ARM and COF are well known for their prominent biological activities and unusual chemical structures; however, the logic of their biosynthesis has long been poorly understood. Actually, the new insights into the ARM and COF pathway will not only enrich the biochemical repertoire for interesting enzymatic reactions but may also lay a solid foundation for the combinatorial biosynthesis of this group of antibiotics via a target-directed genome mining strategy.
KEYWORDS: coordinated biosynthesis, aristeromycin, coformycin, actinomycetes, fine-tuned pathway
ABSTRACT
Purine nucleoside antibiotic pairs, concomitantly produced by a single strain, are an important group of microbial natural products. Here, we report a target-directed genome mining approach to elucidate the biosynthesis of the purine nucleoside antibiotic pair aristeromycin (ARM) and coformycin (COF) in Micromonospora haikouensis DSM 45626 (a new producer for ARM and COF) and Streptomyces citricolor NBRC 13005 (a new COF producer). We also provide biochemical data that MacI and MacT function as unusual phosphorylases, catalyzing an irreversible reaction for the tailoring assembly of neplanocin A (NEP-A) and ARM. Moreover, we demonstrate that MacQ is shown to be an adenosine-specific deaminase, likely relieving the potential “excess adenosine” for producing cells. Finally, we report that MacR, an annotated IMP dehydrogenase, is actually an NADPH-dependent GMP reductase, which potentially plays a salvage role for the efficient supply of the precursor pool. Hence, these findings illustrate a fine-tuned pathway for the biosynthesis of ARM and also open the way for the rational search for purine antibiotic pairs.
IMPORTANCE ARM and COF are well known for their prominent biological activities and unusual chemical structures; however, the logic of their biosynthesis has long been poorly understood. Actually, the new insights into the ARM and COF pathway will not only enrich the biochemical repertoire for interesting enzymatic reactions but may also lay a solid foundation for the combinatorial biosynthesis of this group of antibiotics via a target-directed genome mining strategy.
INTRODUCTION
Nucleoside antibiotics constitute a large family of important microbial natural products bearing diverse biological activities, such as antibacterial, antifungal, antiviral, and antitumor activities (1–4). Structurally, this family of antibiotics highlights distinctive moieties, derived from simple building blocks, either nucleosides or nucleotides of primary origin (1). The biosynthesis of nucleoside antibiotics often follows a succinct logic of sequential modifications of nucleosides or nucleotides and leads to complex structures (2, 3, 5). Because of their unique chemical diversity, nucleoside antibiotics yield leads/analogs with novel structural features and diverse bioactivities (6, 7).
Purine-based nucleoside antibiotics, which include the pentostatin (PTN)-related compounds 2′-Cl PTN, coformycin (COF), adecypenol, and carbocyclic COF (Fig. 1A and B) (8, 9), share an unusual heterocyclic 1,3-diazepine core in which an additional carbon is inserted between C-6 and N-1 of the purine ring. Metabolic labeling experiments indicated that the “extra carbon” was derived from C-1 of d-ribose (10). Previous studies have revealed that the biosynthesis of PTN and Ara-A (arabinofuranosyladenine) is governed by a single gene cluster but arises from independent pathways and employs an unusual protector-protégé strategy; i.e., PTN may protect Ara-A from deamination by the housekeeping adenosine deaminase (8). More recently, a similar strategy has been shown for 2′-Cl PTN and 2′-amino-2′-deoxyadenosine (produced by Actinomadura sp. strain ATCC 39365) (Fig. 1A) (11), as well as by the fungus-derived PTN and cordycepin pair (produced by Cordyceps militaris and other fungal strains) (12). In addition, we have previously shown that a short-chain dehydrogenase and a SAICAR (phosphoribosylaminoimidazolesuccinocarboxamide) synthetase are highly conserved for the biosynthesis of the heterocyclic 1,3-diazepine ring of the PTN-related compounds (8).
FIG 1.
Chemical structures of relevant antibiotics. (A) Structure of the related (potential) antibiotic pairs. Except for ARM and carbocyclic COF (blue dashed box), the antibiotic at the top together with the one shown below constitutes an antibiotic pair which is concomitantly produced by a single microorganism. The formycin A and COF pair is individually produced by Streptomyces kaniharenesis ATCC 21070 and Nocardia interforma ATCC 21072. The Ara-A and PTN pair is produced by Streptomyces antibioticus NRRL 3238. The 2′-amino dA (2′-amino-2′-deoxyadenosine) and 2′-Cl PTN pair is produced by Actinomadura sp. ATCC 39365. Multiple nucleoside antibiotics, including ARM, NEP-A, NEP-D, Carbo-I, and COF, are produced by M. haikouensis and S. citricolor. Green dashed boxes highlight the antibiotics that are produced by a single strain, either M. haikouensis or S. citricolor. (B) Chemical structures of NEP-A, Carbo-I, NEP-D, and adecypenol.
Neplanocin A (NEP-A), aristeromycin (ARM), adecypenol, and carbocyclic COF (Fig. 1A and B), in the carbocyclic purine nucleoside group of antibiotics (9, 13), are distinguished by a unique five-membered cyclitol moiety that is linked to adenine via an N-glycosidic bond. ARM displays several interesting biological properties, including the inhibition of AMP synthesis in mammalian cells and the blocking of cell division and elongation in rice plants (14). NEP-A has notable antiviral and antitumor activities (15), but both NEP-A and ARM have a very limited spectrum of antibacterial activities (16, 17). Mechanistically, these two antibiotics act as competitive inhibitors of S-adenosyl-l-homocysteine (SAH) hydrolase (14, 15). Metabolic feeding experiments by the Parry group showed that the origin of the cyclitol moiety is from d-glucose, from which the C-2 and C-6 carbons are utilized to construct the carbocyclic ring via a carbon-carbon bond and the adenine ring is directly from adenosine (14, 15). Subsequently, Hill et al. identified two key intermediates of the NEP-A and ARM biosynthetic pathway and tentatively proposed a concise pathway (18). Although the ARM gene cluster from Streptomyces citricolor NBRC 13005 (S. citricolor herein) has been identified and a plausible biosynthetic pathway has been proposed (19), several aspects remained to be determined.
In the present study, we report that Micromonospora haikouensis DSM 45626 (M. haikouensis herein) (20) and S. citricolor are new producing strains of multiple nucleoside antibiotics and further show that the ARM and COF gene clusters feature genetic organization diversity. We also demonstrate that NEP-A and ARM biosynthesis includes an irreversible catalytic step for the final assembly and is associated with a fine-tuned pathway. This will lay a solid foundation for the combinatorial biosynthesis of this group of antibiotics and for the rational discovery of novel purine nucleoside antibiotic pairs.
RESULTS
Discovery of M. haikouensis and S. citricolor as producers of multiple nucleoside antibiotics.
The enzymes for the biosynthesis of the PTN-related purine nucleoside antibiotics (Fig. 1A), including a short-chain dehydrogenase and a SAICAR synthetase, are highly conserved and widely distributed (8), and they are therefore potential probes for the discovery of new purine nucleoside antibiotic pathways from available microbial genomes. We then took advantage of the target enzymes, PenB (short-chain dehydrogenase) and PenC (SAICAR synthetase) from the PTN pathway (8), to conduct a BLASTP search against the NCBI database, leading to the discovery of multiple potential gene clusters for the biosynthesis of PTN-related antibiotic pairs. These include the mac gene cluster from M. haikouensis (Fig. 2A; Table 1), in which two genes correspondingly code for the candidate proteins GA0070558_12452 (designated MacM, 60% identity to PenC) and GA0070558_12451 (designated MacN, 49% identity to PenB). Further examination of the surrounding region revealed genes whose products are highly homologous and match those of the ARM biosynthetic pathway in S. citricolor (Fig. 2A), suggesting that the gene cluster (mac) is potentially involved in the biosynthesis of ARM- and PTN-related compounds.
FIG 2.
Genetic organization and verification of the ARM and COF biosynthetic gene cluster. (A) Genetic organization of the ARM and COF gene cluster. The mac gene cluster is responsible for ARM and COF biosynthesis in M. haikouensis, while two separate gene clusters (ari and com) independently govern the biosynthesis of ARM and COF in S. citricolor. (B) Engineered production of ARM, NEP-A, and COF in S. aureochromogenes CXR14. (C) LC-MS analysis of the target metabolite of COF produced by related strains. Std, the authentic standards of ARM and NEP-A; haikou-WT, wild-type strain of M. haikouensis; CXR14/63B3, S. aureochromogenes CXR14 containing 63B3 cosmid; CXR14/pJTU2463b, S. aureochromogenes CXR14 containing pJTU2463b as a negative control.
TABLE 1.
Deduced functions of the open reading frames in the mac gene cluster
| Proteina | No. of amino acids | Protein function | Homolog, origin | Identity, similarity (%) | Accession no. |
|---|---|---|---|---|---|
| MacV | 217 | NAD-dependent epimerase/dehydratase | Ari15, S. citricolor | 56, 64 | BAV57070 |
| MacW | 478 | SAH hydrolase | ASC68_24850, Devosia sp. Root105 | 38, 53 | KQU93055 |
| MacX | 220 | ATP phosphoribosyl-transferase | Sare_0588, Salinispora arenicola CNS-205 | 58, 71 | ABV96516 |
| MacA | 334 | Dehydrogenase | Ari1, S. citricolor | 42, 58 | BAV57056 |
| MacB | 364 | Five-membered cyclitol-phosphate synthase | Ari2, S. citricolor | 81, 88 | BAV57057 |
| MacC | 216 | Short-chain dehydrogenase | Ari3, S. citricolor | 50, 61 | BAV57058 |
| MacD | 249 | Short-chain dehydrogenase | Ari4, S. citricolor | 62, 77 | BAV57059 |
| MacE | 139 | Enamine deaminase | Ari5, S. citricolor | 64, 78 | BAV57060 |
| MacF | 344 | Ribokinase | Ari6, S. citricolor | 55, 66 | BAV57061 |
| MacG | 415 | SAH hydrolase | Ari7, S. citricolor | 81, 90 | BAV57062 |
| MacH | 474 | Phosphoglucomutase | Ari8, S. citricolor | 53, 62 | BAV57063 |
| MacI* | 272 | Adenosine phosphorylase | Ari9, S. citricolor | 76, 87 | BAV57064 |
| MacJ | 422 | SAH hydrolase | Ari10, S. citricolor | 71, 80 | BAV57065 |
| MacK | 407 | SAH hydrolase | Ari11, S. citricolor | 67, 77 | BAV57066 |
| MacL | 230 | Phosphoglycerate mutases | Ari12, S. citricolor | 61, 71 | BAV57066 |
| MacM | 243 | SAICAR synthetase | PenC, S. antibioticus NRRL 3238 | 60, 74 | AKA87338 |
| MacN | 239 | Short-chain dehydrogenase | PenB, S. antibioticus NRRL 3238 | 49, 64 | AKA87339 |
| MacO | 430 | HAD family hydrolase | Krac_6662, Ktedonobacter racemifer DSM 44963 | 34, 50 | EFH85441 |
| MacP | 248 | Phosphatase | Ari14, S. citricolor | 58, 74 | BAV57069 |
| MacQ* | 330 | Adenosine deaminase | Ari17, S. citricolor | 35, 48 | BAV57071 |
| MacR* | 488 | GMP reductase | N864_01465, Intrasporangium chromatireducens Q5-1 | 58, 72 | EWT05858 |
| MacS | 325 | Epimerase | Amir_4843, Actinosynnema mirum DSM 43827 | 42, 57 | ACU38670 |
| MacT* | 272 | Adenosine phosphorylase | Ari9, S. citricolor | 49, 64 | BAV57064 |
| MacU | 383 | MFS transporter | Ari13, S. citricolor | 63, 75 | BAV57068 |
*, function was confirmed in vitro. MacV corresponds to the locus tag GA0070558_12467 (NCBI accession number SCF06629.1), and likewise, MacU matches to that of accession no. GA0070558_12444 (GenBank: SCF06432.1).
To gain evidence whether the candidate mac gene cluster is active at the transcriptional level, we performed reverse transcription (RT)-PCR analysis of the target genes macM and macN. As expected, the results indicated that the gene cluster is indeed transcribed (see Fig. S1A in the supplemental material), implying that M. haikouensis is capable of producing related nucleoside antibiotics. To confirm this, we analyzed metabolites of this strain by liquid chromatography-mass spectrometry (LC-MS). A distinctive [M+H]+ ion for COF (m/z 285.1182) was observed (Table S1), rather than PTN or other related molecules (Fig. 1A and B). Tandem mass spectrometry (MS/MS) analysis indicated that the major fragment ions generated were at m/z 134.9773, 153.0337, and 267.0721 (Table S1), fully consistent with the fragmentation pattern of the authentic COF standard. In addition, characteristic UV peaks were detected for NEP-A and ARM (Fig. 2B; Fig. S1B and C). LC-MS analysis of these showed the distinctive [M+H]+ ions of NEP-A (m/z 264.1085) and ARM (m/z 266.1238), whose fragmentation patterns match those of the authentic standards of NEP-A and ARM, respectively (Table S1). NEP-A and ARM are both adenosine analogs that are prone to deamination by host adenosine deaminases, suggesting that the M. haikouensis strain is also likely to produce the deaminated products of NEP-A and ARM, including NEP-D and Carbo-I (Fig. 1B). We accordingly reanalyzed the metabolites for the characteristic [M+H]+ ions of NEP-D (m/z 265.0927) and Carbo-I (m/z 267.1082). MS/MS analysis showed that the [M+H]+ fragment ions arising from these were produced at m/z 136.9908 and 247.2916, as well as m/z 136.9264 and 249.0855 (Table S1), correspondingly identical to those of NEP-D and Carbo-I (Fig. 1B).
Similarly, trace amounts of NEP-D and Carbo-I were also detected from the metabolites of S. citricolor (Table S1), a well-characterized ARM and NEP-A producer. Hence, this strain is also a potential producer of a PTN-related compound(s). In confirmation of this, LC-MS analysis of metabolites from this strain showed a distinctive [M+H]+ ion of COF (m/z 285.1188) (Table S1) but not those for PTN or other related antibiotics (Fig. 1A and B). This was also evident from the fragmentation pattern corresponding to that of the COF standard (Table S1). All these data verify that both M. haikouensis and S. citricolor are the producers of multiple nucleoside antibiotics.
Reconstitution of the ARM and COF pathways in a heterologous host.
To reconstitute the ARM and COF pathways in a heterologous host, 63B3, a cosmid containing the whole mac gene cluster, was screened from the genomic library of M. haikouensis. It was then introduced into Streptomyces aureochromogenes CXR14 (CXR14 herein), which was previously shown to be an appropriate host for the heterologous production of PTN-related purine nucleoside antibiotic pairs (Table 2) (8). The resultant recombinant was confirmed by PCR and fermented for further metabolite analysis. High-performance liquid chromatography (HPLC) analysis of the CXR14/63B3 recombinant showed target NEP-A and ARM peaks, which correspond to those of standards (Fig. 2B). Further LC-MS analysis of these peaks showed characteristic NEP-A and ARM [M+H]+ ions at m/z 264.1105 and m/z 266.1254, respectively (Table S1). In addition, the MS/MS fragmentation patterns of the two ions were correspondingly consistent with those of the authentic ARM and NEP-A standards (Table S1), but these ions were absent from the metabolites of CXR14/2463b, a control strain without the mac gene cluster. LC-MS also indicated that the CXR14/63B3 strain is able to produce COF, evident from a [M+H]+ ion at m/z 285.1184 (Fig. 2C) and associated fragment ions at m/z 134.9610, 152.9790, and 267.1675 and similar to those for COF produced by M. haikouensis (Table S1). All of this demonstrates that a single mac gene cluster spanning 25.3 kb is responsible for the biosynthesis of ARM, NEP-A, and COF.
TABLE 2.
Strains, plasmids, and cosmids used in this study
| Strain, plasmid, or cosmid | Description | Reference or source |
|---|---|---|
| Strains | ||
| M. haikouensis DSM 45626 | Wild-type producer of ARM, NEP-A, and COF | This study |
| S. citricolor NBRC 13005 | Wild-type producer of ARM, NEP-A, and COF | 19 |
| S. aureochromogenes | ||
| CXR14 | An industrial polyoxin producer with the entire polyoxin gene cluster deleted | 34 |
| CXR14/63B3 | CXR14 strain containing 63B3 | This study |
| CXR14/63B3ΔmacI | CXR14 strain containing 63B3ΔmacI | This study |
| CXR14/63B3ΔmacT | CXR14 strain containing 63B3ΔmacT | This study |
| CXR14/63B3ΔmacQ | CXR14 strain containing 63B3ΔmacQ | This study |
| CXR14/63B3ΔmacW | CXR14 strain containing 63B3ΔmacW | This study |
| CXR14/63B3ΔmacG | CXR14 strain containing 63B3ΔmacG | This study |
| CXR14/63B3ΔmacJ | CXR14 strain containing 63B3ΔmacJ | This study |
| CXR14/63B3ΔmacK | CXR14 strain containing 63B3ΔmacK | This study |
| CXR14/12H2 | CXR14 strain containing 12H2 | This study |
| CXR14/2463b | CXR14 strain containing pJTU2463b | This study |
| E. coli | ||
| DH10B | F− mcrA Δ(mrr-hsdRMS-mcrBC) ϕ80d lacZΔM15 ΔlacX74 deoR recA endA1 araD139 Δ(ara, leu)7697 galU galK λ− rpsL nupG | Gibco-BRL |
| BW25113/pIJ790 | λRED (gam beta exo) cat araC rep101 | 21 |
| ET12567(pUZ8002) | F− dam-13::Tn9 dcm-6 hsdM hsdR recF143 zjj-202::Tn10 galK2 galT22 ara-14 pacY1 xyl-5 leuB6 thi-1 pUZ8002 | 31 |
| BL21(DE3)ΔrihC | BL21(DE3) strain with rihC mutated | This study |
| Plasmids | ||
| pEASY-Blunt | pUCori lacZ f1 ori neo Amp | TransGen Biotech |
| pOJ446 | aac(3)IV SCP2 reppMB1* attФC31 oriT | 35 |
| pJTU2463b | The site of SCP2 in pOJ446 was replaced by int and attP | 36 |
| pET28a | neo reppMB1; T7 promoter | Novagen |
| pET26b | neo reppMB1; T7 promoter | Novagen |
| pET28a/macI | pET28a derivative carrying an NdeI-EcoRI fragment containing macI | This study |
| pET28a/macT | pET28a derivative carrying an NdeI-EcoRI fragment containing macT | This study |
| pET26b/macQ | pET26b derivative carrying an NdeI-HindIII fragment containing macQ | This study |
| pET28a/macR | pET28a derivative carrying an NdeI-EcoRI fragment containing macR | This study |
| pET28a/udp | pET28a derivative carrying an NdeI-EcoRI fragment containing udp | This study |
| Cosmids | ||
| 63B3 | The cosmid from M. haikouensis containing the entire mac gene cluster | This study |
| 63B3ΔmacI | The cosmid 63B3 with macI in-frame deleted | This study |
| 63B3ΔmacT | The cosmid 63B3 with macT in-frame deleted | This study |
| 63B3ΔmacW | The cosmid 63B3 with macW in-frame deleted | This study |
| 63B3ΔmacG | The cosmid 63B3 with macG in-frame deleted | This study |
| 63B3ΔmacJ | The cosmid 63B3 with macJ in-frame deleted | This study |
| 63B3ΔmacK | The cosmid 63B3 with macK in-frame deleted | This study |
| 63B3ΔmacQ | The cosmid 63B3 with macQ in-frame deleted | This study |
| 12H2 | The cosmid from S. citricolor containing the entire com gene cluster | This study |
The ARM producer S. citricolor also harbors a COF gene cluster (com) which is located at an unlinked locus.
Although S. citricolor is shown to be a COF producer in the present study, the genes for the biosynthesis of this antibiotic were missing in the surrounding region of the ari gene cluster (Fig. 2A) (19), suggesting that they are probably distributed over the S. citricolor genome. To locate the COF gene cluster, we sequenced the genome of S. citricolor and used the conserved enzymes, including MacM (SAICAR synthetase) and MacN (short-chain dehydrogenase), as probes, allowing the identification of a candidate gene cluster (com) (Fig. 2A).
To correlate the com gene cluster to COF biosynthesis, 12H2, a cosmid housing the complete com gene cluster, was screened from the S. citricolor genomic library and subsequently conjugated into the heterologous host, S. aureochromogenes CXR14 (Table 2). After validation by PCR, the recombinant (CXR14/12H2) was fermented for metabolite production and analyzed by LC-MS. The metabolites of CXR14/12H2 were characterized by a [M+H]+ ion at m/z 285.1189 (Fig. 3A), plus fragment ions at m/z 134.9322, 152.9562, and 266.9613, consistent with the fragmentation pattern of COF produced by S. citricolor (Fig. 3B to D). Hence, these combined data demonstrate that the target com gene cluster is responsible for COF biosynthesis in S. citricolor.
FIG 3.
Heterologous expression of the com gene cluster in S. aureochromogenes CXR14. (A) Extract ion chromatography (EIC) analysis of COF produced by S. citricolor and S. aureochromogenes CXR14/12H2. (B) Fragmentation pattern of COF. (C) LC-MS/MS analysis of COF produced by S. citricolor. (D) LC-MS/MS analysis of COF from the target metabolites produced by S. aureochromogenes CXR14/12H2. citri-WT, the sample of the S. citricolor wild-type strain; CXR14/12H2, the sample of S. aureochromogenes CXR14 containing 12H2; CXR14/pJTU2463b, the sample of S. aureochromogenes CXR14 containing pJTU2463b as a negative control; 12H2, a cosmid containing the whole com gene cluster.
Comparative analysis of the gene cluster diversity for ARM and COF biosynthesis in M. haikouensis and S. citricolor.
Comparative analysis of the target gene clusters in M. haikouensis and S. citricolor indicated that they generally match (the identity of the corresponding enzymes encoded by the gene clusters ranges from 35% to 81%) but also contain some particular genes (Fig. 2A). Of these, genes macWXRS are unique to the mac gene cluster (Fig. 2A; Table 1), and they may play additional roles in ARM or COF biosynthesis in M. haikouensis. The product of macX shows 58% identity to Sare_0588 (ATP phosphoribosyltransferase, HisG homolog), and macR and macS are individually annotated as an IMP dehydrogenase and epimerase. Likewise, three dispensable genes, including orf1 (encoding a transporter) (Fig. 2A; Table 1) and comEF (coding for a phosphohydrolase and an MFS transporter, respectively) (Fig. 2A, Table 3), are present only in the ari or com gene cluster.
TABLE 3.
Deduced functions of open reading frames in the com gene cluster
| Protein | No. of amino acids | Protein function | Homolog, origin | Identity, similarity (%) | Accession no. |
|---|---|---|---|---|---|
| ComF | 403 | MFS transporter | OV450_6573, Actinobacteria bacterium OV450 | 64, 75 | KPI33463 |
| ComE | 336 | Metal-dependent phosphohydrolase | PenF, S. antibioticus NRRL 3238 | 39, 49 | AKA87335 |
| ComA | 231 | SAICAR synthetase | MacM, M. haikouensis | 59, 71 | SCF06495 |
| ComB | 229 | Short-chain dehydrogenase | MacN, M. haikouensis | 57, 67 | SCF06487 |
| ComC | 140 | Hypothetical protein | MacO, M. haikouensis | 34, 44 | SCF06476 |
| ComD | 224 | Phosphatase | MacO, M. haikouensis | 33, 46 | SCF06476 |
The in silico analysis also showed that there are four genes, macWGJK, coding for S-adenosyl-l-homocysteine (SAH) hydrolases in the mac gene cluster, whereas the corresponding ari gene cluster includes only three of these (ari7, ari10, and ari11) (Fig. 2A; Table 1). Since the com gene cluster in S. citricolor is entirely responsible for COF biosynthesis, the four genes macXMNO accordingly have a role in COF biosynthesis in M. haikouensis (Fig. 2A; Tables 1 and 3). Hence, this implies that the COF and ARM natural products arise from independent biosynthetic pathways.
MacI and MacT function as unusual phosphorylases, catalyzing an irreversible step for the tailoring assembly of NEP-A and ARM.
The mac gene cluster contains two genes, macIT, that likely encode a deduced 5′-methylthioadenosine phosphorylase (Table 1). To investigate their functional roles in vivo, both were independently mutated on the 63B3 cosmid by a PCR-targeting technology (Fig. S2A and B) (21), and the resulting 63B3 variants were introduced into S. aureochromogenes CXR14, respectively. After verification, the recombinants CXR14/63B3ΔmacI and CXR14/63B3ΔmacT were fermented for metabolite analysis and were shown to have abolished ARM and NEP-A production, whereas the recombinant CXR14/63B3 (positive control) is capable of producing both antibiotics (Fig. 4A). This demonstrates the essential roles of these two genes for the biosynthesis of ARM and NEP-A.
FIG 4.
Functional investigation of MacI and MacT in the biosynthesis of ARM and NEP-A. (A) Extract ion chromatography (EIC) analysis of the metabolites produced by S. aureochromogenes CXR14/63B3 and its variants. 63B3, the sample from the strain of S. aureochromogenes CXR14 containing 63B3; 63B3ΔmacI, the sample from the strain of S. aureochromogenes CXR14 containing 63B3ΔmacI. The other two samples are correspondingly designated. (B) Scheme of the MacI- and MacT-catalyzed reaction. R-1-P, ribose-1-phosphate. (C) HPLC traces of the MacI- and MacT-catalyzed reactions using adenosine as the substrate. Std, the authentic standards of adenine and adenosine; MacI, the MacI-catalyzed reaction; MacT, the MacT-catalyzed reaction; N.C, the reaction without enzyme added as a negative control. (D) HPLC analysis of the MacI- and/or MacT- and UDP-coupled reactions. UDP hydrolyzes uridine to provide ribose-1-phosphate, which is used as a substrate by MacI and MacT. Std, the authentic standards of related compounds; MacI, the MacI- and UDP-coupled reaction; MacT, the MacT- and UDP-coupled reaction; MacI+MacT, the MacI-, MacT-, and UDP-coupled reaction, with MacI and MacT added simultaneously; N.C, the UDP reaction with adenine added as a negative control. UDP, uridine phosphorylase (accession no. ACT45511). R-1-P, ribose-1-phosphate.
Further investigation indicated that both MacI and MacT individually display 76% and 49% identity, respectively, to Ari9, a member of purine nucleoside phosphorylase superfamily proteins (Table 1). This superfamily of enzymes is widely distributed among all kingdoms of life, normally catalyzing a reversible reaction using purine (analog) and ribose-1-phosphate as the substrates (22). We tentatively proposed that these two enzymes are involved in the tailoring assembly step during the biosynthesis of ARM and NEP-A. To validate this, we overexpressed and purified the two proteins (MacI and MacT) (Fig. S2C) and tested their activities in vitro, initially using adenosine and phosphate as the substrates (Fig. 4B). HPLC analysis of the products showed that the MacI or MacT reactions could generate a distinctive peak for adenine, whose identity was verified by LC-MS analysis. The enzyme-negative reaction could not generate this target peak, establishing that MacI and MacT could utilize adenosine and phosphate as the substrates to form adenine (Fig. 4C; Fig. S2D to F). To determine if the MacI and MacT reactions are reversible for ARM and NEP-A, as are most of this family of enzymes, we tested the activity of MacI and MacT using ARM, NEP-A, and phosphate as the substrates. The results showed that ARM and NEP-A are not recognized as the substrates (Fig. S2G and H). Furthermore, to evaluate if MacI and MacT are responsible for the tailoring assembly, we used adenine and the compound 14 (see Fig. 7B) analog ribose-1-phosphate generated by a coupled uridine phosphorylase reaction as the substrates (Fig. 4B; Fig. S2C). The results indicated that the reactions of MacI and/or MacT could generate a characteristic peak for adenosine (Fig. 4D) under these reaction conditions. The identity of adenosine was further confirmed by LC-MS analysis (Fig. S3A to D). Taken together, these data suggest that MacI and MacT function as unusual phosphorylases, catalyzing an irreversible reaction for the tailored assembly of NEP-A and ARM but harboring a reversible enzymatic activity of hydrolyzing adenosine to supply adenine as precursor.
FIG 7.
Proposed biosynthetic pathways for COF (A) and for NEP-A and ARM (B).
MacQ is just an adenosine-specific deaminase.
The bioinformatic analysis indicated that MacQ possesses 35% identity to an annotated adenosine deaminase (ADA) Orf2 in the ari gene cluster (Table 1), but the functional role of the enzyme is as yet unassigned. To investigate the possible functional role of MacQ in vivo, we mutated macQ on the cosmid 63B3 via a PCR-targeting technology (Fig. S4A) (21); the resultant cosmid variant was conjugated into the CXR14 strain. After fermentation, metabolites from the recombinant CXR14/63B3ΔmacQ strain were submitted to LC-MS analysis. Unexpectedly, the results showed that the ARM production of the CXR14/63B3ΔmacQ recombinant was dramatically decreased in comparison to that of CXR14/63B3 (Fig. S4B and C), suggesting that macQ plays a positive role for ARM production.
We also deduced that MacQ is likely responsible for self-resistance by deamination of ARM and NEP-A to alleviate the toxicity to host cells, as proposed previously (19). MacQ was expressed and purified from Escherichia coli (Fig. S5A) and assayed in vitro using either ARM or NEP-A as the substrate. However, and unexpectedly, MacQ was not capable of catalyzing the deamination of ARM or NEP-A (Fig. S5B and C). This suggested that MacQ may function as an adenosine deaminase (Fig. 5A). Adenosine was therefore tested as the substrate and showed that the MacQ reaction gave rise to a distinctive inosine peak (Fig. 5B), validated by LC-MS analysis, whereas this is absent for the MacQ-negative reaction (negative control) (Fig. S5D and E). More than that, the activity of MacQ can be inhibited by the COF analog PTN (Fig. 5B). These data demonstrate that MacQ is an adenosine-specific deaminase and probably not directly involved in the self-resistance to ARM and NEP-A by the host cell, as has been recently proposed. An alternative suggestion is that MacQ functions in a fine-tuned pathway to modulate the potential “excess adenosine” stress on the producing cells (there are four SAH hydrolase genes in the mac gene cluster).
FIG 5.
In vitro characterization of MacQ as an adenosine-specific deaminase. (A) Scheme of the MacQ-catalyzed reaction. (B) HPLC analysis of the MacQ-catalyzed reaction. Std, the authentic standards of PTN, adenosine, and inosine. PTN (pentostatin) is a MacQ inhibitor. +1 mM PTN, the MacQ reaction using adenosine and 1 mM PTN as the substrate; +0.1 mM PTN, the MacQ reaction using adenosine and 0.1 mM PTN as the substrate; +0.01 mM PTN, the MacQ reaction using adenosine and 0.01 mM PTN as the substrate; no PTN, the MacQ reaction using adenosine as the substrate but without PTN added; N.C, the reaction without enzyme added as a negative control.
Biochemical characterization of MacR as an NADPH-dependent GMP reductase.
MacR shows 58% identity to N864_01465 of Intrasporangium chromatireducens Q5-1 (Table 1), an annotated GuaB1 family IMP dehydrogenase-related protein that is generally involved in GMP biosynthesis (23). MacR (Fig. S6A) was therefore tested in vitro using IMP as the substrate, although the expected product, XMP, was not detected, suggesting that MacR is not an IMP dehydrogenase. However, closer examination of the conserved domain of MacR showed a relationship to GMP reductase family proteins. An in vitro assay of MacR using GMP as the substrate and NADPH as a cofactor generated a distinctive IMP [M+H]+ ion at m/z 349.0535, with MS/MS fragment ions at m/z 136.8778, 232.9607, and 330.9223 (Fig. S6B to D), conforming to the fragmentation pattern of the IMP authentic standard. This IMP [M+H]+ ion was absent from the MacR-negative reactions (negative control) (Fig. 6B). We subsequently evaluated the cofactor specificity for MacR, and LC-MS analysis confirmed the specificity for NADPH, so that NADH was not recognized as a cofactor to maintain the enzymatic activity (Fig. 6B). Hence, these data establish that MacR functions as an NADPH-dependent reductase catalyzing GMP to IMP.
FIG 6.
In vitro characterization of MacR as an NADPH-dependent reductase. (A) Scheme of the MacR-catalyzed reaction. (B) HPLC analysis of the MacR-catalyzed reaction. Std, the authentic standards of IMP and GMP; NADPH+, the MacR-catalyzed reaction with NADPH as a cofactor and GMP as the substrate; NADH+, the MacR-catalyzed reaction with NADH as a cofactor and GMP as the substrate; no cofactor, the MacR-catalyzed reaction with GMP as the substrate but without a cofactor added; N.C, the reaction (with GMP and NADPH added) but without enzyme added as a negative control.
DISCUSSION
The advent of rapid next-generation DNA sequencing has revolutionized the process for the discovery of natural products from microorganisms (24, 25). Carbocyclic natural products include many interesting chemotypes, including the antitumor antibiotic pactamycin, the chitinase inhibitor allosamidin, and the carbocyclic purine nucleoside antibiotics (13). Generally, purine nucleoside natural products are synthesized by a nonmodular assembly line (2), which complicates the identification of their biosynthetic gene clusters. In this respect, the discovery of new ARM and COF producer strains (see Fig. S7 in the supplemental material) opens the way for the rational search for newer purine antibiotic pairs.
The COF-related antibiotics, including PTN, 2′-Cl PTN, carbocyclic COF, and adecypenol, share common enzymatic steps for the construction of the heterocyclic 1,3-diazepine ring (1, 9), and in this report we have confirmed these assignments. Three enzymes, MacX (ATP phosphoribosyltransferase), MacM (SAICAR synthetase), and MacN (short-chain dehydrogenase), are needed for COF biosynthesis. MacX catalyzes the initial condensation of ATP and PRPP (phosphoribosyl pyrophosphate) to produce compound 1, which is sequentially converted to compound 2 by HisE (phosphoribosyl-ATP pyrophosphatase), HisI (phosphoribosyl-AMP cyclohydrolase), and HisA (phosphoribosylisomerase) enzymes from the histidine pathway (Fig. 7A). Compound 2 undergoes modification by MacM to form compound 3, which is then sequentially dephosphorylated and dehydrogenated to yield the end product COF (Fig. 7A) (8, 11).
Previous labeling studies have established that ARM is derived from d-glucose (14), and more recently, it was shown to arise from d-fructose-6-phosphate (F6P) (19). Investigation of the mac gene cluster led to the identification of 20 genes (except for macXMNO) that are involved in ARM biosynthesis. We propose that ARM biosynthesis is initiated by MacB (Ari2 homolog), a MIPS (myo-inositol-1-phosphate synthase)-like enzyme that catalyzes the conversion of F6P to form compound 7. This is followed by hydrolysis (catalyzed by MacP and MacL) to give compound 8 and continuous reductions to generate compound 10 (Fig. 7B). Subsequently, compound 10 undergoes sequential epimerization and reduction by the enzymes MacV/MacS and MacD to give rise to compound 12 (Fig. 7B). Compounds 11 and 12 have been previously isolated from mutants of S. citricolor and confirmed to be the essential intermediates for ARM biosynthesis (18). Once formed, compound 12 is activated by phosphorylation and serves as the primary substrate for the final assembly of NEP-A. Two enzymes, MacF (ribokinase) and MacH (phosphoglucomutase), are proposed to catalyze the sequential reactions to form compounds 13 and then 14 (Fig. 7B).
The adenine ring in both ARM and NEP-A originate from adenosine (14, 15). Four SAH hydrolases (MacW, MacG, MacJ, and MacK) and two phosphorylases (MacI and MacT) are found in the mac gene cluster as suitable candidates for the adenine supply and final assembly of NEP-A (Fig. 7B; Fig. S8). SAH hydrolases catalyze the reversible hydrolysis of SAH to produce adenosine through a well-characterized oxidoreduction mechanism (26). We therefore propose that SAH is first hydrolyzed to adenosine, followed by further hydrolysis to adenine under the catalysis of MacI and MacT (Fig. 7B). Subsequently, MacI and MacT catalyze an unusual irreversible synthesis of NEP-A by the assembly of adenine and compound 14 (Fig. 7B). Earlier work has shown that ARM biosynthesis involves a final reduction step, with NEP-A as the substrate (27), and in the present study we therefore propose that MacA (or other related enzymes) is responsible for the final tailoring reduction (Fig. 7B).
It is very interesting to ask why the ARM pathway includes more SAH hydrolases in M. haikouensis than it seemingly needs. In this respect, it is notable that ARM and NEP-A are both potent SAH hydrolase inhibitors that could reversely undermine the biosynthesis of ARM and NEP-A by feedback inhibition of the activities of SAH hydrolases (Fig. 8) (13). The additional SAH hydrolases might therefore collaborate to maintain the efficient synthesis of ARM and NEP-A (Fig. S8).
FIG 8.
A proposed fine-tuned pathway associated with ARM and NEP-A biosynthesis. Blue arrows indicate that the reaction is found in the ARM pathway, and the red line (the T-like inhibition symbol) signifies the specific enzymatic step capable of being inhibited by a certain end product. ADSS, adenylosuccinate synthetase; ADSL, adenylosuccinate lyase; MAT, methionine adenosyltransferase; MT, methyltransferase.
The role of the adenosine deaminase (MacQ) in ARM and NEP-A biosynthesis is also unclear (28). In the present study, MacQ is shown to be an adenosine-specific deaminase. An excess of adenosine is generally recycled to produce dATP by the cells, which will exert a direct feedback inhibition of the ribonucleotide reductase under this stress circumstance and hence lead to a termination of DNA synthesis (29). Thus, the role of MacQ might be to fine-tune the pathway to prevent the buildup of the potential excess adenosine (Fig. 8). The inhibitory function of COF on MacQ may act as the safeguard for the protection of ARM and NEP-A from deamination by host ADAs (Fig. 8). IMP, the enzymatic product of MacR (GMP reductase), could be further recycled to AMP and other related products required for ARM biosynthesis (Fig. 8). Hence, an intricate control strategy has developed to modulate the biosynthesis of ARM.
In summary, we report the discovery of new producers of multiple nucleoside antibiotics by a target-oriented genome mining strategy and have delineated the gene cluster diversity of ARM and COF in Actinomycetes. We propose a fine-tuned pathway associated with the biosynthesis of these antibiotics, notably MacI and MacT catalyzing an irreversible reaction for the tailoring assembly of ARM and NEP-A, MacQ acting as an adenosine-specific deaminase, and MacR functioning as a dedicated GMP reductase for the supply of IMP pool. We anticipate that the deciphering of the genetic and enzymatic diversities for ARM and COF biosynthesis will accelerate the discovery of novel purine-related nucleoside antibiotic pairs.
MATERIALS AND METHODS
General materials and methods.
Strains, plasmids, and cosmids used in this study are described in Table 2, and primers are listed in Table 4. General methods employed in this work are in accordance with the standard protocols of Green and Sambrook (30) or Kieser et al. (31).
TABLE 4.
PCR primers used in this study
| Primer | Sequence (5′–3′) |
|---|---|
| mac-hrdB F | CGACTACACCAAGGGCTTCA |
| mac-hrdB R | TCAGCTCGATGACCTGGAA |
| macNidF | AAGCGGGCTGTCGTATTGC |
| macNidR | CCGGTCGTAACTCGATCTCA |
| macMidF | CGAGGTCATCGTCAAGAACG |
| macMidR | CAAGGCAGAAGTCCCACAGA |
| Haikou upidF | CGAGATCATCACGCACCCG |
| Haikou upidR | CTCCACGCCGAACCAGAAC |
| Haikou downidF | GTCATCCCGACGTACAACCG |
| Haikou downidR | CCACGCCGACCGACTAAAT |
| citri wk-id1F | CCAGTTCCAGGGCACGAT |
| citri wk-id1R | GGTAGTTGTCCAGCAGGTTC |
| citri wk-id3F | GACCTCTGGGACTTCTGCC |
| citri wk-id3R | CGACTCCACGCTCATCTTGT |
| macI PCRtgtF | CCGTCGGACGTGCCTTACCAGGCCAACCTGTGGGCGCTGTCTAGAGCTATTCCAGAAGT |
| macI PCRtgtR | GTAGGAGAGATTCATGTAGCAGAGCTGGAGCTCGCGTGCACTAGTCTGGATGCCGACG |
| macT PCRtgtF | CCGGCGAACGTGGCGGCACTCAAGGAGCTCGGCGCACGTTCTAGAGCTATTCCAGAAGT |
| macT PCRtgtR | CAGGTCCTCCGAAATTACGCCGTAGTCGGTGCAGAAGGAACTAGTCTGGATGCCGACG |
| macW PCRtgtF | ACAAGCACCGGGTTGTGGCGGCTGGGCAAGCTGCCTGACTCTAGAGCTATTCCAGAAGT |
| macW PCRtgtR | TTGGATCGCGCCGACGTGTGCCGGCCGGGGCGGCGGCGAACTAGTCTGGATGCCGACG |
| macG PCRtgtF | TACGGCTGCGACCGTGAGACCTTCTACCAGCAGGTGCAGTCTAGAGCTATTCCAGAAGT |
| macG PCRtgtR | CATCACCTCGGCCGGATGTGCCTCCGCCGCGGACTGGCCACTAGTCTGGATGCCGACG |
| macJ PCRtgtF | GTCTTCGGCCGGCGCGGAATGACCACCGCCGAGGTCGGTTCTAGAGCTATTCCAGAAGT |
| macJ PCRtgtR | CTGCTCCGTCGACACGGTCCGCGCCTGAAGGTCGGCCAGACTAGTCTGGATGCCGACG |
| macK PCRtgtF | GACCTGCGCAAGGGTGTCGAAGAGGTCCTGCTCACCTGGTCTAGAGCTATTCCAGAAGT |
| macK PCRtgtR | GGTCAGCACCTCCGAACCGGGATGCGCTGCCGGCATTCGACTAGTCTGGATGCCGACG |
| macQ PCRtgt2F | GAAGTGGCGCGAAACAACGACATCCGACTGCCTGCGGACTCTAGAGCTATTCCAGAAGT |
| macQ PCRtgt2R | GATCCGGCGCGGCCCATACGGCAGGACTTCCCGGACAGAACTAGTCTGGATGCCGACG |
| macIexF | GTCCATATGTCCAGAGCCGAGACG |
| macIexR | GGAATTCTCACGTCCGGGCTCCCGT |
| macTexF | GTCCATATGATCGACCTAGGAATC |
| macTexR | GGAATTCCTACCGATGCCCCTCCGT |
| macRexF | GTCCATATGGATCAGCAACGGTTT |
| macRexR | GGAATTCTCATCGGATGGGCATCGTC |
| macQOP-eF | GTCCATATGGTTGAGGGTAGCGCA |
| macQOP-eR | CCCAAGCTTGCTGCGCACCAGATC |
| UDPexF | GTCCATATGTCCAAGTCTGATGTTT |
| UDPexR | CCAAGCTTACAGCAGACGACGCGCC |
| macI idF | CGTCATCGACACTCCCTTCG |
| macI idR | AAACGGCCTCCGCCTCTTC |
| macT idF | GGATCGTCGAGACACCCTA |
| macT idR | CTCCGTATTCAGCCAGCAC |
| macW idF | CTCAATTCGATGCCGGTGTT |
| macW idR | CGGTGCGTTCGTGATGGT |
| macG idF | CCGCTGTCCACGAAGGAT |
| macG idR | TGAGGTACTGCTGCTGACCC |
| macJ idtF | AGCCAGATGCCGTTGCTC |
| macJ idtR | TGACCTCGCTCGGATTGC |
| macK idtF | ACCGAGCAGATGGTCGTGC |
| macK idR | CCAGGTTGGCGATGTTGC |
| macQ idF | CACGCTGCCGATGATGGA |
| macQ idR | CGTCGCAGGTCGGTGTTGA |
Enzymes, chemicals, and reagents.
All of the restriction enzymes and T4 DNA ligase used in this study were the products of New England Biolabs. The chemicals and reagents were purchased from Sigma-Aldrich, Thermo Scientific, or J&K Scientific. The standards of ARM and NEP-A were individually purchased from Santa Cruz Biotechnology, Inc., and Cayman Chemical.
Sequence analysis of M. haikouensis and S. citricolor.
The sequences of the M. haikouensis genome (accession no. FMCW01000024.1) and ARM biosynthetic gene cluster (accession no. LCO54541.1) were downloaded from the GenBank database. Sequencing of the S. citricolor genome was performed on an Illumina HiSeq 2500 machine. The raw data were processed to render the resulting clean reads, which were assembled by Velvet software (v1.2.07) to obtain the scaffold. After that, Glimmer 3.0 software was used for the genome annotation. Accurate bioinformatic analysis of the target DNA region was performed on the basis of the online programs FramePlot 4.0beta (http://nocardia.nih.go.jp/fp4/) and 2ndFind (http://biosyn.nih.go.jp/2ndfind/).
Transcriptional analysis of the targeted genes (macM and macN) by RT-PCR.
For RT-PCR analysis of the target gene cluster, the fermentation samples were individually collected at 1 day (24 h), 2 days (48 h), and 3 days (72 h), and the total RNA was extracted with TRIzol reagent (Life Technologies) (32) from 1 ml biomass of the cells. The quantity and quality of the extracted RNAs were assessed by a nanodrop spectrophotometer (Thermo Scientific), and 5 μg total RNA was treated with 1 μl RNA-free DNase I (Thermo Scientific) at 37°C for 1 h. RNA (1 μg) was reverse-transcribed into cDNA using the RevertAid first-strand cDNA synthesis kit (Thermo Scientific) primed with a random hexamer primer per the manufacturer's instructions. The reverse transcription reaction product was directly used for PCR amplification with the related pair of primers (mac-hrdBF/mac-hrdBR, macNidF/macNidR, and macMidF/macMidR). The PCR products were then evaluated by gel electrophoresis.
Genomic library construction and screening for M. haikouensis and S. citricolor.
The genomic libraries of M. haikouensis and S. citricolor were constructed on the basis of the standard protocol using EPI300-T1R as the host cell and pJTU2463b as the vector. For the screening of the positive cosmids from the genomic libraries of M. haikouensis, a narrow-down PCR approach was performed with the following PCR primers: Haikou upidF/Haikou upidR, Haikou downidF/Haikou downidR, and macMidF/macMidR (33). Likewise, the cosmid 12H2 was screened from the library of S. citricolor NBRC 13005 with primers citri wk-id1F/citri wk-id1R and citri wk-id3F/citri wk-id3R.
In-frame deletion of the target mac genes by PCR-targeting technology.
For in-frame deletion of the target genes, the kanamycin resistance cassette (neo) from SuperCos1 was amplified with primers (Table 4) and then correspondingly recombined into the target gene in 63B3 by a PCR-targeting strategy (21). Subsequently, the neo cassette was removed in vitro from the cosmid variants by XbaI-SpeI double digestion, followed by religation to generate the in-frame deletion scar of the target genes (Fig. S2A and B, S4A, and S8A; Table S2). The in-frame deletions were finally confirmed by PCR and sequencing analysis with the counterpart primers (Table 4).
Fermentation and detection of related nucleoside antibiotics.
M. haikouensis and S. citricolor were cultivated on YS agar (including 2 g yeast extract, 10 g soluble starch, and 15 g agar per liter, pH 7.3) and an MS plate (31), respectively. For fermentation, a single clone was inoculated in TSB medium and cultivated for 3 days; after that, the cultures (2%, vol/vol) were transferred to fermentation medium (16) and fermented (180 rpm, 28°C) for 5 days. For HPLC and LC-MS analysis, the fermentation beer was processed (with the addition of oxalic acid until pH 3.0 was reached). The HPLC analysis was performed using a Shimadzu LC-20AT instrument equipped with a C18 column (Dikma Diamonsil Plus; 5 μm, 4.6 by 250 mm) with an elution gradient of 5% to 20% methanol–0.15% aqueous trifluoroacetic acid (TFA) over 25 min at a flow rate 0.5 ml/min. LC-MS analysis was carried out on a Thermo Fisher Scientific ESI-LTQ Orbitrap mass analyzer controlled by Xcalibur.
Expression and purification of protein in E. coli BL21(DE3)ΔrihC.
For the overexpression and purification of the target proteins (with MacI as an example), macI was PCR amplified using the primers listed in Table 4 and then cloned into a pEASY-Blunt vector. After confirmation by DNA sequencing, the NdeI-EcoRI engineered fragment was cloned into pET28a at the corresponding sites. Finally, the resulting expression construct was transformed into BL21(DE3)ΔrihC competent cells. Expression and purification of the His6-tagged MacI were conducted according to the protocol of Wu et al. (8). Protein concentration was quantified using a bicinchoninic acid protein assay kit.
Biochemical assays of MacI and MacT coupled with the UDP (uridine phosphorylase) reaction.
For the hydrolysis activity assays, the MacI and MacT reaction mixtures (100 μl), consisting of 50 mM Tris HCl buffer (pH 7.5, 30°C), 2 mM EDTA, 20 mM phosphate buffer (pH 7.5), 1 mM substrate (Nep-A, ARM, or adenosine), and the proteins UDP and MacI or MacT (20 μg each), were incubated at 30°C for 4 h and then immediately terminated by adding an equivalent volume of methanol. Following centrifugation (12,000 rpm, 5 min) to remove the protein pellet, the reaction mixtures were subsequently analyzed by HPLC (Shimadzu LC-20AT) and LC-MS (Thermo LTQ-Orbitrap). HPLC was performed with an elution gradient of 5% to 20% methanol–15 mM TFA over 20 min at a flow rate of 0.5 ml/min. LC-MS/MS analysis was conducted in an electrospray ionization (ESI) trap mass spectrometer in the positive-ion mode with the following parameters: drying gas at 275°C, 10 liters/ml, and nebulizer pressure at 30 lb/in2.
In vitro assay of MacQ.
For the MacQ activity assay, the reaction mixture (100 μl), consisting of 50 mM phosphate buffer (pH 7.5), 1 mM substrate (Nep-A, ARM, or adenosine), 1 mM, 0.1 mM, or 0.01 mM PTN (pentostatin), and 20 μg MacQ, was incubated at 30°C for 2 h and then inactivated by the immediate addition of an equivalent volume of methanol. After the protein pellet was removed by centrifugation (12,000 rpm, 5 min), the reaction supernatant was assayed under the same conditions as described above.
Enzymatic assay of MacR.
For the enzymatic assay of MacR, the reaction mixture (100 μl), consisting of 1 mM IMP, 2 mM NADPH or NADH, 100 mM Tris HCl buffer (pH 8.0), 50 mM KCl, 1 mM dithiothreitol, 2 mM EDTA, and 20 μg MacR, was incubated at 30°C for 4 h, and then the reaction was terminated by the prompt addition of the methanol (an equivalent volume). Subsequently, MacR was detected in the same way as described above.
Accession number(s).
The DNA sequence for the COF gene cluster from S. citricolor is deposited in the GenBank database under accession number KY313601.
Supplementary Material
ACKNOWLEDGMENTS
This work was supported by grants from the National Natural Science Foundation of China (31770041) and the Open Funding Projects of the State Key Laboratory of Bioactive Substance and Function of Natural Medicines (GTZK201701), as well as the State Key Laboratory of Microbial Metabolism (MMLKF16-03).
Footnotes
Supplemental material for this article may be found at https://doi.org/10.1128/AEM.01860-18.
REFERENCES
- 1.Isono K. 1988. Nucleoside antibiotics: structure, biological activity, and biosynthesis. J Antibiot (Tokyo) 41:1711–1739. doi: 10.7164/antibiotics.41.1711. [DOI] [PubMed] [Google Scholar]
- 2.Chen W, Qi J, Wu P, Wan D, Liu J, Feng X, Deng Z. 2016. Natural and engineered biosynthesis of nucleoside antibiotics in Actinomycetes. J Ind Microbiol Biotechnol 43:401–417. doi: 10.1007/s10295-015-1636-3. [DOI] [PubMed] [Google Scholar]
- 3.Winn M, Goss RJ, Kimura K, Bugg TD. 2010. Antimicrobial nucleoside antibiotics targeting cell wall assembly: recent advances in structure-function studies and nucleoside biosynthesis. Nat Prod Rep 27:279–304. doi: 10.1039/B816215H. [DOI] [PubMed] [Google Scholar]
- 4.Chen S, Kinney WA, Van Lanen S. 2017. Nature's combinatorial biosynthesis and recently engineered production of nucleoside antibiotics in Streptomyces. World J Microbiol Biotechnol 33:66. doi: 10.1007/s11274-017-2233-6. [DOI] [PubMed] [Google Scholar]
- 5.Niu G, Tan H. 2015. Nucleoside antibiotics: biosynthesis, regulation, and biotechnology. Trends Microbiol 23:110–119. doi: 10.1016/j.tim.2014.10.007. [DOI] [PubMed] [Google Scholar]
- 6.Shelton J, Lu X, Hollenbaugh JA, Cho JH, Amblard F, Schinazi RF. 2016. Metabolism, biochemical actions, and chemical synthesis of anticancer nucleosides, nucleotides, and base analogs. Chem Rev 116:14379–14455. doi: 10.1021/acs.chemrev.6b00209. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Serpi M, Ferrari V, Pertusati F. 2016. Nucleoside derived antibiotics to fight microbial drug resistance: new utilities for an established class of drugs? J Med Chem 59:10343–10382. doi: 10.1021/acs.jmedchem.6b00325. [DOI] [PubMed] [Google Scholar]
- 8.Wu P, Wan D, Xu G, Wang G, Ma H, Wang T, Gao Y, Qi J, Chen X, Zhu J, Li YQ, Deng Z, Chen W. 2017. An unusual protector-protégé strategy for the biosynthesis of purine nucleoside antibiotics. Cell Chem Biol 24:171–181. doi: 10.1016/j.chembiol.2016.12.012. [DOI] [PubMed] [Google Scholar]
- 9.Bush BD, Fitchett GV, Gates DA, Langley D. 1993. Carbocyclic nucleosides from a species of Saccharothrix. Phytochemistry 32:737–739. doi: 10.1016/S0031-9422(00)95163-X. [DOI] [Google Scholar]
- 10.Hanvey JC, Hardman JK, Suhadolnik RJ, Baker DC. 1984. Evidence for the conversion of adenosine to 2′-deoxycoformycin by Streptomyces antibioticus. Biochemistry 23:904–907. doi: 10.1021/bi00300a017. [DOI] [PubMed] [Google Scholar]
- 11.Gao Y, Xu G, Wu P, Liu J, Cai YS, Deng Z, Chen W. 2017. Biosynthesis of 2′-chloropentostatin and 2′-amino-2′-deoxyadenosine highlights a single gene cluster responsible for two independent pathways in Actinomadura sp. strain ATCC 39365. Appl Environ Microbiol 83:e00078-. doi: 10.1128/AEM.00078-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Xia Y, Luo F, Shang Y, Chen P, Lu Y, Wang C. 2017. Fungal cordycepin biosynthesis is coupled with the production of the safeguard molecule pentostatin. Cell Chem Biol 24:1479–1489. doi: 10.1016/j.chembiol.2017.09.001. [DOI] [PubMed] [Google Scholar]
- 13.Jenkins GN, Turner NJ. 1995. The biosynthesis of carbocyclic nucleosides. Chem Soc Rev 24:169–176. doi: 10.1039/cs9952400169. [DOI] [Google Scholar]
- 14.Parry RJ, Bornemann V. 1985. Biosynthesis of aristeromycin—elucidation of the origin of the adenine and cyclopentane rings. J Am Chem Soc 107:6402–6403. doi: 10.1021/ja00308a047. [DOI] [Google Scholar]
- 15.Parry RJ, Bornemann V, Subramanian R. 1989. Biosynthesis of the nucleoside antibiotic aristeromycin. J Am Chem Soc 111:5819–5824. doi: 10.1021/ja00197a048. [DOI] [Google Scholar]
- 16.Kusaka T, Yamamoto H, Shibata M, Muroi M, Kishi T. 1968. Streptomyces citricolor nov. sp. and a new antibiotic, aristeromycin. J Antibiot (Tokyo) 21:255–263. doi: 10.7164/antibiotics.21.255. [DOI] [PubMed] [Google Scholar]
- 17.Yaginuma S, Muto N, Tsujino M, Sudate Y, Hayashi M, Otani M. 1981. Studies on neplanocin-A, new anti-tumor antibiotic. 1. Producing organism, isolation and characterization. J Antibiot (Tokyo) 34:359–366. [DOI] [PubMed] [Google Scholar]
- 18.Hill JM, Jenkins GN, Rush CP, Turner NJ, Willetts AJ, Buss AD, Dawson MJ, Rudd BAM. 1995. Revised pathway for the biosynthesis of aristeromycin and neplanocin a from d-glucose in Streptomyces citricolor. J Am Chem Soc 117:5391–5392. doi: 10.1021/ja00124a035. [DOI] [Google Scholar]
- 19.Kudo F, Tsunoda T, Takashima M, Eguchi T. 2016. Five-membered cyclitol phosphate formation by a myo-inositol phosphate synthase orthologue in the biosynthesis of the carbocyclic nucleoside antibiotic aristeromycin. Chembiochem 17:2143–2148. doi: 10.1002/cbic.201600348. [DOI] [PubMed] [Google Scholar]
- 20.Xie QY, Qu Z, Lin HP, Li L, Hong K. 2012. Micromonospora haikouensis sp nov., isolated from mangrove soil. Antonie Van Leeuwenhoek 101:649–655. doi: 10.1007/s10482-011-9682-y. [DOI] [PubMed] [Google Scholar]
- 21.Gust B, Challis GL, Fowler K, Kieser T, Chater KF. 2003. PCR-targeted Streptomyces gene replacement identifies a protein domain needed for biosynthesis of the sesquiterpene soil odor geosmin. Proc Natl Acad Sci U S A 100:1541–1546. doi: 10.1073/pnas.0337542100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Zang Y, Wang WH, Wu SW, Ealick SE, Wang CC. 2005. Identification of a subversive substrate of Trichomonas vaginalis purine nucleoside phosphorylase and the crystal structure of the enzyme-substrate complex. J Biol Chem 280:22318–22325. doi: 10.1074/jbc.M501843200. [DOI] [PubMed] [Google Scholar]
- 23.Buey RM, Ledesma-Amaro R, Velazquez-Campoy A, Balsera M, Chagoyen M, de Pereda JM, Revuelta JL. 2015. Guanine nucleotide binding to the Bateman domain mediates the allosteric inhibition of eukaryotic IMP dehydrogenases. Nat Commun 6:8923. doi: 10.1038/ncomms9923. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Tao W, Yurkovich ME, Wen S, Lebe KE, Samborskyy M, Liu Y, Yang A, Liu Y, Ju Y, Deng Z, Tosin M, Sun Y, Leadlay PF. 2016. A genomics-led approach to deciphering the mechanism of thiotetronate antibiotic biosynthesis. Chem Sci 7:376–385. doi: 10.1039/C5SC03059E. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Bachmann BO, Van Lanen SG, Baltz RH. 2014. Microbial genome mining for accelerated natural products discovery: is a renaissance in the making? J Ind Microbiol Biotechnol 41:175–184. doi: 10.1007/s10295-013-1389-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Reddy MC, Kuppan G, Shetty ND, Owen JL, Ioerger TR, Sacchettini JC. 2008. Crystal structures of Mycobacterium tuberculosis S-adenosyl-l-homocysteine hydrolase in ternary complex with substrate and inhibitors. Protein Sci 17:2134–2144. doi: 10.1110/ps.038125.108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Parry RJ, Jiang YJ. 1994. The biosynthesis of aristeromycin-conversion of neplanocin-A to aristeromycin by a novel enzymatic reduction. Tetrahedron Lett 35:9665–9668. doi: 10.1016/0040-4039(94)88354-8. [DOI] [Google Scholar]
- 28.Nakamura H, Koyama G, Iitaka Y, Ono M, Yagiawa N. 1974. Structure of coformycin, an unusual nucleoside of microbial origin. J Am Chem Soc 96:4327–4328. doi: 10.1021/ja00820a049. [DOI] [PubMed] [Google Scholar]
- 29.Johnston JB. 2011. Mechanism of action of pentostatin and cladribine in hairy cell leukemia. Leuk Lymphoma 52(Suppl 2):S43–S45. doi: 10.3109/10428194.2011.570394. [DOI] [PubMed] [Google Scholar]
- 30.Green MR, Sambrook J. 2002. Molecular cloning: a laboratory manual, 3rd ed Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. [Google Scholar]
- 31.Kieser T, Bibb MJ, Chater KF, Butter MJ, Hopwood DA. 2000. Practical Streptomyces genetics, 2nd ed John Innes Foundation, Norwich, United Kingdom. [Google Scholar]
- 32.Chomczynski P. 1993. A reagent for the single-step simultaneous isolation of RNA, DNA and proteins from cell and tissue samples. Biotechniques 15:532. [PubMed] [Google Scholar]
- 33.Cheng L, Chen W, Zhai L, Xu D, Huang T, Lin S, Zhou X, Deng Z. 2011. Identification of the gene cluster involved in muraymycin biosynthesis from Streptomyces sp. NRRL 30471. Mol Biosyst 7:920–927. doi: 10.1039/C0MB00237B. [DOI] [PubMed] [Google Scholar]
- 34.Qi J, Wan D, Ma H, Liu Y, Gong R, Qu X, Sun Y, Deng Z, Chen W. 2016. Deciphering carbamoylpolyoxamic acid biosynthesis reveals unusual acetylation cycle associated with tandem reduction and sequential hydroxylation. Cell Chem Biol 23:935–944. doi: 10.1016/j.chembiol.2016.07.011. [DOI] [PubMed] [Google Scholar]
- 35.Bierman M, Logan R, O'Brien K, Seno ET, Rao RN, Schoner BE. 1992. Plasmid cloning vectors for the conjugal transfer of DNA from Escherichia coli to Streptomyces spp. Gene 116:43–49. doi: 10.1016/0378-1119(92)90627-2. [DOI] [PubMed] [Google Scholar]
- 36.Zhai L, Lin S, Qu D, Hong X, Bai L, Chen W, Deng Z. 2012. Engineering of an industrial polyoxin producer for the rational production of hybrid peptidyl nucleoside antibiotics. Metab Eng 14:388. doi: 10.1016/j.ymben.2012.03.006. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.