a catabolic cluster (Xie et al., 2003). It is still unclear how these novel aldoxime dehydratases function, although initial in vitro studies have been performed (Xie et al., 2003).

For the microbial metabolites saframycin A and cyanocycline A, biosynthesis most probably takes place through a different pathway as their production was increased in certain fermentation experiments through the addition of sodium cyanide to the culture medium (Arai et al., 1980; Gould et al., 1993). It is unclear whether these strains biosynthesize cyanide de novo, or whether they sequester it from their environment. In the case of borrelidin, we have shown here that the carbon attached to C12 of the macrolide ring originates from methylmalonyl-CoA rather than from inorganic cyanide.

Previously, (3) was isolated as a minor metabolite from borrelidin-producing strains (Wakabayashi et al., 1996). One early hypothesis for nitrile formation, based on (3) being a biosynthetic intermediate of borrelidin, was that an amide synthetase could convert (3) to the corresponding amide, which would be followed by tautomerization and dehydration. This is a process reminiscent of that described recently for cyanide ligand formation in vitro (Reissmann et al., 2003). However, this potential pathway is ruled out by our findings that no gene encoding an amide synthetase is present within the bor gene cluster and, additionally, by the biotransformation data, which confirm that (3) is actually a shunt metabolite.

The formation of the nitrile moiety in borrelidin involves at least two different gene products: a cytochrome P450 (product of borI) and a PMP-dependent transaminase (product of borJ). The involvement of these gene products in nitrile formation has been demonstrated by insertional inactivation of these genes and the accumulation of borrelidin analogues lacking a nitrile moiety by the resulting mutants. Inactivation of borI led to the production of a new compound (2) with a methyl group replacing the nitrile at C12. This confirmed the data obtained by sequence analysis and stable isotope feeding experiments, showing that the carbon atom of the nitrile moiety arises from the methyl group of the methylmalonyl-CoA extender unit incorporated during polyketide chain elongation.

On the basis of the experiments described here, we propose a biosynthetic route for the formation of the nitrile moiety of borrelidin (Fig. 5). The C12-methyl carbon of (2) is first oxidized by BorI to introduce an allylic hydroxyl group at C12a to give (4). A similar reaction takes place during the biosynthesis of the macrolide tylosin in a reaction catalysed by the TylI cytochrome P450 (Merson-Davies and Cundliffe, 1994). Conversion to the C12-formyl derivative (5) could then occur by a further oxidation catalysed by BorI to generate a gem-dihydroxyl species with subsequent dehydration. It is likely that such processes occur widely for the formation of aldehyde and keto groups during secondary metabolism (e.g. during the production of the C18-formyl group by TylI during tylosin biosynthesis), and this mechanism has been postulated for the function of aromatase (Akhtar and Wright, 1991 and references cited therein). Other possibilities for this conversion include the action of an oxidoreductase, such as BorK, acting directly upon the allylic alcohol of (4). The involvement of BorK in nitrile formation is suggested given the data presented above, and it may be the case that BorI is capable of performing the oxidation in the absence of BorK, but that this dehydrogenase provides a more efficient pathway. The next step is anticipated to be BorJ-catalysed transamination of (5) in order to introduce a nitrogen atom at C12a, in the form of an amine, through a PMP-mediated process. The putative product amine (6) is, we suggest, then converted to the aldoxime (8), most probably via the N,N-dihydroxy species (7) and, finally, dehydrated to give borrelidin. As described above, there is significant precedent for such a pathway during cyanogenic glucoside biosynthesis in plants, and some evidence that such chemical steps occur during microbial metabolism. In the context of borrelidin biosynthesis, our data described above lead us to suggest that all the P450-dependent hydroxylations as well as some dehydration reactions are performed by the single enzyme, BorI.

image

Figure 5. Proposed pathway for the generation of the nitrile moiety as a post-polyketide step in borrelidin biosynthesis.

Download figure to PowerPoint

Such a scenario, with a single cytochrome P450 catalysing at least six separate chemical steps, would be highly unusual, especially as a number of these steps would have to occur after the activity of the transaminase, BorJ, had modified one of the biosynthetic intermediates produced by BorI. However, the single enzyme CYP79A1 is apparently able to catalyse three separate oxidative steps (including a decarboxylation), as well as an E to Z isomerization, during the biosynthesis of the cyanogenic glycoside dhurrin in Sorghum bicolor (Kahn et al., 1999). Furthermore, during the biosynthesis of cyanogenic glycosides in Sorghum, metabolic flow through CYP79A1 and CYP71E1 is highly channelled, with the majority of the intermediates in the pathway not being accumulated or detectable in vivo (Kahn et al., 1999; Celenza, 2001). Indeed, nitrile metabolism in plants is greatly facilitated when the nitriles are produced directly from their corresponding aldoximes at the catalytic site of the enzyme in a channelled reaction (Møller and Seigler, 1999). This is in keeping with the situation observed for borrelidin biosynthesis as, even after extensive and careful analysis of wild-type S. parvulus Tü4055 and Tü113 fermentations, using both HPLC and LCMS methods, we were unable to observe any intermediates of borrelidin biosynthesis.

The data presented here indicate that the nitrile group of borrelidin arises from oxidation of the carbon-12a, which itself originates from the methyl group of a methylmalonyl-CoA PKS extender unit. Additionally, gene insertional inactivation experiments have clearly implicated the gene products BorI and BorJ in the process of nitrile formation. Further study using purified enzymes in vitro will be required in order rigorously to test our proposal that BorI is a multifunctional cytochrome P450 which, along with the transaminase BorJ, is sufficient for the conversion of preborrelidin (2) into borrelidin.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Strains, culture conditions and plasmids

Streptomyces parvulus Tü4055 (borrelidin producer) and mutants were grown routinely on tryptone soya broth (TSB). For protoplast formation, strains were grown in R5 liquid medium (Kieser et al., 2000). Regeneration of protoplasts after transformation was carried out on solid R5 agar plates using standard procedures (Kieser et al., 2000); after regeneration, the clones were grown on agar plates containing A medium (MA) for sporulation (Fernández et al., 1998). E. coli XL1-Blue (Bullock et al., 1987) and DH10B (Gibco BRL) were used for subcloning. pSL1180 (Amersham Biosciences) and pEM4 (Quirós et al., 1998) were used for gene replacement and gene expression respectively. pEFBA was used as a donor of the apramycin resistance gene aac(3)IV (Blanco et al., 2001), and pLHyg as a donor of the hygromycin resistance gene hyg (Olano et al., 2004). When antibiotic selection of transformants was needed, 100 µg ml−1 ampicillin, 20 µg ml−1 tobramycin (when selecting for the aac(3)IV gene in E. coli), 25 µg ml−1 apramycin, 50 µg ml−1 thiostrepton or 50 µg ml−1 hygromycin (200 µg ml−1 in R5 solid media) was used.

DNA manipulation

Plasmid DNA preparations, restriction endonuclease digestions, alkaline phosphatase treatments, ligations and other DNA manipulations were performed according to standard procedures for E. coli (Sambrook et al., 1989) and Streptomyces (Kieser et al., 2000).

PCR amplification

Amplification of borI was carried out using the following oligoprimers: BP4501 5′-CGTATGCATGGCGCCATGGA-3′ (NsiI site underlined) and BP4502 5′-AGCCAATTGGTG CACTCCAG-3′ (MfeI site underlined).

Amplification of borJ was carried out using the following oligoprimers: BNHT1 5′-GTCATGCATCAGCGCACCCG-3′ (NsiI site underlined) and BNHT2 5′-GTGCAATTGCCCTGG TAGTC-3′ (MfeI site underlined).

Polymerase chain reaction (PCR) conditions were: 97°C, 5 min; 30 cycles of 95°C, 30 s, 70°C (for borI) or 65°C (for borJ), 45 s and 72°C, 3 min and a final extension cycle at 72°C, 10 min. DNA polymerase (New England Biolabs) and 2.5% dimethyl sulphoxide (DMSO) were used for all amplifications. The DNA template used for all the amplifications was cosmid Bor19B9 (Olano et al., 2004).

Insertional inactivation

borI.  The borI gene was amplified as described above using oligoprimers BP4501 and BP4502. The 2.3 kb PCR product was purified, digested with NsiI–MfeI and subcloned into pSL1180 digested with NsiI–EcoRI, generating pSLI. The aac(3)IV gene from pEFBA was then subcloned, as an EcoRI fragment, into the unique EcoRI site located within the borI coding region to generate pSLIa. To introduce a second marker into the plasmid, the hyg gene from pLHyg was subcloned as a PstI–SpeI fragment into the NsiI–SpeI sites of pSLIa, leading to pSLIr1.

borJ.  For inactivation of borJ, oligoprimers BNHT1 and BNTH2 were used. The 2.7 kb PCR product was purified, digested with NsiI–MfeI and subcloned into NsiI–EcoRI-digested pSL1180, generating pSLJ. The aac(3)IV gene was then subcloned as a SpeI–BamHI fragment from pEFBA into the AvrII–BglII sites of pSLJ, thus removing 453 bp from borJ and leading to the generation of pSLJa. Then, the hyg gene was subcloned as a PstI–SpeI fragment into the previous construct digested with NsiI–SpeI, thus generating pSLJr1.

The final constructs (pSLIr1 and pSLJr1) were used to transform protoplasts of S. parvulus Tü4055. Transformants in which a double cross-over allows the replacement of the wild-type allele from the chromosome by the in vitro mutated one were selected for their resistance to apramycin and their sensitivity to hygromycin.

Gene expression

borI.  A 2.1 kb NsiI–AvrII fragment containing borI was recovered from pSLI and subcloned into the PstI–XbaI sites of the bifunctional (Streptomyces–E. coli) expression vector pEM4 under the control of the constitutive promoter of the erythromycin resistance gene, ermE*p, and together with the NheI–SpeI fragment from pLHyg containing the hyg gene. Both fragments were subcloned in the same orientation generating pborIH.

borJ.  A 2.4 kb Nsi I –SphI fragment from pSLJ containing borJ was subcloned into the PstI–XbaI sites of pEM4 and, together with the hyg gene as a SphI–SpeI fragment from pLHyg, both fragments were subcloned in the same orientation as the transcription of the genes. The final construct was designated pborJH.

borK.  A 2.2 kb Bgl II (blunt-ended)–NsiI fragment from pSLK (Olano et al., 2004) was subcloned together with a 1.6 kb PstI–SpeI fragment from pLHyg containing the hyg Gen into PstI-digested (treated with the Klenow fragment) pEM4 and XbaI. In the final construct pborKH, borK is under the control of the ermE*p.

borI+-borJ.  A 1.9 kb HindIII–KpnI fragment from pborIH was subcloned into the HindIII–BamHI sites of pEM4 together with a 1.7 kb KpnI–BamHI fragment from pborJH. In this final construct pborIJ, borI and borJ are transcribed under the control of the ermE*p.

borI+-borJ+-borK.  A 2.2 kb BglII–XbaI fragment from pSLK was subcloned into the BglII–NheI sites of pborIJ. In this way, in the final construct pborIJK, the three genes are together and under the control of the ermE*p.

borI+-borK.  pborIJK was digested with BglII and, after Klenow polymerase treatment, was religated generating pborIK. This treatment causes a frameshift mutation in borJ, thus inactivating this gene.

Feeding experiments

Incorporation of borrelidin intermediates and racemic trans-cyclopentane-1,2-dicarboxylic acid by different recombinant strains was carried out as follows. The strains were grown on R5A solid medium (Fernández et al., 1998) on small squares (2 × 2 cm) on square Petri dishes. Each compound (5 µg) was incorporated to the medium and, after 5 days of incubation at 28°C, the agar was extracted with 1 ml of ethyl acetate. After evaporation of the solvent, the residue was dissolved in 50 µl of ethyl acetate and analysed by HPLC.

Descriptions of the feeding experiments of [1,2-13C2]-acetate and [2,3-13C2]-propionate to S. parvulus Tü113, isolation of the resultant labelled borrelidin and 13C-NMR spectra of the enriched samples are available in Supplementary material.

HPLC, mass spectra analysis and structure elucidation

Detection of borrelidin and related compounds was performed routinely by HPLC on a reversed-phase column (Symmetry C18, 4.6 × 250 mm; Waters), with acetonitrile and 0.1% trifluoroacetic acid (TFA) in water as solvents. A linear gradient from 10% to 100% acetonitrile in 30 min, at a flow rate of 1 ml min−1, was used. Detection and spectral characterization of peaks were performed with a photodiode array detector and millenium software (Waters); bidimensional chromatograms were extracted at 257 nm. HPLC analysis was also conducted on a base-deactivated reversed column (Hypersil C18-BDS, 3 µm particle size, 150 mm × 4.6 mm), using 10% acetonitrile with 90% water as mobile phase A and 90% acetonitrile with 10% water as mobile phase B, both containing 10 mM ammonium acetate and 0.1% TFA. A linear gradient from 25% B to 100% B over 15 min was applied at 1 ml min−1 on an Agilent 1100. Borreldins were quantified by measuring the UV absorbance at 258 nm, while recording data over 190–600 nm. LCMS analysis was carried out using the same chromatography, substituting 0.1% TFA for 0.15% formic acid. Electrospray ionization was applied to 1/10th of the elutant, using an Esquire 3000+ instrument and monitoring m/z from 50 to 1000, while switching between positive and negative modes.

Details on the production, purification and structural elucidation of the two compounds isolated in this work is given as Supplementary material.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

We wish to thank Professor Craig Townsend for access to the NocL sequence and useful comments on the manuscript. This research was supported by grants from the Spanish Ministry of Science and Technology (BIO2000-0274; BMC2003-00478) and from the Plan Regional de Investigación del Principado de Asturias (GE-MED01-05). We thank Obra Social Cajastur for financial support to C.O.

Supplementary material

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Table S1. Stable isotope incorporation study of borrelidin biosynthesis by administering sodium [1,2-13C2]-acetate and sodium [2,3-13C2]-propionate to Streptomyces parvulus.

Table S2. Production, purification and structural elucidation of 12-desnitrile-12-methyl-borrelidin.

Table S3. Production, purification and structural elucidation of 12-desnitrile-12-carboxyl-borrelidin.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information
  • Akhtar, M., and Wright, J.N. (1991) A unified mechanistic view of oxidative reactions catalysed by P-450 and related Fe-containing enzymes. Nature Prod Rep 0: 527551.
  • Anderson, B.F., Herlt, A.J., Rickards, R.W., and Robertson, G.B. (1989) Crystal and molecular structures of two isomorphous solvates of the macrolide antibiotic borrelidin: absolute configuration determination by incorporation of a chiral solvent in the crystal lattice. Aust J Chem 42: 717730.
  • Arai, T., Takahashi, K., Nakahara, S., and Kubo, A. (1980) The structure of a novel antitumor antibiotic, saframycin A. Experientia 36: 10251027.
  • Arima, K., Imanaka, H., Kousaka, M., Fukuda, A., and Tamura, G. (1965) Studies on pyrrolnitrin, a new antibiotic. I. Isolation and properties of pyrrolnitrin. J Antibiot 18: 201204.
  • Berger, J., Jampolsky, L.M., and Goldberg, M.W. (1949) Borrelidin, a new antibiotic with anti-Borrelia activity and penicillin enhancement properties. Arch Biochem 22: 476478.
  • Betlach, M.C., Kealey, J.T., Ashley, G.W., and McDaniel, R. (1998) Characterization of the macrolide P-450 hydroxylase from Streptomyces venezuelae which converts narbomycin to picromycin. Biochemistry 37: 1493714942.
  • Bibb, M.J., White, J., Ward, J.M., and Janssen, G.R. (1994) The mRNA for the 23S rRNA methylase encoded by the ermE gene of Saccharopolyspora erythraea is translated in the absence of a conventional ribosome-binding site. Mol Microbiol 14: 533545.
  • Blanco, G., Pérez-Patallo, E., Braña, A.F., Rohr, J., Méndez, C., and Salas, J.A. (2001) Identification of a sugar flexible glycosyltransferase from Streptomyces olivaceus, the producer of the antitumor polyketide elloramycin. Chem Biol 8: 253263.
  • Bullock, W.O., Fernández, J.M., and Short, J.M. (1987) XL1-Blue: a high efficiency plasmid transforming recA Escherichia coli strain with beta-galactosidase selection. Biotechniques 5: 376.
  • Celenza, J.L. (2001) Metabolism of tyrosine and tryptophan – new genes for old pathways. Curr Opin Plant Biol 4: 234240.
  • Del Vecchio, F., Petkovic, H., Kendrew, S.G., Low, L., Wilkinson, B., Lill, R., et al. (2003) Active-site residue, domain and module swaps in modular polyketide synthases. J Ind Microbiol Biotechnol 30: 489494.
  • Dickinson, L., Griffiths, A.J., Mason, C.G., and Mills, R.F. (1965) Anti-viral activity of two antibiotics isolated from a species of Streptomyces. Nature 206: 265268.
  • Fernández, E., Weissbach, U., Sánchez Reillo, C., Braña, A.F., Méndez, C., Rohr, J., and Salas, J.A. (1998) Identification of two genes from Streptomyces argillaceus encoding glycosyltransferases involved in transfer of a disaccharide during biosynthesis of the antitumor drug mithramycin. J Bacteriol 180: 49294937.
  • Fleming, F.F. (1999) Nitrile-containing natural products. Nature Prod Rep 16: 597606.
  • Fouces, R., Mellado, E., Diez, B., and Barredo, J.L. (1999) The tylosin biosynthetic cluster from Streptomyces fradiae: genetic organisation of the left region. Microbiology 145: 855868.
  • Funahashi, Y., Wakabayashi, T., Semba, T., Sonoda, J., Kitoh, K., and Yoshimatsu, K. (1999) Establishment of a quantitative mouse dorsal air sac model and its application to evaluate a new angiogenesis inhibitor. Oncol Res 11: 319329.
  • Gaisser, S., Lill, R., Staunton, J., Mendez, C., Salas, J.A., and Leadlay, P.F. (2002) Parallel pathways for oxidation of 14-membered polyketide macrolactones in Saccharopolyspora erythraea. Mol Microbiol 44: 771781.
  • Gould, S.J., He, W., and Cone, M.C. (1993) New cyanocyclines from a cyanide-treated broth of Streptomyces lusitanus. J Nat Prod 56: 12391245.
  • Hayashi, T., Okutomi, T., Suzuki, S., and Okazaki, H. (1983) Inhibition of nucleic acid biosynthesis in procaryotic and eucaryotic cells by cyanocycline A. J Antibiot 36: 12281235.
  • Haydock, S.F., Aparicio, J.F., Molnar, I., Schwecke, T., Khaw, L.E., Konig, A., et al. (1995) Divergent sequence motifs correlated with the substrate specificity of (methyl) malonyl-CoA: acyl carrier protein transacylase domains in modular polyketide synthases. FEBS Lett 374: 246248.
  • Hida, T., Muroi, M., Tanida, S., and Harada, S. (1994) Structures of dnacin A1 and B1, new naphthyridinomycin-type antitumor antibiotics. J Antibiot 47: 917921.
  • Kahn, R.A., Fahrendorf, T., Halkier, B.A., and Moller, B.L. (1999) Substrate specificity of the cytochrome P450 enzymes CYP79A1 and CYP71E1 involved in the biosynthesis of the cyanogenic glucoside dhurrin in Sorghum bicolor (L.) Moench. Arch Biochem Biophys 363: 918.
  • Kato, Y., Nakamura, K., Sakiyama, H., Mayhew, S.G., and Asano, Y. (2000a) Novel heme-containing lyase, phenylacetaldoxime dehydratase from Bacillus sp. strain OxB-1: purification, characterization, and molecular cloning of the gene. Biochemistry 39: 800809.
  • Kato, Y., Ooi, R., and Asaon, Y. (2000b) Distribution of aldoxime dehydratase in microorganisms. Appl Environ Microbiol 66: 22902296.
  • Kawamura, T., Liu, D., Towle, M.J., Kageyama, R., Tsukahara, N., Wakabayashi, T., and Littlefield, B.A. (2003) Anti-angiogenesis effects of borrelidin are mediated through distinct pathways: threonyl-tRNA synthetase and caspases are independently involved in suppression of proliferation and induction of apoptosis in endothelial cells. J Antibiot 56: 709715.
  • Keller-Schierlein, W. (1967) Composition of the antibiotic borrelidin. Helv Chim Acta 50: 731753.
  • Kelly, W.L., and Townsend, C.A. (2002) Role of the cytochrome P450 NocL in nocardicin A biosynthesis. J Am Chem Soc 124: 81868187.
  • Kieser, T., Bibb, M.J., Buttner, M.J., Chater, K.F., and Hopwood, D.A. (2000) Practical Streptomyces Genetics. Norwich: The John Innes Foundation.
  • Kishi, K., Yazawa, K., Takahashi, K., Mikami, Y., and Arai, T. (1984) Structure–activity relationships of saframycins. J Antibiot 37: 847852.
  • Kiyasu, T., Nagahashi, Y., and Hoshino, T. (2001) Cloning and characterization of biotin biosynthetic genes of Kurthia sp. Gene 265: 103113.
  • Kuo, M.S., Yurek, D.A., and Kloosterman, D.A. (1989) Assignment of 1H and 13C NMR signals and the alkene geometry at C-7 in borrelidin. J Antibiot 42: 10061007.
  • Maehr, H., and Evans, R.H. (1987) Identity of borrelidin with treponemycin. J Antibiot 40: 14551456.
  • Merson-Davies, L.A., and Cundliffe, E. (1994) Analysis of five tylosin biosynthetic genes from the tyllBA region of the Streptomyces fradiae genome. Mol Microbiol 13: 349355.
  • Møller, B.L., and Seigler, D.S. (1999) Biosynthesis of cyanogenic glycosides, cyanolipids, and related compounds. In Plant Amino Acids: Biochemistry and Biotechnology. Singh, B.K., (ed.). New York: Marcel Dekker, pp. 563609.
  • Naruse, N., Yamamoto, S., Yamamoto, H., Kondo, S., Masuyoshi, S., Numata, K., et al. (1993) β-cyanoglutamic acid, a new antifungal amino acid from a streptomycete. J Antibiot 46: 685686.
  • Nevill, D., Nyfeler, R., and Sozzi, D. (1988) CGA142705: a novel fungicide for seed treatment. Proc Brighton Crop Prot Conf Pests Dis 1: 6572.
  • Nishimura, H., Katagiri, K., Sato, K., Mayam, M., and Shimaoka, N. (1956) Toyocamycin, a new anti-candida antibiotic. J Antibiot 9: 6062.
  • Olano, C., Wilkinson, B., Moss, S.J., Braña, A.F., Méndez, C., Leadlay, P.F., and Salas, J.A. (2003) Evidence from engineered gene fusions for the repeated use of a module in a modular polyketide synthase. Chem Commun 2003: 27802782.
  • Olano, C., Wilkinson, B., Sánchez, C., Moss, S., Sheridan, R., Math, V., et al. (2004) Biosynthesis of the angiogenesis inhibitor borrelidin by Streptomyces parvulus Tü4055: cluster analysis and assignment of functions. Chem Biol 11: 8797.
  • Otoguro, K., Ui, H., Ishiyama, A., Kobayashi, M., Togashi, H., Takahashi, Y., et al. (2003) In vitro and in vivo antimalarial activities of a non-glycosidic 18-membered macrolide antibiotic, borrelidin, against drug-resistant strains of Plasmodia. J Antibiot 56: 727729.
  • Paetz, W., and Nass, G. (1973) Biochemical and immunological characterization of threonyl-tRNA synthetase of two borrelidin-resistant mutants of Escherichia coli K12. Eur J Biochem 35: 331337.
  • Poulos, T.L., Finzel, B.C., and Howard, A.J. (1987) High-resolution crystal structure of cytochrome P450cam. J Mol Biol 195: 687700.
  • Quirós, L.M., Aguirrezabalaga, I., Olano, C., Méndez, C., and Salas, J.A. (1998) Two glycosyltransferases and a glycosidase are involved in oleandomycin modification during its biosynthesis by Streptomyces antibioticus. Mol Microbiol 28: 11771185.
  • Read, T.D., Salzberg, S.L., Pop, M., Shumway, M., Umayam, L., Jiang, L., et al. (2002) Comparative genome sequencing for discovery of novel polymorphisms in Bacillus anthracis. Science 296: 20282033.
  • Reeves, C.D., Murli, S., Ashley, G.W., Piagentini, M., Hutchinson, C.R., and McDaniel, R. (2001) Alteration of the substrate specificity of a modular polyketide synthase acyltransferase domain through site-specific mutations. Biochemistry 40: 1546415470.
  • Reissmann, S., Hochleiter, E., Wang, H., Paschos, A., Lottspeich, F., Glass, R.S., and Böck, A. (2003) Taming of a poison: biosynthesis of the NiFe-hydrogenase cyanide ligands. Science 299: 10671070.
  • Rodríguez, A.M., Olano, C., Mendez, C., Hutchinson, C.R., and Salas, J.A. (1995) A cytochrome P450-like gene possibly involved in oleandomycin biosynthesis by Streptomyces antibioticus. FEMS Microbiol Lett 127: 117120.
  • Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989) Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.
  • Simpson, T.J. (1997) Application of isotopic methods to secondary metabolic pathways. Topics Curr Chem 195: 148.
  • Singh, S.K., Gurusiddaiah, S., and Whalen, J.W. (1985) Treponemycin, a nitrile antibiotic active against Treponema hyodysenteriae. Antimicrob Agents Chemother 27: 239245.
  • Stassi, D., Donadio, S., Staver, M.J., and Katz, L. (1993) Identification of a Saccharopolyspora erythraea gene required for the final hydroxylation step in erythromycin biosynthesis. J Bacteriol 175: 182189.
  • Trefzer, A., Salas, J.A., and Bechthold, A. (1999) Genes and enzymes involved in deoxysugar biosynthesis in bacteria. Nature Prod Rep 16: 283299.
  • Tsuchiya, E., Yukawa, M., Miyakawa, T., Kimura, K.I., and Takahashi, H. (2001) Borrelidin inhibits a cyclin-dependent kinase (CDK), Cdc28/Cln2, of Saccharomyces cerevisiae. J Antibiot 54: 8490.
  • Wakabayashi, T., Kageyama, R., Naruse, N., Yokoyama, Y., Watanabe, Y., and Dobashi, K. (1996) New physiologically active substance. Japanese patent application, JP09227549A. CHEMABS 127: 277250S.
  • Wakabayashi, T., Kageyama, R., Naruse, N., Tsukahara, N., Funahashi, Y., Kitoh, K., and Watanabe, Y. (1997) Borrelidin is an angiogenesis inhibitor: disruption of angiogenic capillar vessels in a rat aorta matrix culture model. J Antibiot 50: 671676.
  • Watanabe, I., Nara, F., and Serizawa, N. (1995) Cloning, characterization and expression of the gene encoding cytochrome P-450sca-2 from Streptomyces carbophilus involved in production of pravastatin, a specific HMG-CoA reductase inhibitor. Gene 163: 8185.
  • Weber, J.M., Leung, J.O., Swanson, S.J., Idler, K.B., and McAlpine, J.B. (1991) An erythromycin derivative produced by targeted gene disruption in Saccharopolyspora erythraea. Science 252: 114117.
  • Xie, S.-X., Kato, Y., Komeda, H., Yoshida, S., and Asano, Y. (2003) A gene cluster responsible for alkylaldoxime metabolism coexisting with nitrile hydratase and amidase in Rhodococcus globerulus A-4. Biochemistry 42: 1205612066.
  • Yonaha, K., Nishie, M., and Aibara, S. (1992) The primary structure of ω-amino acid: pyruvate aminotransferase. J Biol Chem 267: 1250612510.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Table S1. Stable isotope incorporation study of borrelidin biosynthesis by administering sodium [1,2-13C2]acetate and sodium [2,3-13C2]propionate to Streptomyces parvulus. Table S2. Production, purification and structural elucidation of 12-desnitrile-12-methyl-borrelidin (2). Table S3. Production, purifation and structural elucidation of 12-desnitrile-12-carboxyl-borrelidin (3).

FilenameFormatSizeDescription
MMI_4090_sm_tableS1.doc44KSupporting info item
MMI_4090_sm_tableS2.doc45KSupporting info item
MMI_4090_sm_tableS3.doc45KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.