Site-selective incorporation of alpha- and beta-amino acid derivatives : towards new gramicidin S-based bactericides

(1)

towards new gramicidin S-based bactericides

Knaap, M. van der

Citation

Knaap, M. van der. (2010, September 8). Site-selective incorporation of alpha- and beta- amino acid derivatives : towards new gramicidin S-based bactericides. Retrieved from https://hdl.handle.net/1887/15935

Version: Corrected Publisher’s Version

License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden

Downloaded from: https://hdl.handle.net/1887/15935

Note: To cite this publication please use the final published version (if applicable).

(2)

α- and β-Amino Acid Derivatives:

Towards New Gramicidin S-Based Bactericides

Mat th ij s v an d er K n aa p

(3)

Site-Selective Incorporation of α- and β- Amino Acid Derivatives: Towards New

Gramicidin S-Based Bactericides

Proefschrift

ter verkrijging van

de graad van Doctor aan de Universiteit Leiden

op gezag van Rector Magnificus prof. mr. P.F. van der Heijden, volgens besluit van het College voor Promoties

te verdedigen op woensdag 8 september 2010 klokke 15.00 uur

door

Matthijs van der Knaap Geboren te Vlaardingen in 1981

(4)

Prof. dr. G.A. van der Marel Copromotor: Dr. ing. M. Overhand

Overige Leden: Prof. dr. A. van Belkum (Erasmus Universiteit) Prof. dr. J. Brouwer

Dr. G.M. Grotenbreg (Singapore National University) Prof. dr. J.C.M. van Hest (Radboud Universiteit) Prof. dr. J. Lugtenburg

This research is supported by the Dutch Technology Foundation STW, which is the applied science division of NWO, and the Technology Programme of the Ministry of Economic Affairs (project number 07109)

(5)

(6)

General Introduction

1.1HOST-DEFENSE PEPTIDES

Virtually all organisms employ host-defence peptides as the first line of defence against invading pathogens. A wide variety of peptides from mammalian, bacterial, insects, plants and other life forms, have been isolated and analysed.^[1-4] Many of these peptides are cationic and amphiphilic (CAPs) in nature and they act by lysing the cell membrane of the invading pathogen. This process can be summarised as follows: the first step entails electrostatic attraction between the positive charges on the peptide and the negatively charged biomolecules on the surface of the pathogenic cell, such as phospholipids and teichoic acids.

After localisation on the cell membrane hydrophobic amino acid residues interact with the fatty tails in the phospholipid bilayer, which results in disruption of the membrane integrity and ultimately cell death. Membrane disruption can occur in various ways, depending on the membrane composition, peptide concentration and nature of the peptide.^[5-9] Though the membrane lytic mechanism appears to be the main mode of action, some CAPs play a role in immunomodulation, thereby contributing to the protection of the host cell in an additional way.^[2,6] Bacteria from the Bacillus family are known to produce a wide variety of peptides with

(12)

[10]

high membrane disrupting ability. Prominent examples are gramicidin S, gramicidin A, tyrocidine, subtilisin and polymixin.^[10]

Gramicidin S (GS, 1)^[11,12] was first isolated from Bacillus brevis in the 1940s.^[13,14] and it is highly active against Gram-positive bacteria, but also able to kill some Gram-negative bacteria and certain fungi. Unfortunately, GS is incapable of discriminating between bacterial and

mammalian cells, thus rendering GS toxic to humans. The first article^[13] on gramicidin S describes how World War II soldiers with gunshot wounds, inflamed burns, shot wounds and other injuries were treated with a solution of the peptide. Though survival rates were not mentioned, it was observed that inflammations were well treatable with a solution of GS, resulting in the rapid disappearance of bacteria in and around the wound. In more recent years GS has found application in topical treatment of ear infections (in combination with polymixin, framycitin and oxytetracycline),^[15] in eye drops, as a spermicidal vaginal contraceptive and to treat genital ulcers.^[16,17] Besides gramicidin S has been identified as a potent anti-HIV agent.^[16] Recently it was discovered that GS is a potent inhibitor of cytochrome bd-type quinol oxidase, widely distributed in bacterial terminal oxidase, but absent in eukaryotic cells.^[18,19]

In view of the increasing number of multiply resistant pathogens, it is vital that new antibiotics will be developed. Ideally, new antibiotics should have different cellular targets than those on which the current antibiotics act. Prevention of bacterial resistance is another important topic which should be addressed in the design of new broadly applicable antibiotics. CAPs are interesting lead compounds in this respect, as they target the cell membrane, in stead of a specific gene product.^[1,2,17] The general idea is that it will unlikely that

NH HN

NH O

O

O HN NH HN NH

O

H2N O O O

N O

NH2

HN

O N

O

1, Gramicidin S

(13)

[11]

bacteria build up resistance against such membrane active compounds, as it would demand extreme alterations in the structure and composition of cell membranes.^[4,20] With its cyclic structure and the presence of D-amino acids, gramicidin S shows high stability against proteases, though there are some proteins known that specifically hydrolyse GS.^[21-23] A main focus of the research on GS is the preparation of derivatives displaying strong bactericidal properties, but low hemolytic activity. By introducing slight alterations in the secondary structure and amino acid composition of GS the activity against bacteria and human cells may be modulated.

1.2 STRUCTURE OF GRAMICIDIN S

Gramicidin S (GS, 1) is cyclic decameric peptide. The primary structure is cyclo-(Pro-Val- Orn-Leu-^DPhe)2 and is nonribosomally synthesised by gramicidin synthetase.^[24] It contains the D-phenylalanylprolyl dipeptides, which form type II´ β-turns. The strands in between the two turns are hydrophobic and cationic adopting an extended β-sheet conformation, which is stabilised by four interstrand hydrogen bonds, resulting in the alignment of valine and leucine to one face of the peptide, and the ornithines side chains on the opposite side. This amphiphilicity (spatial segregation between hydrophobic and hydrophilic) seems to be a prerequisite for the action of GS, as described before. It needs to be stressed that the rigid β- hairpin structure as a whole is at the basis of the high bactericidal activity of GS and not the individual amino acids. Alteration in the turn, for example, will have a marked influence on the activity too. Because GS forms such a tight β-hairpin structure, it is an interesting object to modify and to assess the influence on the secondary structure and activity. For example several reports have appeared on the incorporation of dipeptide isosteres as β-turn mimetics^[25-32] or peptide bond mimics.^[33,34]

(14)

[12]

1.3 MODIFICATIONS

As mentioned before, introducing alterations in the primary and secondary structure of GS may result in peptides with an improved biological profile, i.e. high antibacterial activity and low toxicity against human cells. Literature has shown many modifications of GS in order to achieve this goal. The majority of the older literature on GS is relevant in the context of structure-activity relationships. The following paragraphs summarise the modifications of gramicidin S that have been described in literature over the past fifty years with an eye on the purpose of the GS derivative. Where applicable the antibacterial activity and hemolytic activity are indicated. In general the hemolytic activity, a measure for toxicity against human cells, is lowered upon decrease in bactericidal activity and vice versa. Few examples are known in which a dissection in bacterial and hemolytic activity was seen, in the sense that peptides with high activity against bacteria and low hemolytic activity were obtained.^[35,36]

1.3.1STRAND

The strands of gramicidin S contain four hydrophobic and two cationic residues, which are positioned in such a way that the cyclic peptide becomes amphiphilic. Changes in the nature of the hydrophobic and hydrophilic residues will likely result in dramatic changes in biological activity.

1.3.1.1VALINE

Table 1 sums up the alterations that have been made on gramicidin S with respect to the valine residue. Complete removal of the hydrophobic side chain (glycine derivative 2) resulted, as expected in total loss of activity,^[37] while a methyl group (3)^[37,38] instead of the iso- propyl group mainly restored the antibacterial activity.^[37] Substituting the valines by leucine (4)^[39,40] reduced the antibacterial activity about fifty percent. The reason for this was not discussed in the original article,^[39] but a possible explanation is the destabilisation of the β- strand, resulting in a less amphiphilic peptide. Altmann et al.^[40] demonstrated that gramicidin

(15)

[13]

S synthetase is able incorporate leucine at the valine position, while it is not able to substitute valine at the leucine position.^[41] Izumiya and co-workers^[42] and Hodges and co-workers^[43]

made asymmetric GS derivatives by substituting one valine for a cysteine. The derivative in which the thiol was still protected with a para-methoxybenzyl (5) showed a reasonable amount of membrane activity, but the deprotected and dimerised construct (6) lost all activity.^[42] Besides gramicidin S, B. Brevis produces other peptides in minor quantities, namely in which valine is once or twice substituted by a 2-aminobutyric acid residue (7)^[41,44] and these derivatives proved to be nearly as active as GS itself.^[44,45] Inversion of the stereochemistry of valine (8)^[46,47] results in steric clash between the iso-propyl side chain and the neighbouring amino acids, with a drop in amphiphilicity and antibacterial activity as a result. The same is true for the α,α-dimethyl substituted aminoisobutyric acid derivative 9.^[48] Disturbance of the amphiphilicty is also expected when a cationic residue is incorporated in the place of valine, like ornithine 10 and this expectation is indeed reflected in the biological activity of this derivative.^[49] Several fluorinated derivatives of GS have been described, like 11,^[50] 12^[51] and 13^[52] in the position of valine. Arai et al.^[50] incorporated hexafluorovaline with the objective to modify the antibacterial activity and to use the fluorine atoms in ¹⁹F NMR spectroscopy. Much of the antibacterial activity was unfortunately lost. Ulrich and co-workers^[51] showed that 12 is a sensitive probe to study the interaction of this GS analogue with membranes and they found

NH O

NH O S MeO

NH O 1 +++ 2 --- 3 ++ 4 ++ 5 + 6 -- 7 ++/++^a

NH O

NH O F3C CF3

NH O F

NH O

8 -- 9 -- 10 -- 11 - 12 + 13 14

Table 1 Substitutions for valine. Behind the compound number the antibacterial activity is given (+++ means highest activity, --- means no activity).

a Activities for the symmetric and asymmetric substitutions, respectively

(16)

[14]

minor loss in activity as well, compared to the wild type GS. Peptide 13 was used to demonstrate the effectiveness of a new solid phase synthesis approach towards cyclic decameric peptides.^[52] No NMR studies or biological activities were reported though.

Additionally, phenylalanine derivative 14^[53] was only synthesised as demonstration of synthesis efficiency and flexibility.

1.3.1.2ORNITHINE

The role played by ornithine (Orn) in the membrane activity is well established.^[12] The partly positive charges of the side chain amine are electrostatically attracted to negatively charged biomolecules on the surface of the cell membrane, like phospholipids and teichoic

NH O NH2

NH O HN H

O

NH O NHAc

NH O HN

O

n

NH O COOH

1 +++ 15 --/+â 16 - 17 --- 18 --/+â 19 ---/-â

NH O HN

O HOOC

NH O NH2

NH O HN N

20 --/-^a 21 ++ 22 ++ 23 -/+++^a 24 ++ 25 +++

NH O HN

O NH2

HO

NH O HN

O NH2

HOOC

NH O HN

O NH

NH O HN

O NH2

NH2

26 +++ 27 +++ 28 - 29 +++ 30 + 31 +++

NH O HN

O NH2

NH2

NH O HN

O NH2

HN N

32 ++ 33 +++ 34 ---/-^a

Table 2 Substitutions for ornithine (part 1). Behind the compound number the antibacterial activity is given (+++ means highest activity, --- means no activity).

(17)

[15]

acids. It is not a surprise then, that when the positive charge is removed the activity of GS drops dramatically. Formylation (15),^[54] acetylation (16)^[55-58] or acylation (17)^[58] with longer carbon chains rendered GS devoid of any membrane activity (Table 2).^[57,59] Substituting one ornithine for a serine (18) diminished the biological activity by about a half,^[60] while the symmetric serine substituted derivative lost nearly all activity.^[61,62] Negatively charged side chains, as in 6 (symmetric and asymmetric)^[60] and 20 (symmetric and asymmetric)^[63] gave comparable results. Although the diaminobutanoic acid- (21),^[64,65] lysine- (22)[27,41,43,53,64-68] and histidine (22)^[69] derivatives were all active, they did not outperform GS in terms of eliminating bacteria. Monomethylation of the δ-amine of Orn resulted in a peptide (24) that was slightly less active against B. subtilis, but slightly more active against S. aureus.^[70] In the same article Yamada et al. described the synthesis of GS derivatives in which the ornithine side chains were covalently bridged by alkyl spacer through the δ-amines. A propylene spacer proved to be optimal, as the activity was completely retained, compared to GS. When the spacer was lengthened to butyl, or pentyl the activity dropped. Trimethylation, thereby yielding the diammonium species 25,^[56,71-73] resulting in a permanent positive charge, did not alter the activity of the molecule as compared to GS. Izumiya and co-workers^[63] and Yagi et al.^[74] described a series of compounds in which the side chains of ornithine were acylated with unprotected amino acids, resulting in GS derivatives with two (26-29) or more (30-33) positive charges. All of these compounds showed good to moderate activity. Inversion of the chemistry from L-ornithine (1) to D-lysine (34) made the antibacterial activity drop 10-fold for the asymmetric derivative and resulted in complete loss of membrane activity when both stereocentres were inverted.

Several other ornithine substitutions were reported as well (Table 3). For these no antibacterial activities were reported, as they were synthesised for different purposes. Arginine derivative 35 was used to probe the activity of Penicillinase of B. cereus.^[75,76]

Dinitrophenylated ornithine derivative 36^[77] found use in the generation of antibodies against the dinitrophenyl hapten. Besides, 36 was also used in ORD (optical rotatory dispersion) studies to elucidate the structure of GS,^[78] which was not certain at the time. Nitrobenzoylated

(18)

[16]

GS 37 was prepared to be treated with hydrogen and palladium on charcoal, followed by K3FeCN6 to produce an actinocyl-GS derivative, which was not able to inhibit RNA synthesis or show anti-tumour activity.^[79] Cysteine derivative 38 found application in CD (circular dichroism) studies in order to study the stereochemistry of disulfide bonds.^[80-82]

Diphthaloylated-GS 39 was used as an NMR marker to investigate the secondary structure of GS.^[83] Spin-labelled GS 40^[84] was used in EPR (electron paramagnetic resonance) studies while GS with two N-(4-nitrobenz-2-oxa-1,3-diazole)-moieties (41)^[85] was used to probe the lateral diffusion of the peptide in lipid multibilayers and GS with two acryloyl handles 42 was

NH O HN

NH2

NH

NH O HN

NO2

NH O N O

O

35 36 37 s r39 39

NH O HN

O

N O NHBoc

NH O

40 41 42 43 44

NH O HN

NH2

O

NH O HN

O

HgOH

NH O N

HO R1

R2

NH O N

N

NH O HN

N O

45 46 47a: R1 = R2 = H 47b: R1 = Me; R2 =

CH2OH

48 49

NH O

N N

2

NH O

HN Porphyrin O

NH O Ph2P

S

50 51 52 53 54

55

Table 3 Substitutions for ornithine (part 2).

(19)

[17]

used in a polymeric sorbent for proteinases.^[86] Kawai and co-workers^[87] were able to crystallise GS derivative 43, confirming its β-hairpin structure by Röntgen diffraction. The mono-alanine substitution (44) was made for synthetic reasons only^[38] and citrulline-GS (45) was identified as a minor constituent of GS isolated from Bacillis brevis.^[41]

The chiral environment of GS has been recognised as a suitable ligand for heterogeneous chiral catalysis. Several GS-metal complexes have appeared in the past decades. Mercury complex 46^[88] and the copper complexes of 47a^[89,90] and 47b^[91] were prepared however for spectroscopic purposes and elucidation of the secondary GS structure respectively. Bipyridyl- GS 48 was complexed to zinc(II), cobalt(II), nickel(II) and copper(II), without significant alteration in the overall peptide structure, with the promise to be used as artificial metalloprotein.^[92] Complexes with divalent zinc were prepared with 49,^[93] 50^[94] and 51^[95,96] as ligands. Zinc complex of 50 was shown to catalyse the hydrolysis of phosphodiester bonds 6500 times, while for the other two complexes no catalysis was reported. Potential phosphine based GS ligands 52-54 were reported by Lammertsma and co-workers,^[97] but the phosphines were left protected with sulphur and metal complexes were not reported. Finally carborane- GS complex 55^[98] was synthesised as a possible compound in boron neutron capture therapy.

1.3.1.3L^EUCINE

Relatively little attention has been paid to modifications of the leucine residues, possibly because the hydrophobicity of this residue appears to be vital for the activity of GS. The validity of this assumption, became apparent after the preparation of glycine and alanine derivatives 56^[99] and 57 (Table 4).^[38,99] Derivative 56 had lost all antibacterial activity and the

NH O

NH O CF3

1 +++ 56 --- 57 - 58 --- 59 60 + 61

Table 4 Substitutions for leucine. Behind the compound number the antibacterial activity is given (+++ means highest activity, --- means no activity).

(20)

[18]

activity of 57 was reduced to half the activity of GS, illustrating the importance of sufficient hydrophobicity at this position. Introduction of an aminomyristic acid residue with a long fatty tail (58) depleted the peptides of all activity against B. subtilis, S. aureus and E. coli.^[100]

The hydrophobic pyrenylalanine was introduced at the leucine position (59) to probe the antiparallel β-sheet structure in cyclic peptides by CD spectroscopy, but no antibacterial activity of this derivative was reported.^[101] As was the case for valine, this position has been used to introduce fluorine containing amino acids (60^[51,102] and 61^[52]). Interestingly introduction of the 4-F-phenylglycine (60) at the leucine position did not interfere with the antibacterial activity of the cyclic peptide and 60 was as active as GS towards several bacterial strains. The viability of the substitution towards maintenance of the structure was shown as well. Both fluorine containing GS derivatives 60 and 61 may be used in ¹⁹F NMR spectroscopy in order to probe the mechanism of action the cationic cyclic peptide towards biomembranes.

1.3.2TURN

The ^DPhe-Pro dipeptide is very important in view of the overall conformation of gramicidin S. Before the establishment of crystal structures of GS,^[103] Crowfoot Hodgkin and Oughton^[104] proposed a secondary structure for the cyclic decapeptide in which ^DPhe-Pro adopted a turn-structure. As such, these two residues have been an interesting target for introducing modifications and correlate these with changes in the overall conformation of the peptide.

1.3.2.1 D-PHENYLALANINE

The D-amino acid in the turn is without doubt important for the overall structure of gramicidin S. Substitution by an ^L-amino acid disturbs the structure of the peptide, as demonstrated by the incorporation of ^LPhe (76)[47,105-109] and ^LAla (77)^[109-111] (Table 5). For peptide 76 it was found that most activity against Gram-positive bacteria was lost.

Interestingly, the activity against E. coli was slightly higher than GS. For the alanine derivative all activity was lost. Using α,α-dimethyl amino acid α-aminoisobutyric acid 78^[110] or glycine

(21)

[19]

(79),[65,106,109-112] yielded peptides that lost both structure and activity. The structure was maintained by using (Z)-ΔPhe (80)^[59] or (Z)-ΔAla (81)^[113] and in the case 80 antibacterial activities were obtained that resembled those of GS, while the activity for 81 was reduced to about a half. An N-methylated ^D-phenylalanine residue (82)^[114] was used to aid in the residue assignment of the ¹⁵N NMR spectra of GS, but no activity or structure were reported. Several derivatives with aliphatic D-amino acids have been synthesised, namely Ala (83),[38,53,65-67,110,111]

Val (84),^[115] Leu (85)[67,113,115] and Cha (cyclo-hexylalanine, 86).[66,67,110,116,117] Alanine derivative 83 was only very weakly active against bacteria, while larger aliphatic analogues 84 and 85

NH O

1 +++ 76 77 --- 78 --- 79 -- 80 +++

N O

NH O

81 + 82 83 -- 84 ++ 85 ++ 86 -

NH O H2N

O

87 --- 88 -- 89 -- 90 + 91 92 ++

NH O F

NH O Cl

NH O S

NH O

93 94 95 +++ 96 97 98

NH O

N 99 100 +++ 101 ++ 102 ++ 103 ++ 104 + O

NH O H2N

NH O HN N

NH O N

105 - 106 - 107 108 --- 109 ++/+++^a

Table 5 Substitutions for D-phenylalanine. Behind the compound number the antibacterial activity is given (+++ means highest activity, --- means no activity).

(22)

[20]

were active against Gram-positive strains.^[110] The most bulky aliphatic amino acid cyclo- hexylalanine yielded a peptide (86) that was considerably less active than gramicidin S.^[110]

DAsp derivative 87 was synthesised by Wishart et al.^[43] to validate their synthetic approach towards GS and its derivatives, but no activity of this peptide is known. Octanol-water partition was determined of ^DSer derivative 88.^[43,67] The biological activity of 88 was determined to be rather low,^[66,118] but could be fully restored by benzylation of the alcohol.^[118]

This same trend was observed for ^D-tyrosine: bacteria were not very sensitive towards the action of the derivative with the free phenol 89,[66,67,119-122] while benzylation to give 90,^[121]

delivered a peptide with a comparable antibacterial activity profile as GS. Peptide 89 appeared to be half as hemolytic as GS itself.^[122] Of the many other para-substituted derivatives reported (91,^[88,123] 92 and 93^[120,124]) only of 92 was reported that the activity was slightly diminished.^[124]

In addition, Aarstad et al.^[120] applied a series of amino acid (86, 89, 92-98) in the biosynthesis of GS derivatives by gramicidin synthase to determine the efficiency of incorporation. It was found that tyrosine (89) and thiophenalanine (96) were incorporated most efficiently (60 and 58% respectively). D-Pyrenylalanine-GS (99) was synthesised by Izumiya and co-workers^[125]

by a novel cyclisation-cleavage solid-phase synthesis method and proposed to be an interesting compound to probe the activity of GS and its derivatives, but the actual antibacterial activity was never assessed. In 2009 Andreu, Cativiela and co-workers published an article in which several aromatic ^D-amino acids were incorporated into GS (100-104) and tested for their activity against bacteria and leukocytes.^[36] All derivatives appeared to be active against Gram-positive bacteria in the low micromolar range, but bicyclic amino acid residue 104 delivered a peptide that was significantly less hemolytic. Finally basic residues were also

introduced at the ^DPhe position. Incorporation of ^DTrp (95)[43,120,122] produced a peptide that was extremely active against both Gram-positive bacteria and red blood cells, but hardly against Gram-negative cells. The D-diaminopropionic acid residue (105)[109,113,126] destabilised the overall structure of GS and the activity was diminished concomitantly, but was restored when the amine was protected with a Cbz-group.^[113] Introducing more distance between the positively charged amine and the backbone, as with ^DLys (106),^[127] gave similar results with

(23)

[21]

respect to activity. For histidine derivative 107, which combines aromaticity and hydrophilicity, biological data were not reported,^[43] but pyridylalanine derivatives 108 and 109 compared well with GS with respect to their antibacterial activity, but with diminished toxicity.^[35]

1.3.2.2PROLINE

Table 6 presents the substitutions that have appeared in literature over the past five decades with respect to the proline residue. The type II̕ β-turn is often built up from a D-Xaa-Pro sequence, but it is known that instead of proline an N-alkylated amino acid will also result in the desired turn. One of the first modification of proline reported is substitution by sarcosine (N-methylglycine, 110). This modification appears to yield a cyclic decapeptide with a structure comparable to GS, but slightly less rigid.^[128] The antibacterial activity was slightly

NH O

1 +++ 110 ++ 111 +/+^a 112 --

NH O 113 +++

114 ++

114 ++ 115 – 116a: S ++

116b: R +++

N O HN

O H2N

117a: S --- 17b: R --

118a: S -- 118b: R ---

119a: S -- 119b: R ---

120a: S + 120b: R +++

N O HN

O H2N

121a: S +++

id=": R ++

122a: S +++

122b: R ++

123

Table 6 Substitutions for proline. Behind the compound number the antibacterial activity is given (+++ means highest activity, --- means no activity).

(24)

[22]

lower, but in the range of GS itself.^[129] Many papers have appeared in which glycine was used to substitute proline (111). Some of these papers deal with the synthesis of the GS analogues only.^[130-133] For instance, Matsuura et al.^[134] prepared both symmetric and asymmetric glycine modified GS. These modifications yielded analogues with significantly less ordered structure in ethanol, but no appreciable reduction in antibacterial activity against B. subtilis^[135,136] and S.

aureus^[134] was observed. Structurally interesting as well is the substitution with β-alanine (112),^[65,137] which contains an additional methylene in its backbone. Both activity and structure were lost upon incorporation of this residue, which indicates distortion of the turn and consequently of the amphiphilic nature of GS. Attachment of a fluorine atom to the pyrrolidine ring of proline yielded a compound (113)^[138] that was even more active against Baccilus subtilis than GS. Unfortunately the stereochemistry of the fluorine substituent was not indicated. Somewhat surprising is the retention of structure and activity upon substituting proline for leucine (114),^[139] as destabilisation of the β-turn and disturbance of the amphiphilicity of the molecule was predicted. The expected drop in activity is observed when the stereochemistry proline is inverted (115).^[140] The low activity was ascribed to the change in conformation upon epimerisation of the proline residue. Independently Grotenbreg et al.^[121] and Kawai and co-workers^[141,142] made use of 4-azidoproline. Grotenbreg demonstrated that this modification of proline in GS (116a,b) did not alter the antibacterial profile, for either diastereomer. Upon reduction however to the amine (117a,b), the antibacterial activity was completely lost and activity was not restored by Cbz-protection of the amine (118ab), or by acylation with succinic anhydride (119a,b). Acylating the amine of 117a or 117b with an amino acid (^DPhe (120a,b), ^LPhe (121ab), or Lys (122a,b)) delivered GS derivatives that recovered their antibacterial activity. Especially the peptides with R configuration at the aminoproline centre proved to be active against several Gram-positive bacterial strains. The positive charge(s) on the amino acids appear to aid in the amphiphilic structure as they segregate on the side of the ornithines. Alanine derivative (123) was synthesised only to prove the flexibility of a new on resin cyclisation procedure towards GS and its derivatives.^[38]

(25)

[23]

1.4 OUTLINE OF THE THESIS

The next three Chapters deal with gramicidin S derivatives containing α-amino acids only.

Chapter 2 describes the synthesis of a GS derivative that has been asymmetrically

functionalised with a nitro-functionality in the turn. The nitro group were reduced and several functional groups were attached, such as aromatic and aliphatic acyl groups (Scheme 1). The peptides obtained were assessed for their structure by NMR and X-ray structures were obtained from two derivatives. Besides, the new molecules were tested for their membrane disrupting activity against several bacterial strains and against red blood cells.

In Chapter 3 it is described how the triazole moiety may be used in the turn to mimic the aromatic ring of phenylalanine. Earlier studies have shown that both hydrophobicity and aromaticity influence the biological activity of GS. Therefore a series of GS-triazoles with hydrophobic, aromatic substituents were prepared and analysed for their structure by NMR and for their activity against bacterial membranes. The key molecule in this Chapter is the

Scheme 1 Nitro-GS derivative described in Chapter 2 and its functionalisation.

NH HN

NH O

O

OH N N H HN NH

O

BocHN O O O

N O

NHBoc

HN

O N

O

NO2

1. Reduction 2. Functionalisation

3. Deprotection 12 GS derivatives

124

Scheme 2 Synthetic approach applied in Chapter 3 to obtain various triazoles with different substituents and regiochemistry.

(26)

[24]

alkynyl derivative (Scheme 2, 125), which was synthesised on large scale by a combination of solid-phase and solution phase chemistry. Peptide 125 was subjected to two different catalytic 1,3-dipolar cycloaddition procedures to give the triazoles, expected to have different regiochemistry with respect to the triazole ring.

Chapter 4 focused on three different known antibacterial cyclic peptides with various ring

sizes (Scheme 3), namely one with 8 amino acids in the ring (126), one with ten (GS, 1) and one with twelve (gratisin, 127). Several hybrid analogues were prepared that featured parts of the three different peptides. These peptides were analysed for their structure and antibacterial activity. To prepare the peptides with ring sizes varying from eight to twelve amino acids, a synthetic approach was developed in which all steps, including the cyclisation reaction, were performed on the solid support.

The last three Chapters of this Thesis deal with β-amino acids and their incorporation in Scheme 3 Two starting structures for the design of new amphiphilic cyclic peptides, as presented in Chapter 4.

Scheme 4 Synthetic scheme towards α-substituted β-amino acids, as described in Chapter 5.

O N

O

Bn R

O TiCl4,DiPEA

O N

O

Bn R

O

Bn N Bn N

Bn Bn

EtSLi R SEt

O

Bn N Bn

R OH O

Bn N Bn Hg(CF3COO)2

H2O

H2,Pd/C

R O O

H3N FmocOSu

NaHCO3

O OH R FmocHN

128 129 130 131 132

133 134

(27)

[25]

cyclic peptides. First the necessary β²-amino acids were synthesised and Chapter 5 presents a new approach towards their preparation. The acylated Evans’ chiral oxazolidinone 128 was engaged in the synthesis, which was transformed into its titanium enolate and reacted with dibenzyliminium ion 129. Engagement of the lithium salt of ethanethiol to remove the chiral auxiliary from 130 proved to work smoothly. The complete synthetic scheme is outlined in Scheme 4.

Chapter 6 shows the application of the β²-amino acids that were outlined in Chapter 5.

Both β²- and β³-amino acids were incorporated in the strand of GS (Scheme 5), thus making a combination of α- and β-amino acids in a β-strand (135). Various NMR techniques were used to probe the secondary structure of the synthesised peptides. Additionally the crystal structure of one of the peptides is presented in this Chapter. The peptides were designed in such a way that they would be amphiphilic and the new peptides were therefore assessed for their

Scheme 5 Target molecules in of Chapter 6. Here a combination of α- and β-amino acids is made within a cyclic peptide.

Scheme 6 Design of the peptides that were prepared and evaluated in Chapter 7.

NH N

H

HN H

N O

O O

O HN HN

NH NH

O

O O

O N

N O

O

NH2

H2N 136 R1

R1

R2

(28)

[26]

biological activity.

Finally, Chapter 7 elaborates further on this theme, by inversion of the stereochemistry of the ornithines in 135. In β-strands entirely composed of α-amino acids it would be impossible to incorporate D-amino acids, due to steric hindrance between the nearby amino acids.

However, due to the incorporation of β-amino acids it was possible to shift the side chains adjacent to the ornithine (Scheme 6, 136), thereby avoiding a steric clash between the side chains. NMR was used to gain insight into the secondary structure and the obtained peptidic compounds were tested against a range of bacteria to probe their membrane activity.

1.5REFERENCES

[1] R.E.W. Hancock, Lancet 1997, 349, 418-422.

[2] R.E.W. Hancock, H.-G. Sahl, Nat. Biotechnol. 2006, 24, 1551-1557.

[3] C. Walsh, Antibiotics: Action, Origins, Resistance, ASM Press, Washington, D.C., 2003.

[4] R.E.W. Hancock, Lancet Infect. Dis. 2001, 1, 156-164.

[5] B. Bechinger, Crit. Rev. Plant Sci. 2004, 23, 271-292.

[6] K.A. Brogden, Nat. Rev. Microbiol. 2005, 3, 238-250.

[7] Z. Oren, Y. Shai, Biopolymers 1998, 47, 451-463.

[8] Y. Park, K.S. Hahm, J. Biochem. Mol. Biol. 2005, 38, 507-516.

[9] Y. Shai, Biochim. Biophys. Acta 1999, 1462, 55-70.

[10] E. Katz, A.L. Demain, Bacteriol. Rev. 1977, 41, 449-474.

[11] T. Kato, N. Izumiya, in Chemistry and Biochemistry of Amino Acids, Peptides, and Proteins, Vol.

2 (Ed.: B. Weinstein), Marcel Dekker, Inc., New York, 1974, pp 1-31.

[12] N. Izumiya, T. Kato, H. Aoyagi, M. Waki, M. Kondo, Synthetic Aspects of Biologically Active Cyclic Peptides: Gramicidin S and Tyrocidins, Halsted, New York, 1979.

[13] G.F. Gause, M.G. Brazhnikova, Nature 1944, 154, 703.

[14] R.L.M. Synge, Biochem. J. 1945, 39, 363-367.

[15] C.J. Wadsten, C.A. Bertilsson, H. Sieradzki, S. Edstrom, Arch. Otorhinolaryngol. 1985, 242, 135- 139.

[16] A.S. Bourinbaiar, C.F. Coleman, Arch. Virol. 1997, 142, 2225-2235.

[17] J.P. Bradshaw, Biodrugs 2003, 17, 233-240.

[18] T. Mogi, K. Kita, Cell. Mol. Life Sci. 2009, 66, 3821-3826.

[19] T. Mogi, H. Ui, K. Shiomi, S. Omura, K. Kita, FEBS Lett. 2008, 582, 2299-2302.

[20] A.K. Marr, W.J. Gooderham, R.E.W. Hancock, Curr. Opin. Pharmacol. 2006, 6, 468-472.

[21] T. Kurotsu, M.A. Marahiel, K.-D. Müller, H. Kleinkauf, J. Bacteriol. 1982, 151, 1466-1472.

[22] T. Maeda, M. Takagi, T. Imanaka, J. Ferment. Bioeng. 1993, 75, 173-177.

[23] M. Yukioka, Y. Saito, S. Otani, J. Biochem. 1966, 60, 295-302.

[24] M.A. Marahiel, T. Stachelhaus, H.D. Mootz, Chem. Rev. 1997, 97, 2651-2673.

[25] A.C. Bach, J.A. Markwalder, W.C. Ripka, Int. J. Peptide Protein Res. 1991, 38, 314-323.

(29)

[27]

[26] W.C. Ripka, G.V. De Lucca, A.C. Bach, R.S. Pottorf, J.M. Blaney, Tetrahedron 1993, 49, 3609- 3628.

[27] D. Andreu, S. Ruiz, C. Carreño, J. Alsina, F. Albericio, M.Á. Jiménez, N. de la Figuera, R.

Herranz, M.T. García-Lopez, R. González-Muñiz, J. Am. Chem. Soc. 1997, 119, 10579-10586.

[28] G.M. Grotenbreg, A.E.M. Buizert, A.L. Llamas-Saiz, E. Spalburg, P.A.V. van Hooft, A.J. de Neeling, D. Noort, M.J. van Raaij, G.A. van der Marel, H.S. Overkleeft, M. Overhand, J. Am.

Chem. Soc. 2006, 128, 7559-7565.

[29] G.M. Grotenbreg, A.E. Christina, A.E.M. Buizert, G.A. van der Marel, H.S. Overkleeft, M.

Overhand, J. Org. Chem. 2004, 69, 8331-8339.

[30] G.M. Grotenbreg, M. Kronemeijer, M.S.M. Timmer, F. El Oualid, R.M. van Well, M. Verdoes, E.

Spalburg, P.A.V. van Hooft, A.J. de Neeling, D. Noort, J.H. van Boom, G.A. van der Marel, H.S.

Overkleeft, M. Overhand, J. Org. Chem. 2004, 69, 7851-7859.

[31] G.M. Grotenbreg, M.S.M. Timmer, A.L. Llamas-Saiz, M. Verdoes, G.A. van der Marel, M.J. van Raaij, H.S. Overkleeft, M. Overhand, J. Am. Chem. Soc. 2004, 126, 3444-3446.

[32] G.M. Grotenbreg, M.D. Witte, P.A.V. van Hooft, E. Spalburg, P. Reiss, D. Noort, A.J. de Neeling, U. Koert, G.A. van der Marel, H.S. Overkleeft, M. Overhand, Org. Biomol. Chem. 2005, 3, 233-238.

[33] J. Xiao, B. Weisblum, P. Wipf, J. Am. Chem. Soc. 2005, 127, 5742-5743.

[34] J. Xiao, B. Weisblum, P. Wipf, Org. Lett. 2006, 8, 4731-4734.

[35] K. Yamada, S. Shinoda, H. Oku, K. Komagoe, T. Katsu, R. Katakai, J. Med. Chem. 2006, 49, 7592-7595.

[36] C. Solanas, B.G. de la Torre, M. Fernández-Reyes, C.M. Santiveri, M. Ángeles Jimenez, L. Rivas, A.I. Jiménez, D. Andreu, C. Cativiela, J. Med. Chem. 2009, 52, 664-674.

[37] M. Kondo, N. Izumiya, Bull. Chem. Soc. Jpn. 1967, 40, 1975-1980.

[38] X.Z. Bu, X.M. Wu, N.L.J. Ng, C.K. Mak, C.G. Qin, Z.H. Guo, J. Org. Chem. 2004, 69, 2681-2685.

[39] M. Kondo, N. Izumiya, Bull. Chem. Soc. Jpn. 1970, 43, 1850-1854.

[40] M. Altmann, H. Koischwitz, J. Salnikow, H. Kleinkauf, FEBS Lett. 1978, 93, 247-250.

[41] P. Thibault, D. Faubert, S. Karunanithy, R.K. Boyd, C.F.B. Holmes, Biol. Mass Spectrom. 1992, 21, 367-379.

[42] K. Sato, K. Ueda, M. Kondo, H. Aoyagi, N. Izumiya, Bull. Chem. Soc. Jpn. 1978, 51, 1830-1834.

[43] D.S. Wishart, L.H. Kondejewski, P.D. Semchuk, B.D. Sykes, R.S. Hodges, Lett. Peptide Sci. 1996, 3, 53-60.

[44] S. Nozaki, I. Muramatsu, J. Antibiot. 1984, 37, 689-690.

[45] S. Nozaki, I. Muramatsu, Bull. Chem. Soc. Jpn. 1985, 58, 331-335.

[46] M. Tamaki, T. Okitsu, M. Araki, H. Sakamoto, M. Takimoto, I. Muramatsu, Bull. Chem. Soc.

Jpn. 1985, 58, 531-535.

[47] M. Tamaki, M. Takimoto, S. Nozaki, I. Muramatsu, J. Chromatogr. 1987, 413, 287-292.

[48] M. Kondo, M. Kimura, K. Sato, H. Horimoto, Bull. Chem. Soc. Jpn. 1987, 60, 1391-1397.

[49] Y. Soejima, A. Hashiguchi, N. Izumiya, Biosci. Biotech. Biochem. 1994, 58, 826-829.

[50] T. Arai, T. Imachi, T. Kato, H.I. Ogawa, T. Fujimoto, N. Nishino, Bull. Chem. Soc. Jpn. 1996, 69, 1383-1389.

[51] S. Afonin, R.W. Glaser, M. Berditchevskaia, P. Wadhwani, K.-H. Guhrs, U. Möllmann, A.

Perner, A.S. Ulrich, ChemBioChem 2003, 4, 1151-1163.

[52] P. Wadhwani, S. Afonin, M. Ieromino, J. Buerck, A.S. Ulrich, J. Org. Chem. 2006, 71, 55-61.

[53] X. Wu, X. Bu, K.M. Wong, W. Yan, Z. Guo, Org. Lett. 2003, 5, 1749-1752.

[54] B.R. Srinivasa, J. Antibiot. 1982, 35, 571-573.

(30)

[28]

[55] Y.A. Ovchinnikov, V.T. Ivanov, V.F. Bystrov, A.I. Miroshnikov, E.N. Shepel, N.D. Abdullaev, E.S. Efremov, L.B. Senyavina, Biochem. Biophys. Res. Commun. 1970, 39, 217-225.

[56] E.M. Krauss, S.I. Chan, Biochemistry 1984, 23, 73-77.

[57] T. Katsu, H. Kobayashi, T. Hirota, Y. Fujita, K. Sato, U. Nagai, Biochim. Biophys. Acta 1987, 899, 159-170.

[58] Y. Shimohigashi, S. Ono, H. Sakamoto, H. Yoshitomi, M. Waki, M. Ohno, Chem. Lett. 1993, 671-672.

[59] Y. Shimohigashi, H. Kodama, S. Imazu, H. Horimoto, K. Sakaguchi, M. Waki, H. Uchida, M.

Kondo, T. Kato, N. Izumiya, FEBS Lett. 1987, 222, 251-255.

[60] M. Tamaki, S. Akabori, I. Muramatsu, Int. J. Peptide Protein Res. 1996, 47, 369-375.

[61] M. Tamaki, M. Takimoto, S. Sofuku, I. Muramatsu, J. Antibiot. 1980, 33, 105-106.

[62] W. Danders, M.A. Marahiel, M. Krause, N. Kosui, T. Kato, N. Izumiya, H. Kleinkauf, Antimicrob. Agents Chemother. 1982, 22, 785-790.

[63] H. Yonezawa, K. Tomokiyo, N. Izumiya, Rep. Fac. Sci., Kagoshima Univ. (Math., Phys. & Chem.) 1987, 71-76.

[64] M. Waki, O. Abe, R. Okawa, T. Kato, S. Makisumi, N. Izumiya, Bull. Chem. Soc. Jpn. 1967, 40, 2904-2909.

[65] T. Kato, M. Waki, S. Matsuura, N. Izumiya, J. Biochem. 1970, 68, 751-753.

[66] T. Katayama, K. Nakao, M. Akamatsu, T. Ueno, T. Fujita, J. Pharm. Sci. 1994, 83, 1357-1362.

[67] P. Buchwald, N. Bodor, Proteins 1998, 30, 86-99.

[68] J. Alsina, F. Rabanal, C. Chiva, E. Giralt, F. Albericio, Tetrahedron 1998, 54, 10125-10152.

[69] K. Nonaka, K. Sato, S. Terada, T. Kato, N. Izumiya, Bull. Chem. Soc. Jpn. 1978, 51, 2127-2130.

[70] K. Yamada, K. Ando, Y. Takahashi, Y. Oda, H. Yamamura, S. Araki, K. Kobayashi, R. Katakai, F.

Kato, M. Kawai, J. Peptide Res. 1999, 53, 611-617.

[71] E.M. Krauss, S.I. Chan, J. Am. Chem. Soc. 1982, 104, 6953-6961.

[72] M. Kawai, N. Fukuta, N. Ito, M. Ohya, Y. Butsugan, M. Maruyama, Y. Kudo, J. Chem. Soc., Chem. Commun. 1986, 955-956.

[73] M. Kawai, M. Ohya, N. Fukuta, Y. Butsugan, K. Saito, Bull. Chem. Soc. Jpn. 1991, 64, 35-41.

[74] Y. Yagi, S. Kimura, Y. Imanishi, Int. J. Peptide Protein Res. 1990, 36, 18-25.

[75] A.K. Saz, D.L. Lowery, Biochem. Biophys. Res. Commun. 1965, 19, 102-107.

[76] A.K. Saz, D.L. Lowery, Biochem. Biophys. Res. Commun. 1964, 16, 525-529.

[77] P.C. Montgomery, J.H. Rockey, A.R. Williamson, Proc. Natl. Acad. Sci. USA 1972, 69, 228-232.

[78] U. Nagai, M. Kurumi, T. Umemura, Tetrahedron 1972, 28, 4959-4964.

[79] A.B. Mauger, R. Wade, J. Chem. Soc. C 1966, 1406-1408.

[80] U. Ludescher, R. Schwyzer, Helv. Chim. Acta 1972, 55, 2052-2060.

[81] U. Ludescher, R. Schwyzer, Helv. Chim. Acta 1971, 54, 1637-1644.

[82] R. Schwyzer, Angew. Chem. 1972, 84, 902-903.

[83] R. Schwyzer, U. Ludescher, Biochemistry 1968, 7, 2519-2522.

[84] V.T. Ivanov, A.I. Miroshnikov, L.G. Snezhkova, Y.A. Ovchinnikov, A.V. Kulikov, G.I.

Likhtenstein, Khim. Prir. Soedin. 1973, 91-98.

[85] E.-S. Wu, K. Jacobson, F. Szoka, A. Portis, Biochemistry 1978, 17, 5543-5550.

[86] G.N. Balandina, G.I. Lavrenova, L.D. Uzhinova, E.A. Chekhovskikh, N.A. Platé, V.M. Stepanov, Khim. Prir. Soedin. 1986, 215-218.

[87] K. Yamada, M. Unno, K. Kobayashi, H. Oku, H. Yamamura, S. Araki, H. Matsumoto, R.

Katakai, M. Kawai, J. Am. Chem. Soc. 2002, 124, 12684-12688.

Site-selective incorporation of alpha- and beta-amino acid derivatives : towards new gramicidin S-based bactericides

towards new gramicidin S-based bactericides

α- and β-Amino Acid Derivatives:

Towards New Gramicidin S-Based Bactericides

Mat th ij s v an d er K n aa p

Site-Selective Incorporation of α- and β- Amino Acid Derivatives: Towards New

Gramicidin S-Based Bactericides

Table of Contents

General Introduction