Antibiotic-AMP conjugates are synthetized by anchoring conventional antibiotic to an AMP or CPP through a suitable bifunctional linker. The peptide has two points of attachment, the N-terminus or the C- terminus, even if the selective use of a side chain could be, in theory, taken into consideration. In general, the linker can be classified as a stable covalent linker or cleavable stimuli-responsive linker. With cleavable stimuli-responsive linkers the AMP and the antibiotic could each act independently upon entering the bacterial cell, targeting their respective sites. Conversely, if the conjugate molecules remain intact, they function as a single, multimodal antibacterial compound. This allows them to bind to and affect their targets simultaneously, with dynamics that may differ from those of the individual components. Probably, the most problematic task for the synthesis of these conjugates is to find the right point of attachment on the antibiotic, since they usually possess many reactive functional groups in their scaffolds. Accordingly, the following sections of this chapter are organized considering the antibiotic scaffold.
4.1. Vanconomycin-AMP Conjugates
Vancomycin, exhibiting one of the strongest bindings known for low-molecular-weight organic compounds with the D-Ala-D-Ala motif of the cell wall precursor lipid II, was initially considered the drug devoted to treat antibiotic resistant bacteria being immune to the development of resistance [
65,
66,
67]. However, 30 years after its discovery different vancomycin-resistant strains, such as
Enterococcus faecium (VRE), vancomycin-intermediate and resistant
Staphylococcus aureus (VISA and VRSA) have been observed for which the discovery and development of novel antibiotics are urgently needed. In this contest, since clinical resistance to vancomycin took a lot of time to arise, the modification of this glycopeptide could be a successful strategy. From a synthetic chemical perspective, the selective functionalization of vancomycin could seem very difficult due to the presence of many functional group. However, it has been shown that there are for functional groups, referred to as point of attachments, which can be exploited for the selective functionalization due to their unicity or particular reactivity (
Figure 2). First, the carboxylic acid at the C-terminus of the peptide sequence is the only carboxy functional group present in the vancomycin scaffold, thus can be selectively reacted with amines upon worthy activation. Also the amino functional group at the N-terminus can be selectively coupled with activated carboxylic acids but only when it is not methylated (R = H, norvancomycin). Indeed, in vancomycin (R = Me) there is a more reactive amino group in the glycosyl moiety (vancosamine) that is more reactive in the presence of activated carboxylic acids because less sterically congested. Finally, the fourth point of attachment is the resorcinol aromatic carbon in ortho position of the two hydroxy groups that, being very electron rich, readily undergo electrophilic aromatic substitution with iminium salts in situ produced by reaction of formaldehyde and primary amines.
One way to face vancomycin-resistant strains is to modulate the structure of the drug to increase membrane binding and selectivity eventually enhancing drug concentration at the target site [
94]. Accordingly, a library of vancomycin derivatives, referred to as vancapticins
1, was designed by coupling the free carboxylic acid on the glycopeptide with different cationic peptides, mostly polilysines, having distinct lipophilic membrane-insertive elements (MIE) tethered at the N-terminus through two linkers, one of them built on a cleavable disulfide bond (
Figure 3). Structure-activity relationship studies (SAR) revealed that vancapticins
1 possess enhanced membrane affinity, which boosts their effectiveness against MRSA and various other Gram-positive bacteria. Additionally, vancapticins
1 retain their potency against strains that are resistant to traditional glycopeptides.
The same strategy, namely tethering polycationic peptides to vancomycin to fight antibacterial resistance, have been exploited for the synthesis of two vancomycin-polyarginine conjugates [
95]. Exploiting again the reactivity of the free carboxylic acid, vancomycin was tethered to the N-terminus of D-octaarginine (r8) through not cleavable aminohexanoic acid (Ahx) linker obtaining the conjugate
2 (
Scheme 1A) which could have a stronger affinity for the surface of cell membrane and enhanced cell permeability facilitating the action of vancomycin in arresting the cell wall synthesis and giving to vancomycin the access to intracellular binding targets. Octaarginine R8
4 was prepared trough solid phase peptide synthesis (SPPS) and coupled with Cbz-Ahx-OH
3 in solution leading to the formation of r8-Ahx-r8
5 after hydrogenolysis of the Cbz protecting group (
Scheme 1B). Intermediate
5 was finally coupled to vancomycin producing the final conjugate vacomycin-Ahx-r8
2, which resulted to be much more active than vancomycin by orders of magnitude against difficult-to-treat MRSA populations, such as biofilms and persister cell, maintaining comparable minimal inhibitory concentration (MIC) against vancomycin-resistant Gram-positive organisms such as VISA and VRE.
Following the same rational, different vancomycin-polyarginine conjugates were synthetized at four distinct point of attachments, namely the free carboxylic acid (V
C), the carbon in orto position to the hydroxy groups of the resorcinarene ring (V
R), the N-methylammino function on the leucine residue (V
N), and the free amino function on the glycosyl frame (V
V) (in
Scheme 2A the structure of the most performant V
N derivative FU002
6 is represented) [
96]. Apart from the site of attachment, the structures of the conjugates are very similar, being composed of the vancomycin antibiotic
7, a heterobifunctional cross linker and hexaarginine tagged with Cys at the C-terminus
9. The lead candidate FU002
6 was prepared by site specific coupling of vancomycin with sulfo-succinimidyl 4-(N-maleimidomethyl)cyclohaxane-1-carboxylate (sulfo-SMCC) providing intermediate
8 which was clicked in solution with Cys-(Arg)
6-NH
2 9 affording FU002
6 (
Scheme 2B). FU002
6 showed a remarkable increased activity against the most important types of vancomycin-resistant bacteria having additional mechanism of action beyond the interaction with the D-Ala-D-Ala moiety responsible for the cell-wall synthesis and superior pharmacokinetics.
With the aim to increase the cell-permeability of vancomycin but also to exploit the synergistic effect of two antibiotics belonging to different classes, Adams et al. conjugated vancomycin to amphiphilic AMPs Hectate (Hec) which is an amphiphilic peptide with a net positive charge and α-helix predominant conformation (
Figure 4) [
97]. Exploiting the reactivity of the free carboxylic acid on vancomycin and without the use of any linker, vancomycin and Hec were coupled in solution producing Van-Hec conjugate
10 [
98]. The synergistic effect of vancomycin conjugate to Hec, differently from the two antibiotics alone cause the disruption of the bacterial cell wall integrity resulting very active against wild type MRSA and VRSA.
Inspired by the vancomycin conjugates showed before and considering that Gram-negative strains are challenging to fight due to the presence of the impermeable lipopolysaccharide (LPS)-rich outer membrane [
99], a novel series of conjugates have been recently designed and synthetized by tethering vancomycin to antimicrobial LPS binding peptides which have been previously demonstrated exhibiting strong effect against Gram-negative bacteria [
100]. Actually, a library of 80 conjugates, referred to as vancomycin-LPS binding peptide conjugates (VPCs) has been synthetized exploiting all the four points of attachment on the vancomycin scaffold, different chemical inert bifunctional linkers, comprising alkyl and PEG linkers, and a collection of 6 LPS binding peptides [
101]. After a first generation of conjugates where short peptides were tethered through click azide-alkyne reaction which showed modest MICs against Gram-negative strains, even if better activity that vancomycin against
E. faecium Gram-positive strain, a second generation of VPCs was prepared, the structure of the most active of which, conjugate VPC
11, is showed in
Scheme 3A. For this second generation of conjugates a different chemical strategy was chosen, namely functionalization of the vancomycin core
7 with bifunctional linkers
12 leading to the formation of intermediate
13 having a maleimide moiety at the other end of the vancomycin point of attachment, which was clicked in solution with Cys-functionalized LPS binding peptides
14 previously prepared on SP (
Scheme 3B). Some of these conjugates were further functionalized on a different vancomycin point of attachment with lipophilic tails to study the synergistic affect of the latter and LPS binding peptides when tethered to vancomycin scaffold. The results in terms of MICs of VPNs compared to vancomycin against Gram-positive and Gram-negative strains showed an increase of activity against VRE and Gram-negative strains such as AB1157 (
E. Coli),
A. baumannii, PA01 (
P. Aeruginosa), and
K. pneumonia, showing that modification with LPS binding peptides (and further lipophilic tails) alters the antimicrobial profile of vancomycin in the fighting Gram-negative bacteria.
4.2. lactams-AMP Conjugates
With the evolution of many microorganisms that developed resistance to β-lactam antibiotics manly due to the widespread diffusion of β-lactamase enzymes, a huge effort has been devoted, and it is still ongoing, by the scientific community in the quest of new derivatives which eventually lead to new generations of these antibiotics. Since the β-lactam ring, the main feature for their activity, is very labile, the vast majority of the β-lactams analogues arose from chemical modification of the pharmacophoric scaffold rather then total synthesis [
40]. Actually, both the penicillin scaffold and cephalosporin scaffold posses two functional groups, namely the amino and the carboxylic acid groups, easily to be derivatized upon protection of the other (
Figure 5). Moreover, cephalosporin possesses a further point of attachment consisting in the hydroxy group at the 3’ position.
One of the drawbacks that limits the use AMPs as therapeutic agents is their toxicity mainly caused by the presence of different cationic moieties. Indeed, it has been shown that after blocking the amino groups of polymyxin E as methane sulphonate the resulting prodrug can be used systematically [
102]. Inspired by the prodrug concept, the first β-lactam-AMP conjugate was synthetized by linking cephalotin to D-Bac8c(Leu2,5), an enantiomeric derivative of AMP Bac8c where the two D-isoleucine amino acids in position 2 and 5 are substituted with D-leucine, producing a conjugate, cephalotin- D-Bac8c(Leu2,5)
15, which has reduced net positive charge due to the presence of a carboxylate (
Scheme 4A) [
103]. The point of attachment chosen was the hydroxy group on commercially available 7-aminocephalosporanic acid (7-ACA). Accordingly, 7-ACA
19 was first deacetylated to obtain the required free 3’-OH group by using mild tetrabutylammonium hydroxide (TBAOH) at low temperature to avoid the β-lactam amide hydrolysis, then reacted with thienylacetyl chloride producing the amide intermediate
20 (
Scheme 4B). Next, after protection of the carboxylic acid as diphenylmethyl ester, the OH group was converted into the corresponding tetrachloroethyl carbamate
21, which was finally transformed in
22 by reaction with propargyl amine followed by the ester protecting group cleavage in acidic condition. Resin bound D-Bac8c(Leu2,5)-NH
2 16 was prepared by standard SPPS according to the Fmoc/
tBu protecting groups strategy and converted to D-Bac8c(Leu2,5)-N
3 18 by diazotransfer reaction with imidalole-1-sulfonyl azide hydrochloride and cleavage from the resin. Finally, click reaction between
22 and
18 was performed in solution leading to the formation to the target cephalotin- D-Bac8c(Leu2,5)
15. The conjugate
15 is considered a prodrug since the carbamate moiety of the conjugate acts as cleavable linker when the β-lactam ring is hydrolyzed in the presence of β-lactamases, delivering the free peptide.
However, even if the final conjugate 15 could be potentially used in systemic therapies preventing toxicity issues, it resulted to have a slightly lower MIC than the parent free peptide against E. Coli and MRSA probably due to a lower uptake.
A second exploited point of conjugation on β-lactam antibiotics is the free amino function on the β-lactam ring. Accordingly, Wade et al. investigated the possibility to link the N-terminus of three cationic AMPs, namely MSI-78, CA(1-7)M(2-9)NH
2 and des-Chex1-Arg20, to the amino function of 7-ACA
19 and cephalosporins precursor 7-aminodesacetoxycephalosporanic acid (7-ADCA)
29 directly on solid phase through glutaric acid linker, producing 6 AMP-β-lactam conjugates
23-28 (
Scheme 5A) [
104]. Accordingly, by protecting the amino function of 7ACA and 7-ADCA as NH-Boc carbamate and the carboxylic acid as Fmoc-ester they obtained intermediates
30 that were selectively Boc-deprotected generating derivatives
31 that can be readily used in SPS. The AMPs were grown on Rink resin and after Fmoc-deprotection of the last amino acid of
32, they were coupled with glutaric anhydride leading to the formation of resin-bound peptides
33 which were coupled with
31, cleaved form the resin producing
34 and finally deprotected at the carboxylic function of the β-lactams in solution affording conjugates
23-28. The activity of these conjugates was measured against different nosocomial pathogens and only in one case did the conjugate MSI-78-ACA-
25 and MSI-78-ADCA
26 revealed synergistic effect against
A. baumannii and
MDR A. baumannii 156.
Another small library of 4 β-lactam antibiotics-AMPs conjugates was synthetized exploiting the free amino group on the β-lactam ring and a stimuli responsive disulfide linker [
105]. The rationale behind the design of such conjugates is to exploit the ability AMPs to cross the inner and outer bacterial cell membrane of Gram-negative bacteria to help the β-lactam antibiotic to reach his targets after the cleavage of the disulfide linker in the periplasm and cytosol [
106]. Accordingly, ampicillin (Amp), herein used as a model β-lactam antibiotic, was tethered at either the N- and C-terminus of two AMPs having different characteristics, namely membrane-disrupting magainin analogue 2P2-2 that was developed by the same group [
107] and proline-rich oncocin which is able to cross the inner and outer membranes without membrane lysis [
108], producing the four conjugates
35-38 represented in
Scheme 6A. Amp
43 was functionalized with 3-(2-pyridyldithio)propionic acid
N-succinimidyl ester (PDPS) to obtain intermediate
44 which was coupled in solution with the two AMPs which were previously synthetized through SPPS and tagged with the required Cys either at the N- and C-terminus, yielding the target Amp-AMP conjugates
35-38 having cleavable disulfide linker. The MIC values of the four conjugates, along with undecorated AMPs, physical mixtures of AMP + Amp, and conjugates built with non-cleavable thioether linker were evaluated against Gram-negative bacteria
E. Coli BW 25113 and
A. baumannii ATCC 19606 and Gram-positive
Staphylococcus epidermidis ATCC 12228. Interestingly, derivative Amp-SS.9P2-2
37 showed significantly increased activity against Amp-resistant
A. baumannii and no cytotoxicity against HEK cells, whereas the oncocin-conjugates did not show enhanced antimicrobial activity probably due a lower membrane permeability induced by the introduction of the Cys tag on the peptide sequence. It is worth noting that conjugate
37 is the first β-lactam-AMP conjugate that showed remarkably increased activity against ampicillin-resistant Gram-negative bacteria highlighting the efficiency of an approach based on tethering suitable antibiotics with ad hoc AMPs through a cleavable linker.
4.3. Aminoglycoside-AMP Conjugates
Aminoglycosides have a broad spectrum of activity against pathogenic bacteria, which has led to their extensive use and occasional misuse over the past seventy years, establishing them as a valuable class of antibiotics. However, their clinical effectiveness is hampered by the toxic side effects associated to their use, in particular nephrotoxicity and ototoxicity, and by evolution of different mechanisms of resistance developed by bacteria, encompassing the decrease in AGs uptake and emergence of aminoglycoside-modifying enzymes (AMEs) [
53,
55]. Much effort has been dedicated, and it is still ongoing, to the chemical modification of the natural aminoglycosides to face a renaissance of these antibiotics. As glycopeptide antibiotics, also the chemical functionalization of aminoglycosides could seem tricky at a first sight due to the presence of many identical functional groups, i.e. hydroxy groups and amino groups that have clearly the same reactivity. However, aminoglycosides have primary alcohols or amino groups bonded to primary carbons that could be exploited for a selective functionalization because less sterically congested (
Figure 6).
The first aminoglycoside-AMP conjugate appeared in literature, i.e. Pentobra
45 (
Scheme 7A) was designed to target persister bacterial cells and to combat anaerobic bacterium
Propionibacterium Acnes (P. acnes) which are difficult to treat since charged antibiotics are not able to penetrate into the largely lipophilic sebaceous membrane [
109]. To increase the bacterial permeability of tobramycin without missing its ribosomal activity, Pentobra
45 was designed by linking tobramycin to a short 12mer AMP with ability to selectively permeate bacterial membrane, through a succinyl linker [
110,
111]. Accordingly, after Boc-protection of the amino groups of tobramycin, the primary hydroxy group of
46 was selectively functionalized with succinic anhydride leading to the formation of intermediate
47 which was coupled in SP to the N-terminus of the AMP producing after cleavage from the resin Pentobra
45 (
Scheme 7B). The conjugate showed high activity against
E. coli and
S. aureus persister cells and a wide range of
P. acnes thanks to the synergic effects, namely membrane activity and inhibition of protein synthesis [
110], along with no adverse effect and anti-inflammatory activity [
111].
In the attempt to increase the potentiality of Pentobra
45 in terms of accumulation in bacteria by increasing membrane permeability and limiting the action of the efflux systems activated by the bacterial species, the same group synthetized a collection of four new kanamycin-AMP conjugates, MAAP02-05
50-53, where the peptide transporter sequence is modified according to sequence principles based on quantum mechanical models for membranes-permeating peptides (
Scheme 8A) [
112]. Probably due to the low yield obtained in the functionalization of tobramycin for the synthesis of Pentobra
45, a new synthetic pathway has been employed. Accordingly, the primary hydroxy group of Boc-protected tobramycin
46 was selectively transformed in tosylate upon treatment with tosyl chloride in pyridine and substituted by an azide leading to the formation of intermediate
54 which was clicked with Fmoc-NH protected propargyl-alanine affording conjugate
55 (
Scheme 8B). The free carboxylic acid was used to anchor
55 to 2-cholotrytyl chloride resin (CTC) where the AMP peptide was grown through Fmoc-strategy providing after cleavage the synthesis of two 13-mer and two 12-mer tobramycin-AMP conjugates, referred to as MAAPCs
50-
53. The MAAPCs demonstrated good selectivity for bacterial cell membranes over mammalian cell membranes and do not cause significant hemolysis of human red blood cells. They also exhibit superior antibacterial activity against actively growing Gram-negative
E. coli compared to Gram-positive
S. aureus. Among them, MAAPC05
53, along with pentobra
45, exhibit the highest inner membrane permeability, which correlates well with antimicrobial activity against persisters, showing much better activity than tobramycin alone.
With the same rational, namely to increase the ability of aminoglycosides to cross bacterial cell membranes, aminoglycoside-AMP conjugates were designed to fight bacterial pathogens encompassing MRSA,
Salmonella,
Mycobacterium, and
Brucella, which are internalized within mammalian cells macrophages [
113,
114,
115]. Since aminoglycosides along with other antibiotics are characterized by insufficient membrane permeability within macrophages and suffer drug efflux, kanamycin was tethered to a modified proline-rich cell penetrating peptide with intrinsic, nonmembrane lytic antimicrobial activity targeting intracellular pathogenic bacteria [
116], through both a cleavable disulfide linker or non-cleavable alkyl linker, generating P14kanS
58 and P14kanC
59, respectively (
Scheme 9A) [
117]. Boc-protected kanamycin
60 was reacted with 4,4’-dithiobutyric acid or sebacic acid giving rise to the formation of mixtures of isomers form which compounds
61 and
62, respectively, were isolated and fully characterized by NMR spectroscopy. The obtained intermediates
61,
62 were coupled in SP to the N-terminus of the proline-rich AMP generating after cleavage from the resin the target conjugates P14kanS
58 and P14kanC
59, respectively (
Scheme 9B). Very interestingly, P14kanS
58 was more potent than P14kanC
59, P14LRR AMP and non-covalent mixture of kanamycin and P14LRR against different Gram-negative and Gram-positive bacteria, encompassing intracellular pathogens. Since the conjugates did not lyse membranes, as demonstrated by monitoring β-galactosidases release form
E.coli after addition of the conjugates, these results demonstrated the synergistic effect of the antibiotics which can operate when the disulfide bond is cleaved in the reductive environment inside the cell and kanamycin is released. Moreover, in a successive work, P14kanS
58 has proved to have potent antimicrobial activity against ESKAPE pathogens along with anti-inflammatory activity and a great ability to treat biofilms [
118].
Another proline-rich antimicrobial peptide (PrAMP) is Bac7, and in particular the segments Bac7(1-16) and Bac7(1-35), the 16-mer and 35-mer N-terminal segments that showed comparable antimicrobial activity than the parent full peptide [
119,
120]. These functional fragments are referred to as bacteria-penetrating peptides (BPPs) since they cross the bacterial inner membrane via the SbmA transporter without permeabilizing the membrane at active concentrations to eventually interact with the target ribosome and inhibit protein synthesis. To target bacterial ribosomes with two distinct synergistic mechanisms, Bac7(1-16) and Bac7(1-35) fragments were tethered to tobramycin with a cleavable disulfide linker that would release the active components in the intracellular reductive environment (
Scheme 10A) [
121]. Resin bound Cys
66, which was obtained by anchoring FmocNH-Cys(Tr)-OH to the Rink resin, was Fmoc deprotected and coupled with succinyl Boc-trobamycin
47, obtained as described in
Scheme 7B, leading to the formation of tobramycin-Cys conjugate
67 after cleavage from the resin (
Scheme 10B).
67 was reacted with 2,2’-dithiopyridine to yield
68 that was submitted to conjugation with Bac7(1-15)[Cys
16]NH
2 and Bac7(1-35)[Cys
36]NH
2 producing the final hybrid antibiotics mTob-Bac7(1-15)[Cys
16]NH
2 64 and mTob-Bac7(1-35)[Cys
36]NH
2 65, respectively. The resulting conjugates showed activity against strains to which tobramycin and the Bac7 segments were inefficient, such as clinically isolated Gram-negative bacteria strains
E.coli and
P. aetuginosa and others Gram-negative species (
A. baumanii and
S. enteridis), proving that the conjugation strategy is rewarding even if the real mechanism of action is not yet clear.
The antibacterial activity of PrAMPs depends also on the propensity of such peptides to assume more stable secondary conformations that have been shown very important for their ability to permeate and destabilize the bacterial cell membrane. A common strategy to stabilize the secondary structure of peptides, other than the introduction of prolines in the sequence, is the peptide stappling [
122], a technique that has been successfully exploited in the design of active stapled antimicrobial peptides (StAMPs) [
123]. A peptide that has witnessed an improvement in terms of proteolytic stability and antibacterial activity thanks to the stabilization of its helical structure upon hydrocarbon stapling is anoplin [
124]. Recently, anolplin and stapled anoplin have been tethered to amikacin and neomycin through both non-cleavable triazole linker and a cleavable disulfide linker generating a small library of two not cleavable neomycin-anoplin conjugates, namely Neo-anoplin
69 and Neo-anoplin[2-6]
70, two cleavable neomycin- anoplin conjugates, i.e. Neo-SS-anoplin
71 and Neo-SS-anoplin[2-6]
72 and two not-cleavable amikacin-anoplin conjugates, i.e. Amk-anoplin
73 and Amk-anoplin[2-6]
74, which structures are reported in
Scheme 11A [
125].
For the synthesis of these conjugates, the presence of only one primary hydroxy group on both the aminoglycosides neomycin and amikacin was exploited. As an example, neomycin was first Boc protected at the amino functions, then reacted with bulky triisopropylsulfonyl chloride (TIPS-Cl) to selectively transform the primary hydroxy group in sulfate which was transformed into the corresponding azide
75 by nucleophilic substitution (
Scheme 11B). Azide
75 can be either clicked with anoplin or anoplin[2-6] functionalized at the N-terminus with dec-9-ynoic acid for the synthesis of the not-cleavable derivatives
69, 70, respectively, or reacted with thiourea followed by 2-mercaptopryridined to afford intermediate
76, which were reacted in solution to anoplin or anoplin[2-6] functionalized at the N-terminus with Cys to afford cleavable Neo-SS-anoplin
71 and Neo-SS-anoplin[2-6]
72, respectively. In this case, the conjugates obtained, regardless the nature of the linker and the structure of the AMP were only slightly more active, or as active as the corresponding components, showing no synergistic effect
The improvement of the uptake of aminoglycosides to make them able to fight intracellular bacterial infections has been explored also by combination with CPPs. In particular two peptides, α1H and α2H which are two α-helices responsible for the penetration ability of the bacterial effector protein YopM into eukaryotic cell [
126], have been tethered to gentamycin trough a not-cleavable linker producing two conjugates, namely α1H-gentamycin
77 and α2H-gentamicin
78 along with a third conjugate synthetized by anchoring gentamycin to the well-known polyarginine Tat peptide, i.e. Tat-gentamycin
79 (
Scheme 12A) [
127]. Accordingly, by reacting gentamycin
80 with cross linker succinimidyl-4-(N-maleimidomethyl)cyclohexane-1-carboxylate (SMCC) the intermediate
81 was obtained as the main product (
Scheme 12B).
81 was then clicked trough thiol-maleimide chemistry to Cys-modified Tat, α1H and α2H peptides leading to the formation of the final conjugates
77-79, respectively.
Both α1H, α2H and Tat peptides were able to promote cellular internalization of gentamycin since the corresponding conjugates 77-79 were active against multiple intramolecular Gram-negative pathogenic bacteria, such as E. coli K1, Salmonella enterica and Shigella flexneri.
4.4. Miscellanous
Apart for β-lactams, vancomycin and aminoglycosides, also other classes of antibiotics have been used to build antibiotic-AMP conjugates.
Inspired by the observation that when fluoroquinolones are administered with AMPs the resulting cocktail showed a synergistic effect broadening the antibacterial spectrum of the antibiotics along with a decreasing therapeutic dose that would result in a lower adverse reactions [
128,
129], Toh et al. reasoned that a similar result could be obtained by linking levofloxacin to indolicin, an AMP with a broad spectrum of activity also against Gram-negative and Gram-positive bacteria [
129], through a labile ester linkage or a more stable amide linker. The conjugation produced two AMP-levofloxacin conjugates, namely the prodrug levo-O-indolicidin
82 and the corresponding amide derivative levo-N-indolicidin
83, which ability to cross the outer membrane of bacteria could be higher than that of levofloxacin due to the present of the highly lipophobic peptide (
Scheme 13A) [
130]. The two conjugates
82,
83 were synthetized through SPPS by reacting the free carboxylic acid of levofloxacin
86 with the N-terminus of the peptide tagged with glycolic acid or Gly
84,
85 respectively (
Scheme 13B). While the physical mixture of indolicin and levofloxacin was slightly more active compared to both antibiotics, in particular against
B. subtilis ATCC 6633, the conjugates did not show the same effect.
Levofloxacin, along with another fluoroquinolone ciprofloxacin, was also tethered to a different AMP, namely HLopt2 which is an antimicrobial analogue of HLP-2 a segment of Lactoferrin with potent antimicrobial activity against both Gram-negative and Gram-positive bacteria [
131]. The three conjugates were designed and synthetized to increase the permeability of the fluroquinolone antibiotics thanks to the ability of HLopt2 to destroy the bacterial cell trough pore formation mechanisms (
Figure 8) [
132]. All the conjugates were synthetized trough SPPS, LVX-HLopt2-NH
2 87 by anchoring the carboxylic acid of levofloxacin to the N-terminus of HLopt2, CIP-CH2CO-HLopt2-NH
2 88 by coupling the secondary amine of ciprofloxacin with the N-terminus of HLopt2 previously functionalized with bromoacetic acid, while CIP-Cys-SS-HLopt2-NH
2 89, the only conjugate with a stimuli-responsive linker, by formation of a disulfide bond between Cys-modified ciprofloxacin and Cys residue linked to the N-terminus of HLopt2 [
133]. Interestingly, all the conjugates showed increased activity along with low toxicity to mammalian cell and very low hemolytic activity, being CIP-Cys-SS-HLopt2-NH
2 89 the most active against
S. aures due to the reducing environment that trigged the disulfide bridge cleavage with the corresponding release of the two antibiotics.
Other than to increase the permeability of small molecule antibiotics, the conjugation strategy with AMP could be exploited to overcome the non-specificity of potent broad-spectrum antibiotics which suffer severe toxic side effects [
134]. For instance, chloramphenicol (CAP) is one of the most effective broad-spectrum antimicrobial agents which clinical use was hampered by its high risk of bone marrow toxicity [
135]. CAP, being lipid-soluble, diffuses through the bacterial cell membrane and reversibly binds to the L16 protein of the 50S subunit of bacterial ribosomes. This binding prevents the transfer of amino acids to growing peptide chains, likely by suppressing peptidyl transferase activity, thereby inhibiting peptide bond formation and subsequent protein synthesis. To overcome its no specificity, chloramphenicol was tethered to UBI
29-41 which is a cationic AMP highly investigated for its capacity to bind bacteria with high affinity [
136], through a not cleavable glutaric linker, leading to the formation of CAP- UBI
29-41 90 (
Scheme 14A) [
137].
CAP 91 was reacted with glutaric anhydride yielding a mixture of products due to the indiscernible reactivity of the two hydroxy groups, from which intermediate 92 was isolated in around 50% yield. After that, 92 was coupled in solution with commercially available UBI29-41 yielding, after HPLC purification, CAP-UBI29-41 conjugate 90. Gratifyingly, in vitro studies demonstrated that CAP-UBI29-41 90 has enhanced antibacterial effects on S. aureus and E. coli., also showing significantly reduced toxicity to normal cells compared to CAP. Most importantly, this result was obtained also in bacteria-bearing mouse models, indicating that UBI29-41 is an ideal targeting ligand for constructing antibacterial agents for bacteria-targeting therapy.
Differently form CAP, selectivity is not a big issue for the use of macrolides, since their mechanism of action depends on their affinity to the so-called “macrolide binding site” which allows them to inhibit selectively translation in bacteria. However, this class of antibiotics suffers antibiotic resistance due to the ability of bacteria to modify the target binding site [
59]. To fight antibacterial resistance, a huge body of work has been devoted to chemically modify the different scaffolds of macrolides. For instance, the modification of the 4’- and 4’’-hydroxyl groups of the mycaminose moiety of desmycosin (DES) leads to analogues able to fight antibacterial resistance [
138]. With the same aim, DES was conjugated to fragments of oncocin, an AMP which activity depends on the interaction with binding site that overlaps with the binding site of macrolides [
108,
139]. DES-oncocin
93 (
Scheme 15A) was synthetized starting from tylosin antibiotic
94 which was acetylated and hydrolyzed under acidic condition producing DES derivative
95 having the hydroxyl group in 4’ position unprotected, thus ready to be selectively functionalized (
Scheme 15B) [
140]. Accordingly,
95 was coupled with Boc-γ-aminobutyric acid (GABA) linker affording, after deprotection, intermediate
96 that was coupled with three different oncocin-fragments, the longer being Boc-Val-Asp(tBu)-Lys(Boc)-Pro-Pro-Tyr(tBu)-OH previously prepared through SPPS, yielding the target conjugate
93 after deprotection of all the side chain.
The resulting conjugates showed activity against some macrolide-resistant bacteria strains by binding to the E. coli 70S ribosome, thus inhibiting bacterial protein synthesis and suppressing bacterial growth.