Advances in the Chemical Biology of Desferrioxamine B - American

Subscriber access provided by READING UNIV

Review

Advances in the Chemical Biology of Desferrioxamine B (DFOB) Rachel Codd, Tomas Richardson-Sanchez, Thomas J Telfer, and Michael P. Gotsbacher ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.7b00851 • Publication Date (Web): 28 Nov 2017 Downloaded from http://pubs.acs.org on November 29, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Chemical Biology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 55 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

Advances in the Chemical Biology of Desferrioxamine B (DFOB)

Rachel Codd*, Tomas Richardson-Sanchez, Thomas J. Telfer, Michael P. Gotsbacher

School of Medical Sciences (Pharmacology), The University of Sydney, New South Wales 2006, Australia

Corresponding author. Tel.: +61 2 9351 6738. E-mail address: [email protected] (R. Codd).

1 ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract: Desferrioxamine B (DFOB) was discovered in the late 1950s as a hydroxamic acid metabolite of the soil bacterium Streptomyces pilosus. The exquisite affinity of DFOB for Fe(III) identified its potential for removing excess iron from patients with transfusiondependent hemoglobin disorders. Many studies have used semi-synthetic chemistry to produce DFOB adducts with new properties and broad-ranging functions. More recent approaches in chemical biology have revealed some nuances of DFOB biosynthesis and discovered new DFOB-derived drugs and radiometal imaging agents. The current and potential applications of DFOB continue to inspire a rich body of chemical biology research focused on this bacterial metabolite.


Page 2 of 55

Page 3 of 55 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Desferrioxamine B (DFOB, 1) is a linear trihydroxamic acid siderophore (Figure 1) that was first characterized in the late 1950s as a metabolite of Streptomyces pilosus, a soil bacterium isolated in Rome, Italy.1 Many species of terrestrial and marine actinomycetes have since been shown to produce DFOB, often in concert with a suite of siderophore analogues.2-6 Siderophores are produced by bacteria to acquire Fe(III), which is available in aqueous solution under aerobic and pH neutral conditions at very low concentrations (~10–18 M) (Ksp Fe(OH)3 = 10–39 at pH 7).7,8 Siderophore biosynthesis is initiated when intracellular iron concentrations fall below approximately 1 µM, which is the threshold for sustaining growth of non-pathogenic and pathogenic bacteria.9-11 DFOB and other siderophores act as high affinity shuttles to sequester Fe(III) from an inorganic or biological source. The Fe(III)siderophore complex is returned to the parent cell following selective recognition by cell surface proteins, which triggers an importation cascade.9-16 DFOB has special status within the siderophore class, due to its long-standing clinical impact upon patients with transfusion-dependent blood disorders.17 These patients are administered DFOB on a close-to-daily basis, either alone or in combination with one or two synthetic agents, to remove excess iron that accumulates from the haemolysis of red blood cells. The three bidentate hydroxamic acid functional groups embedded along the DFOB backbone wrap around Fe(III) to form a stable 1:1 Fe(III):DFOB complex (logβ110 = 30.5),18,19 called ferrioxamine B (FOB) (Figure 1). DFOB harbors a number of characteristics that make it attractive for chemical biology research. Its capacity to coordinate Fe(III) and other metal ions11,18-28 (Table 1) confers potential upon DFOB-based agents for applications that include, but are not limited to: (i) diseases associated with metal ion dyshomeostasis, including iron overload disease and others; (ii) radiometal-based imaging; (iii) metal ion sensing and remediation; (iv) inhibiting metal ion-dependent processes, such as dysfunctional metalloproteins, or bacterial iron



Page 4 of 55

starvation as an antibiotic mechanism; and (v) the delivery of other antibiotics to the cell through exploiting native Fe(III)-siderophore-mediated uptake pathways. The development of such agents is made possible by the terminal amine group of DFOB, which is not involved in metal ion coordination, and allows small molecules, antibiotics, and other biological entities to be grafted onto DFOB using simple chemistry to produce derivatives with new properties. In addition, the modest cost of DFOB and its solubility in water and other organic solvents support these studies. These facets have underpinned a large body of research on DFOB and the chemistry and properties of more than 60 semi-synthetic derivatives. In the realm of biology, the availability of sequenced bacterial genomes has supported research into better understanding DFOB biosynthesis and evolution,6,29-32 with additional studies focused on its enzyme-mediated catabolism by bacterial symbionts.33-35

Table 1. Logarithmic Stability Constants for Complexes of DFOB Metal ion

logKa

Reference

Metal ion

logKa

Reference

Fe(III)

30.5

18,19

Yb(III)

16.0

18

Ga(III)

28.1

19

Ni(II)

10.9

18

Al(III)

24.1

19

Co(II)

10.3

18

In(III)

20.6

19

Zn(II)

10.1

18

La(III)

10.9

18

Cd(II)

7.9

18

Cu(II)

14.1

18

Our group has been inspired by these rich foundational studies of DFOB and has undertaken research to further contribute to knowledge of aspects of its biosynthesis, catabolism and the production of semi-synthetic derivatives. This defines the central scope of this review, which will open with the clinical history of DFOB, some structural studies, and 4 ACS Paragon Plus Environment

Page 5 of 55 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


then move to describe the chemical biology of three aspects of DFOB: its biosynthesis, its enzyme-mediated fragmentation, and its use in semi-synthesis (Figure 1). For each topic, descriptions of foundational studies are dispersed with more recent advances in the field.

Insert Figure 1 near here

Figure 1. Desferrioxamine B (DFOB, 1) as a scaffold for chemical biology, and its complexation with Fe(III) to form ferrioxamine B (FOB).

HISTORICAL ASPECTS The extraordinary Fe(III) chelating properties of DFOB36,37 were quickly recognised to have significant therapeutic potential, which accelerated its use for the treatment of secondary iron overload associated with β-thalassemia, and other conditions of chronic iron burden.38,39 Early studies undertook the gram-scale production of FOB from S. pilosus fermentation to evaluate its potential as an agent to deliver (rather than remove) iron to patients with iron-deficiency anaemia. This trial showed that FOB was rapidly excreted in urine with no evidence of iron released through dissociation, which was an early indicator of the magnitude of the Fe(III)-ligand affinity. This prompted the reverse idea of using DFOB as a free ligand to sequester excess Fe(III) from patients with accumulated iron, such as those with haemochromatosis and β-thalassaemia major.39,40 Before this idea could be tested, it was necessary to determine the protein targets of DFOB, since it would not be clinically useful if DFOB removed iron from haemoglobin, particularly for patients with iron-deficiency anaemia. The clinical progression of DFOB was ultimately assured, since it was found to remove iron only from the iron-storage proteins ferritin and haemosiderin, and not from ironporphyrin haemoglobin or the iron-transport protein transferrin.41 The first haemochromatosis



patient was treated with DFOB in 1961, with positive results leading to expanded clinical trials shortly thereafter for the treatment of iron overload associated with β-thalassemia.39,42 What is notable about the clinical progression of DFOB is the remarkably short timeline from the discovery of the producing organism S. pilosus (ETH 11686, ATCC 19797) in 1958,1 to the isolation and characterisation of FOB and DFOB in 1960,36,43 and its attaining status as the first-line therapy for β-thalassemia in the mid-1960s in the UK.17,39 Although DFOB is amenable to small-scale chemical synthesis using solution44,45 and solidphase chemistry46 approaches, the most efficient means of industrial-scale production remains bacterial fermentation. The mesylate salt of DFOB (Desferal®) appears on the current World Health Organization (WHO) Model List of Essential Medicines and for about 30 years was the only therapy available to patients with secondary iron overload disease.47,48 It remains in use as a single or combination therapy with two other synthetic agents, deferasirox and deferiprone, for iron overload disease, with iron chelation therapy supporting a significant improvement in life-span and reduced morbidities and mortality from iron-induced cardiomyopathy in these patients.17,49-55 DFOB is also used in cases of acute iron poisoning, which occurs most often in children who accidentally ingest iron supplement tablets.56 The clinical and biological importance of FOB prompted attempts to obtain structural information as a single molecule and in complex with protein binding partners, as detailed in the following section.

X-RAY CRYSTALLOGRAPHY OF FOB The cation of FOB has been co-crystallised with Mg(H2O)5(EtOH) and (ClO4)3 as counter ions, with the X-ray crystallographic data showing FOB crystals as a racemic mixture of the ∆-N-cis,cis and Λ-N-cis,cis coordination isomers (Figure 2),57 with the naming system for these isomers as previously described.58,59 The formation of FOB as a racemic mixture was


Page 6 of 55

Page 7 of 55 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


also evident from X-ray crystal structures in supramolecular complexes with crown ethers.60 Two X-ray crystal structures of FOB bound to high-affinity iron-siderophore transport proteins, namely the periplasmic protein FhuD from gram-negative Escherichia coli61 or the functionally equivalent lipoprotein FhuD2 from gram-positive Staphylococcus aureus62 have been solved at a resolution of 1.97 Å or 2.4 Å, respectively. The membrane-bound lipoproteins FhuD1 and FhuD2 bind FOB and other Fe(III)-hydroxamic acid siderophores with high affinity, with the KD value for FOB bound to FhuD1 or FhuD2 measured at 0.9 µM or 0.05 µM, respectively.63,64 As a member of the class III binding proteins, FhuD2 features two globular domains that define the FOB binding cavity separated by an alpha helical region. There are only small conformational changes in FhuD2 upon FOB binding, with several amino acid residues at the C-terminal domain (Arg199, Trp197) providing contacts to the complex.62 The extended methylene groups of the ligand interact with a cluster of aromatic amino acid residues, including Tyr110, Tyr130, Phe186, Tyr191, Tyr193, Trp197, Trp225, Trp279 and Tyr280. Each of these X-ray crystal structures show that the terminal amine group is oriented away from the protein, which supports the veracity of studies that use fluorophores or biotin as ancillary fragments appended to DFOB to establish FOB distribution and/or protein binding partners.


Figure 2. X-ray crystallographic data from the FOB cation isolated as a racemic mixture of the (a) ∆-N-cis,cis or (b) Λ-N-cis,cis coordination isomer;57 or as co-crystallized with: (c) FhuD from E. coli (PDB: 1K2V) (overlaid with structure in (a)), or (d) FhuD2 from S. aureus (PDB: 4FIL) (overlaid with structure in (b)). The protein bound structures of (c) or (d) are shown in (e)61 or (f),62 respectively.



In the FhuD structure, FOB is bound as the ∆-N-cis,cis coordination isomer (Figure 2c), while in the FhuD2 structure, it is bound as the Λ-N-cis,cis enantiomer (Figure 2d), suggesting an enantioselective effect is in operation during FOB-protein recognition. The FOB cation in FhuD (E. coli) and FhuD2 (S. aureus) is anchored in the protein cavity by a number of hydrogen bond interactions, with those from a proximal arginine residue of highest-rank importance. In the FOB-FhuD complex, this residue (Arg-84) is located more closely to the N-terminus, while in the FOB-FhuD2 complex, the cognate residue (Arg-199) is located closer to the C-terminus. An approximate alignment of FhuD and FhuD2 according to the respective N- and C-termini and FOB binding cavities shows a difference in the orientation of the key arginine residue, which extends from the ‘north’ in FhuD (Figure 2e) or from the ‘south’ in FhuD2 (Figure 2f). This differential orientation of the arginine residue involved in key hydrogen bonding interactions could be a factor underlying the enantioselective binding of the ∆-N-cis,cis- or Λ-N-cis,cis FOB coordination isomer to FhuD or FhuD2, respectively. The uptake of specific enantiomeric forms of metal-siderophore complexes has been reported in Mycelia sterilia,65 and in Pseudomonas aeruginosa and P. fluorescens.66 Other studies have informed understanding of siderophore-mediated Fe(III) uptake using synthetic retro-analogues of DFOB containing hydroxamic acid groups in a reverse orientation to the natural molecule.67-70

BIOSYNTHESIS OF DFOB Foundational Studies. Similar to the trimeric hydroxamic acid macrocycle desferrioxamine E, the biosynthesis of DFOB occurs by pathways independent of nonribosomal peptide synthetases.14,29-32,71,72 In the first committed step in the biosynthesis, L-lysine undergoes decarboxylation, as catalysed by pyridoxal 5ʹ-phosphate (PLP)-dependent lysine 8 ACS Paragon Plus Environment

Page 8 of 55

Page 9 of 55 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


decarboxylase (LDC),73-75 to form 1,5-diaminopentane (common name: cadaverine) (Figure 3a). Recent work has shown that the functional LDC (DesA) of S. coelicolor encoded by the desA gene is a homologue of L-2,4-diaminobutyric acid decarboxylase and is distinct from other types of LDCs.74-76 1,5-Diaminopentane next undergoes mono-N-hydroxylation, as catalysed by FAD-dependent monooxygenase (DesB), with the oxygen atom in the transfer reaction sourced from molecular oxygen.77 Acylation of N-hydroxy-1,5-diaminopentane with either acetic acid or succinic acid catalysed by acyl coenzyme A transferase (DesC), produces N-acetyl-N-hydroxy-1,5-diaminopentane (AHDP) or N-succinyl-N-hydroxy-1,5diaminopenane (SHDP), respectively. The final condensation reactions of the units of the endo-hydroxamic acid amino carboxylic acid monomers is catalysed by a NTP-dependent siderophore synthetase (DesD) of the IucA and/or IucC type to form the two amide bonds present in DFOB.31


Figure 3. Biosynthesis (a) or catabolism (b) of DFOB.

Sequence of Steps Catalysed by DesD. Recent work delineated the sequence of steps in the final stages of DFOB biosynthesis catalysed by DesD. Two possible pathways existed for the two DesD-catalysed condensation steps among the one AHDP monomer and the two SHDP monomers that comprise DFOB. One pathway entailed the initial conjugation of AHDP and SHDP to form an AHDP-SHDP heterodimer, with the subsequent conjugation of a second SHDP unit to give trimeric AHDP-SHDP-SHDP (DFOB). The alternative pathway was the initial formation of an SHDP-SHDP homodimer, with AHDP conjugated in the subsequent step to form the same trimeric product. To distinguish between these two pathways, the



native DFOB producer S. pilosus was cultured in medium supplemented with 1,4-diamino2(E)-butene (E-DBE) as a non-native substrate that could compete against native 1,5diaminopentane during DFOB assembly.


Figure 4. Left-hand side: Cartoon representation of unsaturated analogues of DFOB (1) produced in cultures of S. pilosus supplemented with E-DBE containing one E-DBE insert (3 isomers, 2–4), two E-DBE inserts (3 isomers, 5–7) or three E-DBE inserts (unique species, 8). Right-hand side: Experimental data (LC-MS traces) of DFOB analogues produced in S. pilosus culture grown in the absence (native system) (upper panel) or presence of E-DBE (lower panel).

The study established that the biomass of S. pilosus cultures in medium containing EDBE to 20 mM was similar to control cultures, and that total siderophore production was also comparable. Analysis of the distribution of DFOB-type siderophores showed that in addition to DFOB (Figure 4, 1), an additional seven siderophores were produced.78 This number arose from the biosynthesis of two sets of constitutional isomers that contained either one E-DBE unit or two E-DBE units. Three constitutional isomers were possible in the case of the exchange of one 1,5-diaminopentane unit for one E-DBE unit, with the non-native E-DBE inserted at the N-acetylated region, the internal region, or at the amine-containing region of DFOB. A binary nomenclature system was invoked to describe these isomers as uDFOA1[001] (2), uDFOA1[010] (3) or uDFOA1[100] (4) (Figure 4), where ‘u’ indicted unsaturation, ‘A1’ indicated the presence of a single unit of a four-methylene-containing diamine, and the digits in square brackets indicted either a native (0) or non-native (1)


Page 10 of 55

Page 11 of 55 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


substrate in the position from left to right correlating with the N-acetyl, internal, or amine region of DFOB. The second set of constitutional isomers arose from the exchange of two 1,5-diaminopentane units for two E-DBE units: uDFOA2[011] (5), uDFOA2[110] (6) or uDFOA2[101] (7). The seventh compound resulted from the complete exchange of 1,5diaminopentane for E-DBE: uDFOA3[111] (8). Precursor-directed biosynthesis was used in earlier work in cultures of Erwinia amylovora to generate constitutional isomers of desferrioxamine G1, a siderophore closely related to DFOB.79 The dominant constitutional isomer from the uDFOA1 series was uDFOA1[001] (2), which was resolved from the mixture of analogues as a unique species (Figure 4). The other two isomers from this series uDFOA1[010] (3) and uDFOA1[100] (4) co-eluted in the LC trace. Signature MS/MS fragmentation patterns of each species allowed the determination of the relative concentrations as uDFOA1[001] (2) >> uDFOA1[100] (4) > uDFOA1[010] (3). Dimer precursors were also detected in the supernatant, with dDFX[00-] >>> dDFX[-00] (where ‘d’ denotes dimer, and ‘-’ denotes a vacant position). These data were consistent with the notion that the DesD-catalysed biosynthesis involves: (i) the activation of an SHDP monomer; (ii) conjugation of the activated SHDP monomer to AHDP to form a AHDPSHDP heterodimer; (iii) the activation of a second SHDP monomer; and (iv) the conjugation of the second activated SHDP monomer to give the AHDP-SHDP-SHDP trimeric product, DFOB.78 A factor that may be driving this sequence of steps is the relative ease of activating a SHDP monomer, compared to a SHDP-SHDP homodimer. These types of precursordirected biosynthesis experiments provide new insight into the subtleties of DFOB biosynthesis and a pathway to access new analogues. The unsaturated diamine substrates harbor additional interest, since the unsaturated end products can be used for downstream semi-synthetic chemistry to increase structural diversity of DFOB. This is a research avenue being actively prosecuted in our laboratory.



ENZYME-MEDIATED CATABOLISM (REVERSE BIOSYNTHESIS) OF DFOB Our goal to contribute to a better understanding of DFOB biosynthesis was running in parallel with considering how DFOB might be broken down into smaller fragments. This anabolic-catabolic nexus of DFOB as a natural product was interesting to us in the context of a new approach in drug design termed ‘reverse biosynthesis’. A significant proportion of research efforts based around natural products-based drug discovery focuses upon elucidating the pharmacophore of the structurally complex parent and undertaking its chemical synthesis from smaller building blocks. Other approaches have used chemical or enzyme-mediated approaches to modify a region of a natural product in a site-selective fashion, followed by installing different chemical motifs using semi-synthetic chemistry.80 Reverse biosynthesis poses the notion that the entire scaffold of a structurally complex natural product could be broken down in a non-site-selective fashion to give a pool of smaller fragments, that by virtue of the unusual functional groups, topologies and chiral centres contained in the parent, could themselves be useful drug leads. This is an attractive notion, since the parent natural product, with a particular activity profile, could deliver a pool of new chemical entities with activities against different targets. The veracity of reverse biosynthesis would ideally be explored using a mild enzyme-mediated fragmentation of a natural product, since the lability of natural products could limit the use of chemical fragmentation methods. DFOB presented an ideal construct for this work, since there was knowledge of its enzyme-mediated catabolism by an amidase of Niveispirillum irakense,34 which has recently been reclassified from the original Azospirillum irakense.81 As a trihydroxamic acid, DFOB would liberate smaller hydroxamic acid fragments that could act as potential inhibitors of metallo-enzyme drug targets, such as the Zn(II)-containing histone deacetylases (HDACs) and Fe(III)-containing 5-lipoxygenase (5-LO), known to be inhibited by hydroxamic acid-based compounds.


Page 12 of 55

Page 13 of 55 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Previous work established the hydrolysis of DFOB by N. irakense,34 with this bacterium isolated from the rhizosphere of rice crops, which evidently co-existed with strains of DFOB-producing actinomycetes. In the absence of glucose, N. irakense could hydrolyse DFOB to provide succinic acid as an alternative source of carbon for growth. The amidase from a different DFOB-degrading organism, classified as a Rhizobium loti-like bacterium, was determined to be a serine protease-like amidase.82 These amidases have been found to be active only towards DFOB and are unable to hydrolyse FOB.83 The coding of TonBdependent receptors on the N. irakense genome suggests the bacterium could garner dual benefit from DFOB as an alternative source of carbon, and as a source of iron through FOB xenosiderophore uptake pathways. Our laboratory corroborated the majority of the results from the original work, with only subtle differences in one aspect of the catabolite profile, which may have been due to differences in experimental conditions. The N. irakense-mediated hydrolysis of DFOB (Figure 3b) can occur at the amide bond proximal to the N-acetyl group, to liberate firstround catabolites AHDP and the homodimer SHDP-SHDP, with the latter subject to secondround hydrolysis to give two SHDP units. Third-round hydrolysis of each SHDP unit would furnish one equivalent of succinic acid (SUC) as a carbon source, and one equivalent of 5(hydroxyamino)pentanamine (HAPA), with this cadaverine-type catabolite a potential nitrogen source for the bacterium. First-round hydrolysis of the amide bond proximal to the terminal amine group of DFOB gives the heterodimer AHDP-SHDP and one equivalent of SHDP. In our studies, AHDP-SHDP was hydrolysed at the hydroxamic acid carbonyl region to give dimeric AHDP-SUC and HAPA, with the former yielding AHDP and SUC upon third-round hydrolysis. The hydrolysis of AHDP-SHDP at the amide bond was also likely taking place to give monomers AHDP and SHDP. These catabolites were detected using liquid chromatography-mass spectrometry (LC-MS) from sub-samples of the supernatant of



N. irakense cultures that were deplete of glucose and supplemented with DFOB.84 The catabolism of DFOB followed a sigmoidal decay, with its consumption complete at about 5472 h (Figure 5a, inset). The sigmoidal decay of DFOB was well correlated with the sigmoidal growth of AHDP, which reached a maximum concentration at about 100 h. AHDP was a terminal catabolite and was stable in the mixture when re-analysed one week after DFOB supplementation. The other catabolites showed Gaussian behaviour, with different rates of formation and degradation. The catabolite mixture represented a combinatorial pool of potential drug leads with activity profiles predicted to depend upon the time of harvest. As determined by LC-MS with total ion current (TIC) and extracted ion chromatogram (EIC) detection modes, the catabolite profile from the harvest at 48 h (Figure 5a, b), was more complex than the harvest at 168 h (Figure 5c, d). Further, since many catabolites contained an amine group (Figure 3b), there arose the possibility of conducting semi-synthetic chemistry to diversify the library of compounds. The activity of the catabolite pool harvested at t = 48 h and t = 168 h was tested against HDAC (Figure 5e, f) and 5-LO (Figure 5g, h). From these systems, the t = 168 h pool showed a concentration-dependent activity against 5-LO (Figure 5h), which identified AHDP as a likely lead, since it was present in this mixture as the major and terminal catabolite. A combinatorial chemistry approach was undertaken with the more complex catabolite pool at t = 48 h, which was reacted with 1,8-naphthalic anhydride (NA) in a microwave synthesizer to produce a library of scriptaid-like compounds. Scriptaid is a known hydroxamic acid HDAC inhibitor (Figure 5).85-87 The isolated compounds, DFOB, AHDP and AHDP-NA, were screened against HDAC (Figure 5i-k) or 5-LO (Figure 5l-m). The AHDP-NA construct N-(5(1,3-dioxo-1H-benzo[de]isoquinolin-2(3H)-yl)pentyl)-N-hydroxyacetamide was identified as an inhibitor of 5-LO (IC50 = 60 µM), which was 28 times more selective towards 5-LO than HDAC. This study demonstrated that DFOB, which has no activity against HDAC or 5-LO


Page 14 of 55

Page 15 of 55 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(Figure 5i, l), could be subject to enzyme-mediated catabolism to generate catabolites that themselves could act as new inhibitors (Figure 5h) or could be converted using combinatorial chemistry to more potent leads (Figure 5n).


Figure 5. N. irakense-mediated catabolism of DFOB, as monitored using LC-MS (a-d), and the activity (n = 3) of the catabolite pool (CP) harvested at t = 48 h or 168 h against HDAC (e, f) or 5-LO (g, h). DFOB and its catabolites were detected by EIC as a function of time (a, inset). Subsamples of the DFOB-supplemented culture of N. irakense were analysed using TIC detection at t = 48 h (a) or t = 168 h (c), with the corresponding profiles of individual catabolites (EIC detection) shown in (b) or (d), respectively. Note: the right-hand axis in (d) is 10× left-hand axis. The activity (n = 3) of DFOB, the terminal catabolite AHDP and a semi-synthetic derivative AHDP-NA were measured against HDAC1 (i-k) or 5-LO (l-n). Significance: *p

Advances in the Chemical Biology of Desferrioxamine B - American

Recommend Documents