3′-O-β-Glucosyl-4′,5′-didehydro-5′-deoxyadenosine Is a Natural Product of the Nucleocidin Producers Streptomyces virens and Streptomyces calvus

3′-O-β-Glucosyl-4′,5′-didehydro-5′-deoxyadenosine 13 is identified as a natural product of Streptomyces calvus and Streptomyces virens. It is also generated in vitro by direct β-glucosylation of 4′,5′-didehydro-5′-deoxyadenosine 12 with the enzyme NucGT. The intact incorporation of oxygen-18 and deuterium isotopes from (±)[1-18O,1-2H2]-glycerol 14 into C-5′ of nucleocidin 1 and its related metabolites precludes 3′-O-β-glucosyl-4′,5′-didehydro-5′-deoxyadenosine 13 as a biosynthetic precursor to nucleocidin 1.

Nucleocidin 1 is a modified nucleoside antibiotic isolated originally from Streptomyces calvus. 1 It is of structural interest as a natural product because it contains a fluorine atom and a sulfamyl moiety, two functional groups that are exceedingly rare in nature. 2 This has to be contrasted with the widespread use of selective fluorination and to a lesser extent sulfamylation, in medicinal chemistry and drug discovery programs. 3Genome mining has allowed several additional Streptomyces strains to be identified which also have the ability to produce nucelocidin 1 and various related metabolites in culture. 4These organisms include Streptomyces virens, which is a good producer of nucelocidin 1.Only a handful of fluorine-containing metabolites are known, the most notable of which is fluoroacetate 2, a toxin found in a wide range of plants and bacteria. 5A bacterial fluorination enzyme (fluorinase) that converts S-adenosyl-L-methionine (SAM) to 5′-fluorodeoxyadenosine (5′-FDA 3) is known to be involved in bacterial fluoroacetate 2 biosynthesis; however, that enzyme is not involved in nucleocidin biosynthesis. 6There is no such fluorinase gene encoded in any of the genomes of the nucleocidin producers, and the site of the fluorine atom at the 4′-carbon of the ribose is inconsistent with the chemistry of that enzyme.
In an effort to shed light on nucelocidin 1 biosynthesis, we and others have identified additional fluorinated metabolites associated with nucleocidin 1. 7 These include 4′-fluoroadenosine 4 and F-Met I 5 and F-Met II 6, the latter two of which are β-glucosylated at the 3′-O hydroxy group.7a Most recently, the acetylated glucose derivatives 7 and 8 were also reported.7b We and others 7,8 have carried out extensive gene knockout experiments probing the putative nucleocidin biosynthetic gene cluster.We reported 8 recently that at least 11 genes are required for sulfamylation, and four genes appear crucial to fluorination.Interestingly when sulfamylation ability was disabled in selected KO strains, this did not always adversely affect F-Met-I 4 production, and conversely some gene KOs that disabled fluorination still resulted in 5′-O-sulfamylated natural products such as 9 and 10.Some of these findings are reinforced in the studies of Zechel et al. 7c Consistent with the observed decoupling of these biosynthetic motifs, we have found that the corresponding nonfluorinated, but sulfamylated metabolites 9 and 10 are also co-produced with nucleocidin 1 and its analogues in the wild-type producing strains of S. calvus and S. virens at a level of about 10−15% of the fluorometabolites.7b

■ RESULTS AND DISCUSSION
We now report the isolation of 3′-O-β-glucosyl-4′,5′didehydro-5′-deoxyadenosine 13 from both S. calvus and S. virens.3′-O-β-Glucosyl-4′,5′-didehydroadenosine 13 was observed initially by mass spectrometry (412.17Da) from the total ion chromatogram of crude natural product extracts of S. calvus.In order to confirm the structure of 13 by isolation, cultures of both S. calvus and S. virens were grown to maturity (see SI), and nucleocidin 1 and its related metabolites such as 5 and 6 were isolated as previously described 7 after adsorption onto charcoal and then washing with acetone.HPLC fractionation guided by MS-MS analysis was used to identify 3′-O-β-glucosyl-4′,5′-didehydro-5′-deoxyadenosine 13, and the metabolite was further isolated in low microgram amounts after two rounds of preparative HPLC.High-resolution mass spectrometry (HRMS) gave an accurate mass for 13 ([M + H] + = 412.1451m/z, C 16 H 22 N 5 O 8 + , Figure S3).Sufficient material (around 0.2 mg) was isolated to be able to record 1 H NMR. In addition, a reference sample of 13 was prepared from synthetic 12 following a previously reported synthesis. 9ith synthetic 12 in hand, we explored its enzymatic 3′-Oglucosylation using a previously identified 7a glucosyl transferase, NucGT, which is associated with the nucleocidin 1 biosynthetic gene cluster (BGC).NucGT from S. calvus and S. virens is already shown to have the ability to β-glucosylate the 3′-OH group of nucleocidin 1, adenosine, and defluoronucleocidin 9 to generate 6, 10, and 11, respectively.Compound 12 was assayed as a potential substrate for NucGT, and it generated a product that proved to be identical to 13 by HPLC retention time and the 1 H NMR of the isolated product was essentially identical to semisynthetic 13.The data are listed in Table 1.The alignment of the two 1 H NMR spectra is shown in the SI (Figure S4).
A study of the reaction kinetics for NucGT with 12 and UDP-glucose to generate 13 evaluated key kinetic parameters (K m and V max ), and the data presented in Figure 1 unexpectedly indicated that 12 is a measurably more efficient substrate overall than adenosine.
The presence of 13 in the culture medium and the ability of NucGT to convert 12 to 13 presented the possibility that either 13 or 12 may be a biosynthetic precursor to nucleocidin 1.For example, epoxidation of the 4′,5′-exocyclic double bond and then epoxide ring opening with fluoride ion at C-4′ presented a plausible strategy for fluorination at C-4′ of the ribose with concomitant formation of the necessary C−O bond at C-5′.It follows that if this C−O bond is introduced late in the biosynthesis from molecular oxygen, then the oxygen atom would not derive from the original ribose, but instead would arise from an oxidizing enzyme supplied by molecular oxygen. 10Thus, we decided to carry out an incubation experiment with (±)[1-18 O,1-2 H 2 ]-glycerol 14.Glycerol incorporates into the pentose phosphate pathway and is known to contribute an intact C−O bond to ribose at C-5′−O. 11Thus, the successful incorporation of both deuterium and oxygen-18 isotopes from glycerol would disqualify 13 and 12 as biosynthetic intermediates to nucleocidin 1, as retention of an intact C-5′−O bond from glycerol would disqualify an origin by oxidation of the exomethylene double bond.
It is notable that there are clear heavy isotope (deuterium)induced fluorine signals ∼0.22 ppm upfield of F-Met I 5 and F-Met II 6 at approximately 1−2% of the unlabeled signal.This is entirely consistent with the incorporation of two deuterium atoms into C-5′ as illustrated in Scheme 1 and previously established 11 in glycerol supplementation experiments, and it is a clear indication of intact incorporation of the deuteriums from (±)[1-18 O,1-2 H 2 ]-glycerol 14 into C-5′ of the fluorometabolites.It was important now to determine if the oxygen-18 atom was also incorporated, although this could not be determined directly by 19 F{ 1 H} NMR as the chemical shifts induced by 18 O over 16 O are just too small to be recorded over three bonds.
The intact incorporation of 18 O along with both deuterium atoms of 14 was, however, confirmed by LC-HRMS and FT-ICR MS analyses.Metabolite extracts from (±) [1-18 O,1-2 H 2 ]-glycerol 14-supplied S. virens were semipurified on HPLC.The fractions containing fluorometabolites 5 and 6 were analyzed by LCMS and LC-HRMS, and isotope fine structure analysis was conducted on FT-ICR MS particularly to explore the intensity of any [M + 4 + H] + ions associated with the fluorometabolites.The [M + 4 + H] + abundance for 6 was 2.6% (14 supplied) relative to the parent unlabeled molecular ion, where it was 0.0% in the control (unlabeled glycerol added, Figure S17).The value of 2.6% is approximately twice that indicated for deuterium incorporations by 19 F{ 1 H} NMR; however there are two sites for glycerol incorporation into the β-glucosylated metabolites; these are the ribose ring and the βglucose moiety itself.Therefore, daughter ion fragmentation analysis was conducted in an FT-ICR experiment to deconvolute the incorporation of the isotopes into the ribose and glucose moieties.This analysis indicated that there was a 1.0% [M + 4 + H] + abundance of heavy isotope detected from the daughter ion 369.0842 m/z of F-Met II 6 by FT-ICR MS analysis.This ion contains the ribose moiety but no longer has the β-glucosyl moiety.The level of enrichment is at a level consistent with the observed 19 F NMR incorporations, and the difference indicates incorporations of the isotope also into the β-glucose moiety of the parent molecule.The intact

■ CONCLUSIONS
In summary, we have identified 3′-O-β-glucosyl-4′,5′-didehydro-5′-deoxyadenosine 13 as a metabolite of both S. calvus and S .v i r e n s .S u p p l e m e n t a t i o n e x p e r i m e n t s w i t h (±)-[1-18 O,1-2 H 2 ]-glycerol 14 indicated the intact incorporation of both deuteriums and the oxygen-18 atom into C-5′ of F-Met II 6, and thus 13 or deglucosylated 12 does not appear to be a biosynthetic precursor of nucleocidin 1 and its related fluorometabolites, as this would be inconsistent with oxygen-18 retention from glycerol.NucGT is able to convert dehydroadenosine 12 to 13 in vitro, although it is not clear if this is relevant metabolically, as 12 could not be identified in S. virens extracts.Such dehydroadenosines have not previously been reported as microbial natural products, although 4′,5′didehydro-5′-deoxyadenosine 12 has been detected analytically within excreted metabolites as a disease 13 and dietary 14 biomarker in mammalian metabolism.In addition, 12 has been identified as a photodegradation product of adenosyl cobalamin in vitro, 15 and also 4′,5′-didehydro-5′-deoxyadenosine 12 is interconverted with adenosine in vitro by the action of purified S-adenosyl-L-homocysteine hydrolase, 16 so free dehydroadenosine 12 may get processed that way if 13 is deglucosylated.

■ EXPERIMENTAL SECTION
General Experimental Procedures.Room temperature refers to 18−25 °C.Air-and moisture-sensitive reactions were carried out under an atmosphere of argon in oven-dried glassware.All evaporations and concentrations were performed under reduced pressure (in vacuo) with a Buchi Rotavapor R-200.The freeze-drying was performed under vacuum by a Christ Alpha 1-2 LDplus −55 °C freeze-dryer.All reagents were purchased from commercial suppliers and were used without further purification unless otherwise stated.Anhydrous solvents (DCM, THF, Et 2 O) were obtained from an MBraun MB SPS-800 solvent purification system by passage through two drying columns and dispensed under an argon atmosphere.All microbiological work was carried out in a Gallenkamp laminar flowhood, using standard sterile techniques.Glassware and consumables for biological operations were sterilized by autoclaving, flaming, or wiping with 75% ethanol before using.Sterilized consumables were used as supplied.Media were sterilized by 121 °C, 15 min autoclaving.Cell cultures were incubated in a temperature-controlled incubator (New Brunswick Scientific).Centrifugation of 20 mL to 1 L was processed by a Beckman Avanti centrifuge.A Hettich Mikro 200 benchtop centrifuge was used for microcentrifugation.
Nuclear Magnetic Resonance (NMR) Spectroscopy.NMR spectra were recorded at 298 K on a Bruker Advance II 400, Advance III HD 500, or Advance III HD 700 instrument. 1 H and 13 C NMR spectra were recorded in a deuterated solvent as the lock and the residual solvent as the internal standard. 19F NMR spectra were recorded by using CFCl 3 as an external reference.Chemical shifts are reported in parts per million (ppm), and coupling constants (J) are reported in hertz (Hz).The abbreviations for the multiplicity of the proton, carbon, and fluorine signals are as follows: s singlet, d doublet, dd doublet of doublets, ddd doublet of doublet of doublets, t triplet, dt doublet of triplets, q quartet, m multiplet, br s broad singlet.
LC-MS Analysis.Extracts from culture media were freeze-dried, resuspended in 50% acetonitrile/water to about 1−5 mL, and centrifuged (21300g) for 10 min to remove precipitates.These samples were analyzed at the Mass Spectrometry Facility at the University of St Andrews using ThermoFisher Xcalibur Orbitrap instrument in positive ion mode.Due to low abundance of metabolites, some samples were partially purified by HPLC; the majority of the acetonitrile/water elution fractions were collected, and after removal of the solvent, the dry extracts were resuspended in water.
High-resolution electrospray ionization spectra were acquired on a Bruker MaXis II ESI-Q-TOF-MS instrument connected to a Dionex 3000 RS UHPLC instrument fitted with an ACE C4-300 RP column (100 × 2.1 mm, 5 μm, 30 °C).The metabolites were eluted with a linear gradient of 5−100% MeCN containing 0.1% formic acid over 30 min.The mass spectrometer was operated in positive ion mode with a scan range of 200−3000 m/z.Source conditions: end plate offset at −500 V; capillary at −4500 V; nebulizer gas (N 2 ) at 1.8 bar; dry gas (N 2 ) at 9.0 L min −1 ; dry temperature at 200 °C.Ion transfer conditions: ion funnel RF at 400 Vpp; multiple RF at 200 Vpp; quadrupole low mass at 200 m/z; collision energy at 8.0 eV; collision RF at 2000 Vpp; transfer time at 110.0 μs; prepulse storage time at 10.0 μs.MS data were analyzed using Bruker DataAnalysis.
FT-ICR high-resolution MS data were acquired on a 12T SolariX 2XR Fourier transform−ion cyclotron resonance instrument equipped with electrospray (ESI) ionization (Bruker Daltonics).RP-HPLCpurified samples were infused at 2 μL/min, and spectra were acquired between 280 and 4000 m/z using 4 MWord data collection.Using these conditions, mass resolution of ca.1,000,000 was achieved, allowing isotope fine structure analysis.For fragmentation experiments, individual species were isolated using the quadrupole, and tandem MS was performed using collision-induced dissociation (CID) by applying a collision energy of 15−25 V.
Seed Culture of Streptomyces calvus and Streptomyces virens.The seed culture was performed in TSBY liquid medium composed of tryptone soy broth (3%, w/w), sucrose (10.3%, w/w), and yeast extract (0.5%, w/w).The seed cultures of S. calvus and S. virens were obtained by inoculating 50 μL of spores into 50 mL of TSBY, and the culture was allowed to grow at 28 °C for 2 days (50 mL of medium, in a 250 mL conical flask with shaking at 180 rpm).
Fermentation Culture.A mass of the mycelium of S. calvus or S. virens was obtained by inoculating a sterilized, defined medium (100 mL in a 500 mL conical flask) with the seed culture obtained above (inoculate with 2 mL per 100 mL), and the culture was allowed to grow at 28 °C, 180 rpm for 8 days.The defined medium (1 L) was made with tap water, corn steep liquor (12.5 g), mannitol (10 g), sodium chloride (2 g), diammonium phosphate (2 g), monopotassium phosphate (1.5 g), magnesium sulfate heptahydrate (0.25 g), Hoagland's salt solution (1 mL), and potassium fluoride solution (7.5 mL, 0.5 M).
Kinetic Study.Enzymatic activity was assayed at 37 °C by monitoring the production using analytical HPLC (Shimadzu SPD-20A detector at 254 nm coupled with a SIL-20A HT autosampler).The glucosyltransferase (0.567 μM) was incubated with various concentrations of substrates, MgCl 2 (100 mM), and a saturating concentration of UDP-glucose (17.7 mM) in Tris-HCl buffer (50 mM, pH 7.8), in a final volume of 0.25 mL.An aliquot (100 μL) was denaturalized with phenol/chloroform at various time points (1 or 2 min) and then instantly cooled on ice.Precipitated protein was then removed by centrifugation (13 000 rpm, 10 min at 4 °C), and the sample was filtered with a PTFE filter (0.22 μm, Fisherbrand).The eluant was injected into analytical HPLC to determine the level of the products against a standard curve.Each sample was injected three times to obtain the average value.Kinetic parameters were obtained by Michaelis−Menten fitting of the initial velocity against substrate concentrations using Prism 8.0.
Pulse Supplementation Experiment.Cultures of S. calvus or S. virens (100 mL) were shaken at 30 °C, labeled glycerol was added after 2 days, and then the same quantity was added every day for the next 6 days.The final concentration of labeled glycerol was 8.8 mM.After 8 days of fermentation, the cells were discarded after centrifugation and the supernatant was extracted by charcoal.The crude extract was analyzed by 19 F{ 1 H} NMR (500 MHz, D 2 O, 4000 scans) to detect fluorometabolites.