Lecture Notes - University of Crete 1 of 47 Lecture Notes: Siderophores and Iron Metabolism - Structures, Functions, Role in Infection and Potential as a Novel Class of Antibiotics

Page 1 of 47

Lecture Notes:

Siderophores and Iron Metabolism -

Structures, Functions, Role in Infection and Potential as

a Novel Class of Antibiotics

Prof. Dr. Berthold F. Matzanke

Isotopes Laboratory TNF

University of Lübeck

Ratzeburger Allee 160

D-23538 Lübeck

Germany

Tel. +49-(0)451-5004140

Fax: +49-(0)451-5004139

e-mail: [email protected]

Page 2 of 47

1.Introduction 3 2. Structures of Siderophores 2.1. Hydroxamates 2.2. Catecholates 2.3 Carboxylates 2.4. Mixed ligands

4 5 7 8 9

3. Coordination Geometry of Siderophores 3.1Geometrical and optical isomers 3.2UV/Vis and CD spectra 3.3Band assignment of chromic siderophores 3.4 Ligand denticity

10 11 12 13 13

4. Equilibrium Thermodynamics of Siderophore Iron Binding 4.1 Ligand protonation and complex formation constants 4.2 Iron exchange kinetics 4.3 Redox chemistry of siderophores

14 14 17 17

5. Siderophore Transport in Microorganisms 5.1. Siderophore receptors 5.2 Specificity and stereospecificity of siderophore recognition 5.3 Transport mechanisms 5.4Reductive removal from iron siderophores in vivo 5.4a. Bacterial iron metabolism analyzed by in situ Mössbauer spectroscopy 5.5 Siderophore uptake regulation 5.6 Siderophore secretion

19 20 22 24 26 27 29 32

6. Siderophores in Medicine 6.1 Iron overload diseases, ß-thalassemia 6.2 Siderophores and infection 6.3 Siderophore-based antibiotics 6.4 Siderophores and MRI 6.5 Iron chelators and cancer

32 32 32 33 37 37

7. Additional functions of siderophores 7.1 Siderophores as iron storage compounds 7.2 Other applications of siderophores

38 38 38

8. References 39

Abbreviations AL= alcaligin; BC= bisucaberin; DHBS = dihydroxybenzoylserine; DMB = N,N-dimethyl-2,3-dihydroxybenzamide; Feent = ferric enterobactin; LICAMS = N,N′,N″-tris(2,3-dihydroxy-5-sulfobenzoyl)-1,5,10-triazadecane; MECAM = N,N′,N″-tris(2,3-dihydroxy-5-benzoyl)-1,3,5-tris(aminomethyl)benzene; MECAMS = N,N′,N″-tris(2,3-dihydroxy-5-sulfobenzoyl)-1,3,5-tris(aminomethyl)benzene; men = N-methyl-l-menthoxyacet-hydroxamato; OMR = outer membrane receptor; RA = rhodotorulic acid; Sid = siderophore; TREN = tris(2-aminomethyl)amine; TRIMCAM = N,N′,N″-tris(2,3-dihydroxy-5-benzoyl)-1,3,5-tricarbamoylbenzene; TRIMCAMS = N,N′,N″-tris(2,3-dihydroxy-5-sulfobenzoyl)-1,3,5-tricarbamoylbenzene.

Page 3 of 47

1 Introduction

For all plants and animals, and for virtually all microbes, with the exception of some Lactobacilli and a Borrelia species1, life without iron is impossible. A multitude of essential enzymes bind iron in their active centers. Therefore, up to 105 Fe-ions are typically required in key metabolic processes of a single bacterial cell. Why iron has gained such an eminent role in the course of biological evolution remains open to speculation.11 Though iron is the fourth most abundant element in the Earth's crust, it is present under aerobic conditions at nearly neutral pH in the form of extremely insoluble minerals like hematite, goethite, and pyrite or as polymeric oxidehydrates, carbonates, and silicates which severely restrict the bioavailability of this metal. In response to this, microorganisms secrete high-affinity iron-binding compounds called siderophores (Greek: σιδηρος=iron, φορέας=carrier).3-10 In any natural environment of microbial activity, siderophores are present. Significant amounts of siderophores can be extracted from soil and from fresh or salt water. Even some foods contain siderophores.8 Biosynthesis of siderophores is executed on a cellular level by a set of enzymes specific for the respective siderophore. The corresponding genes are located on the chromosome or on a plasmid.12 Expression of these genes is controlled by the amount of cellularly available iron12-15. Therefore, microbial metabolic products (mainly secondary metabolites) can be classified as siderophores, if

(i) they exhibit iron chelating capability, (ii) they participate in active transport across the cell membrane(s) and (iii) their biosynthesis is regulated by the intracellular iron level.

During the last decade we witnessed an exceptional progress in the field of iron-transport research3-6. More than 500 naturally occurring siderophores have been isolated and characterized (Section 2)3,4, and the discovery of new siderophores is continuing at a good rate. The structural features of siderophores are diverse. The ligating groups contain oxygen atoms of hydroxamate, catecholate, α-hydroxy carboxylic and salicylic acids, or oxazoline and thiazoline nitrogen. Siderophores display a selectivity for iron which is taken into account in the corresponding complex stability constants that are higher with Fe3+ than with Al3+, and with bivalent cations like Ca2+, Cu2+ or Zn2+. The physical properties of siderophores yield important information on biological mechanisms involving siderophore iron complexes. Fundamental data on solution thermodynamics (Section 4.1), electrochemistry (Section 4.3), and kinetic studies (Section 4.2) have been used in the search for likely intracellular iron release mechanisms (Section 5). Moreover, sufficient thermodynamic and kinetic data facilitate an estimation of the advantages of certain siderophores over others in their competition for iron. Siderophore uptake in microorganisms is, in general, a receptor-dependent process (Section 5.1). A diversity of high-affinity receptors has evolved either to recover iron loaded endogenous siderophores that have been excreted to scavenge iron from the environment, or to utilize xenosiderophores or iron sequestering agents of a host. Crystal structures have been solved of a variety of siderophore receptors and of other components of the high-affinity, energy-dependent siderophore-transport systems as well (Section 5.1). The permeation of cell walls or bacterial membranes by siderophores is in most microbes a highly specific process requiring an array of up to eight proteins. The advent of modern molecular biology delivered a cornucopia of methods enabling high-yield production of specific gene products relevant to siderophore-synthesis and -transport, analyses of structure-function relationships (employing site directed mutagenesis), and detailed insights into the regulation of the corresponding processes. Nevertheless, siderophores may exhibit both optical and geometrical isomers (Section 3). Synthesis of Cr3+ or Rh3+ siderophores enables isolation and characterization of these isomers (Section 3.3). Stereochemically well-characterized isomers are an indispensable prerequisite for studying the specificity of siderophore-mediated iron uptake in microorganisms (Section 5.2). A variety of cellular iron-release mechanisms

Page 4 of 47

from siderophores have been characterized, the majority of which involve reduction of ferric iron either on the membrane level or in the cytoplasm(Section 5.4). Reduction potentials of ferric siderophore complexes vary between -700-and -150mV . In particular at the low potential end below -450mV special biological strategies of reductive iron removal are required because these potentials are too negative for typical cellular reductases. The molecular nature of the overall transport process is complex (Section 5.3) and strictly regulated (section 5.5).

In vivo Mössbauer spectroscopic investigations on the time course of siderophore mediated iron assimilation revealed, that

(i) iron removal from siderophores is a fast and in most cases a reductive process (ii) few main metabolites of iron metabolism are found (iii) a ferrous iron complex, ferrochelatin, is present in all microorganisms analyzed (iv) siderophores function as iron storage compounds in various fungi (Section 6.1).

One siderophore, ferrioxamine B, serves as a detoxifier in iron overload diseases and in the treat-ment of ß-thalassemia. This drug is also employed clinically for removal of aluminum from the body (Section 6.2) Siderophores may serve as MRI imaging agents and as templates for novel classes of antibiotics. Invading microorganisms exposed to circulating blood produce siderophores to compete for iron with the human transport protein transferrin, thus constituting one aspect of virulence and pathogenicity (Section 6.3). Siderophores and siderophore analogs play also a role as basic models for actinide chelators in order to remove these metals from the environment or from a contaminated body . 2 Structures of Siderophores

As with secondary metabolites, various and complex chemical structures are typical within the siderophores, preventing their unequivocal and universal classification. Since the biosyntheses and structural features of siderophores are diverse, a classification scheme will be to some extent arbitrary. Criteria may include the producing organisms (bacteria, fungi, plants), the nature of the backbone (peptidic or non-peptidic, cyclic or open chain), or the nature of the chelating group. Despite the considerable structural variation found in the siderophores, their common feature is to form six-coordinate complexes with iron (III) of great thermodynamic stability. The ligating groups contain the oxygen atoms of hydroxamate, catecholate, α-hydroxy-carboxylic acids, and α-keto-carboxylic acids.

Page 5 of 47

In addition, siderophores with diverse FeIII ion binding groups were isolated, like salicylic acid, oxazoline and thiazoline nitrogen, and even negatively charged nitrogen (in the case of maduraferrin). Comprehensive reviews on the structural variety of siderophores are available3, 7-9, 16 Therefore we shall focus on salient structural features of siderophores. NMR studies of most siderophores have been part of the general chemical and structural characterization. The metal is generally removed from the complex prior to NMR spectroscopy, because FeIII causes severe line broadening of the NMR signals. To obtain spectra of the siderophore in the metal-chelated form, usually the diamagnetic Ga3+ or the Al3+ siderophore analogs are employed.16-19 1H and 13C NMR data of a variety of hydroxamate-type siderophores have been tabulated.20

Investigations on the biosynthesis of siderophores represent a major activity in the field.12 Considerable effort has been devoted to the chemical synthesis of natural siderophores, enantiomeric siderophores, and completely synthetic siderophore analogs. These topics will not be covered here and the reader is referred to the corresponding literature.8,10,21-24 Both, the pathways of siderophore biosyntheses as well as their chemical partial or total synthesis are central building blocks for the design of novel classes of antibiotics.

2.1 Hydroxamate-Type Siderophores

Hydroxamate group-bearing siderophores are mainly synthesized by fungi and Gram-positive filament-forming bacteria (streptomycetes). In fungal systems the hydroxamic acid chelating group is commonly derived from acylated Nδ-acyl-Nδ-hydroxy-L-ornithine.

The ferrichromes comprise one large family of hydroxamate siderophores (Figure 1a) and were isolated from low-iron cultures of many fungi.25 With few exceptions, ferrichromes possess cyclic hexapeptide backbones in which one tripeptide is linked to a second tripeptide of Nδ-acyl-Nδ-hydroxy-L-ornithine. Linear derivatives of the latter tripeptide form backbones of various antibiotics, termed albomycines. Crystal structures have been determined of several ferrichromes.26, 27 In all structures the iron coordination site is on one side of the molecule, the coordination of the metal is Λ-cis, and the conformation of the amino acids is L. A β(II) bend and a β(I) bend of the cyclic peptide skeleton is found.

Ferrioxamines, typical constituents of culture broths of Actinomycetes, occur as both linear and cyclic compounds containing 1-amino-5-hydroxyaminopentane (N-hydroxycadaverine) and succinic

C

NH

NH

CNH C

NH

O

O

C

NH

C NHC

R2R1

N CR3

OO

N C

OO

R3

N

CR3

O

OFe

O

O

OO

Fig. 1a Ferrichromes ferrichrome: R1=R2=H, R3=CH3 Ferricrocin: R1= R2=CH2=HH, R3=CH3

Page 6 of 47

acid as building blocks (Figure 1c). A cyclic trimer of succinyl-(N-hydroxycadaverine), is named ferrioxamine E. In some cases the pentane moiety is replaced by a butane carbon skeleton (putrescine). The most prominent representative of this siderophore family, desferrioxamine B (Figure1) , has become the drug of choice for the treatment of transfusional iron overload (Section 6.2).29 The crystal structure of ferric ferrioxamine B has been published recently.30 Certain derivatives of the ferrioxamines display antibiotic activity and therefore have been designated as ferrimycins.31

A B C

Fig 1B: (A) fusarinines (n = 1, fusarinine; n = 3, R = acetyl, cyclic, triacetylfusarinine); (B) coprogens (R1 = H, R2 = COMe, R3 = R4 = isopentenol, coprogen; R1 = H, R2 = COMe, R3 = R4 = Me, neocoprogen II); (c) ferrioxamines (X = NH2, m = n = 5, R = Me, ferrioxamine B; X = NH (cyclic), m = n = 5, R = (CH2)2CO-[X], ferrioxamine E);;

Typical tetradentate representatives of the hydroxamate-siderophore family include rhodotorulic

acid,33 dimerum acid,20 bisucaberin,34 alcaligin35 and putrebactin36. Bisucaberin (Figure 1d), a cyclic dimer of succinyl-(N-hydroxycadaverine), which sensitizes tumor cells to macrophage-mediated cytolysis, was isolated from the supernatant of cultures of the marine bacterium Alteromonas haloplanktis.34. Putrebactin is a cyclic dimer of succinyl-(N-hydroxyputrescine) whereas rhodotorulic acid is a linear tetradentate chelator assembled of two Nδ-(res)-Nδ-hydroxyornithine units which are cyclized to form a diketopiparazine ring.33 The crystal structure of the Fe-alcaligin complex disclosed a monobridged topology and a Fe2L3 stoichiometry at near neutral pH.37

Page 7 of 47

Bisucaberin

2.2 Siderophores with Catecholate Ligands

Under conditions of iron deficiency, many bacteria excrete siderophores exhibiting phenolate or 2,3-dihydroxybenzoate (DHB) iron binding groups. In 1970, enterobactin (also known as enterochelin), the first tricatechol siderophore, was isolated from culture fluids of E. coli, Aerobacter aerogenes, and Salmonella typhimurium (Figure 2a).38 Enterobactin is the cyclic triester of DHBS (2,3-dihydroxybenzoylserine) exhibiting extraordinary features: (i) an extremely high complex formation constant, (ii) a redox potential too low for physiological reductants, (iii) a strong pH-dependence of (i) and (ii), and (iv) a trilactone backbone the cleavage of which lowers the redox potential. Due to these exceptional properties relevant to its physiological reactions, enterobactin is one of the most intensively analyzed siderophores. Recently, enterobactin was also isolated from Gram-positive bacteria39 as well as the enterobactin homolog corynebactin from Corynebacterium glutamicum and Bacillus subtilis40,41 All other known tris-catecholate siderophores exhibit a linear backbone based on spermidine or norspermidine. Parabactin, N4-(2,3-dihydroxybenzene-3-methyloxazoline-2-carboxamidyl)-N1,N8-bis(2,3-dihydroxybenzoyl) spermidine, is produced by Paracoccus denitrificans (Figure 2b). Exposure to acid destroys the oxazoline ring, producing a threonyl moiety.42 Other members of the linear tris-bidentate catecholate-type siderophores include vibriobactin,7 vulnibactin,7 fluvibactin,7 protochelin7, and salmochelin-243. Salmochelin is unique, because the backbone of the DHBS-chelating units is made up by glucose( see chapter pathogenicity).

A B

NH

NH

O

O

N

N OH

OH

O

RO

R

OH

CH3

A:

Fig. 35. (a) Rhodotorulic acid: R = CH3

(b) Dimerum acid: R = A

Page 8 of 47

C D

Figure 2 Structures of representative catecholate and mixed-ligand siderophores: (a) enterobactin; (b) parabactin; (c) mycobactins (R1 = various alkyl chains, R2 = R3 = R5 = Me or H, R4 = alkyl chains or H;3 (d) fluorescent chromophore of pseudobactins and pyoverdins

2.3 Carboxylate-Type Siderophores

A great variety of siderophores exhibit carboxylate and hydroxy donor groups. Many of these siderophores belong to the mixed ligand group. In the nineties of the last century, however, a completely novel class of siderophores was detected whose members neither possess hydroxamate nor phenolate ligands. Rather, FeIII iron binding is achieved exclusively by α-hydroxycarboxylates and carboxylates. The late discovery of these compounds is mainly based on the fact that they are colorless, therefore requiring novel methods for screening and isolation (chromazurole test and iron nutrition bioassays).44 These siderophores are found in the kingdom of bacteria as well as in the realm of fungi. A very hydrophilic complex, termed staphyloferrin A, was isolated from Staphylococcus hyicus (Figure 3b).45 The molecule is composed of two moles of citrate linked by ornithine. Besides Staphyloferrin A,45 also staphyloferrin B,44 vibrioferrin,45 and rhizoferrin48

contain citric acid building blocks. Rhizoferrin has been isolated from culture filtrates of Rhizopus and other members of the class of Zygomycetes (Figure 3c).48 In this compound, two citric acid residues are linked to diaminobutane, resulting in N1,N4-bis(1-oxo-3-hydroxy-3,4-dicarboxybutyl)diaminobutane. Although citrate is an intracellular primary metabolite, it may be regarded - from an evolutionary view - as the simplest siderophore from which the above mentioned siderophores, as well as the mixed ligand compounds schizokinen, arthrobactin, and aerobactin (section 2.4) have evolved.

Rhizobium meliloti, capable of fixing atmospheric nitrogen when symbiotically associated with certain legumes, excretes and utilizes rhizobactin DM4 (Figure 3a), N2-[2-{(1-carboxymethyl)amino}ethyl]-N6-(3-carboxy-3-hydroxy-1-oxopropyl)lysine.50 Phytosiderophores, detected in root washings of gramineous plants, represent a separate subclass of the carboxylate-type of siderophores. For eaxample, the phytosiderophores nicotianamin, mugineic, avenic and distichonic acid are produced and utilized by cereals, like barley, wheat, rye oat etc.51,52

Page 9 of 47

Figure 3 Structures of carboxylate siderophores: (a) rhizobactin; (b) staphyloferrin; (c) rhizoferrin

2.4.Siderophores with mixed ligands and heterocyclic chelating groups

Fiftyfive years ago the very first siderophore, mycobactin, was isolated by the crystallization of the aluminum complex.53 Mycobactins from Gram-positive Mycobacteria and the closely related nocobactins from Nocardia embody a series of lipid-soluble siderophores located in the lipid-rich boundary layers of these bacteria (Figure 2c).54 The X-ray structure revealed that iron binding in mycobactins is accomplished by two hydroxamates, a phenolate group, and oxazoline nitrogen. It is generally assumed that in oxazoline and thiazoline containing donor-deficient siderophores, the imine-N will participate in ferric ion complexation. These heterocycles result physiologically from an enzymatic cyclization of cysteinyl, seryl, or threonyl side chains. The imine-type of Fe-coordination is a common feature of the siderophores pyochelin, yersiniabactin, anguibactin, and acinetobactin (Figure 4).3,7 A variety of fluorescent chromopeptide siderophores, termed pseudobactins and pyoverdins, are synthesized by Pseudomonas species.57 The chromophores, derived from 2,3-diamino-6,7-dihydroxyquinoline (Figure 2d), are linked to a peptide chain exhibiting either two hydroxamate groups or one hydroxamate and one α-hydroxycarboxylate group.

A B

C

OHOHN

O CH3

N

O

NH

N

OH

OH

N

S

S

N

HO

CH3

CH3

N

S

CH3

COO

H2

+

_

OH

OH

S N

NO

NHN

HO

Page 10 of 47

Fig. 4 Structures of siderophores with heterocyclic ligands: acinetobactin (a), anguibactin (b), yersiniabactin (c) and maduraferrin (Note: iron binding nitrogen of hexahydropyridazine-3corboxylic acid is negatively charged)

3 Coordination Geometry of Siderophores

The stereochemistry of siderophores is a very important aspect of their role in receptor mediated iron uptake, since it has been shown that very subtle discrimination by microbial iron transport systems takes place between siderophore isomers. In fact, uptake of siderophores by microorganisms shows – at least in part - stereospecific preferences. The geometry at the metal center of a trihydoxamate complex is shown in Fig. 5a Figure 5a The geometry and dimensions of the iron-coordination octahedron in a natural tris(hydroxamate) complex, triacetylfusarinine. (ACS from M. B. Hossain et al.66)

N NH

O

O O

HN

N

O HCH2OHN

NOO

CH3

N

H

H Fe

O

H

O

Figure 27. Maduraferrin

Page 11 of 47

3.1 Geometrical and Optical Isomers

Upon metal chelation, a tris complex with a bidentate ligand forms a more or less distorted coordination octahedron via three five-membered chelate rings. According to IUPAC rules,63 the planes of these five-membered rings may form a right-handed (∆) propellor or a left-handed (Λ) propellor (Figure 5). If the ligand is not optically active, one will find a racemic mixture of ∆ and Λ optical isomers. If the ligand is chiral (e.g. D), two complexes are again possible, DΛ and D∆. However, they are not formed in equal amounts because they have different standard free energies of formation. Exclusively, one optical isomer may be formed if there is a large thermodynamic advantage for this isomer (as, for instance, in the case of ferrichromes where LΛ is exclusively observed).

Figure 5 : Λ and ∆ optical configuration at the metal center (top) and eight geometrical isomers of a FeIII∆-trichelate complex, involving three unsymmetrical bidentate ligands attached to an asymmetric backbone. Not shown is the set of eight Λ diastereomers. R, X correspond to unique functional groups,e.g. amino terminus of ferrioxamine B(R) or diketopiperazine ring of coprogen (X)26

Many siderophores are hexadentate ligands with three asymmetrical bidentate functional units at-tached to an asymmetrical backbone. These comp-lexes reveal a rather complicated stereochemistry. The bidentates are not equivalent and there will be 23 = 8 cis and trans isomers. Each can have a Λ and ∆ metal environment and when the siderophore is made from chiral residues there are 16 diastereo-mers, none of which is enantiomeric to one another. A selection of such diastereomers is shown in Figure 5. Nomenclature for the geometrical isomers of hydroxamate tris-bidentates was outlined by Leong and Raymond.10, 64 No general rule can be applied for an absolute assignment of the chelate ring sequence. However, if the structure exhibits a unique functional group, this group can be utilized to define the chelate ring sequence. In the case of ferrioxamine B or D1, the unique functional group is the N-terminus (a). In the case of the coprogens, this group is a diketopiperazine ring placed between rings 1 and 2

Page 12 of 47

. (b) Looking down the C3 axis the sequence of the chelate rings 1, 2, and 3 corresponds to the rotation direction, i.e. clockwise for Λ-isomers and counter-clockwise for ∆-isomers. (c) If the ring 1 has the carbon atom of the hydroxamate group below the nitrogen, it is denoted ‘C’; if the reverse is true, it is called ‘N’. (d) For rings 2 and 3, each is called cis or trans depending upon whether it has the same or opposite relative orientation with respect to the coordination axis as does ring 1.

The X-ray crystallographic analysis of all ferrichrome siderophores yielded Λ-C-cis,cis configurations of the coordination octahedron.65 Neocoprogen I adopts a ∆-C-trans,trans configuration in the crystal structure.64 N,N′,N″-Triacetylfusarinine crystallizes as either the ∆ or Λ isomer, depending on the solvent system used.66 Achiral siderophores such as ferrioxamine E and ferrioxamine D1 crystallize as racemic mixtures of ∆ and Λ isomers.67,68 Similarily, the crystal structure of Ferrioxamine B displays a racemic mixture of Λ-N-cis,cis and ∆-N-cis,cis,30 although various geometrical isomers of the inert Cr-complex are present in solution (see section 3.3) .

A multitude of siderophores achieve iron chelation via catecholates. Unlike hydroxamate, catecholate is a symmetric, bidentate ligand. Thus there are no geometrical isomers of simple tris(catecholate) metal complexes. The chirality at the metal center of (bis(catecholato)-siderophores) was determined in amonobactins, where a slight ∆ preference in Fe2L3 stoichiometry was observed whereas the corresponding FeL(H2O)2–complexes in the low-pH region are achiral.69 Enterobactin displays a ∆-cis configuration, dictated by the asymmetric centers (L) of the trilactone ring.70 Synthetic enterobactin analogs composed of tris(2-aminomethyl)amine (TREN) as anchors and amino acids linking the anchor to catechol units also form Fe3+ complexes of preferential ∆-cis configuration when L-amino acids are used. These complexes are stabilized by intramolecular H-bonds from the catechoylamides CONH to the catecholate oxygens, as in enterobactin. When the possibility of this type of H-bond is eliminated by replacing the amide proton CONH by an amide N-methyl (CONMe), the chiral preference of the complex is inverted from the ∆-cis to the Λ-cis configuration.71 Surprisingly, the recently detected trilactone siderophore corynebactin forms a Λ-ferric complex. The addition of a glycin spacer and the methylation of the trilactone ring as compared with enterobactin are sufficient to favor the opposite chirality.72, 73 3.2 UV/Vis and CD Spectra of Ferric Siderophores

The d5 electronic configuration of Fe3+ rules out any crystal field stabilization energy (CFSE) and makes the complexes relatively labile with respect to isomerization and ligand exchange in aqueous solution. Furthermore, high-spin Fe3+ has no spin-allowed d–d transitions. Therefore the UV/Vis and CD spectra of iron(III) microbial iron chelates arise from the coulombic interaction between the positively charged metal ion and the negatively charged oxygen atoms (charge-transfer transition), which is not as readily interpreted as are ligand-field (d–d) transitions.

Although transitions in high-spin iron(III) complexes are extremely weak, the determination of the metal center chirality of Fe3+ complexes in solution is possible through comparison of the solution and solid-state CD spectra. However, the correlation of the rotary power with left-handed or right-handed helical stereochemistry requires an absolute assignment based on crystal structure data which is usually determined employing the Bijvoet method for anomalous dispersion of Cu-Kα radiation by the Fe3+ ion in the crystalline solids. This correlation could be established for triacetylfusarinine and neurosporin.74, 66 This correspondence extends into the other Fe3+ complexes and shows that the CD spectra of the iron(III) complexes in solution can be used for determinations of the metal center chirality. Fe3+ complexes will have the Λ configuration (at least predominantly) if the CD band in the region of the absorption maximum (400–500 nm for hydroxamates) has a positive

Page 13 of 47

sign. The visible and CD spectral parameters of siderophores in aqueous solution are summarized in Table 1. Table 1 Visible and CD spectral characteristics of aqueous siderophores and of a model complex Siderophore λmax (nm)

(ε [M−1 cm−1]) λmax (nm) ∆ (∆ε [M−1 cm−1])

λmax (nm) Λ (∆ε [M−1 cm−1])

Ferrichrome 425 (2895) 360 (−3.7) 465 (2.4) Ferricrocin 434 (2460) 290 (−3.78) 360 (−1.62) 450 (+2.47) Ferrioxamine B 428 (2800) Coprogen 434 (2820) 375 (+2.1) 474 (−1.26) N,N′,N″-Triacetylfusarinine

370 (+3.25)

467 (−2.04) Fe2RA3 425 (2700)b 372 (+2.73) 464 (−1.41) Fe(benz)3

a 435 (4910) 350 (+2.3) 350 (−2.8) 452 (−1.5) 455 (+1.1) Enterobactin 495 (5600)b 553(−2.2)∆-cis Corynebactin 545(+1.7) Agrobactin 505 (4100) Λ-cis Parabactin 512 (3300) Λ-cis Pseudobactin 400 (15 000) 400 (+2.0) 436 (−0.8) 502 (+0.3) Pseudobactin A 400 (2000) Neurosporin 360 (−4.8) 465 (+4.5) a In acetone solution. b At pH 7.

3.3 UV/Vis and CD Band Assignments of Chromic Siderophores

An alternative method that enables the classification of geometrical and optical isomers is accomplished by substituting the Fe3+ ion with kinetically more inert d3 Cr3+ or d6 Rh3+ ions.75 Since the substituted ions have almost the same ionic radii and the same charge as Fe3+, their complexes show a high degree of structural similarity with the corresponding Fe3+ complexes, as demonstrated by the crystal structures of model compounds.76, 77 The d-electron configurations of Cr3+ and Rh3+ grant significant CFSE for kinetic inertness and provide well characterized d–d transitions with distinct UV/Vis and CD spectra. For details, please refer to the literature

3.4 Ligand denticity and metal center binding mode

The structure of the metal center is also depending on the denticity of the ligand. Tetradentate ligands are of particular interest because they are not able to achieve the octahedral coordination geometry in a 1:1 stoichiometry. The dihydroxamate siderophores rhodotorulic acid (RA),

Page 14 of 47

bisucaberin (BC) and alcaligin (AL) were analyzed in some detail. 88,89 The binding mode of these complexes is pH-dependent. At low pH(<5) a cationic species of the type [FeL(H2O)2]+ was observed. In order to completely encapsulate the metal a binuclear complex of Fe2L3 stoichiometry is required which was shown to be formed at near neutral pH. This complex may exhibit either a monobridged or a tribridged binding mode of the bis-hydroxamate ligands. 88,89 At high pH a µ-oxo-bridged alcaligin-complex [Fe2L2O2] is present.89 Like the bis-hydroxamates, amonabactins, a series of four bis-catecholate siderophores form 2:3 metal:ligand complexes at high pH and in excess of ligand. At lower pH, a 1:1 Fe3+:ligand stoichiometry is found, again like in the corresponding hydroxamate complexes.90

4 Equilibrium Thermodynamics of Siderophore Iron Binding

The formation constants of siderophore complexes define the thermodynamic limits for the conditions in which siderophores can compete for iron and extract it from a weaker substrate. Superior propagation conditions will be provided for that organism from which the superior complexing agent was secreted. In addition, these constants are a measure of the siderophore ligand's selectivity for aqueous Fe3+ in the presence of competing metal ions and its ability to solubilize Fe and prevent precipitation by hydrolysis. The tenacity of Fe3+ binding may determine the mechanism of metabolic turnover of the metal in microorganisms by means of ligand exchange, reductive removal, or ligand destruction.

Definitions: the equilibria of metal-ligand stability constants are expressed by a standard convention as ßmlh for the reaction mM+lL+hH=MmLlHh where M is metal, L is ligand and H is proton(s).ß110 means: the ligand is fully deprotonated and forms a 1:1 complex with the metal. 4.1 Ligand Protonation and Complex Formation Constants

Solutions of hydroxamate- or catecholate-containing siderophores are strongly absorbing with characteristic spectra (see Table 1 and Fig.6) which can be utilized for spectrophotometric determination of the complex formation constant. Iron(III) hydroxamates absorb in the visible region, producing a broad absorption band in the 420–440 nm region. Iron(III) catecholates exhibit pH-dependent absorption maxima. Unfortunately, the overall Fe3 + ion complex formation constants

Page 15 of 47

cannot be determined directly at neutral pH, because the extremely high stability of siderophore complexes precludes direct measurements of the equilibrium of interest, which would yield the desired formation constant for a tris-bidentate siderophore complex, β110 (see equation (2) ).10,16

Fe3+ + Ln- FeL3-n ß110 = [FeL3-n] / [Fe3+][Ln-] (2) One method of circumventing this problem is the spectrophotometric measurement of competition

for the metal by another thermodynamically well-characterized ligand, typically EDTA. To convert the resultant proton-dependent equilibrium constant into the conventional formation constant, it is necessary to know the ligand protonation constants.10

Ligand protonation constants (pK) of the iron-binding groups of siderophores were determined by spectrophotometric and potentiometric titration. Hydroxamic acids containing siderophores are fairly weak acids with pKa values ranging from 7.6 to 9.9.10 Catechols are very weak diprotic acids with widely separated protonation constants (pK1 between 7.3 and 9.2, pK2 between 11.5 and 13).10 In general, the greater the basicity of a chelator, the greater its affinity for FeIII. However, when the pKa of the ligand is substantially greater than the physiological pH (7.4), proton competition will considerably decrease the concentration of the basic form of the ligand, thus reducing iron binding.

Although the conventional form of tabulation, formation constants are not meaningful alone in judging the relative ability of ligands to compete with one another for FeIII at a given pH. This is due to differences in ligand protonation constants (their number and absolute value), which define the amount of free, uncomplexed, and unprotonated ligand L in aqueous solution. To have a more direct ranking of the ligands under physiologically relevant conditions, so-called pM values have been used. Here pM is defined as −log[M(H2O)n]m+, calculated from the constant at pH 7.4, an assumed ligand concentration of 10 µM, and metal concentration 1 µM.10 Stability constants and pM values of selected siderophores are listed in Table 4. The constants indicate unusually high specificity of the ligands for Fe3+. The reason for this is the high charge and small size of the FeIII ion, which makes it a hard acid. This, in turn, requires the ligand to be a hard base and charged oxygen atoms as occur in catecholates and hydroxamates behave as hard basic ligands.

A comparison of the stability constants of the naturally occurring siderophores uncovers a difference of 17 orders of magnitude between enterobactin (K ~ 1049)96,97 and the most stable hydroxamate complex, ferrioxamine E.9,10 Using the more comparable pM values, enterobactin remains still eight orders of magnitude more effective than ferrioxamine E. Enterobactin has the highest affinity for FeIII ion of any biological iron chelator tested so far. Table 4 Stability constants and redox potentials of selected natural FeIII siderophore complexesa Siderophore logβ110

b pMc logKMHLd E1/2 (mV vs. NHE)

Ferrioxamine E 32.5 27.7 – – Coprogen 30.2 27.5 0.5 −447 Ferrioxamine B 30.5 26.6 1.0 −468 Aerobactin 22.5 23.3 – −336 Enterobactin 49 35.5 4.80 −750 DHBS – – – −350 Parabactin – – – −673 Parabactin A – – – −400 a All solutions are aqueous. b See text. c pM = −log[Fe(H2O)6

3−] when [Fe]T = 10−6 M, [L] = 10−5 M, pH 7.4. d KMHL = [FEHL4−n]/ [H+][FeL3−n].

Page 16 of 47

Fig. 6 UV-Vis and CD spectrum of ferricrocin in aqueous solution

Page 17 of 47

The protonation constants KMHL of iron(III) trihydroxamate siderophore complexes are in the

range of 1, indicating that these complexes are stable even under very acidic conditions. In contrast, tricatecholate complex structures are strongly pH dependent. The pH dependence reflects changes of the iron environment. In enterobactin, the metal is coordinated by three catecholate dianions at pH 7. Raymond and co-workers proposed that with decreasing pH the catecholate oxygens are protonated stepwise concomitant with a bonding shift from catecholate to a salicylate mode of binding, resulting eventually in the neutral complex [Fe3+(H3ent)]0 (path 1 in Figure 9).97 For enterobactin and the synthetic analog MECAM, the following complex protonation constants have been evaluated: KMHL = 4.89 and 7.08, respectively, KMHL = 3.15 and 5.6, respectively.97,98 Raymond and co-workers corroborated their initial proposal, employing spectrophotometric titrations,97 pH-dependent Mössbauer spectra,99 an IR study,100 and a 2H NMR investigation.101

Figure 9 Possible protonation schemes of tris(catecholate) metal complexes. In path 1 the metal complex undergoes a series of two overlapping one-proton steps to generate a mixed salicylate–catecholate coordination. Further protonation results in the precipitation of a tris(salicylate) complex (e.g. enterobactin, MECAM). This differs from path 2, in which a single two-proton step dissociates one arm of the ligand to form a bis(catecholate) chelate. Path 3 incorporates features of paths 1 and 2. In this model the metal again undergoes a series of two overlapping one-proton reactions. However, unlike the case of path 1, the second proton displaces a catecholate arm, which results in a bis(catecholate) metal complex

Page 18 of 47

4.2 Iron Exchange Kinetics

In early papers on chelate exchange of iron(III) siderophore complexes it was assumed that the FeIII ion is readily exchanged, due to the known kinetic lability of high-spin FeIII-ion binding. More recent investigations disclosed, however, that ligand exchange reactions are slow and involve stepwise ligand dissociation and protonation, and ternary complex formation.10,102 From exchange kinetics of the form shown in equation (3)

55FeL + L' 55L' + L (3) an approximate linear-free-energy relationship between the equilibrium competition constant and

the second-order rate constant for iron removal from ferrichrome, ferricrocin, and coprogen to ferrioxamine B has been derived.10 Taking triacetylfusarine C as competing ligand, 63% 59Fe was removed from iron(III) citrate after 0.5 h, and even 82% from iron(III) rhodotorulate but only 4% from ferrichrome A.103 In addition, it was shown that backbone sequence changes in siderophore structures can influence Fe3+ lability, but the largest effect is observed on going from a cyclic to an open-chain structure.104 55Fe exchange kinetics of the form of equation (4)

55FeL + FeL' FeL + 55FeL' (4)

between ferrioxamine B and ferrichrome A were measured at equimolar concentrations (4.0 mM) and 5% excess of desferriferrioxamine B at pH 7.4. The t1/2 for exchange is very slow (220 h).10 In contrast, the exchange reaction between ferrioxamine B and a synthetic tris-catecholate complex exhibits a half-life of approximately 10 hours for equimolar concentrations (1.0 mM) of the competing complexes with no excess free ligand. From a mechanistic point of view the characteristic feature of siderophore iron-exchange processes is the unraveling of one ligand from the inner coordination sphere, concomitant with complexation by the incoming ligand. This occurs via the formation of a ternary complex between metal ion and entering and leaving ligands. The general features of these exchange processes conform to ideas formulated in earlier studies of multidentate ligand exchange.105 Dissociation of iron from its siderophore complex is a proton driven process and and its first step is relatively fast at low pH (t1/2 ≅ 2-5ms)106, 107 This step normally involves the dissociation of a bidentate moiety from the fully formed complex (k1 = 3.8x102M-1 s-1) Further dissociation is much slower (second order rate constant k3 = 2.3 x 10-2 M-1 s-1).108 Proton driven ligand dissociation from tetradentate siderophore complexes (Fe2L3) exhibit relatively fast rate constants (k3= 0 1.5 x 102 M-1s-1 for alcaligin and k3 = 6.8 x 102 M-1 s-1 for rhodoturulic acid).107 The resulting tetracoordinated complex is relatively stable and requires a drop in pH below 1 for further dissociation. In the case of the hexadentate siderophore ferrioxamine B dissociation kinetics disclosed an extreme kinetic and thermodynamic stability at physiological pH.108 In contrast to hexadentate siderophores, tetradentate siderophores can undergo ligand dissociation by multiple pathways. Intermediates of different structures, depending on environmental conditions (pH, Fe:L ratio), have been identified by electrospray ionisation mass spectrometry(ESI-MS). The multiplicity of interconvertible species may play a role in cellular recognition, and in the formation of ternary complexes representing a likely prerequisite for biological reduction.107, 109 4.3 Redox Chemistry of Siderophores

In microbial iron assimilation, one mechanism for the release of iron from siderophores is the enzymatic reduction to the FeII state. Siderophore stability constants are much lower for Fe2+, which

Page 19 of 47

has a lower charge-to-radius ratio. Moreover, ligand exchange reactions for high-spin FeII ion are much faster than for FeIII ion. Stability constants of ferrous siderophores are experimentally difficult to obtain. Limiting i.e. pH-independent redox potentials can be utilized, however, to describe the electrochemical and chemical equilibria between fully coordinated Fe3+ and Fe2+-siderophore complexes and the uncomplexed Fe(H2O)6

3+ and Fe(H2O)62+, respectively in a simple model as

described in equation (5) 115 E0

Fe(III)Sid/Fe(II)Sid = E0aq - 59.15 log( ßFeIII

110 / ßFeII110) (5)

where E0aq is the redox potential of hexaaquated iron (E0

aq = +770 mV/NHE).116 Given the redox potential of the siderophore, and the stability constant of the ferric siderophore complex, the stability constant of the ferrous siderophore complex can be calculated. Based on these considerations Fe2+-siderophore stability constants were determined. Fe2+-siderophores are some 20 orders of magnitude less stable than their Fe3+ counterparts.

Table 4 illustrates the redox potentials of various siderophores obtained by cyclic voltammetry.

The values of most of the hydroxamates are within the range of typical biological reductants such as NADH-dehydrogenase [−320 mV (NHE)].( section 5.4). However, the redox potential of tris-catecholate siderophores such as enterobactin and parabactin are considerably more negative than those of hydroxamates. Raymond and co-workers have also measured the redox potentials of Fe3+ complexes of synthetic tricatechols and found them to be in the range −0.8 to −1.1 V (NHE).9,

102,117,118 In the case of enterobactin the enormous stability of the FeIII complex, its high specificity for Fe3+, and low affinity for Fe2+ result in the Fe3+/Fe2+ formal potential being highly negative. The full coordination of Fe3+ by all three catechol groups requires the loss of six protons. Thus there is a strong dependence of the stability of the iron(III) enterobactin complex and its reduction potential on pH. The formal electrochemical potentials for FeIII/II enterobactin are (at pH noted) −0.99 (> 10.4), −0.79 (7.4), −0.57 (6.0), +0.17 (4.0) V (vs. NHE).99,119 The redox equilibria have been discussed controversially for some years but there now seems to be agreement on the formulation of the redox chemistry. Below pH 6, protonation of the FeIII complex and dissociation of the FeII become significant and the electrochemical reaction is not well defined. Therefore at pH 4 the potential has been estimated assuming no complexation of the FeII ion by enterobactin.99 Mössbauer spectra revealed that, in a methanolic solution, FeIII enterobactin undergoes protonation and a ligand-based redox reaction with the formation of a semiquinone/FeII couple under acidic conditions.120 The reaction becomes irreversible if the solution is allowed to stand under aerobic conditions. The mechanism is described as shown in equations (6) and (7)

[Fe3+(ent)]3- + nH+ → [Fe3+[Hnent)]n-3 (methanol) (6) H+ O2

[Fe3+[Hnent)]n-3 ⇒ [Fe2+[Hnent)]n-3 ⇒ polymeric quinones + Fe2+ (7)

Page 20 of 47

5 Siderophore Transport in Microorganisms

Siderophore-mediated iron uptake in microorganisms is both a receptor- and an energy-dependent

process.4,5 Moreover, many receptor–siderophore interactions are very specific.121-125 Therefore, unraveling the molecular basis of siderophore recognition by the cell continues to represent a major topic of siderophore research. Both, the transport mechanisms, and the closely related processes of metal release from the complex are of comparable importance. Various excellent books and review articles have been published on iron transport and iron transport agents in microorganisms.4-6,8,9,126-

128 At this point it should be mentioned that recently a series of specific microbial iron transport systems have been discovered which do not require siderophores in order to fulfill their central task of solving the iron-supply problem. These systems include various haem transporters129,130, lactoferrin transport131 and microbial transferrin132. In addition, direct mobilization of metals from a mineral surface by membrane protein might be another strategy of iron acquisition in microbes.133 Merely salient features of siderophore transport will be presented in the following sections. Within the past two decades our knowledge on bacterial iron transport systems exploded. In particular, the Gram-negative bacterium Escherichia coli has been analyzed in great detail by means of genetic methods. Therefore we shall mainly focus on the model organism, E. coli.

Gram-negative bacteria like E. coli are surrounded by two membranes, the outer membrane and the cytoplasmic membrane (Figure 10).134 Between these two layers is the compartment known as the periplasmic space, which may make up 10–40% of the total cell volume, depending on the osmotic strength of the external medium. Solutes from the medium have to diffuse, or be tranported, across these three regions to reach the cytoplasm. Pores large enough to allow diffusion of siderophores through the outer membrane pose a threat to the organism, because they would also allow entry of noxious substances like,e.g., detergents and antibiotics. Therefore, receptor proteins arose parallel to siderophores during the course of evolution exhibiting extraordinary specificities for individual siderophores or a small group thereof. Figure 10 depicts the proteins involved in siderophore transport in E. coli. These proteins are essential constituents of the different transport systems.

Page 21 of 47

Figure 10 A schematic drawing of the cell envelope of E. coli consisting of the cytoplasmic membrane, the periplasm, and the outer membrane. Various proteins are shown, sets of which represent specific siderophore-transport systems. Outer membrane receptors (OMR) shown here are FepA (enterobactin), IutA (aerobactin), Fec A (FeIII dicitrate), FhuA (ferrichrome), and FhuE (coprogen, FeIII rhodotorulate, and ferrioxamine B). FoxA (ferrioxamine B), is not a receptor of E. coli, but of the closely related Salmonella. Not shown here are the receptors Fiu and Cir (FeIII and Feo, a transport system for FeII and Feo, a transport system for FeII. Esterbonds of enterobactin are hydrolized by Fes leading to different degradation products (DHBS)n. Details are discussed in Section 5.1.

5.1 Siderophore receptors For E. coli, seven Fe3+ transport systems have been characterized.5,135 At the level of the outer

membrane, these seven transport systems are determined by specific receptors (the corresponding gene products are mentioned in parentheses) which selectively recognize the siderophores enterobactin (FepA), FeIII (DHBS)n (Fiu and Cir; n indicates 3 possible linear degradation products of enterobactin),43 aerobactin (IutA), FeIII dicitrate (FecA), ferrichrome and closely related ferricrocin and ferrichrysin (FhuA), and coprogen, FeIII rhodotorulate, and ferrioxamine B (FhuE). In addition, a transport system for FeII iron has been detected (Feo).136 Only enterobactin and, in some E. coli strains, aerobactin are synthesized by the organism itself. The residual siderophores transported in E. coli are xenosiderophores produced by different microorganisms. The enormous metabolic activity aimed at iron aquisition underscores the central role that this metal plays for survival. In general, similar transport systems were identified in other Gram-negative bacteria.137 In Gram-positive bacteria, which lack an outer membrane, the receptor protein is anchored by a covalently linked lipid. The receptor proteins are synthesized in response to iron limitation in such amounts that they become major outer membrane proteins with copy numbers between 10 000 and 100 000 per cell.

In the past few years crystal structures have been published of three of the seven outer-membrane-siderophore receptors from E.coli, namely of FepA,137,138 FhuA,139,140,141 and FecA142,143. FepA and FhuA represent similar monomeric transmembrane proteins that are composed of 22 antiparallel ß-strands of approximately 70Å height (see Figure 11)143. The right- handed twist of the ß-strands produce an elliptical-shaped barrel with a diameter of 35x47Å thus constituting a trans-membrane pore. 143 Large extracellular loops extend approx 35 Å (in the case of FepA) above the surface of the outer membrane.144 These loops are predominantly hydrophilic but contain in addition a large number of aromatic amino acids (Tyr,Phe) and are believed to represent the region of initial siderophore-contact. The globular, N-terminal domain of the receptor protein inserts from the periplasmic side into the ß-barrel functioning as a plug/cork (Figure 11).

Fortunately, X-ray structures of various FhuA-substrate complexes have been determined recently allowing a more detailed analysis145,146. Surprisingly, siderophore- binding does not trigger big changes at the extracellular side of the receptor. Some alterations are detectable in the periplasmic loops and very few in the barrel and extracellular loops. Transition from the unliganded to siderophore-liganded conformation induces local and allosteric transitions illustrating an induced-fit binding mechanism. Changes occur primarily in the N-terminal domain. The most visible change is the unwinding of the first helix (α1) in the N-terminal domain. 144 The mode of receptor-binding to ferrichrome and to other ferrichrome analogs discloses that it is the iron chelating hydroxamic acid portion and key groups on the peptide backbone of the siderophore which are preferentially recognized (“binding pocket”). It is noteworthy that this mode of recognition was predicted already 15 years earlier based on siderophore-transport and -inhibition studies employing modified

Page 22 of 47

siderophores, siderophore analogs and Fe,Ga, Cr and Rh-complexes thereof (see section 5.2). Unfortunately, a crystal structure of Fe-ent and of the Fe-ent-FepA complex is still missing. However, a recently published structure of a Fe-ent complex with the human serum protein lipocalin NGAL allows us to imagine what kind of interaction can be expected for the FepA-Feent reptor complex.147 The crystal structure disclosed a novel type of substrate binding consisting of simple ionic and concomitantly of cation-π interactions. The ionic interaction is achieved by intercalating of positively charged side chains (lysine, arginine) of the NGAL-protein between the ligand catecholates.147

The outer membrane receptors exhibit two functional states. In an unenergized conformation the siderophore binds to the receptor, but is not transported through the membrane. The energized state of the receptors requires a periplasmic protein, named TonB, which serves as a coupling device between the the cytoplasmic membrane and the outer membrane receptors.148-151 . It is assumed that the above mentioned unwinding of helix α1 of the N-terminal domain facilitates the interaction of the outer-membrane siderophore-receptor complexes with TonB

Page 23 of 47

Figure 11 Crystal structure of the OM-siderophore receptor FhuA (Ferguson et al.139) with the bound siderophore ferrichrome (emphasized by a space filling atom presentation). The plot is based on PDB file 1QFF, taken from the NCBI data base using the program PYMOL. Resolution of the structure is 2.70 Å, and the space group is P61. The upper side of the figure corresponds to the extracellular side of the outer membrane . See chapter 5.1 for a detailled discussion.

5.2 Specificity and Stereospecificity of Siderophore Recognition Probed by Metal and Ligand Substitution

Recognition of siderophores may be dependent on different parts and structural features of the

molecule: (a) the molecule as an entity; (b) the geometry and chirality of the backbone; (c) chirality or geometry at the metal center; (d) peripheral groups. As described in section 5.1 Crystal structures of siderophore receptor proteins are now available. However, the molecular nature of the siderophore-recptor interactions is still not fully understood. In an early approach, the structural elements responsible for specific recognition were traced out by using compounds which are related to the natural siderophore, i.e., by employing homologous siderophores, synthetic enantiomers, or synthetic analogs.121-125 Moreover, diastereoisomers or geometrical isomers might affect such discriminations. Substitution of FeIII iron for kinetically inert ions such as Cr3+ or Rh3+ enables separation of isomers and studies of their receptor specificity in vivo employing transport and inhibition studies. Based on metal-substituted siderophores and separated geometrical isomers the sterospecificity and mechanisms of uptake were studied in E. coli,121,122 Rhodotorula minuta,125 Streptomyces pilosus,80, 123 and Neurospora crassa.81, 124

In E. coli the FeIII complexes of some synthetic analogs of enterobactin can act as iron sources and detailed kinetic and inhibition studies enabled a more precise description of the molecular groups essential for recognition by Fep A (Figure12 and 13).121,122 FeEnt carries a ∆ absolute configuration at the metal center, and although the synthetic FeIII enantio-enterobactin does not promote growth in E. coli,82 subsequent studies revealed that both enantiomers are transported.161 Therefore, chirality is not required for receptor-binding. On the other hand, enterobactin uptake is effectively inhibited by the structural analog Fe–MECAM (Figure 12) and Rh–MECAM122 indicating that the trilactone backbone is also not required for recognition.

In addition, substrates which are modified at the aromatic rings (MECAMS, LICAMS) or at the amide bond connecting the 2,3-hydroxybenzoyl groups with the backbone (TRIMCAM) are ineffective inhibitors (Figure 12, 13).121 From these experiments it was concluded that the part of the molecule recognized by the receptor is the metal-binding end, but the ligand to which these functionalities are attached is not important in receptor recognition. Actually, an unsubstituted triscatechol iron center and the amide linkage is essential for recognition. These studies clearly show that modified siderophores and synthetic models are excellent tools in determining specificity of siderophore uptake on a molecular level.

Stereospecific recognition in Rhodotorula minuta was probed with separate Λ-trans, ∆-trans, and ∆-cis chromium(III) RA as well as with synthetic enantio-RA.83 The RA uptake system discriminates between RA and its enantiomeric form. It also discriminates between Λ-trans and ∆-trans isomers of Cr2(RA)3, as shown by inhibition of FeIII RA uptake83 and, finally, changes in the backbone have only little effect on uptake rates.125 In Streptomyces pilosus the racemic C-cis,cis, N-cis,cis, and trans isomers of Cr3+ desferriferrioxamine B (see Table 2) inhibited equally well the 55Fe-ferrioxamine B uptake, suggesting that no differentiation between cis and trans geometrical isomers occurs in this system.80 Coprogen is the species-specific siderophore synthesized by Neurospora crassa. For Cr3+ desferricoprogen, two fractions were separated by HPLC, a cis complex (20% excess Λ) and a mixture of trans complexes (20% excess ∆) (see Table 2).81 Although this separation of optical and

Page 24 of 47

geometrical isomers was not complete, inhibition studies suggest a predominant recognition of the ∆-trans isomers. In summary, these investigations revealed that the metal-center chirality and the shape of the metal-center portions of siderophores are of central importance for molecular recognition by receptors in the cell envelope.

Figure 12 Inhibition of [55Fe]enterobactin uptake in E. coli by structural analogs. The initial uptake rate (five minute interval) of [55Fe]enterobactin was determined with and without inhibitor. The synthetic analogs of enterobactin, Fe–MECAM (circles), Fe–TRIMCAM (squares) and Fe–3,4-LICAMS (triangles) were added at various concentrations as potential inhibitors to the transport assay. The [55Fe]enterobactin uptake rate (36 pmol mg−1 min−1 at 0.5 µM complex concentration) without any inhibitor is defined as 100%. Only Fe MECAM is an efficient inhibitor of Fe enterobactin uptake. See text for discussion. (Reproduced American Society of Microbiology from D. J. Ecker et al.121).

Fig. 12

Fig. 13

Figure 13 (a) Inhibition of 2 µM [59Fe]enterobactin by [Rh(cat)3]3−. Control (closed circles) represents uptake of 1 µM label with no inhibitor added. Inhibition experiments: K3[Rh(cat)3]3− added at t = 6 min in 10-fold excess (open squares) or 100-fold excess (open triangles). (b) Inhibition of 2 µM [59Fe]enterobactin by [Rh(DMB)3]3−. Control (closed circles) represents uptake of 1 µM label with no inhibitor added. K3[Rh(DMB)3]3− added at t = 6 min in 50-fold excess (open squares). (Reproduced American Chemical Society from D. J. Ecker et al.122).

Page 25 of 47

5.3 Transport mechanisms After Fe3+ siderophores have docked at the binding pocket of the outer membrane receptors of E.

coli translocation into the periplasmic space is mediated by the TonB-complex. Once released into the periplasm, siderophores are rapidly bound by the specific periplasmic binding proteins FhuD(hydroxamate siderophores)162, FepB(enterobactin)163, and FecB(ferric dicitrate)164. FhuD, for example, exhibits a broad substrate specifity for a variety of hydroxamte siderophores including ferrichrome, coprogen, aerobactin, ferrioxamine B, shizokinen, rhodotorulic acid, and the antibiotic albomycin. The dissociation constants of FhuD with these siderophores range from 0.3 to 5.4 µM

165,166 X-ray structures of FhuD and FhuD-siderophore complexes167--169 explain this broad substrate specificity. Ferrichrome bound to FhuD is exposed at the protein surface thus enabling fewer constraints to ligands than a deep cavity does.

The siderophore-loaded periplasmic siderophore-carrier proteins eventually transfer the siderophores to so-called ABC-Transporter systems (from ATP-binding cassette)170-172 assembled of two proteins, one to span the membrane acting as a permease and a second one which can hydrolyse ATP to provide the energy for transport. The transmembrane permeases are FhuB for hydroxamates, FepDG for enterobactin, and FecCD for ferric dicitrate. The corresponding ATP-hydrolyzing proteins at the cytoplasmic side of the cytoplasmic membrane are named FhuC, FepC, and FecE. Ferric siderophores are released from the transport system at the cytoplasmic side of the cytoplasmic membrane. Eventually, iron is rapidly released from the siderophore complex via reduction (details in section 5.4) The molecular mechanism of siderophore transport across the cytoplasmic membrane is not yet fully understood.173 A completely different iron transport mechanism has been suggested for an OM-transport of pyoverdin via the FpvA receptor in Pseudomonas aeruginosa.174 The ligand exchange step occurs at the cell surface and involves the exchange of iron from a ferric pyoverdin to an iron-free pyoverdin strongly bound to the receptor FpvA. This mechanism suggests an increase of the iron uptake rate, with increasing concentration of iron-free siderophore. A similar transport system has been identified in the fresh water bacterium Aeromonas hydrophila.175 However, in contrast to P. aeruginosa, the Aeromonas receptor recognizes and transports an extremely broad range of siderophores, with chelating groups as varied as catecholate, hydroxamate or hydroxopyridoxinate.

In fungal systems, only little is known about the number and spatial organization of the proteins of the siderophore transport systems. Employing the modern methods of molecular biology some information is now available on yeasts and Ascomycetes.176-179 Most is known on the siderophore transporters from Saccharomyces cerevisiae because it is the only fungus the whole genome of which is known. The yeast transporter family – as far as it is known - includes Taf1 (triacetylfusarinine C), Arn1 (ferrichromes), Enb1 (enterobactin), and Sit1 (ferrioxamine and ferrichrome transporter).180

In general, two different types of siderophore-transport mechanisms are operative in fungi.8–10,181 One major path comprises transport across cell membrane(s) into the cytoplasm. In the cytoplasm, FeIII is removed from the complex by a specific or unspecific ferrireductase and FeII is transferred to an intracellular FeII carrier, the structure of which has not yet been determined. Finally the metal-free siderophore is resecreted for additional solubilization–assimilation cycles. Coprogen transport into the ascomycete Neurospora crassa and ferrichrome transport into the yeast Ustilago sphaerogena are typical representatives of this mechanism. The most prominent feature of the second transport mechanism is the lack of membrane permeation by the iron complex. The extracellular siderophore complex releases the metal to a membrane-associated iron shuttle system either by ligand exchange or by a reduction. A ligand exchange mechanism was found for FeIII rhodotorulate transport in Rhodotorula pilimanae125 (Figure 14) and a reductive removal is typical of the ferrichrome A

Page 26 of 47

transport in U. sphaerogena. In principle, redox-relevant aspects of transport mechanisms can be determined employing GaIII,

RhIII and CrIII desferrisiderophores. The redox potentials of these metal-substituted siderophores are outside the range observed for biological reductants. If these complexes show lack of intracellular accumulation, this could imply a reductive step during uptake. If Ga is transported and Cr is not, the uptake process should involve a ligand exchange step because GaIII complexes are kinetically labile.80,83 Detailed information on transport mechanisms and steps of iron metabolization have also been derived from in vivo Mössbauer spectroscopy (Figure 14).182-185

Figure 14 Mössbauer spectra of the frozen aqueous siderophore solution of [57Fe]rhodotorulate, Fe2(RA)3, (a), and of the yeast Rhodotorula minuta (b–d) at 4.2 K in a magnetic field Happ = 20 mT perpendicular to the γ-rays. Spectrum (b) corresponds to frozen Rhodotorula minuta cells, grown in a salt medium supplied with 15 µM 57Fe2(RA)3 and harvested at OD660 = 0.85. The spectrum exhibits various components. The magnetically split subspectrum is due to Fe2(RA)3 (a), as can be demonstrated by subtraction (c). This indicates that Fe2(RA)3 is present in the cell. However, the metal is exchanged at the membrane level as shown by uptake studies with Cr- and Ga-substituted RA. (c) shows the residual spectrum (b) after stripping. Only two species can be observed. This metabolic pattern is similar to those found in Gram-negative bacteria.180,182 Part (d) represents the Mössbauer spectrum of R. minuta at OD660 = 0.8 grown in iron-depleted medium. 57Fe2(RA)3 (15 µM) was added for 30 min. The 57Fe2(RA)3 contribution is 68% of the absorption area; the residual contribution is solely due to an Fe2+ high-spin species which seems to be the first iron metabolite after uptake

Page 27 of 47

5.4 Reductive removal of iron from siderophores in vivo Siderophores exhibit an affinity for Fe2+ which is some 20 orders in magnitude lower than for

Fe3+.106 The kinetics of ligand exchange of high-spin Fe2+ are much faster than for high-spin Fe3+.186 Therefore, reduction of ferric siderophores accompanied by ligand exchange is an excellent mechanism for intracellular iron release. In fact, one main route of bacterial iron acquisition includes active transport of ferric siderophores across cell membranes, followed by intracellular iron removal from the siderophore carrier via reduction. 183, 185,187,189 The deferrated siderophore is rapidly excreted.189 Many flavin reductases show the property to reduce a broad spectrum of [Fe3+]-siderophores in vitro. However, surprisingly it is not known which enzymes reduce the [Fe3+]-siderophore in the cell.187 Moreover, a substantial number of trihydroxamate-siderophores possess very negative redox potentials well below -320mV at neutral pH precluding reduction via NADH or NADP. In a very recent investigation, the Fe2S2-protein, FhuF, of E. coli has been identified as a siderophore reductase exhibiting a redox potential E1/2 of –310 ± 25 mV vs NHE.190

Although the redox potential of FhuF is in the same range as NADH, FhuF is capable of reducing ferrioxamine B (see Table 4) in vitro and in vivo.190 In fact, it has been postulated that it is the redox potential of the ternary reductase-ferric-siderophore complex rather than the redox potential of the free ferric siderophore that is important when considering reduction as a process involved in iron release.191 In general, one has to bear in mind that a thermodynamically unfavorable process can nevertheless be driven if steady-state conditions are applied. In addition, the pH could influence the redox potential of a siderophore. For example, the midpoint potential of ferrioxamine B is –468 mV at pH 7.4, but increases to –180 mV at pH 4.5.192

A different strategy of intracellular iron release was found for the catechol siderophores enterobactin and parabactin. The pH 7 redox potential values for these siderophores are beyond the range of normal physiological reducing agents (see Table 4). This suggests that cellular iron release may occur by a reductive mechanism only after chemical transformation of the FeIII siderophore to a form with an appropriate redox potential; an alternative process would be a significant drop in pH (e.g. in the acidic periplasm) to bring about an increase in FeIII siderophore redox potential. In fact, an esterase was found in the cell extracts of E. coli that cleaves the ester backbone of FeIII enterobactin.193 In addition, it was demonstrated that the redox potential of the hydrolysis product Fe(DHBS)3 (dihydroxybenzoylserine = DHBS) is −350 mV and within the range available from biological reductants (Table 4). Therefore, a mechanism of enterobactin-mediated iron uptake was proposed which involved an initial ester cleavage to metabolize iron. However, uptake studies with synthetic analogs which lack ester units and which are not susceptible to hydrolysis showed growth response of B. subtilis and of E. coli (Figure 9).194 Thus the esterase mechanism for in vivo iron release from FeIII enterobactin became uncertain. An in vivo Mössbauer study of 57Fe(ent) and 57Fe(MECAM) uptake in the aro− mutant E. coli K12 RW193 has helped to address this problem.195 Whereas both Fe(ent) and Fe(MECAM) are taken up as the intact complex, the rates of metabolic utilization are drastically different (70% of enterobactin-bound iron within 140 min, but only traces after two hours of Fe(MECAM) uptake). This is clear evidence for a merely slow, i.e. non-specific, utilization of MECAM-bound iron. Above all, this finding substantiates the esterase mechanism of iron removal from enterobactin. Since no appreciable amounts of Fe(DHBS)3 could be detected in the enterobactin experiments, any esterase activity and reductive ligand exchange must be closely correlated. The positive growth promotion with Fe(MECAM) uptake must be interpreted as being due to a small amount of nonspecifically metabolized complex which is still sufficient for normal growth rates of E. coli.195 Although the existing data provide a consistent model, there continues to be confusion on this subject. Even a recent major microbiolical review196 incorrectly assigned the major route of iron release from enterobactin.

Page 28 of 47

Another illustration of a ligand-degradation process may be seen in the case of parabactin. Parabactin also has too negative a redox potential for iron removal via a reductive mechanism. However, upon hydrolysis of the oxazoline group of parabactin to produce parabactin A the redox potential is shifted to a more positive value that is within the physiological range for FeIII reduction (Table 4). Therefore, in vivo oxazoline ring cleavage followed by FeIII reduction may be involved in metabolic iron removal in Paracoccus denitrificans.

Finally, it has to be pointed out that the formation of bulk “free” Feaq2+ in the cytoplasm of

microorganisms is not to be expected in vivo under "healthy” conditions. Feaq2+ is Fenton-active and

thus deleterious to cells. Therefore, it is likely that the ferrous iron is bound by a chaperone-like structure or other novel structural elements to enable sheltered intracellular transfer of this ion. In fact, an oligomeric sugar phosphate was detected in E. coli197 that binds ferrous iron (cytoplasmic concentration: approximately 500 µM) intracellularly released from ferrichrome (E. coli185) or ferrioxamine E (Pantoea agglomerans183). This compound was named ferrochelatin.198

5.4a Bacterial Iron metabolism analyzed by in situ Mössbauer spectroscopy

time δ(mms-1) ∆EQmms-1 percentage

component 1 [Fe3+(06)] 30 min 0.47(4) 0.85(5) 13.6 60 min 0.53(4) 0.88(4) 13.3 90 min 0.49(4) 0.91(4) 29.6 OD578=1.0 0.43(2) 0.97(2) 50.1

component 2 [Fe2+(O6)] 30 min 1.28(2) 3.10(2) 86.4 60min 1.24(2) 3.06(2) 86.7 90 min 1.25(2) 3.04(2) 70.4 OD578=1.0 1.23(2) 2.86(3) 49.9

Fig. 15: Mössbauer spectra measured at 4.3 K, in a field of 20mT perp to the γ-beam: A) 57Fe-ferrioxamine E (fox E) in frozen aque-ous solution. Cells were grown in iron defi-cient media to OD578= 0.8 and then incuba-ted with 57Fe-fox E. B) Pantoea, 30 min in-cubation C)Pantoea 60 min incubation, D) Pantoea 90 min incubation

Conclusions: ferrioxamine E is not accumulated intracellularily. The main metabolite formed initially is an octahedral ferrous high spin compound ([Fe2+(OX)6]. This requires iron reduction i.e. reduc-tase activity. The ferrous iron spe-cies was isolated and represents an oligomeric sugar phosphate (R. Böhnke, BF Matzanke, Biometals 8, 223-30 1995)A second ferric ion species is growing in, the nature of which is not completely clear yet. Probably, it corresponds to an [4Fe-4S]-cluster (S=0).

Page 29 of 47

Iron metabolization kinetics in E. coli: In some strains of E. coli (endogeneous siderophore: enterobactin) an array of major iron metabolites can be discriminated by Mössbauer spectroscopy thus allowing a kinetics of the iron metabolite pattern. In this study an enterobactin negative mutant (entC) was employed and iron transport occured via ferricrocin

Question Sample Mutant Conditions MB1 entC Addition of 57Fe-ferricrocin with the inoculum A MB2 wt Addition of 57Fe-ferricrocin with the inoculum

MB3 entC Addition of 57Fe-ferricrocin with the inoculum heavy shaking = high po2 B

MB4 entC Addition of 57Fe-ferricrocin with the inoculum no shaking = low po2

MB5 entC Addition of 57Fe-ferricrocin 30min before harvest C MB6 entC Addition of 57Fe-ferricrocin 180min before harvest MB7 entC Addition of 57Fe-3,4 DHB with the inoculum D MB8 entC Addition of 57Fe-2,3 DHB with the inoculum MB9 entC Cell debris from MB4 MB10 entC Membrane fraction from MB4 E MB11 entC Cytoplasmic fraction from MB4

Fig.16: Contribution of main iron metabolites af-ter ferricrocin-mediated iron uptake in E.coli de-termined by in situ Möss-bauer spectroscopy. .

Page 30 of 47

The general pattern of Mössbauer spectra is the same as in Erwinia. However in certain E.coli strains and under specific conditons, additional main iron metabolites are observed (bacterial ferritin Ftn, and a rubredoxin-like component. MB 1 represents the fit results derived from the steady state spectrum of entC and MB2 of the wild-type strain. The results are similar and 4 different metabolites can be discriminated: A putative [4Fe-4S]-cluster, [Fe(II)(OX)6], bacterial ferritin (Ftn) and a small amount of [Fe(II)S4] (rubredoxin-like). In MB4, the total iron level is lower than under highly oxygenated conditions (MB4), because it contains less Fe(II) and ferritin. The Fe(II) contribution in the highly oxygenated sample MB3 is one of the highest in this study. This is an indication for regulation patterns different of current concepts. Samples MB5 and MB6 represent the kinetics of iron metabolization and are compared to the steady state sample MB1. Again, in MB5 over 80% of total iron within the short incubation time (30min) corresponds to [Fe(II)(OX)6], indicating that reductive decomplaxation dominates the metabolization process. After 3h (MB6), the metabolite pattern shifted dramatically. MB6 displays the highest iron accumulationin in this study, which is reflected by the largest amount of Fe(II) and ferritin. In contrast, rubredoxin and 4Fe-4S levels are slightly lower than under steady state conditions. This is most likely due to the still high activity of the iron uptake apparatus and a slower cellular turnover at the stationary growth phase (OD578=1,2). As a consequence, excessive iron is rapidly stored in ferritin in order to circumvent oxygenic stress associated with the Fenton reaction. After 3 hours of incubation, the metabolic pattern is obviously significantly different to the metabolic state of MB1. Different iron chelates (MB7, MB8) also affect the iron metabolite pattern indicating differences in the regulatory tuning.

In summary, these studies demonstrates that Mössbauer spectroscopy is a powerful tool for deciphering distribution- and regulation patterns of main iron metabolites in microorganisms and probably also in eukaryotic cells 5.5 Siderophore uptake regulation

The genes coding the proteins of siderophore transport systems and the enzymes of siderophore biosynthesis and a variety of additional genes scattered through the chromosome are regulated by the iron requirement of the organism. A protein, named Fur (ferric uptake regulation) has been identified which is responsible for the transcriptional regulation of these genes. The fur gene product acts as a classical repressor requiring Fe2+ as an activator.199-201 When the internal iron concentration is high, Fur forms a complex with FeII iron, exhibiting a strong binding affinity to DNA sequences (fur boxes) in the promoter regions of all genes which are negatively regulated by iron.202 The Fur binding prevents transcription of these genes. The crystal structure of Fur from Pseudomanas aeruginosa uncovered a functional dimer, each subunit exhibiting an N-terminal DNA-binding domain and a C-terminal dimerization domain.202 Two metal centers were found, a Zn2+-, and a Zn2+/Fe2+-binding site.202,203 The Zn-binding site plays probably an important role in maintaining the integrity of the structure. At the second site Zn is exchangeable with Fe2+ and hence represents the regulatory site. Binding of ferrous iron leads to local conformational changes, which in turn probably alter the DNA-binding of the DNA-binding domain. Under conditions of iron deficiency, FeII iron is removed from Fur, the binding affinity to the DNA recognition sequences is lost, and Fur-dependent genes are transcribed. In addition, it has been demonstrated by in vivo Mössbauer spectroscopy that very high intracellular levels of FeII iron are maintained even under extreme growth conditions in fur− mutants, whereas wild-type strains show high levels of FeIII iron. 185,204 Therefore it has been concluded that fur also regulates the oxidation state of the main components of iron metabolism in E. coli. It is now well established that a major subclass of Fur-like proteins is mainly involved in cellular iron homeostasis. But in addition, it can function in acid tolerance and protection against oxidative stress 205-208

Page 31 of 47

Page 32 of 47

Ferricrocin uptake in E.coli wildtype W3110 and in a ∆fur-mutant.

Ferricrocin (not diluted)

E.coli K12 W3110

1

2

12

Ferricrocin uptake in E.coli wildtype W3110: δ1= 0.43 mms-1∆EQ1= 1.05 mms-1, 47% δ2= 1.28 mms-1∆EQ2= 2.89 mms-1, 53% ∆fur-mutant: δ1= 0.45 mms-1∆EQ1= 1.01 mms-1, 24% δ2= 1.29 mms-1∆EQ2= 2.95 mms-1, 76%

Lack of repressor Fur reduces dramatic-ally metabolic turnover of Fe2+ aquired from the siderophore ferricrocin by ra-pid reduction employing the siderophore reductase FhuF( see section 5.5)

Page 33 of 47

5.6 Siderophore secretion Siderophore secretion systems have been identified in only a few microorganisms so far. The

exporters that were found or suggested to be involved in siderophore release belong to efflux pumps of major facilitator superfamily (MFS); the resstance, nodulation, and cell division superfamily(RND); and the ATP-binding cassette superfamily(ABC).[287,228,28,69,72,152]

6. Siderophores in Medicine 6.1 Iron overload diseases, ß-thalassemia

In the treatment of β-thalassemia and certain other anemias, periodic whole blood transfusions are required.211-214 Since there is no specific physiological mechanism for the excretion of iron in man, continued transfusion therapy leads to a steady buildup of iron in the body and to deposition of iron in a number of organs, resulting in tissue damage and premature death. These iron excesses, as well as the primary iron overload diseases such as hemochromatosis and hemosiderosis, and accidental iron poisoning, require the removal of iron from the body, especially from the liver. The current drug of choice for the treatment of transfusional iron overload is the trihydroxamate siderophore desferrioxamine B. The long-term treatment with desferrioxamine B is generally well tolerated but it has two major drawbacks: lack of oral efficacy and short lifetime in the serum. Because of these limitations, there has been a wide search for new drugs with increased efficacy. Ideally, such an agent should be orally active and resistant to degradation or metabolism in the gastrointestinal tract, the bloodstream, the liver, and the kidneys. Furthermore, the drug should be effective at concentrations much lower than the gram amounts per day presently used for desferrioxamine B, should show a relatively long retention time in the body, and should have no adverse side effects. The design of such drugs continues to use siderophores as a principal model.213-215

Desferrioxamine B has also found therapeutic application for various pathological conditions due to aluminum overload.216-218 Accumulation of this toxic metal is frequently observed in chronically dialyzed patients who have lost the ability to clear via renal excretion. Desferrioxamine B has also been recommended for the diagnosis of such an overload state. 6.2 Siderophores and infection

Iron is abundant in the human body, but it is bound to intracellular and extracellular components (transferrin, lactoferrin, ferritin; hemo-proteins) This strict iron homeostasis leads to a free serum concentration of about 10-24M. Thus, a multitude of microorganisms, among them important human and animal pathogens, are restricted in multiplication. Under these conditions the iron transport systems of the invading microorganism are expressed.233,234. Therefore, siderophore production contributes to bacterial virulence.230,231 Reduced virulence of mutants deficient in siderophore production have been reported for Yersinia pestis, the causative agent of bubonic plague232 and Vibrio vulnificus, a cholera-related pathogen associated with septicaemia arising from eating infected shell fish.

Mammalian defense mechanisms: 1. During the acute-phase reaction occurring upon inflammation, iron storage is enhanced

leading to a reduction in serum iron levels, termed hypoferemia of infection. 2. Immunoglobulins against siderophores: Besides an unconfirmed report on enterobactin

specific immunoglobulins no observation of siderophores acting as real immunogens have been made

3. Serum albumin does not efficiently impair enterobactin-mediated iron acquisition..

Page 34 of 47

4. Siderophore sequestration: The protein siderocalin (also known as uterocalin,lipocalin 2[Lcn2], NGAL, or 24p3) is an acute-phase protein in mammals.[99,193] NGAL, binds tightly Fe-enterobactin bacillibactin, carboxymycobactins and parabactin (see section 5.1).147 NGAL is a potent bacteriostatic agent under iron limiting conditions. Therefore, NGAL very likely participates in the antibacterial iron depletion strategy of the innate immune system. A glycosilation of enterobactin at two catechol C-5 positions efficiently impairs the sequestration by sidercalin, thus representing an efficient antidefense of pathogenic Salmonella enterica.

An alternative to siderophore modificationb or self-restriction to certain compartments(

Mycobacterium tuberculosis) the establishment of completely novel siderophores as virulence factors is a third strategy that allows circumvention of siderophore sequestration. Certain invasive strains of E. coli, Shigella flexneri K.pneumoniae responsible for infections in vertebrates, harbor plasmids that enhance their virulence. One set of proteins encoded by these plasmids was identified as the transport system for the mixed-ligand siderophore aerobactin. It has been demonstrated in various studies that the ability to synthesize and transport aerobactin enhances virulence of bacteremic strains. Iron is not only an essential nutrient for invading microorganisms but may also serve as a regulatory signal influencing the expression of a variety of virulence factors. It is likely that iron, like calcium and constant temperature, is one of the environmental signals which serve to regulate the expression of genes required for survival and virulence in the host. There are several bacterial toxins that are derepressed under the conditions of low-iron stress typically found in a host. These include exotoxin A in Pseudomonas aeruginosa, the phage-mediated diphtheria toxin in Corynebacterium diphtheriae, the tetanus toxin of Clostridium tetani, and the shiga toxin of Shigella dysentheriae. In sum , bacterial pathogens have evolved sophisticated high-efficiency iron acquisition mechanisms to secure a suitable iron supply, whereas the mammalian hosthas evolved an equally sophisticated countermechanism to restrict the essential iron supply to the pathogen.

6.3 Siderophore-based antibiotics Given the growing resistance eg of Mycobacterium tuberculosis and certain strains of Pseudomonas against antibiotic drugs currently at hand there is an urgent need for completely new classes of antibiotics. In many pathogens multi drug resistance (MDR) is observed and merely a limited development of novel drugs takes place. Therefore, one of the pressing goals to increase public health preparedness is the development of new chemotherapeutics against conventional and unconventional bacterial targets. Siderophores have tremendous therapeutic potential that is barely tapped.

A B

Page 35 of 47

A) Trends of antibacterial drug resistance: incidence of methicillin resistant S.aureus (▲) fluoroquinolone resistant Ps aeruginosa (○) and of vancomycin resistant enterococci (■). B) Trend of antibacterial drug development (approved in the USA). Taken from L.E.N Quadri, Infectious Disorders-Drug Targets 2007, 7, 230-237. “Trojan Horse” antibiotics Some naturally occurring antibiotics employing a siderophore backbone have been isolated: Albomycins, derivatives of ferrichrome block protein biosynthesis by inhibiting seryl-tRNA synthetases after hydrolysis of seryl-thioribosyl pyrimidine from the siderophore carrier. Danomycins and salmycins produced by actinomycetes contain the trihydroxamate danoxamine connected with an aminoglycoside moiety. The compounds are highly active protein synthesis imhibitors in gram positive bacteria such staphylococci and streptococci. The main drawback for therapeutical use is the high frequency of resistant siderophore transport mutants. Recent research, however, has shown that semisynthetic siderophore-drug conjugates exhibit reduced susceptibility to resistance mechanisms

It has been shown in early inhibition studies as well as in recent crystal structures (see section 5.1 and 5.2) that the backbone portion of siderophores is not predominantly involved in receptor interactions; the carbon skeleton is partially solvent exposed instead. To this portion of the molecule antibiotics may be attached. The resulting mixed compounds would be actively taken up into the cell, where the antibiotic portion of the melecule can kill the cell.226 Indeed, the naturally occurring antibiotics albomycin and ferrimycin utilize this route for action.

Page 36 of 47

A) Albomycins δ1(X=O); δ2(X=NCONH2) ε (X=NH

B) ferrimycin A1 C) Salmycin A1

(R=NOH, n=5)

Page 37 of 47

This “Trojan horse” concept led to the synthesis of a series of siderophore-antibiotics.227-229

Several studies showed that synthetic siderophore-drug conjugation is feasible for creating antibiotics with improved cell permeability and reduced susceptibility to resistance mechanisms such as secretion or enzymatic inactivation. the above mentioned approach might turn out to be of importance in the battle against pathogenic microbes and deserves therefore further intensive investigations.

As with almost any organism, iron is an essential nutrient for Plasmodium falciparum, the parasite

causing malaria. In a series of investigations the antimalarial activities of various iron chelators were scrutinized. Some of these iron chelators exhibit activity in vitro and in animals with experimental plasmodial infections.220 In particular, desferrioxamine B, the only chelator that is widely available for use in humans, has clinical antimalarial activity. The same siderophore was shown to inhibit growth of Trypanosoma brucei, another protozoic parasite causing sleeping sickness in human bloodstream.221 However, in order to assess a siderophore or iron-chelator based therapy, extensive

Iron binding semisynthetic antipseudomonal cephalosporin(tonB-dependent transport)

Page 38 of 47

further investigations are required in this field. Inhibitors of siderophore synthesis

Given the central importance of iron availability for bacterial pathogenicity in mammals, it should be possible to develop anti.-inefectives that target siderophore production or other functions essential to the specific physiology associated with the iron starvation state. One group of bacterial pathogens synthesize –as outlied above (section2)- siderophores exhibiting building blocks of salicylic acid derivative. Important members are carboxymycobactin of M. tuberculosis, yersiniabactin of Yersinia pestis and pyochelin secreted by several pathogenic Pseudomonas species and Bukholderia, an opportunistic human pathogen. Required for the biosynthesis of salicylic acid components are specific synthetases. Humans have no homologs of these enznzymes. A synthetic nucleoside antimicrobial agent namend salicyl-AMS(9) blocks effectively salicylic acid biosynthesis in these bacteria. The agent has no cytotoxicity against mammalian cells in vitro. The therapeutic efficacy of salicyl-AMS remains, however, to be investigated in animal models of infection M.J. Miller et al, Biometals 2009, Vol. 22, 61-75 M. Niethke and M.A. Marahiel, Microbiology amd Molecular Biology Reviews, 2007, Vol.71, 413-451 L.E.N. Quadri, Infectious Disorders-Drug Targets 2007, Vol 7, 230-237. 6.4 Siderophores and MRI

For improved contrast enhancing for magnetic resonance imaging, different paramagnetic ions like Mn2+, Fe3+, and Gd3+ have been used. The Gd3+ is particularly well suited as contrast agent in diagnostoc medical MRI due to its high magnetic moment and favorable electronic relaxation rate, However, Gd3+ is highly toxic at concentrations required for MRI. Therefore, chelators are required that prevent release of the free cation in vivo. Again, siderophores and synthetic analogs thereof serve as principal models for such chelators219 6.5 Iron chelators and cancer

There might be also some therapeutic potential for siderophores and synthetic iron chelators in the treatment of cancers.222,223 One strategy is aimed on the antineoplastic activity of siderophores.224 A second strategy is focussed on chelators for the clearance of non-transferrin bound iron in serum which occurs in cancer therapy as a result of some chemotherapies.225

Amamistatin A and B from an actinomycete.

Page 39 of 47

7 Additional Functions of Siderophores 7.1 Siderophores as Iron Storage Compounds

The predominant class of intracellular iron storage compounds is represented by ferritin in eukaryotes and bacterioferritin in prokaryotes. In various in vivo Mössbauer spectroscopic studies on siderophore uptake in fungi it was realized that siderophores can also function as intracellular iron-storage compounds.182,184,209 In the ascomycete Neurospora crassa, the transport siderophore coprogen represents an intrcellular transient iron pool. A major part of coprogen-bound iron is transferred to a second siderophore, ferricrocin, which represents the long-term iron storage form in this fungus.209 Spores of N. crassa and of Aspergillus ochraceus are also lacking in ferritin-like iron-storage proteins. Rather, ferricrocin constituted 47% of the total iron content in N. crassa. In spores of A. ochraceus, 74% of the iron content was bound by ferrichrome-type siderophores.184 It is safe to conclude that these siderophores are iron-storage peptides in spores. In view of the large siderophore iron-storage pools in spores, the intracellular accumulation of ferricrocin in mycelia seems to warrant an iron pool sufficiently large for sporulation. In the heterobasidiomycetous yeast Rhodotorula minuta, the major intracellular iron pool is FeIII rhodotorulate, and in the smut fungus Ustilago sphaerogena, ferrichrome is found.210 Again, no ferritins or bacterioferritins could be detected. From this the conclusion was drawn that, in various Ascomyctes and Heterobasidiomycetes, intracellular siderophores replace ferritins as iron-storage compounds. This further extends the biological role of siderophores. Whether genes for ferritin or bacterioferritin are present in these organisms remains an open question. However, ferritins definitely do not play a significant role in iron storage of these fungi. 7.2. Other applications of siderophores Siderophores have been found to bind effectively to actinides, such as Th, U, Np, and Pu. 235-237 Siderophores and other naturally occurring ligands may therefore affect actinide mobility in waste repositories and in the environment and may also be used to treat radioactive waste prior to storage or to decontaminate soils and water237-239 Siderophores and siderophore analogs have also been investigated for possible medical use in the decorporation of actinides from the body235, while simple hydoxamates may have uses in the reprocessing of nuclear fuel.

Page 40 of 47

8 References 1. J.E. Posey and F.C. Gheradini, Science, 2000, 288, 1651 2. G. Wächtershäuser, Microbiol. Rev., 1988, 52, 452. 3. G. Winkelmann, H. Drechsel, Microbial Siderophores in ’Biotechnology’, 2nd edition, ed. H.-J.

Rehm, and G. Reed, VCH, Weinheim, 1997, Vol 7, pp 199-246 4. A. Sigel, and G. Sigel, Iron transport and storage in microorganisms, plants, and animals, metal

ions in biological systems Vol. 35., Marcel Dekker, Basel, 1998. 5. G.Winkelmann ed., Microbial Ttransport Systems, Wiley-VCH, Weinheim, 2001 6 Molecular and Cellular Iron Transport, ed D.M. Templeton, and R.F. Barron, Marcel Dekker,

New York, 2002 7 H. Drechsel, and G. Winkelmann, Iron chelation and siderophores, in Transition Metals in

Microbial Metabolism, eds G. Winkelmann, and C.J. Carrano, harwood academic publishers, Amsterdam, 1997, p1-49

8. G. Winkelmann, Handbook of Microbial Iron Chelates, CRC Press, Boca Raton, 1991. 9. B. F. Matzanke, G. Müller-Matzanke, and K. N. Raymond, Siderophore-Mediated IronTransport,

in Iron Carriers and Iron Proteins, ed. T. M. Loehr, VCH, New York, 1989, pp. 1-121. 10. K. N. Raymond, G. Müller, and B. F. Matzanke, Top. Curr. Chem., 1984, 123, 49. 11. J. B. Neilands, and C. Ratledge, Microbial Iron Transport Comounds, in Handbook of

Microbiology, 2nd edition,eds A. I. Laskin and H. A. Lechevalier, CRC Press, Boca Raton, 1982, Vol. 4, pp. 565-574.

12. J. H. Crosa, and C. T. Walsh, Microbiol. Mol. Biol. Rev., 2002, 66, 223. 13. K. Hantke, J. Mol. Microbiol. Biotech. , 2002, 4, 217. 14. N. Baichoo, T. Wang, R. Ye, and J. D. Helmann, Mol. Biol. , 2002, 45, 1613. 15. J. H. Crosa, Microbiol. Mol. Biol. Rev. , 1997, 61, 319. 16. B. F. Matzanke, Structures, Coordination Chemistry and Functions of Microbial Iron Chelates,

in Handbook of Microbial Iron Chelates, ed. G. Winkelmann, CRC Press, Boca Raton, 1991, pp. 15-64.

17. E. Wasielwski, R. A. Atkinson, M. A. Abdallah, and B. Kieffer, Biochemistry , 2002, 41, 12488. 18. R. A. Atkinson, A. L. M. Salah El Din, B. Kieffer, J.-F. Lefèvre, and M. A. Abdallah,

Biochemistry ,1998, 37, 15965. 19. H. Stephan, S. Freund, J.-M. Meyer, G. Winkelmann, and G. Jung, Liebigs Ann. Chem. 1993, 43. 20. M. A. F. Jalal and D. van der Helm, Isolation and Spectroscopic Identification of Fungal

Siderophores, in Handbook of Microbial Iron Chelates, ed G. Winkelmann, CRC Press, Boca Raton, 1991, pp. 235 - 270.

21. J.M. Roosenberg, Y.M. Lin,Y. Lu, and M.J. Miller, Curr. Med. Chem., 2000, 7, 159. 22. S. Dhungana, S. Heggemann, M. S. Gebhard, U. Möllman, and A. L. Crumbliss, Inorganic

Chemistry,2003, 42, 42-50. 23. Dong, J. M. Roosenberg, II, and M. J. Miller, .J Am. Chem. Soc. ,2002, 124, 15001. 24. A. Shanzer,and J. Libman, Biomimetic Siderophores; from Structural Probes to Diagnostic

Tools, in Iron transport and storage in micoorganisms, plants, and animals ed A.Sigel and H. Sigel, 1998, Marcel Dekker, Basel, Vol.35 of series metal ions in biological systems, pp.329-354

25. T. F. Emery and J. B. Neilands, J. Am. Chem. Soc., 1961, 83, 1626. 26. M. A. F. Jalal, M. B. Hossain, and D. van der Helm, Biol. Metals, 1988, 1, 77. 27. D. van der Helm and G. Winkelmann, Hydroxamates and Polycarboxylates as iron Transport

Agents (Siderophores) in Fungi, in Metal Ions in Fungi, eds G. Winkelmann and G. Winge, Dekker, New York, 1994, pp. 39-98.

28. H. Diekmann, Arch. Microbiol., 1967, 58, 1. R. E. Moore and T. Emery, Biochemistry, 1976, 15, 2719.

Page 41 of 47

29. H. Bickel, G. Hall, W. Keller-Schierlein, V. Prelog, E. Vischer, and A. Wettstein, Helv. Chim. Acta, 1960, 43, 2129.

30. Dhungana, S., White, P. S., and Crumbliss, A. L., J. of Biol. Inorg. Chem. ,2001,6, 810. 31. H. Bickel, E. Gäumann, G. Nussberger, P. Reusser, E. Vischer, W. Voser, A. Wettstein, and H.

Zähner, Helv. Chim. Acta, 1960, 43, 2105. 32. M. B. Hossain, M. A. F. Jalal, B. A. Benson, C. L. Barnes, and D. van der Helm, J. Am. Chem.

Soc., 1987, 109, 4948. 33. A Nishio, T., Tanaka, N., Hiratake, J., Katsube, Y., Ishida, Y., and J.Oda, J. Am. Chem. Soc. ,

1988, 110, 8733. 34. K. M. Ledyard, and A. Butler, J. Biol. Inorg. Chem. ,1997, 2, 93. 35. A. Takahashi, H. Nakamura, T. Kameyama, S. Kurasawa, H. Naganawa, Y. Okami, S. Takeuchi,

and H. Umezawa, J. Antibiotics, 1987, 40, 1671. 36. C. L. Atkins and J. B. Neilands, Biochemistry, 1968, 7, 3734. 37. Z. Hou, K. N. Raymond, B. O'Sullivan, T. W. Esker, T. Nishio, Inorg. Chem . ,1998, 37, 6630. 38. J. R. Pollack and J. B. Neilands, Biochem. Biophys. Res. Commun., 1970, 38, 989. 39. H.-P. Fiedler, P. Krastel, J. Müller, K. Gebhardt, and A. Zeeck, FEMS Microbiol. Lett., 2001,

196, 147. 40. H. Budzikiewicz, A. Boessenkamp, K. Taraz,A. Pandey, and J.M. Meyer, Z. Naturforsch, 1997,

52, 551. 41. J.J. May, T.M. Wendrich, and M.A. Marahiel, J. Biol. Chem., 2001, 276, 7209. 42. G. H. Tait, Biochem. J., 1975, 146, 191. 43. K. Hantke, G. Nicholson, W. Rabsch, G. Winkelmann, Proc.Natl. Acad. Sci. USA, 2003, 100,

3677. 44. B. Schwyn and J. B. Neilands, Anal. Biochem., 1987, 160, 47. 45. J. Meiwes, H.-P. Fiedler, H. Haag, S. Konetschny-Rapp, and G. Jung, FEMS Microbiol. Lett.,

1990, 67, 201. 46. H. Drechsel, S. Freund, g. Nicolson, H. Haag, O. Jung, H. Zähner, and G. Jung, Biometals, 1993,

6, 185. 47. S. Yamamoto, N. Okujo, T. Yoshida, S. Matsuura, and S. Shinoda, J. Biochem. ,1994, 115, 868. 48. A. Thieken and G. Winkelmann, FEMS Microbiol. Lett., 1992, 94, 37. 49. K. Barbeau, G. Zhang, D. H. Live, A. Butler, J. Am. Chem. Soc. , 2002, 124, 378. 50. M. J. Smith, J. N. Shoolery, B. Schwyn, I. Holden, and J. B. Neilands, J. Am. Chem. Soc., 1985,

107, 1739. 51. K. Nomoto, Y. Sugiura, and S. Takagi, Mugineic acids, Studies on Phytosiderophores, in ‘Iron

Transport in Microbes, Plants and Animals’, eds. G. Winkelmann, D. van der Helm, and J. B. Neilands, VCH, Weinheim, 1987, pp 401-428.

52 .U.W. Stephan, I. Schmidke, V.W. Stephan, and G. Scholz; BioMetals, 1996, 9, 84 53. J. Francis, J. Madinaveitia, H. M. Macturk, and G. A. Snow, Nature, 1949, 163, 365. 54. C. Ratledge, Mycobactins and Nocobactins, in Handbook of Microbiology, 2nd edn., eds. A. I.

Laskin and H. A. Lechevalier, CRC Press, Boca Raton, 1982, Vol. 4, p. 575-581 55. M. A. F. Jalal, Houssain, M. B., van der Helm, D., Sanders-Loehr, J., Actis, L. A., and Crosa, J.

H. (1989) et al., J. Chem. Soc. , 1989, 111, 292. 56. G.J. Sharman, D.H. Williams, D.F. Ewing, and C. Ratledge, Chem. Biol., 1995, 2, 553. 57. M. Abdallah, Pyoverdins and Pseudobactins, in Handbook of Microbial Iron Chelates, ed. G.

Winkelmann, CRC Press, Boca Raton, 1991, pp. 139-154. 58. Stephan, H., Freund, S., Meyer, J.-M., Winkelmann, G., and Jung, G. Liebigs Annalen der

Chemie, 1993, 43.

Page 42 of 47

59. Martinez, J. S., Carter-Franklin, N., J., Mann, E. L., Martin, J. D., Haygood, M. G., and Butler, A., Proc. Acad. Sci. USA , 2003, 100, 3754.

60. Martinez, J. S., Zhang, G. P., Holt, P. D., Jung, H.-T., Carrano, C. J., Haygood, M. G., and Butler, A. Science , 2000, 287, 1245.

61. Nemoto, A., Hoshino, Y., Yazawa, K., Ando, A., Mikami, Y., Komaki, H., Tanaka, Y., and Gräfe, U., J. Antibiot. , 2002, 55, 593.

62. C. J. Carrano, M. Jordan, H. Drechsel, D. G. Schmid, G. Winkelmann, Biometals , 2001, 14, 119..

63. Inorg. Chem., 1970, 9, 1. 64. M. B. Hossain, M. A. F. Jalal, B. A. Benson, C. L. Barnes, and D. van der Helm, J. Am. Chem.

Soc., 1987, 109, 4948. 65. C. L. Barnes, D. L. Eng-Wilmot, and D. van der Helm, Acta Crystallogr., 1984, C40, 922. 66. M. B. Hossain, D. L. Eng-Wilmot, R. A. Loghry, and D. van der Helm, J. Am. Chem. Soc., 1980,

102, 5766. 67. M. B. Hossain, M. A. F. Jalal, and D. van der Helm, Acta Crystallogr., 1986, C42, 1305. 68. M. B. Hossain and D. van der Helm, Acta Crystallogr., 1983, B39, 258. 69. J. R. Telford, K. N. Raymond, Inorganic Chemistry 37, 4578-4583 (1998). 70. S. Isied and K. N. Raymond, J. Am. Chem. Soc., 1976, 98, 1763. 71. Y. Tor, J. Libman, A. Shanzer, C. E. Felder, and S. Lifson, J. Am. Chem. Soc., 1992, 114, 6661. 72. M. E. Bluhm, S. S. Kim, E. A. Dertz, K. N. Raymond, J. Am. Chem. Soc. , 2002, 124, 2436. 73. M. E. Bluhm, B. P. Hay, S. S. Kim, E. A. Dertz, K. N. Raymond, Inorg. Chem. ,2002, 41, 5475. 74. D. L. Eng-Wilmot, A. Rhaman, J. V. Mendenhall, S. L. Grayson, and D. van der Helm, J. Am.

Chem. Soc., 1984, 106, 1285. 75. J. Leong and K. N. Raymond, J. Am. Chem. Soc., 1974, 96, 1757. 76. K. Abu-Dari and K. N. Raymond, Inorg. Chem., 1980, 19, 2034. 77. K. N. Raymond, K. Abu-Dari, and S. R. Sofen, Stereochemistry of Microbial Iron Transport

Compounds, in ‘Stereochemistry of Optically Active Transition Metal Compounds’, eds. B. E. Douglas and Y. Saito, ACS Symp. Ser., 1980, 119, 7.

78. K. N. Raymond, in ‘Bioinorganic Chemistry II’, ed. K. N. Raymond, Adv. Chem. Ser., 1977, 162, 33.

79. A. J. McCaffery, S. F. Mason, and R. E. Ballard, J. Chem. Soc., 1965, 2883. 80. G. Müller and K. N. Raymond, J. Bacteriol., 1984, 160, 304. 81. T. D. Y. Chung, B. F. Matzanke, G. Winkelmann, and K. N. Raymond, J. Bacteriol., 1986, 165,

283. 82. J. B. Neilands, T. J. Erickson, and W. H. Rastetter, J. Biol. Chem., 1981, 256, 3831. 83. G. Müller, Y. Isowa, and K. N. Raymond, J. Biol. Chem., 1985, 260, 13921. 84. D. L. Kepert, Inorg. Chem., 1972, 11, 1561. 85. A. Avdeef, and J. P. Fackler, Inorg. Chem., 1975, 14, 2002. 86. E. L. Muetterties, and L. J. Guggenberger, J. Am. Chem. Soc., 1974, 96, 1748. 87. T. J. McMurry, M. W. Hosseini, T. M. Garrett, F. E. Hahn, Z. E. Reyes, and K. N. Raymond, J.

Am. Chem. Soc., 1987, 109, 7196. 88. Carrano, C. J., Cooper, S. R., and Raymond, K. N., J. Am. Chem. Soc. , 1979, 101, 599. 89. Z. Hou, K. N. Raymond, B. O'Sullivan, T. W. Esker, and T. Nishio, Inorg. Chem. , 1998, 37,

6630. 90 J. R. Telford and K. N. Raymond, Inorg. Chem. , 1998, 37, 4578. 91. P. Gütlich, R. Link, and A. X. Trautwein, Mössbauer Spectroscopy and Transition Metal

Chemistry, Springer Verlag, Berlin, 1978. 92. A. Abragam and B. Bleany, Electron Paramagnetic Resonance of Transition Ions, Clarendon

Page 43 of 47

Press, Oxford, 1970. 93. B. F. Matzanke, Mössbauer Spectroscopy of Microbial Iron Uptake and Metabolism, in Iron

Transport in Microbes, Plants and Animals, eds G. Winkelmann, D. van der Helm, and J. B. Neilands, VCH, Weinheim, 1987, pp 251-284.

94. M. Llinas, D. M. Wilson, and M. P. Klein, J. Am. Chem. Soc., 1977, 99, 6846. 95. M. Llinas and K. Wüthrich, Biochim. Biophys. Acta, 1978, 532, 29. 96. L. D. Loomis and K. N. Raymond, Inorg. Chem., 1991, 30, 906. 97. W. R. Harris, C. J. Carrano, S. R. Cooper, S. R. Sofen, A. E. Avdeef, J. V. McArdle, and K. N.

Raymond, J. Am. Chem. Soc., 1979, 101, 6097. 98. W. R. Harris and K. N. Raymond, J. Am. Chem. Soc., 1979, 101, 6534. 99. V. L. Pecoraro, G. B. Wong, T. A. Kent, and K. N. Raymond, J. Am. Chem. Soc., 1983, 105,

4617. 100. V. L. Pecoraro, W. R. Harris, G. B. Wong, C. J. Carrano, and K. N. Raymond, J. Am. Chem.

Soc., 1983, 105, 4623. 101. M. E. Cass, T. M. Garrett, and K. N. Raymond, J. Am. Chem. Soc., 1989, 111, 1677. 102. A. L. Crumbliss, Aqueous Solution Equilibrium and Kinetic Studies of Iron Siderophore and

Model Siderophore Complexes, in ‘Handbook of Microbial Iron Chelates’, ed. G. Winkelmann, CRC Press, Boca Raton, 1991, pp 177-234.

103. J. P. Adjimani and T. Emery, J. Bacteriol., 1987, 169, 3664. 104. M. Akiyama, A. Katoh, and T. Mutoh, J. Org. Chem., 1988, 53, 6089. 105. D. C. Olson and D. W. Margerum, J. Am. Chem. Soc., 1963, 85, 297. 106. H. Boukhalfa and A.L. Crumbliss, Biometals , 2002, 15, 325. 107. H. Boukhalfa, T.J. Brickmann, S.K. Armstrong, and A.L. Crumbliss, Inorg. Chem. , 2000, 39,

5591. 108. M. Birus, Z. Bradic, G. Krznaric, N. Kujundzic, M. Pribanic, P.C. Wilkins, and R.G. Wilkins, Inorg. Chem, 1987, 26, 1000. 109. I. Spasojevic, H. Boukhalfa, R.D. Stevens, and A.L. Crumbliss, Inorg. Chem. , 2001, 40, 49. 110. A. Shanzer, J. Libman, S. Lifson, and C. E. Felder, J. Am. Chem. Soc., 1986, 108, 7609. 111. S. S. Eaton, J. R. Hutchinson, R. H. Holm, and E. L. Muetterties, J. Am. Chem. Soc., 1972, 94,

6411. 112. E. L. Muetterties, Acc. Chem. Res., 1970, 3, 266. 113. D. B. Rorabacher and D. W. Margerum, Inorg. Chem., 1964, 3, 382. 114. B. Borgias, A. D. Hugi, and K. N. Raymond, Inorg. Chem., 1989, 28, 3538. 115. I. Spasojevic, S. K. Armstrong, T. J. Brickmann, and A. L. Crumbliss, Inorg. Chem. ,1999, 38,

449. 116. E.H. Swift, and E.A. Butler. Quantitative Measurements and Chemical Equilibria, Freeman,

San Francisco, 1972 117. T. M. Garrett and K. N. Raymond, Pure Appl. Chem., 1988, 60, 1807. 118. S. J. Rodgers, C.-W. Lee, C. Y. Ng, and K. N. Raymond, Inorg. Chem., 1987, 26, 1622. 119. C.-W. Lee, D. J. Ecker, and K. N. Raymond, J. Am. Chem. Soc., 1985, 107, 6920. 120. V. L. Pecoraro, G. B. Wong, T. A. Kent, B. H. Huynh, and K. N. Raymond, J. Am. Chem. Soc.,

1983, 105, 4617. 121. D. J. Ecker, B. F. Matzanke, and K. N. Raymond, J. Bacteriol., 1986, 167, 666. 122. D. J. Ecker, L. D. Loomis, M. E. Cass, and K. N. Raymond, J. Am. Chem. Soc., 1988, 110,

2457. 123. G. Müller, B. F. Matzanke, and K. N. Raymond, J. Bacteriol., 1984, 160, 313. 124. H. Huschka, H.-U. Naegeli, H. Leuenberger-Ryf, K. Keller-Schierlein, and G. Winkelmann, J.

Bacteriol., 1985, 162, 715.

Page 44 of 47

125. G. Müller, S. J. Barclay, and K. N. Raymond, J. Biol. Chem., 1985, 280, 13916. 126. K. Poole and G.A.McKay, Frontiers in Bioscience , 2003, 8, d661. 127.V. Braun and M. Braun, Curr.Opinion Microbiol, 2002, 5, 194 128. K.N. Raymond, E.A. Dertz, and S.S. Kim, Proc. Natl. Acad. Sci. USA , 2003, 100, 3584. 129. C.A. Genco, and D.W. Dixon, Mol. Biol., 2001, 39, 1. 130. P. Arnoux, R. Haser, N. Izadi, A. Lecroisey, M. Delepierre, C. Wandersmann, and M. Czjzek,

Nature Struct. Biol. , 1999, 6, 516. 131. C.N. Cornelissen, Frontiers in Bioscience , 2003, 8, d836. 132. C.H. Taboy, K.G. Vaughan, T.A. Mietzner, P. Aisen, and A.L. Crumbliss, J. Biol. Chem.,

2001, 276, 2719. 133. D. Alexeev, H. Zhu, M.Guo, W.Zhong, D.J. Hunter, W.Yang, D.J. Campopiano, and P.J.

Sadler, Nat. Struct Biol., 2003, 4, 297. 134. H. Nikaido and M. Vaara, Outer Membrane, in ‘Escherichia coli and Salmonella typhimurium -

Cellular and Molecular Biology’, ed. F. C. Neidhard, American Society for Microbiology, Washington, 1987, pp 7-22.

135. V. Braun, K. Hantke, and W. Köster, Bacterial Iron Transport: Mechanisms, Genetics, and Regulation, in ‚Iron transport and storage in micoorganisms, plants, and animals’ ed. A.Sigel and H. Sigel, 1998, Marcel Dekker, Basel, Vol.35 of series ‚metal ions in biological systems, p.67-146.

136. M. Kammler, C. Schön, and K. Hantke, J. Bacteriol., 1993, 175, 6212 137. S.K. Buchanan, B. S. Smith, L. Venkatramani, D. Xia, L. Esser, M. Palnitkar, R. Chakraborty,

D. van der Helm, and J. Deisenhofer, Nature Struct. Biol., 1999, 6, 56. 138. K.C. Usher, E. Özkan, K. H. Gardner, and J. Deisenhofer, Proc. Natl. Acad. Sci. USA, .2001,

98,10676. 139. D. Ferguson, E. Hofmann, J.W. Coulton, K. Diederichs, and W. Welte, Science, 1998, 282,

2215. 140. A.D. Ferguson, V. Braun, H.-P. Fiedler, J. W. Coulton, K. Diederichs, and W. Welte. Protein

Science, 2000, 9, 956. 141. P. Locher, B. Rees, R. Koebnik, A. Mitschler,L. Moulinier, J.P. Rosenbusch, and D. Moras,

Cell, 1998,95, 771. 142. W. Yue, S. Grizot, and S. K. Buchanan, J Mol Biol , 2003, 332, 353.. 143. A. D. Ferguson and J. Deisenhofer, Biochim. Biophys. Acta ,2002,1565, 318. 144. D. van der Helm and R. Chakraborty, Structures of Siderophore Receptors, in ‘Microbial

Transport Systems’, ed. G. Winkelmann, Wiley-VCH, Weinheim, 2001, pp 261-287. 145. A.D. Ferguson, J. Ködding, J.W. Coulton, K. Diederichs, V. Braun, and W. Welte, Structure,

2001, 9, 707 146. A.D. Ferguson, R. Chakraborty, B. S. Smith, L. Esser, D. van der Helm, and J. Deisenhofer.

Science , 2002, 295:1715. 147. D.H. Goetz, M.A. Holms, N.Borregaard, M.E. Bluhm, K.N. Raymond, and R.K. Strong,

Molecular Cell ,2002, 10, 1033. 148. H. Schöffler and V. Braun, Mol. Gen. Genet., 1989, 217, 378. 149. U. Roos, R. E. Harkness, and V. Braun, Mol. Microbiol., 1989, 3, 891. 150. E. Fischer, K. Günter, and V. Braun, J. Bacteriol., 1989, 171, 5127. 151. K. Eick-Helmerich and V. Braun, J. Bacteriol., 1989, 171, 5117. 152. V. Braun, FEMS Microbiol.Rev., 1995, 16, 295. 153. V. Braun, S. I. Patzer, and K. Hantke, Biochimie , 2002, 84, 365. 154. R. A. Larsen, M. G. Thomas, and K. Postle, Mol. Microbiol., 1999, 31, 1809.. 155. A. Sauter, S. P. Howard, and V. Braun, J. Bacteriol., 2003, 185, 5747.

Page 45 of 47

156. J. D. Faraldo-Gomez, G. R. Smith, and M. S. P. Sansom, Biophys. Journal , 2003, 85, 1406. 157. M. Braun, F. Endriss, H. Killmann, and V. Braun, J. Bacteriol. , 2003, 185, 5508. 158. C. S. Klug, S. S. Eaton, G. R. Eaton, and J. B. Feix, Biochemistry , 1998, 37, 9016. 159. T. J. Barnard, M. E. Watson, Jr, and M. A. McIntosh, Mol. Microbiol. , 2001, 41, 527. 160. P. E. Klebba, Frontiers in Bioscience , 2003, 8, s1422. 161. P. Thulasiraman, S.M. Newton,J. Xu, K.N. Raymond, C. Mai, A. Hall, M.A. Montague, and

P.E. Klebba , J Bacteriol. ,1998, 180, 6689. 162. J.W. Coulton, P.Mason, D.R. Cameron, G. Carmel, R. Jean, and H.N. Rode, J. Bacteriol.

1986, 165, 181. 163. M.L. Lundrigan and R.J. Kadner, J. Biol.Chem., 1986, 261, 10797. 164. U. Pressler, H. Staudenmaier, L. Zimmermann, and V. Braun, J. Bacteriol., 1988, 170, 2716. 165. W. Köster and V. Braun, J.Biol.Chem., 1990, 265, 21407 166. M.R. Rohrbach, V. Braun, and W. Köster, J. Bacteriol., 1995, 177, 7186. 167. T. E. Clarke, V. Braun, G. Winkelmann, L. W. Tari, and H. J. Vogel, J. Biol. Chem., 2002, 277,

13966. 168. T. E. Clarke, M. R. Rohrbach, L. W. Tari, H. J. Vogel, and W. Köster, Biometals , 2002, 15,

121. 169. T. E. Clarke, S.-Y. Ku, D. R. Dougan, H. J. Vogel, L. W. Tari, Nature Structural Biology 7,

287-291 (2000). 170. A.L.Davidson, and H. Nikaido, J. Biol.Chem., 1991, 266, 8946. 171. P.-Q. Liu, and G.F.-L. Ames, Proc. Acad. Natl Sci. USA, 1998, 95, 3495. 172.W. Boos, T. Eppler, Prokaryotic Binding Protein-Dependent ABC Transporters, in ‘Microbial

Transport Systems’, ed. G. Winkelmann, Wiley-VCH, Weinheim, 2001, p77-114. 173. V. Braun, and K. Hantke, Mechanisms of Bacterial Iron Transport, in ‘Microbial Transport

Systems’, ed. G. Winkelmann, Wiley-VCH, Weinheim, 2001, pp 289-311 174. I.J. Schalk, C.Hennard, C. Dugave, K. Poole, M.A. Abdallah, and F. Pattus, Mol. Microbiol. ,

2001, 39, 351. 175. A. Stintzi, C. Barnes,J. Xu, and K.N. Raymond, Proc. Natl. Acad. Sci. USA , 2000, 97, 10691. 176. H. Haas, M. Schoeser, E. Lesuisse, J.F. Ernst,W. Parson, B. Abt, G. Winkelmann, and H.

Oberegger, Biochem. J. , 2003, 371, 505. 177. K. Haselwandter and G. Winkelmann, Biometals , 2002, 15, 73. 178. H. Haas, Appl. Microbiol. Biotechnol. , 2003, 62, 316. 179. P. Heymann, M.Gerads, M. Schaller,F. Dromer, G. Winkelmann, and J.F. Ernst, Infect. Immun.

, 2002, 70, 5246. 180. G. Winkelmann, Siderophore Transport in Fungi, in Microbial Transport Systems, G.

Winkelmann ed., VCH, Weinheim, 2001, p 463 -479. 181 G. Winkelmann, D. van der Helm, and J. B. Neilands, ‘Iron Transport in Microbes, Plants and

Animals’, VCH, Weinheim, 1987. 182. B. F. Matzanke, E. Bill, G. I. Müller, G. Winkelmann, and A. X. Trautwein, Eur. J. Biochem.,

1987, 162, 643. 183. B. F. Matzanke, I. Berner, E. Bill, A. X. Trautwein, and G. Winkelmann, Biol. Metals, 1991, 4,

181. 184. B. F. Matzanke, E. Bill, A. X. Trautwein, and G. Winkelmann, J. Bacteriol., 1987, 169, 5873. 185. B. F. Matzanke, G. Müller, E. Bill, and A. X. Trautwein, Eur. J. Biochem., 1989, 183, 371. 186. S.N. Lincoln, and A.E. Merbach, Adv. Inorg. Chem., 1995, 42, 1 187. M. Fontecave, J. Coves, and J.L. Pierre, J. L. Biometals, , 1994, 7, 3–8. 188. L. Rauscher, D. Expert, B.F. Matzanke, and A. Trautwein, J.Biol. Chem., 2002, 277, 2385. 189 A. Hartmann, H.P. Fiedler, and V. Braun, V., Eur. J. Biochem. , 1979, 99, 517.

Page 46 of 47

190. B.F. Matzanke, S. Anemüller, V. Schünemann, A.X. Trautwein, and K. Hantke, Biochemistry, 2004, 43, 1386.

191. J.B. Neilands, J. Biol. Chem. , 1995, 270, 26723.. 192. R. Helman, and G.D. Lawrence, J. Electroanal. Chem. , 1989, 22, 276. 193. T. J. Brickman and M. A. McIntosh, J. Biol. Chem., 1992, 267, 12350. 194. S. Heidinger, V. Braun, V. L. Pecoraro, and K. N. Raymond, J. Bacteriol., 1983, 153, 109 195. B. F. Matzanke, D. J. Ecker, T.-S. Yang, B. H. Huynh, and K. N. Raymond, J. Bacteriol., 1986,

167, 674. 196. C. Ratledge, and L.G. Dover, Annu. Rev. Microbiol., 2000, 54, 881. 197. R. Böhnke, and B.F. Matzanke, Biometals , 1995, 8, 223. 198. B. F. Matzanke, Iron Storage in Microorganisms, in Transition metals in microbial metabolism

(Winkelmann, G., and Carrano, C., eds), harwood academic publishers, 1997, 117-157. 199. K. Hantke, Mol. Gen. Genet., 1984, 197, 337. 200. S. Schäffer, K. Hantke, and V. Braun, Mol. Gen. Genet., 1985, 200, 110. 201. D. W. Griggs and J. Konisky, J. Bacteriol., 1989, 171, 1048. 202. E. Pohl, J.C. Haller, A. Mijovilovich,W. Meyer-Klauke, W., E. Garman, and M.L. Vasil,

Molecular Microbiology , 2003,47, 903. 203 A. Gonzalez de Peredo, C. Saint-Pierre, A. Adrait, L. Jacquemet, J.M. Latour, I. Michaud-Soret,

and E. Forest, Biochemistry, 1999, 38, 8582. 204. B. F. Matzanke, E. Bill, and A. X. Trautwein, Hyperfine Interact., 1992, 71, 1259. 205. K. Hantke, curr. Opin. Microbiol., 2001, 4, 172 206. H.K. Hall, and J.W. Foster, J. Bacteriol, 1996, 178, 5683 207. N. Bsat, A. Herbig, L. Cassillas-Martinez, P. Setlow, and J.D. Helmann, J. D. Mol. Biol.,, 1998,

29, 189. 208. S. Dubrac, and D. Touati, Microbiology , 2002, 148, 147-159 209. B. F. Matzanke, E. Bill, A. X. Trautwein, and G. Winkelmann, Biol. Metals, 1988, 1, 18. 210. B. F. Matzanke, E. Bill, A. X. Trautwein, and G. Winkelmann, Hyperfine Interact., 1990, 58,

2359. 211. C. Hershko, Link, G., and Konijn, A. M. , Cardioprotective Effect of Iron Chelators, in Iron

Chelation Theraphy , ed. Hershko, C., ed, Kluwer Academic / Plenum Publishers, New York, Vol. 509, 1 Ed., 2002, pp. 77-89

212. B. Modell, E. A. Letsky, E. A. Flynn, R. Peto, and D. J. Weatherall, Br. Med. J., 1982, 284, 1081.

213. A. Pietrangelo, Mechanism of iron toxicity, in Iron Chelation Theraphy ed. Hershko, C., ed, Kluwer Academic / Plenum Publishers, New York, Vol. 509, 1 Ed., 2002, pp. 19-43.

214. T.B. Chaston, and D. R. Richardson, Am. J. Hematol., 2003, 73, 200 215. Membranes, ed. A. N. Martonosi, Plenum Press, New York, 1985, Vol. 3 216. P. Ackrill, A. J. Ralston, J. P. Day, and K. C. Hodge, Lancet, 1980, 2, 692. 217. R. A. Romero and J. P. Day, Trace Elem. Med., 1985, 2, 1. 218. H. G. Nebeker and J. W. Coburn, Annu. Rev. Med., 1986, 37, 79. 219. D.M.J. Doble,M. Melchior, B. O'Sullivan, C. Siering, J. Xu, V.C. Pierre, and K.N. Raymond,

K. N. Inorg. Chem. ,2003, 42, 4930. 220. V.R. Gordeuk, and M. Loyevsky, Antimalarial effect of iron chelators, in Iron Chelation

Theraphy , ed. Hershko, C., Kluwer Academic / Plenum Publishers, New York, Vol. 509, 1 Ed., 2002, pp. 251-272.

221. T. Breidbach, S. Scory, R. L. Krauth-Siegel, D. Steverding, Int J Parasitol, 2002, 32, 473.

Page 47 of 47

222. D. R. Richardson, Therapeutic potential of iron chelators in cancer therapy, in Iron Chelation Theraphy , ed C. Hershko, Kluwer Academic / Plenum Publishers, New York, Vol. 509, 1 Ed., 2002, pp 231-249.

223. J. D. Kemp, Histol. Histopathol. , 1997, 12, 291. 224. A. Kicic, A. C. G. Chua, E. Baker, British J. Pharmacol. , 2002, 135, 1393. 225. A. C. A. Chua, H. A. Ingram, K. N. Raymond, E. Baker, Eur. J. Biochem. ,2003, 270, 1689. 226. B.F. Matzanke, U. Möllmann, R. Reissbrodt, V. Schünemann, and A.X. Trautwein, Hyperf.

Interact. , 1998, 112, 123. 227. J.M. Rosenberg,Y.M. Lin, Y. Lu, and M.J. Miller, Curr. Med. Chem., 2000, 7, 159. 228. C. Hennard, Q.C. Truong, J.-F. Desnottes,J.-M. Paris, N.J. Moreau, and M.A. Abdallah, J.

Med. Chem. , 2001, 44, 2139. 229. L. Heinisch, S. Wittmann,T. Stoiber, A. Berg, D. Ankel-Fuchs, andU. Mollmann, J. Med.

Chem. , 2002, 45, 3032. 230. J.J. Bullen, and E. Griffith, Iron and Infection, John Wiley, Chichester, 1999 231. M. Stork, M.Di Lorenzo, T.J. Welch, L.M. Crosa, L. M., and J.H. Crosa, Plasmid, 2002, 48,

222. 232. R.D. Perry, J. Abney, I. Mier, Jr., Y. Lee, S.W. Bearden, and J.D. Fetherston, Adv. Exper. Med.

Biol. , 2003, 529, 275-283 233. E. D. Weinberg, Clin. Physiol. Biochem., 1986, 4, 50. 234. S. M. Payne, CRC Crit. Rev. Microbiol., 1988, 16, 81. 235. P.W. Durbin, N. Jeung, S.J. Rodgers,, P.N. Turowski, F.L. Weitl, D.L. White, and K.N.

Raymond, Radiat. Prot. Dosim., 1989, 26, 351 236. G.N. Stradling, J. Alloys Comp., 1998, 271-273, 72 237. J.C. Renshaw, V. Halliday, G. D. Robson, A. P. J. Trinci, M. G. Wiebe, F. R. Livens, D.

Collison, R. J. Taylor, Appl. Environ. Microbiol. 2003, 69, 3600. 238. C.E. Ruggiero, M.P. Neu, J.H. Matonic, and S.D. Reilly, Actinide Res. Q, 2000, 16 239. H.R. von Gunten, and P. Benes, Radiochim. Acta,1995, 69, 1