Siderophore-Linked Ruthenium Catalysts for Targeted Allyl Ester Prodrug Activation within Bacterial Cells

Abstract

Due to rising resistance, new antibacterial strategies are needed, including methods for targeted antibiotic release. As targeting vectors, chelating molecules called siderophores that are released by bacteria to acquire iron have been investigated for conjugation to antibacterials, leading to the clinically approved drug cefiderocol. The use of small-molecule catalysts for prodrug activation within cells has shown promise in recent years, and here we investigate siderophore-linked ruthenium catalysts for the activation of antibacterial prodrugs within cells. Moxifloxacin-based prodrugs were synthesised, and their catalyst-mediated activation was demonstrated under anaerobic, biologically relevant conditions. In the absence of catalyst, decreased antibacterial activities were observed compared to moxifloxacin versus Escherichia coli K12 (BW25113). A series of siderophore-linked ruthenium catalysts were investigated for prodrug activation, all of which displayed a combinative antibacterial effect with the prodrug, whereas a representative example displayed little toxicity against mammalian cell lines. By employing complementary bacterial growth assays, conjugates containing siderophore units based on catechol and azotochelin were found to be most promising for intracellular prodrug activation.

Keywords: antibacterials; bio-orthogonal; catalysts; prodrugs; siderophores.

Figure 1

Kitamura‐type catalyst for the activation of the allyl carbamate‐protected anticancer drug doxorubicin under biologically relevant conditions.

Figure 2

Chemical structure of the hexadentate catechol‐ and hydroxamate‐based siderophores enterobactin and desferrioxamine, with chelator parts in blue.

Figure 3

Chemical structures of the sideromycin albomycin and Trojan‐horse antibacterial cefiderocol with the antibacterial warheads in black, siderophore mimics in blue.

Figure 4

Chemical structures of Moxi, and its prodrugs N‐moxi and C‐moxi.

Figure 5

Chemical structure of Ru‐1.

Figure 6

Catalyst‐mediated prodrug activation kinetics using Ru‐1 (10 mol%) to activate C‐moxi (100 μM) in 10 % DMSO in aqueous MOPS buffer (pH 7.4) at room temperature, with 5 mM GSH under an aerobic atmosphere (∼45 % overall yield) and an anaerobic atmosphere (∼90 % overall yield), with corresponding final solution appearances shown next to their corresponding endpoints.

Figure 7

Catalytic mechanism proposed by Völker et al.

Figure 8

¹H NMR spectra of Ru‐1 at 10 mM in [D₆]DMSO after 1 min, 6 h, 1 d and 3 d (▪: resonances assigned to the Ru^IV catalyst precursor, ▴: resonances assigned to the corresponding Ru^II intermediate, •: resonances assigned to allyl alcohol).

Figure 9

Micro‐aerobic (2 % O₂) growth of E. coli K12 (BW25113) in just MHII (pH 7.4) or MHII supplemented with 200 μM bpy, 100 μM FeCl₃, and 200 μM bpy + 100 μM FeCl₃, at 37 °C for 24 h.

Figure 10

Dosage–response curves of E. coli K12 (BW25113) overall growth 24 h after substrate addition. Data are normalised to “no addition” controls, for each of Moxi, N‐moxi and C‐moxi, at their varied substrate concentrations under iron‐limited (MHII supplemented with 200 μM bpy), micro‐aerobic (2 % O₂) conditions, at 37 °C.

Figure 11

Chemical structures of siderophore‐linked Kitamura‐type catalysts.

Figure 12

Catalyst‐mediated prodrug activation kinetics in aqueous MOPS buffer (pH 7.4) with 10 % DMSO at room temperature under an anaerobic atmosphere, showing C‐moxi (100 μM) activation to form Moxi for synthesised catalysts Ru‐1, Ru‐s1, Ru‐s2, Ru‐s3, Ru‐s4 and Ru‐s5 at 10 mol % loading, in triplicate.

Figure 13

Dose‐response curves of E. coli K12 (BW25113) overall growth. Overall growth at 24 h, for each of Ru‐1, Ru‐s1, Ru‐s2, Ru‐s3, Ru‐s4, Ru‐s5 and the iron control (FeCl₃) at their varied substrate concentrations under iron‐limited (MHII supplemented with 200 μM bpy), micro‐aerobic (2 % O₂) conditions, in at least technical triplicate.

Figure 14

Catalyst–prodrug co‐addition. a) Schematic representation of antibacterial activity by intra‐ and extracellular prodrug activation during co‐incubation of catalyst and prodrug. Created with biorender.com. b) Overall growth of E. coli K12 (BW25113) grown under iron‐limited (MHII supplemented with 200 μM bpy), micro‐aerobic (2 % O₂) conditions after 24 h with each catalyst–siderophore conjugate at its upper nontoxic concentration with and without C‐moxi (10 μM), and controls for siderophores, C‐moxi and Moxi, in at least technical triplicate.

Figure 15

Evaluating cellular uptake. a) Schematic representation of antibacterial activity as consequence of intracellular prodrug activation, following sequential prodrug incubation, cellular uptake, washing and catalyst addition steps. Created with biorender.com. b) Overall growth of E. coli K12 (BW25113) under iron‐limited (MHII supplemented with 200 μM bpy), micro‐aerobic (2 % O₂) conditions, 18 h after C‐moxi (hatched bars) or DMSO (solid bars) incubation and subsequent addition of substrate. Substrates include Ru‐1, Ru‐s1, Ru‐s2, Ru‐s3, Ru‐s4 and Ru‐s5 at their upper, nontoxic concentrations and controls: Moxi (white) and “no addition” (black). The difference in overall growth between each incubation with and without C‐moxi incubation, is highlighted in a bracket over the corresponding bar charts. Each incubation was carried out in technical triplicate, and each subsequent substrate addition in at least technical triplicate.