Acetyl-CoA-Mediated Post-Biosynthetic Modification of Desferrioxamine B Generates N- and N-O-Acetylated Isomers Controlled by a pH Switch.

Biosynthesis of the hydroxamic acid siderophore desferrioxamine D1 (DFOD1, 6), which is the N-acetylated analogue of desferrioxamine B (DFOB, 5), has been delineated. Enzyme-independent Ac-CoA-mediated N-acetylation of 5 produced 6, in addition to three constitutional isomers containing an N-O-acetyl group installed at either one of the three hydroxamic acid groups of 5. The formation of N-Ac-DFOB (DFOD1, 6) and the composite of N-O-acetylated isomers N-O-Ac-DFOB[001] (6a), N-O-Ac-DFOB[010] (6b), and N-O-Ac-DFOB[100] (6c) (defined as the N-O-Ac motif positioned within the terminal amine, internal, or N-acetylated region of 5, respectively), was pH-dependent, with 6a-6c dominant at pH < 8.5 and 6 dominant at pH > 8.5. The trend in the pH dependence was consistent with the pKa values of the NH3+ (pKa ∼ 10) and N-OH (pKa ∼ 8.5-9) groups in 5. The N- and N-O-acetyl motifs can be conceived as a post-biosynthetic modification (PBM) of a nonproteinaceous secondary metabolite, akin to a post-translational modification (PTM) of a protein. The pH-labile N-O-acetyl group could act as a reversible switch to modulate the properties and functions of secondary metabolites, including hydroxamic acid siderophores. An alternative (most likely minor) biosynthetic pathway for 6 showed that the nonribosomal peptide synthetase-independent siderophore synthetase DesD was competent in condensing N'-acetyl-N-succinyl-N-hydroxy-1,5-diaminopentane (N'-Ac-SHDP, 7) with the dimeric hydroxamic acid precursor (AHDP-SHDP, 4) native to 5 biosynthesis to generate 6. The strategy of diversifying protein structure and function using PTMs could be paralleled in secondary metabolites with the use of PBMs.

The chemical biology and coordination chemistry of putrebactin, avaroferrin, bisucaberin, and alcaligin.

Dihydroxamic acid macrocyclic siderophores comprise four members: putrebactin (putH2), avaroferrin (avaH2), bisucaberin (bisH2), and alcaligin (alcH2). This mini-review collates studies of the chemical biology and coordination chemistry of these macrocycles, with an emphasis on putH2. These Fe(III)-binding macrocycles are produced by selected bacteria to acquire insoluble Fe(III) from the local environment. The macrocycles are optimally pre-configured for Fe(III) binding, as established from the X-ray crystal structure of dinuclear [Fe2(alc)3] at neutral pH. The dimeric macrocycles are biosynthetic products of two endo-hydroxamic acid ligands flanked by one amine group and one carboxylic acid group, which are assembled from 1,4-diaminobutane and/or 1,5-diaminopentane as initial substrates. The biosynthesis of alcH2 includes an additional diamine C-hydroxylation step. Knowledge of putH2 biosynthesis supported the use of precursor-directed biosynthesis to generate unsaturated putH2 analogues by culturing Shewanella putrefaciens in medium supplemented with unsaturated diamine substrates. The X-ray crystal structures of putH2, avaH2 and alcH2 show differences in the relative orientations of the amide and hydroxamic acid functional groups that could prescribe differences in solvation and other biological properties. Functional differences have been borne out in biological studies. Although evolved for Fe(III) acquisition, solution coordination complexes have been characterised between putH2 and oxido-V(IV/V), Mo(VI), or Cr(V). Retrosynthetic analysis of 1:1 complexes of [Fe(put)]+, [Fe(ava)]+, and [Fe(bis)]+ that dominate at pH < 5 led to a forward metal-templated synthesis approach to generate the Fe(III)-loaded macrocycles, with apo-macrocycles furnished upon incubation with EDTA. This mini-review aims to capture the rich chemistry and chemical biology of these seemingly simple compounds.

Dimeric and trimeric homo- and heteroleptic hydroxamic acid macrocycles formed using mixed-ligand Fe(III)-based metal-templated synthesis.

Macrocyclic hydroxamic acids coordinate Fe(III) with high affinity as part of siderophore-mediated bacterial iron acquisition. Trimeric hydroxamic acid macrocycles, such as desferrioxamine E (DFOE), are prevalent in nature, with fewer dimeric macrocycles identified, including putrebactin (pbH2), avaroferrin (avH2), bisucaberin (bsH2) and alcaligin (alH2). This work used metal-templated synthesis (MTS) to pre-assemble complexes between one equivalent of Fe(III) and two equivalents of 4-((4-aminobutyl)(hydroxy)amino)-4-oxobutanoic acid (BBH) or 4-((5-aminopentyl)(hydroxy)amino)-4-oxobutanoic acid (PBH). Following peptide coupling, the respective Fe(III) complexes of pbH2 or bsH2 were formed, which analysed by LC-MS under acidic pH as [Fe(pb)]+ ([M]+, m/zobs 426.1) or [Fe(bs)]+ ([M]+, m/zobs 454.2). The mixed-ligand 1:1:1 Fe(III):BBH:PBH system furnished [Fe(pb)]+ and [Fe(bs)]+, together with chimeric [Fe(av)]+ ([M]+, m/zobs 440.2). The deviation from the expected 1:2:1 distribution of [Fe(pb)]+:[Fe(av)]+:[Fe(bs)]+ to 1:3.2:1.6 suggested the MTS-mediated formation of dimeric macrocycles could be influenced by steric effects in the pre-complex and/or cavity size, as governed by the monomer. 21-Membered avH2 defined the lower boundary of the optimal architecture. Mixed-ligand MTS between Fe(III):PBH-d4:ret-PBH at 1:1.5:1.5, where ret-PBH=3-(6-amino-N-hydroxyhexanamido)propanoic acid, gave four Fe(III)-loaded trimeric hydroxamic acid macrocycles in a distribution of 1.0:3.0:2.9:1.1 that closely matched the expected distribution 1:3:3:1 for a system without any kinetic and/or thermodynamic bias. Apo-macrocycles pbH2, avH2 and bsH2 were produced upon incubation with diethylenetriaminepentaacetic acid (DTPA) and co-eluted with a biosynthetic mixture of the native macrocycles. The work has demonstrated the utility of single- and mixed-ligand MTS for producing a variety of homo- and heteroleptic dimeric hydroxamic acid macrocycles as Fe(III) complexes and free ligands.