Benzotriazole derivatives as alternate O-acetylserine sulfhydrylase substrates to impair cysteine biosynthesis in gram-negative bacteria

Summary

Cysteine biosynthetic pathway, unique to bacteria and absent in the human host, represents a relevant reservoir of potential antimicrobial targets. Cysteine, in fact, is involved in redox homeostasis and is prone to the production of reactive oxygen species, and altering its metabolic equilibrium provokes severe consequences on bacterial fitness and antibiotic response. Here, we developed a library of benzotriazole-derived compounds targeting O-acetylserine sulfhydrylase (OASS)—the enzyme responsible for cysteine synthesis in most pathogenic bacteria—which act as alternate OASS substrates that can impair cysteine production. The library was tested on three clinically relevant gram-negative species, and the three most promising candidates, upon the validation of their mechanism of action in vitro, were demonstrated to have adjuvant effect in combination with selected approved antibiotics. While further optimization is required before their application, these molecules might find application as adjuvants of the antibiotic therapy and help to preserve existing drugs.

Graphical abstract

Highlights

• •Benzotriazole derivatives were synthesized and tested as antibiotic adjuvants
• •Compounds act as false substrates of OASS, impairing cysteine biosynthesis
• •Adjuvant activity enhances colistin efficacy against resistant gram-negative strains
• •Thiol levels in bacteria decrease significantly upon compound treatment

Chemistry; Biological sciences

Introduction

The major urgencies arising from the global upsurge of antimicrobial resistance (AMR) are the need to preserve or restore antibiotic efficacy and the search for new molecules and strategies. In this direction the efforts to explore bacterial metabolism in search of new targets are addressed^1,2; in fact, very little is still known about pathogens’ adaptation to host niches, symbiosis-to-infection switches, and bacterial defense mechanisms. It is evident that the insurgence rate of resistant pathogens is way faster than the discovery of effective compounds, so it is central to preserve already approved antibiotics. From this point of view, the use of antibiotic adjuvants is a supported strategy, which can increase microorganism susceptibility and enhance antibiotic efficacy at lower doses. Virulence factors and homeostasis regulators are targets of primary interest since they do not affect bacterial viability per se but influence the pathogenic switch, limiting the risk of resistance selection. Among the so-called non-essential targets, cysteine anabolism engages specific interest since (1) cysteine homeostasis is intimately connected to bacterial fitness, especially to oxidative stress response, through the production of glutathione (GSH) and other reducing agents³; (2) it is dispensable during some bacterial growth phases⁴; and (3) the biosynthetic machinery is absent in the human host, limiting undesired off-target side effects. Moreover, cysteine becomes toxic to bacteria over a certain concentration range,⁴ whereas it has been proven that the alteration of its biosynthetic pathway influences the antibiotic resistance in both vegetative and swarm cell populations in Salmonella Typhimurium.⁵ These results suggest that inhibitors of cysteine biosynthesis could improve the efficacy of antibiotics, allowing their use at a lower dose and decreasing the spread of resistance. For this reason, up to now, many efforts have been devoted to developing inhibitors of cysteine biosynthesis as a promising strategy for the discovery of antimicrobial adjuvants for different bacterial species.^6,7,8

The last two steps of cysteine biosynthesis involve the formation of O-acetylserine (OAS) from L-serine by serine acetyltransferase (SAT) and the synthesis of L-cysteine from OAS and bisulfide by O-acetylserine sulfhydrylase (OASS, EC 2.5.1.47), an enzyme generally present in bacterial cells in two isoforms known as OASS-A and OASS-B (encoded by cysK and cysM genes, respectively). OASS-A and SAT regulate their enzymatic activities by engaging in the cysteine synthase complex (CSC), whose formation equilibrium depends on a feedback regulation based on OAS and cysteine availability.^9,10,11,12 On the other hand, OASS-B can also exploit thiosulfate as a substrate (EC 2.5.1.144) (Figure 1A) but does not participate in CSC formation.^10,13 The role of the two isoforms is still debated, but in S. Typhimurium, it was observed that ΔcysM strains are bradytrophs under anaerobic conditions, suggesting that OASS-B may sustain cysteine biosynthesis in different growth states.¹⁴ OAS is produced by SAT in the presence of acetyl coenzyme A (Ac-CoA), whose metabolism again relies on cysteine availability; CoA, another thiol-containing molecule, is also exploited to protect sulfhydryl groups on proteins in oxidative stress conditions.¹⁵

Figure 1.

Cysteine biosynthesis in bacteria

(A) The last two steps of L-cysteine biosynthesis are catalyzed by SAT, which produces OAS from L-serine and Ac-CoA, and OASS-A and OASS-B, which use OAS and bisulfide to form L-cysteine. OASS-B can also use thiosulfate as a substrate, with production of sulfocysteine that is subsequently converted to L-cysteine. OAS and its derivatives N-acetylserine (NAS) or N,O-diacetylserine (DAS) are involved in the cys operon transcription.

(B) Catalytic mechanism of OASS ping-pong reaction. The pyridoxal 5′-phosphate cofactor reacts with the first substrate OAS, forming the intermediate α-aminoacrylate with the release of acetate. During the second semi-reaction, bisulfide reacts with α-aminoacrylate to give L-Cys.

In the same context, it has been observed that, when accumulated, OAS acts—directly and/or through its derivative N-acetylserine—as a transcriptional activator of the cys operon transcription factor CysB¹⁶; further, in Salmonella enterica it has been isolated as N-acetyltransferase able to acetylate OAS producing N,O-diacetylserine, an inducer of the cys regulon.¹⁷ Noteworthy, OAS seems to also play a specific signaling role in the formation of biofilms, as evidenced in Escherichia coli and Vibrio fischeri.^18,19 Based on these considerations, an approach aimed at impairing cysteine biosynthesis through the deviation of cysteine anabolic flux may represent an alternative vision to classical inhibition strategies, which hit OASS activity likely leading to OAS accumulation.

OASS catalyzes a ping-pong reaction, in which in the initial step OAS reacts with the pyridoxal 5′-phosphate (PLP) cofactor, forming an α-aminoacrylate intermediate with the concomitant release of acetate. In the subsequent step, sulfide engages in the reaction, leading to the production of L-cysteine and the regeneration of the internal aldimine (Figure 1B).²⁰ The use of alternate OASS substrates (i.e., molecules able to be used in the enzymatic reaction instead of the natural substrates) that compete with bisulfide can sustain the depletion of OAS, avoiding the triggering of secondary metabolic signals but at the same time affecting the cysteine reservoir. Moreover, in the presence of alternate substrates detrimental to bacterial metabolism, the target overexpression does not overcome the effect or induce resistance. An example of OASS alternate substrate is represented by 1,2,4-triazole, a known downregulator of cysteine biosynthesis.²¹ This molecule competes with bisulfide in the OASS-catalyzed reaction acting as a false substrate to produce 1,2,4-triazol-alanine instead of L-cysteine, thus removing OAS from the intracellular pool and inhibiting the feedback regulation in biosynthetic enzyme expression. This behavior is evident as a delayed growth rate.²¹

Recently, during a drug discovery campaign, Lee and coworkers selected some thioacetamide-linked 1,2,3-triazole derivatives for their bacteriostatic activity on selected gram-negative strains and found them to act as OASS substrates and cysteine biosynthesis disruptors.^22,23

Based on these pieces of evidence, we focused our attention on 1H-benzo[d][1,2,3]triazole (BT) as a promising scaffold for inhibitor derivatives acting as alternate substrates on OASS-A and OASS-B. In fact, BT may be considered a preferred scaffold because of its numerous pharmacological activities. Indeed, suitably substituted benzotriazole derivatives boast a wide variety of biological properties, including antiprotozoal, antiviral, and antibacterial ones. In addition, this core appears to be a very interesting scaffold in drug discovery and development processes and is useful for broadly developing structure-activity relationship (SAR) analyses on different classes of pharmacological agents.²⁴ Taking into account these observations and considering that benzo-fused azoles are a class of heterocyclic compounds of interest in medicinal chemistry, in this work we synthesized a small library of BT derivatives to test their effect on cysteine biosynthesis and their potential antibiotic/adjuvant activity on selected gram-negative strains of clinical interest. The three best hits showed a minimum inhibitory concentration (MIC) in the range of 16–32 μg/mL and significantly affected the intracellular thiol reservoir. In addition, their use in combination with colistin, gentamicin, and ciprofloxacin showed interesting and promising synergistic and/or additive effects on E. coli, S. Typhimurium, and Klebsiella pneumoniae.

Results

Library design

The rationale behind developing a compound library of BT derivatives is to design molecules that function as alternate substrates for OASS, able to compete with bisulfide in the reaction with OAS. The compound library comprises 33 molecules with the BT scaffold linked to various aromatic or aliphatic substituents on the benzene ring via an amide bond (Figure 2). It is hypothesized that the nitrogen atom at position 2 of the BT scaffold is crucial for the nucleophilic attack on α-aminoacrylate, leading to the formation of a false product. Given that outer membrane permeability in gram-negative bacteria is a key factor in designing compounds that effectively target intracellular enzymes, derivatives bearing substituents of varying chemical nature and size at position 5 of the BT scaffold were synthesized to optimize physicochemical properties, enhance bacterial cell penetration, and explore the interaction site with the protein (Figure 2). Moreover, to provide preliminary evidence for the mechanism of action of the BT derivatives, selected compounds 3l and 3n were also synthesized with a methylated nitrogen at position 2. According to our working hypothesis, methylation of this nitrogen should inhibit the nucleophilic attack and thereby prevent the formation of the false product and render compounds 17l and 17n inactive.

Figure 2.

BT general scaffold with numbering and chemical structures of derivatives synthesized and tested

Compound 16ai initially demonstrated solubility at 100 mM in 100% DMSO upon resuspension; however, it precipitated irreversibly after freezing and was consequently excluded from further analysis.

Chemistry

To efficiently and rapidly generate a focused library of derivatives bearing aliphatic and aromatic substituents connected via an amide linker at position 5 of the BT scaffold, optimized synthetic strategies were employed. The chemical structures of the synthesized compounds were confirmed by nuclear magnetic resonance (NMR) and mass spectrometry analyses (MS), as described in the chemistry subsection of the materials and methods section. The most straightforward synthetic approach involves the coupling of 1H-benzo[d][1,2,3]triazole-5-carboxylic acid with primary amines in the presence of carbonyldiimidazole (CDI) as an acid activator in dry dimethylformamide (DMF) or tetrahydrofuran (THF). This strategy enables the facile synthesis of a small library of derivatives utilizing primary amine fragments. Similarly, reverse amide derivatives at position 5 can be synthesized by reacting 1H-benzo[d][1,2,3]triazole-5-amine with carboxylic acids bearing the desired substituents.

The reaction of benzotriazole-5-carboxylic acid (1) with the appropriate amine (2a–v) in the presence of CDI in dry DMF resulted in the formation of benzotriazole-5-carboxyamides (3a–v), featuring the desired substituents on the amide group (Scheme 1).

Scheme 1.

Reagents and conditions for the synthesis of compounds 3a–v

I: CDI, DMF, 70°C, overnight (o/n) 46%–74%.

For the synthesis of the final compound 3f, the amine 2f was not commercially available and was, therefore, synthesized through a concise synthetic route. The reaction of ethylenediamine (4) with di-tertbutyl carbamate afforded intermediate 5, which subsequently reacted with 2-chloropyridine to yield intermediate 7. The final step involved the removal of the Boc protecting group through the use of trifluoroacetic acid (TFA) in dichloromethane (DCM), affording the desired amine 2f (Scheme 2).

Scheme 2.

Reagents and conditions for the synthesis of intermediate 2f

II: Di-tertbutyl carbamate, dioxane, r.t., o/n, quant. yield; III: CuI, L-proline, K₂CO₃, DMSO dry, 90 °C, o/n 47%; IV: TFA, DCM, r.t., 3 h, quantitative yield.

Similarly, amines 2j, 2p, and 2q were synthesized following an analogous synthetic protocol. In this approach, 2-aminoethanol (8) was reacted with di-tertbutyl carbamate to afford the intermediate 9. The subsequent reaction of intermediate 9 with the appropriate iodides (10w–y) in the presence of NaH as a base resulted in the formation of intermediates 11w–y. The final step involved Boc group deprotection, yielding the desired amines 2j, 2p, and 2q (Scheme 3).

Scheme 3.

Reagents and conditions for the synthesis of intermediates 2j, p, q

II: Di-tertbutyl carbamate, dioxane, r.t., o/n; V: NaH, THF, 0 °C, 30′, r.t., o/n, 51%–63%; IV: TFA, DCM, r.t., 3 h, dioxane, quant. yield.

For the synthesis of the final compounds 13aa–ae, which contain a tertiary amide, a slightly modified protocol was employed compared to Scheme 1. Specifically, the reaction of compound 1 with the corresponding amines 12aa–ae, in the presence of CDI as an activating agent and THF as the solvent, resulted in the formation of the desired final compounds (Scheme 4).

Scheme 4.

Reagents and conditions for the synthesis of final compounds 13aa–ae

VI: CDI, THF, reflux o/n 46%–74%.

Compounds featuring a reverse amide configuration were also synthesized. The reaction of 5-amino benzotriazole (14) with the appropriately substituted carboxylic acids (15af–ai) resulted in the formation of the desired final compounds (16af–ai; Scheme 5).

Scheme 5.

Reagents and conditions for the synthesis of compounds 16af–ai

I: CDI, DMF, 70 °C, o/n 46%–74%.

To obtain preliminary insights into the mechanism of action, compounds featuring methylated nitrogen on the BT ring were synthesized. The compounds 3l and 3n, prepared according to Scheme 1, underwent a methylation reaction of the nitrogen atom via treatment with NaH and methylene iodide, affording the methylated derivatives 17l and 17n (Scheme 6). Following the N-methylation reaction of BT, it is not immediately clear which of the three nitrogen atoms has been methylated. However, NMR spectroscopy studies, particularly the ¹⁵N NMR spectrum, suggest that the methylation has occurred predominantly at the nitrogen in position 2 (Figure S1).

Scheme 6.

Reagents and conditions for the synthesis of compounds 17l, n

VII: NaH, DMF, CH₃I, 0 °C 30′, r.t. 3 h., 46%–74%.

Microbiological screening and selection of active compounds

The cysteine biosynthetic pathway is highly conserved among Enterobacteriaceae, thus we decided to test the effect of the compounds on three AMR-relevant gram-negative ATCC (American Type Culture Collection) strains, namely, E. coli, S. enterica serovar Typhimurium, and K. pneumoniae, known to express this pathway, and specifically OASS.²⁵ To identify compounds capable of effectively inhibiting bacterial growth, the library was primarily screened by a microdilution assay aimed at determining the MIC. Since cysteine biosynthesis is induced only in cysteine-depletion conditions, the compounds’ activity was assessed in M9, a minimal medium devoid of amino acids. Since in this condition OASS activity is necessary for bacterial proliferation, the effect of inhibiting cysteine biosynthesis can be related to a growth-inhibited phenotype.²⁵ The MIC values of the 32 compounds were calculated using compound dilutions in the range of 0.5–256 μg/mL (Tables 1 and S1). Twelve compounds inhibited E. coli growth; four of them were only able to affect bacterial viability at concentrations higher than 120 μg/mL; conversely, for eight compounds promising MIC values were found in the range 16–64 μg/mL (Table 1). In particular, the compounds 3d, 3l, and 3n revealed in E. coli MICs of 16 μg/mL, 21 ± 8 μg/mL, 21 ± 9 μg/mL, respectively.

Table 1.

MIC of BT derivatives effective in M9 medium on E. coli, S. Typhimurium, and K. pneumoniae

Compound	MIC average (μg/mL)
	E. coli ATCC 25922		S. Typhimurium ATCC 14028		K. pneumoniae ATCC 13883
	M9	MHB	M9	MHB	M9	MHB
3a	64	>256	>256	>256	>256	>256
3c	64	>256	256	>256	256	>256
3d	16	>256	114 ± 28	>256	107 ± 32	>256
3h	121 ± 21	>256	>256	>256	>256	>256
3j	185 ± 67	>256	>256	>256	>256	>256
3k	64	>256	>256	>256	>256	>256
3l	21 ± 8	>256	43 ± 18	>256	128	>256
3n	21 ± 9	>256	128	>256	>256	>256
3s	128	>256	>256	>256	>256	>256
3t	64	>256	>256	>256	>256	>256
13ac	128	>256	>256	>256	>256	>256
16ag	43 ± 16	>256	>256	>256	>256	>256

BT derivatives’ effect was evaluated in the range 0.5–256 μg/mL after 24 h of incubation at 37 °C on E. coli ATCC 25922, S. Typhimurium ATCC 14028, and K. pneumoniae ATCC 13883. The table includes the results on the derivatives that showed activity on the tested strains; the data collected for ineffective compounds are reported in Table S1. None of the compounds showed activity when tested in the rich medium Mueller Hinton Broth (MHB). The data represent the average MIC of nine replicates ± SD; values without error reported an SD equal to 0.

Interestingly, 3l also showed an MIC value of 43 ± 18 μg/mL in S. Typhimurium, whereas higher MIC values of 128 μg/mL and 114 μg/mL were obtained for 3d and 3n. Similarly, compound 3d inhibited K. pneumoniae growth with an MIC equal to 107 ± 32 μg/mL, an effect comparable to compound 3l; on the contrary, 3n-treated wells showed no effect all along the concentration range for this strain. The same compounds, when tested in rich Mueller-Hinton Broth medium, containing cysteine, reversed their inhibition without evident bactericidal or bacteriostatic effect on the growth of the three strains, thus supporting a specific impact on bacterial metabolism (Tables 1 and S1).

Based on these results, compounds 3d, 3l, and 3n, which showed the most promising profile, were subjected to further in vitro studies on recombinant OASS to confirm their activity as alternate substrates.

Activity tests in vitro

A promising inhibition approach should hit both OASS-A and OASS-B isoforms to gain a pervasive response in vivo. Among bacterial species, S. Typhimurium OASS-A and OASS-B (StOASS-A and StOASS-B) have been extensively characterized both from kinetic and structural points of view, and their structures and sequences are highly conserved.^{20,25,26,27,28} Taking advantage of this information, the compounds were tested in vitro on recombinantly produced isozymes from S. Typhimurium. OASS ping-pong reaction can be spectroscopically probed by following the PLP absorption shift associated with the catalytic events. In fact, in the absence of the first substrate the PLP cofactor is linked to the catalytic lysine as internal aldimine, which has an absorption peak centered at 412 nm (Figure 3). When OAS reacts with the internal aldimine, the α-aminoacrylate intermediate is formed, revealed by the disappearance of 412 nm absorption maximum and the appearance of a main peak at 470 nm, together with a less-intense peak at 330 nm.^26,28 When a second substrate able to react with α-aminoacrylate is added to the reaction solution (e.g., bisulfide or, in this study, an active BT derivative), PLP returns in the initial state and its maximum absorbance at 412 nm is restored (Figure 3). The activity and specificity of the compounds identified through microbiological analyses were spectroscopically monitored by following the regeneration of the internal aldimine, which also corresponds to the formation of the false product, starting from the α-aminoacrylate intermediate. The latter was obtained by adding to each of the two OASS isoforms the lowest amount of OAS that ensured a stable α-aminoacrylate signal for at least 20 min (Figure S2 and materials and methods). PLP spectra were measured before (internal aldimine, Figure 3 black curve) and after the addition of OAS (α-aminoacrylate state, Figure 3 red curve) and at defined time intervals after the addition of 1 mM BT derivative (Figure 4). As a reference control, we used 1,2,4-triazole, which can function as an alternate substrate to bisulfide by OASS, catalyzing a triazolylase reaction to form 1,2,4-triazole-1-alanine.^20,21 As expected, StOASS-A can restore the internal aldimine at 412 nm after the formation of α-aminoacrylate in the presence of 1,2,4-triazole (Figure S3B), supporting our assay setup. We also assayed the reaction on StOASS-B, for which no kinetic data with 1,2,4-triazole as an alternate substrate have been reported so far. Our results demonstrated that this isoform is likewise capable of processing the substrate and completing the catalytic cycle (Figures S3B and S4B). As shown in Figure 4, both StOASS-A and StOASS-B can use compounds 3d, 3l, and 3n as second substrates, restoring the internal aldimine. For the isoform B, the reaction in the presence of compounds 3d and 3l seems slower than in the presence of isoform A; however, due to the different stability conditions identified for α-aminoacrylate in the two enzymes, this assay allows only to make qualitative observations without precise comparisons. As a confirmation of the reaction mechanism, we exploited two BT analogs with modification on the triazole ring, i.e., compounds 17l and 17n (analogs of 3l and 3n, respectively), presenting a methyl group linked to N2 (Figure 2). When tested in the presence of OAS, both StOASS isoforms did not close the catalytic cycle independently from the substituents linked to the reactive core, confirming that the reaction involves the benzotriazole moiety (Figures S3 and S4).

Figure 3.

Spectroscopic monitoring of catalytic intermediates of OASS and proposed mechanism of action of BT derivatives

The absorption peak of PLP internal aldimine is centered at 412 nm (black curve); when OAS is added, the PLP-α-aminoacrylate intermediate forms, with a spectroscopic signature at 470 nm (red arrow and curve). In the presence of a second substrate (bisulfide or a BT derivative), the final reaction product is formed and the internal aldimine is restored (412 nm, black arrow and curve).

Figure 4.

Activity of selected BT on StOASS-A and StOASS-B monitored as α-aminoacrylate consumption during time

The spectra of the enzymes were first recorded in the absence (black) and presence (red) of 75 μM OAS for isoform A (A–C) and 750 μM OAS for isoform B (D–F). After the addition of compounds 3d (A, D), 3l (B, E), or 3n, (C, F) at 1 mM, spectra were registered after 1 (yellow) and 10 (blue) min of incubation in 100 mM HEPES, 10% DMSO, pH 7.0 at 25°C.

NMR assignment of 3l alternate product structure

The formation of the alternate product by OASS needs the nucleophilic attack of the BT derivative on α-aminoacrylate. Based on the previous results on 3l and 3n and on their methylated derivatives (Figures 4, S3, and S4), it was possible to reasonably affirm that the reaction involves the BT moiety, as expected. To structurally verify the nature of the enzymatic product, we used compound 3l as the StOASS substrate and studied the resulting reaction mixture by NMR spectroscopy to determine the structure of the alternate product. The reaction was carried out using both StOASS isoforms, to assess that the product resulting from the catalysis of the two isoforms was equivalent. To this purpose, the combination of 1H-1D (Figure 5), 2D Correlation Spectroscopy (COSY), and 1H-13C Heteronuclear Multiple Quantum Coherence (HMQC) allowed for the complete assignment of non-exchangeable protons and relative carbons (see also Figures S5–S7; Table S2). The results obtained for the two isoforms were superimposable, attesting to a comparable mechanism of action.

Figure 5.

1H-1D spectra of 3l alternate product

The product was obtained from OASS-B reaction, lyophilized, and resuspended in D₂O, at 25°C. (Inset top left) Magnification of the 8.15- to 7.8-ppm region; (top center) the theoretic chemical structure of the molecule with atom nomenclature.

Some impurities deriving from the assay mixture were observed, especially in the aliphatic region (5.0–0.5 ppm); residual signals of OAS were identified when adding directly in the tube 300 μM OAS (Figure S5). The further methyl signal at 1.88 ppm was ascribed to acetate elimination from OAS upon enzymatic reaction (Figure S5). To prove that the substitution reaction occurred, a diffusion-ordered spectroscopy experiment on 3l alternate product was acquired. The signals arising from aminopropanoic group (H19a/b and H₂O) exhibited the same diffusion coefficient as the aromatic signals of 3l (Figure S8), confirming that they belonged to a unique molecule. The absence of Nuclear Overhauser Effect/Rotating-frame Overhauser Effect (NOE/ROE) signals (data not shown) between H19a/b-H₂O and H4-H7 of the benzotriazole suggested that the substitution reaction occurred on N2; otherwise the distances between the mentioned above nuclei would have generated an appreciable NOE/ROE signal.

Target engagement confirmation

To further investigate the selectivity of the chosen alternate substrate, we tested their activity in bacterial cells with specific deletions of the genes encoding the putative target proteins. The MIC values of compounds 3d, 3l, and 3n were then calculated on E. coli strains JW2407-1 (ΔcysK743::kan) and JW2414-3 (ΔcysM750::kan), derived from E. coli K12 strain and presenting cassette deletions in the cysK and cysM loci, respectively, coding for OASS-A and OASS-B (E. coli genetic resource center https://ecgrc.net/).²⁹ The MIC values calculated for E. coli ATCC 25922 were consistent with those calculated for the ΔcysM mutant strain; on the other hand, ΔcysK mutant strain neither grew under the same experimental conditions nor in the control wells. It is known that the prominent role in cysteine biosynthesis is played by OASS-A, whereas OASS-B activity has been attributed to support growth in anaerobic conditions.¹⁴ However, MIC microdilution assay was performed in microaerobic conditions and possibly OASS-B could not sustain E. coli ΔcysK mutant growth in these conditions. It is then conceivable to conclude that the action of the BT derivatives is mainly exerted on the OASS-A enzyme in vivo, even though this result cannot be directly attributed to isoform specificity.

To further confirm the metabolic target of the selected BT, MIC values were determined in the presence of L-cysteine or GSH, a molecule metabolically strictly related to cysteine. GSH is a tripeptide formed by glutamate, cysteine, and glycine and can be degraded as a cysteine source for the bacterium, therefore supporting E. coli strains auxotroph for this amino acid.³⁰ When E. coli MG1655 was grown in the presence of compounds 3d, 3l, and 3n and different concentrations of either L-cysteine or GSH, bacterial susceptibility to BT decreased, with the vitality almost completely restored in the presence of 256 μg/mL BT and 5 μg/mL L-cysteine or 100 μg/mL GSH (Table 2). The rescue of the growth inhibition caused by BT derivatives supports a specific action of these molecules on the cysteine biosynthetic pathway.

Table 2.

MIC of compounds 3d, 3l, and 3n on E. coli MG1655 grown in M9 medium compared to the presence of different concentrations of L-cysteine or GSH

Compound	MIC μg/mL
	M9	L-cysteine		GSH
		5 μg/mL	10 μg/mL	10 μg/mL	100 μg/mL
3d	42.67 ± 16.52	>256	>256	85.33 ± 33.05	>256
3l	16	>256	>256	26.67 ± 8.26	>256
3n	64	>256	>256	106.67 ± 33.05	>256

The data represent the average MIC of six replicates ± SD; values without error reported an SD equal to 0.

Evaluation of the effect of compounds on intracellular thiol production

The mechanism of action of the three active compounds is thought to target cysteine production over time, being consumed instead of bisulfide in OASS catalysis. Since cysteine is the precursor of many reducing species, the most abundant of which is GSH (see Ferguson and Booth³¹ and references therein), dysregulation in its synthesis would also affect the intracellular thiol pool. To assess the in vivo impact of compounds 3d, 3l, and 3n, we treated E. coli ATCC 25922 grown in M9 medium at 37°C under agitation, using the compounds at their static MIC values and 3-fold higher concentrations; these values were compatible with E. coli growth and viability under these experimental conditions, including medium agitation and initial inoculum size (see materials and methods and Schuurmans et al.³²). To obtain a sufficient number of cells suitable for thiol determinations during a 20-h time-course monitoring, specific experimental conditions were met: (1) overnight (o/n)-grown cells were diluted 1:1,000 in fresh minimal medium and allowed to grow for 3 h before starting the treatment to clear intracellular thiol accumulation; (2) the treatment started during the log phase (OD₆₀₀ around 0.25); and (3) bacterial cultures were maintained under agitation, favoring culture aeration to promote growth. The concentration of intracellular thiols was then measured at different time points, starting from the beginning of the treatment until 20 h after, and normalized for the total protein content in the samples. As a negative control, 1% DMSO-treated cultures were analyzed, whereas 1,2,4-triazole, a known downregulator of cysteine biosynthesis, which results in a delayed growth rate, was used as a positive reference (Figure 6).²¹ In our experimental setup, 1,2,4-triazole was used at 345 μg/mL (5 mM) concentration, showing statistically significant differences as compared to DMSO-treated culture in the 2- to 6-h interval after the beginning of the treatment (p < 0.1 after 2–4 h and p < 0.001 after 6 h, respectively) (Figure 6D). When E. coli cultures were incubated in the presence of 21 μg/mL 3l or 3n, or 16 μg/mL 3d (MIC values), no significant effect was measured on the intracellular thiol concentration (Figures 6A and 6B); nevertheless, when the cultures were exposed to a 3-fold higher amount of the compounds, growth rate was slower and the thiol concentration decreased significantly starting from 1 to 6 h after treatment; interestingly, the effect of compound 3n was significant almost all along the first 5 h of treatment. Conversely, the effect of compounds 3d and 3l became non-significant after 4 h, even though thiol concentration was still lower than in the control culture (Figures 6C and 6D).

Figure 6.

Growth curves and intracellular thiol analysis

E. coli ATCC 25922 growth curves and intracellular thiol analysis in the presence of compounds 3d (red), 3l (dark yellow), and 3n (green) at the static MICs (A and B) and at 3-fold static MICs (C and D). Cultures grown in the presence of 1% DMSO (white) and 5 mM (345 μg/mL) 1,2,4-triazole (gray) were included as negative and positive controls, respectively. The bars represent the average of at least 3 biological replicates, and the error bars represent the SD. The statistical significance of thiol decrease was determined by a two-way ANOVA test (see STAR Methods). ∗p < 0.1; ∗∗p < 0.01; ∗∗∗p < 0.001.

Cytotoxicity

The toxicity of the compounds was assessed at increasing concentrations following the MIC setup, checking the viability of adult bovine kidney-derived cells (MDBK) grown in vitro. Compounds 3l and 3n demonstrated comparable profiles, as shown in Figure 7. They showed no toxicity at the concentrations at which they inhibit E. coli ATCC 25922 growth in M9 medium, with survival rates above 80% at 32 μg/mL, indicating safe profiles. On the other hand, compound 3d showed a survival rate of around 70% when tested at 16 μg/mL, i.e., the concentration at which it exerts its antibacterial activity, indicating a safety issue.

Figure 7.

Toxicity evaluation of compounds 3d (red), 3l (dark yellow), and 3n (green) on viability of MDBK cells growing in vitro in M9 medium

The percentage bars represents the average of at least six replicates, and the error bars represent the SD for each tested concentration, with respect to the viability in the presence of 1% DMSO.

Combination experiments of most promising BT derivatives with colistin, gentamicin, ciprofloxacin, and ampicillin

The selected compounds were further tested as adjuvants in combination with four broad-spectrum antibiotics, each representing a specific class based on their use and mechanism of action: colistin, which increases membrane permeability; gentamicin, affecting ribosomal protein synthesis; ciprofloxacin, affecting DNA replication; and ampicillin, which targets the cell wall synthesis.

The results reported in Table S3 show a relevant, additive, inhibitory effect on E. coli when colistin is used in combination with compounds 3d, 3l, and 3n at 16 μg/mL and 8 μg/mL, with a 10-fold reduction in colistin MIC. A similar trend was observed in S. Typhimurium, with a synergistic effect with 3l and 3n compounds and an additive effect for 3d. Interestingly, better results were obtained for the association when tested on K. pneumoniae. In this case, the combination led to a large decrease in the MIC of colistin, with a reduction greater than 50-fold when 3l and 3n were tested at 16 μg/mL (Table S3). Their association with colistin showed strong synergistic effects in K. pneumoniae with fractional inhibitory concentration index values of 0.14 and 0.13 for compounds 3l and 3n, respectively; an additive effect was observed for compound 3d under the same conditions.

When the compounds’ effect was evaluated with the presence of gentamicin, no combination exhibited significant activity against E. coli and S. Typhimurium, indicating an indifferent effect at the tested concentrations (Table S3). The result of gentamicin on these strains seems to be in accordance with the effect observed for chloramphenicol, another inhibitor of protein synthesis. In fact, the inhibition of protein synthesis triggers metabolic responses that limit the accumulation of cysteine, which is associated with the increase in H₂S and GSH production in E. coli.³³ In this context, the inhibition of cysteine biosynthesis by BT derivatives would be redundant, resulting in an indifferent effect of the combination with gentamicin.

On the contrary, the most promising results were obtained with K. pneumoniae, against which compounds 3l and 3n notably displayed synergistic activities with gentamicin, leading to a 4-fold reduction in its MIC when tested at 16 μg/mL.

Compelling findings were observed in combination in assays with ciprofloxacin against E. coli, where exceptionally strong synergistic effects were detected (Figure 8; Table S3). Specifically, the combination of compound 3d at 4 μg/mL with ciprofloxacin in this strain resulted in a remarkable 100-fold reduction in ciprofloxacin MIC. Compound 3d exhibited an even greater synergistic effect, achieving a 300-fold MIC reduction. Most notably, compound 3n demonstrated the most pronounced synergy, leading to a more than 500-fold reduction in ciprofloxacin MIC, underscoring its potential as a highly effective adjuvant. More modest results were observed against K. pneumoniae, where an additive effect was noted with compounds 3d and 3l, whereas a synergistic interaction was achieved with compound 3n when tested at 16 μg/mL. In contrast, none of the tested compounds displayed significant activity in combination with ciprofloxacin against S. Typhimurium, indicating an indifferent effect at the tested concentrations (Figure 8).

Figure 8.

Fractional inhibitory concentration (FIC) index of the best combinations among 3d, 3l, and 3n BT derivatives with different antibiotics

Antibiotic-BT derivatives were tested on E. coli ATCC 25922 (magenta), S. Typhimurium ATCC 14028 (violet), and K. pneumoniae ATCC 13883 (light blue). The data represent the FIC index between colistin, gentamicin, ciprofloxacin, or ampicillin and the most effective concentration of BT derivative. An FIC index ≤0.5 is indicative of synergy of action, 0.5 < FIC ≤1 shows additivity of action, and 1 < FIC <4 shows indifference between the couples of considered compounds. The complete set of data is reported in Table S3.

The compounds were also tested in association with ampicillin, with an indifferent outcome in all cases, as reasonably expected based on the ampicillin mechanism of action.

Discussion

Cysteine metabolism represents a still underexplored reservoir of drug targets for the development of antibacterial agents and antibiotic adjuvants. In fact, the homeostatic equilibrium of cysteine—which is produced de novo exclusively in bacteria and plants, and not in animals—is intimately connected to bacterial fitness and infection sustainment. In this framework, a central step in cysteine biosynthesis is carried out by OASS, whose activity regulates not only the availability of the amino acid but also the upstream accumulation of OAS, which in turn regulates the cys operon transcription and biofilm formation. An alternative vision to the classical enzyme activity inhibition is represented by the development of alternate substrates that can be consumed by the bacterium with the dual benefit of preventing the accumulation of both the natural substrates and products. In this paper, we explored the potential of a 1H-benzo[d][1,2,3]triazole scaffold to develop a small library of 33 compounds targeting OASS as alternate substrates competing with sulfide. Each molecule features the BT scaffold linked via an amide bond to various aromatic or aliphatic substituents on the benzene ring. Structural modifications were introduced at position 5 of the BT core to optimize physicochemical properties, to enhance bacterial cell permeability, and to explore the interaction site with the target proteins.

As a first approach, the molecules were tested on microbiological cultures and selected for their ability to impair bacterial growth of E. coli, S. Typhimurium, and K. pneumoniae, opportunistic pathogens prioritized in the search for new antibiotics by the World Health Organization because of their ability to acquire multidrug resistance.³⁴ The choice of these species was dictated by the following reasons: (1) they are important, worldwide distributed, human opportunistic pathogens that often cause resistant infections; (2) the active therapies available for resistant strains are generally toxic and aggressive at the effective dosage; and (3) according to the literature and our previous works in this field,^{3,5,23,25,35,36,37,38,39} there is evidence that the enzymes of the sulfur assimilation pathway and, in particular, OASS, are expressed in the aforementioned pathogens and might be involved in the mechanisms of antibiotic resistance. Besides, OASS in enteric bacteria is generally present in two isoforms, OASS-A and OASS-B. While at a glance the presence of two isoenzymes may seem redundant, their functions differ in terms of substrate specificity (OASS-B can use thiosulfate in addition to bisulfide), conditions of expression (OASS-A is generally more expressed, while the role of OASS-B is still poorly understood), and regulation of cellular processes.¹⁰ Promising inhibitors must, therefore, target both isoforms to avoid functional complementation. After the initial analysis of their effect on bacterial cultures, the compounds 3d, 3l, and 3n were selected for their promising MIC values and underwent in vitro biochemical assays to assess their mechanism of action on both OASS isoforms, taking the extensively characterized S. Typhimurium enzymes as a benchmark for validation. All three compounds were found to be substrates of both OASS-A and OASS-B, supporting the validity of the drug design approach. In particular, the nitrogen atom at position 2 of the BT scaffold was hypothesized to play a crucial role in the nucleophilic attack on OAS, leading to the formation of a false product. To validate this hypothesis, selected derivatives, 3l and 3n, were synthesized with a methylated nitrogen at position 2, which, as expected, inhibited nucleophilic attack and rendered the compounds inactive. In addition, the false product resulting from the reaction between OAS and compound 3l was characterized using NMR spectroscopy, further supporting the proposed mechanism of action (Figures 3 and 5).

To further assess the specific in vivo effect on cysteine metabolism, we evaluated (1) the intracellular thiol concentration over time upon compound exposure in a wild-type E. coli strain, (2) the phenotype recovery in the presence of a cysteine source, and (3) the response in cysK/cysM-deleted strains. Experimental results confirmed the specificity and ability of the compounds to affect the intracellular thiol pool, a condition that is also reflected in a decreased growth rate (Figures 4A and 4C). Since the compounds are false substrates, their effect is limited by their consumption by the enzymes and, if not resupplied, reduces over time; this effect becomes evident after 20-h treatment, when the growth levels are restored and thiol concentration is comparable to the DMSO-treated culture.

The central aim of this research was to test the ability of BT derivative compounds to sustain the efficacy of already approved antibiotics on E. coli, S. Typhimurium, and K. pneumoniae strains. In fact, hitting cysteine metabolism could in principle weaken the bacterial response to antibiotics, allowing lowering of their effective doses. To this purpose, we selected four broad-spectrum antibiotics targeting different bacterial metabolic processes, namely, colistin, gentamicin, ciprofloxacin, and ampicillin, to be tested in combination with compounds 3d, 3l, and 3n. Colistin is considered by the World Health Organization to be a molecule of last resort, used to treat severe infections caused by multi-drug resistant (MDR) and extensively drug resistant (XDR) bacteria. It belongs to the polymixin class of antibiotics, and its bactericidal effect is achieved through the interaction with components of the outer membrane of gram-negative bacteria, leading to its rupture. Colistin is not without serious adverse effects, including nephrotoxicity and neurotoxicity, which have limited its clinical use. However, the spread of MDR and XDR bacteria, particularly those belonging to the ESKAPEE group, has led to the resurgence of its clinical use worldwide. Gentamicin, ciprofloxacin, and ampicillin are widely employed in treating severe infections, but their extensive use has led to the emergence and dissemination of resistance mechanisms, making them ideal candidates for assessing the potential synergistic effects of novel adjuvants.

Overall, the combination results are particularly compelling for K. pneumoniae, a bacterium that exhibits a high level of resistance toward most classes of antibiotics and a leading cause of infections diffused in health care institutions. Interesting synergistic effects have been found in particular for colistin and gentamicin in the presence of compound 3l, with a more than 50- and a 6-fold MIC reduction, respectively, at a safe concentration of BT derivative for eukaryotic cells (Table S3; Figure 7). On the contrary, more modest results were obtained for ciprofloxacin, a fluoroquinolone with antibiotic activity against bacterial DNA replication and repair mechanisms that also elicits oxidative stress as an additional mechanism of action, for which the MIC was halved in the presence of compounds 3d or 3n. Interestingly, even though the MIC values of the compounds were modest or absent per se on K. pneumoniae, their combination with the aforementioned antibiotics at concentrations well below their MICs consistently enhanced antibiotic efficacy.

Particularly remarkable was the strong synergistic effect on ciprofloxacin MIC in E. coli at low concentrations of BT derivatives, evidencing the possibility to develop a very promising combination. Conversely, a similar effect was not visible in S. Typhimurium. For this bacterium, it has been reported that ciprofloxacin MIC decreased more than 180-fold in the presence of 10 mM 1,2,4-triazole in S. Typhimurium cultivated in swarm conditions.³⁹ However, our experimental setup can be more closely associated with swim conditions, under which an indifferent or even antagonist behavior has been already reported in Salmonella for ciprofloxacin/1,2,4-triazole combination.³⁹ Since the checkerboard assay with ciprofloxacin and BT derivatives was performed in minimal medium (M9), the absence of a synergistic effect on S. Typhimurium may be attributed to the lack of expression of resistance factors, which are typically activated in the swarming state.⁴⁰ As mentioned by Turnbull and Surette,^5,39 actively migrating swarming S. Typhimurium cells exhibit increased antibiotic resistance, a shift in central metabolism from catabolic to anabolic growth, and enhanced responsiveness to acylhomoserine lactones. Therefore, under our experimental conditions, the effect of compounds 3d, 3l, and 3n in combination with ciprofloxacin may not be detectable.

Furthermore, Kim et al.^40,41 reported no difference in ampicillin MIC values between the swarm and vegetative states of S. Typhimurium, whereas resistance was expressed in the swarm state against antibiotics targeting intracellular processes. It is then possible to speculate that BT derivatives may target a metabolic pathway unrelated to the cell envelope, resulting in an indifferent combination with ampicillin.

Further considerations can be made regarding colistin’s mechanism of action, which involves altering the membrane permeability. In this context, its combination with BT derivatives yielded positive results for all three strains; it is possible that, under standard conditions, a partial uptake of the compounds is enhanced in the presence of colistin.

These findings underscore the potential of the tested adjuvants to enhance the antibacterial efficacy of colistin, gentamicin, and ciprofloxacin and support the possibility of broadening their application to additional antibiotic classes.

Despite the promising results obtained for compounds 3d, 3l, and 3n, the current stage of investigation remains too preliminary to justify the selection of a single lead candidate. The three lead compounds (3d, 3l, and 3n) share a common benzotriazole core substituted at position 5 with an amide moiety. The key structural variation lies in the heteroaromatic group linked through the amide: 3d bears a thiazole ring, 3l a pyridine, and 3n a thiophene. However, given the comparable activity observed, the influence of the heterocycle may be subtle. The similar IC₅₀ values obtained for the three compounds suggest that these structural differences are unlikely to significantly impact target binding. Instead, the nature of the heteroaromatic moiety may influence physicochemical properties such as solubility or cellular permeability, which could in turn affect the overall activity. Among the three, 3l exhibited the most potent antibacterial activity, showing not only the lowest MIC values against E. coli but also consistent efficacy against Salmonella spp. and, to a lesser extent, Klebsiella spp. Compound 3d demonstrated good antibacterial activity against E. coli, albeit with a less-favorable toxicity profile. Compound 3n displayed an activity profile closely aligned with that of 3l, reinforcing its potential for continued development. Collectively, these findings support the advancement of all three compounds as viable starting points for further optimization. Ongoing studies will be essential to fully assess their pharmacodynamic, toxicological, and pharmacokinetic properties, ultimately enabling the data-driven selection of the most suitable candidate for lead development.

The present results provide compelling evidence that targeting the cysteine biosynthetic pathway via small-molecule adjuvants represents a promising strategy to potentiate antibiotic activity, overcome AMR, and improve therapeutic outcomes.

Limitations of the study

In this paper, we demonstrated the feasibility of exploiting a benzotriazole scaffold to develop potential antibiotic adjuvants against selected gram-negative strains of clinical interest. However, the data collected in this study reflect in vitro experimental conditions and need to be complemented by in vivo observations. In fact, the controlled conditions used during bacterial growth may not fully reflect the complexity and variability of in vivo environments, potentially influencing the physiological responses. To date, the distribution of cysteine-containing compartments across different tissue types, which may play a role in localizing or activating the compound at infection sites, has not been systematically investigated. Additionally, a detailed kinetic description of the compounds’ mechanism of action and membrane permeability studies will support the building of a comprehensive SAR. Finally, an ad hoc experimental setup granting the maintenance of an adequate and sustained supply of the compound will help to assess the biological effects over time.

Resource availability

Lead contact

Requests for further information and resources should be directed to and will be fulfilled by the lead contact Giannamaria Annunziato (giannamaria.annunziato@unipr.it).

Materials availability

Compounds generated in this study will be made available on request, but we may require a payment and/or a completed materials transfer agreement if there is potential for commercial application.

Data and code availability

• •All data reported in this paper will be shared by the lead contact upon request. The ¹H NMR and ¹³C NMR spectra and HRMS data are included in the supplemental and have been deposited at the Science DataBank and are publicly available under DOI: https://doi.org/10.57760/sciencedb.28774.
• •This paper does not report the original code.
• •Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Acknowledgments

This research was granted by the University of Parma through the action Bando di Ateneo 2022 per la ricerca co-funded by MUR-Italian Ministry of Universities and Research-D.M. 737/2021-PNR-PNRR–NextGenerationEU, to G.A. and M.M. This article is also based on work from COST Action EURESTOP, CA21145, supported by COST (European Cooperation in Science and Technology).

The authors thank the facility Centro di Servizi e Misure “G. Casnati,” University of Parma, for its contribution to the analytical determination of the molecules’ structures.

Author contributions

Conceptualization, S.B., M.P., C.S., B.C., G.C., M.M., and G.A.; methodology, S.B., M.P., C.S., J.M.L.D’A., B.C., M.M., and G.A.; formal analysis and investigation, S.B., M.P., C.S., N.M., J.M.L.D’A., F.M., G.Q., O.D.B., and M.M.; writing – original draft preparation, S.B., C.S., G.Q., M.M., and G.A.; writing – review and editing, M.P., N.M., J.M.L.D’A., F.M., O.D.B., C.S., L.R., M.P., S.B., C.S.C., B.C., and G.C.; funding acquisition, M.M. and G.A.; supervision, M.M. and G.A. All the authors read and approved the manuscript.

Declaration of interests

The authors declare that they have no conflict of interest.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Bacterial strains
BL21 Tuner (DE3)	Novagen	N/A
E. coli ATCC 25922	ATCC	N/A
S. enterica subsp. enterica serovar Typhimurium ATCC 14028	ATCC	N/A
K. pneumoniae ATCC 13883	ATCC	N/A
E. coli strains JW2407-1	Coli Genetic Stock Center	Cat# 9916
E. coli strains JW2414-3	Coli Genetic Stock Center	Cat# 9922
E. coli MG1655	ATCC	N/A
Chemicals, peptides, and recombinant proteins
1,1,-Carbonyldiimidazole, ≥90.0%	Sigma-Aldrich	Cat# 8.02301
N,N-dimetilformamide, anhydrous 99.8%	Sigma-Aldrich	Cat# 5.89582
Benzotriazole-5-carboxylic acid, 99%	Sigma-Aldrich	Cat# 304239
2-Amino-5-methylpyridine, 99%	Sigma-Aldrich	Cat# A75684
2-Amino-5-(trifluoromethyl)pyridine, 97%	Sigma-Aldrich	Cat# A236383
2-Thiophenemethylamine, 96%	Sigma-Aldrich	Cat# 220884
2-Aminothiazole, 97%	Sigma-Aldrich	Cat# 123129
2-Amino-3-methylpyridine, 95%	Sigma-Aldrich	Cat# A75633
Cyclopentylamine, 99%	Sigma-Aldrich	Cat# C115002
3-Aminopyridine, 99%	Sigma-Aldrich	Cat# A78209
Cycloheptylamine, 99%	Sigma-Aldrich	Cat# C99604
2-Thiopheneethylamine, 96%	Sigma-Aldrich	Cat# 423270
2-Aminopyridine, ≥99%	Sigma-Aldrich	Cat# A77989
Cyclohexylamine, ReagentPlus®, ≥99.9%	Sigma-Aldrich	Cat# 240648
Thiophen-3-amine, 95%	Alfa Chemistry	Cat# ACM17721061
Amylamine, 99%	Sigma-Aldrich	Cat# 171409
3-(Trifluoromethyl)aniline, ≥99%	Sigma-Aldrich	Cat# A41801
2-Amino-5-methylthiazole, 98%	Sigma-Aldrich	Cat# 380563
Aniline	Sigma-Aldrich	Cat# 8.22256
m-Anisidine, 97%	Sigma-Aldrich	Cat# A88204
1-Adamantylamine, 97%	Sigma-Aldrich	Cat# 138576
Di-tert-butyl dicarbonate, ≥98.0%	Sigma-Aldrich	Cat# 34660
1,4-dioxane, ≥99.0%, contains ≤25 ppm BHT as stabilizer	Sigma-Aldrich	Cat# 360481
Ethylenediamine, for synthesis	Sigma-Aldrich	Cat# 8.00947
Copper(I) iodide, for synthesis	Sigma-Aldrich	Cat# 8.18311
L-proline	Sigma-Aldrich	Cat# P8865
Potassium carbonate, reagent grade, ≥98%, powder, −325 mesh	Sigma-Aldrich	Cat# 347825
Dimethyl sulfoxide, anhydrous ≥99.9%	Sigma-Aldrich	Cat# 5.89569
2-Chloropyridine, 99%	Sigma-Aldrich	Cat# C69802
Trifluoroacetic acid	Sigma-Aldrich	Cat# 8.08260
Dichloromethane. ≥99.5% ACS	Macron Fine Chemicals™	Cat# 4881-26
Ethanolamine, ≥98%	Sigma-Aldrich	Cat# E9508
Sodium hydride, 60% dispersion in mineral oil	Sigma-Aldrich	Cat# 452912
Tetrahydrofurane, ≥99.9%, anhydrous, contains 250 ppm BHT as inhibitor	Sigma-Aldrich	Cat# 5.89570
Iodobenzene, for synthesis	Sigma-Aldrich	Cat# 8.20730
2-Iodopyridine, 98%	Sigma-Aldrich	Cat# 558761
4-Iodopyridine, 96%	Sigma-Aldrich	Cat# 722170
Morpholine, for synthesis	Sigma-Aldrich	Cat# 8.06127
Methylamine solution, 2.0 M in THF	Sigma-Aldrich	Cat# 395056
pyrrolidine, for synthesis	Sigma-Aldrich	Cat# 8.07494
1-Methylpiperazine, for synthesis	Sigma-Aldrich	Cat# 8.05859
Piperidine, for synthesis	Sigma-Aldrich	Cat# 8.22299
5-Amino-1H-benzotriazole	Sigma-Aldrich	Cat# CDS000711
2-Picolinic acid	Sigma-Aldrich	Cat#P42800
2-Thiophenecarboxylic acid, 99%	Sigma-Aldrich	Cat# T32603
Thiazole-2-carboxylic acid	Sigma-Aldrich	Cat# 734748
Benzoic acid, for synthesis	Sigma-Aldrich	Cat# 8.22257
Ethyl acetate, ≥99% TECHNICAL	VWR Chemicals	Cat# 23879.295
Sodium sulfate	Sigma-Aldrich	Cat# 239313
Silica Gel, for thin-layer chromatography	Sigma-Aldrich	Cat# 1.07731
Silica gel Inorganic Sorbent, technical grade, 230–400 mesh	Sigma-Aldrich	Cat# 717185
Methyl iodide, for synthesis	Sigma-Aldrich	Cat# 8.06064
Methanol, ≥99.8%	VWR Chemicals	Cat# K977
Petroleum ether, 40°C–60°C	Apollo Scientific	Cat# APOSOR800019
Dimethyl sulfoxide-d6, 99.5 atom % D	Sigma-Aldrich	Cat# 175943
Sodium Chloride	Sigma-Aldrich	Cat# S9888
HEPES	PanReac Applichem	Cat# A1069,0500
pyridoxal 5′-phosphate hydrate	Sigma-Aldrich	Cat# P9255
Dimethyl sulfoxide Uvasol for spectroscopy	Supelco	Cat# 1.02950
Imidazole	PanReac Applichem	Cat# A1073,0500
Yeast extract	PanReac Applichem	Cat# A1552,1000
Tryptone	PanReac Applichem	Cat# A1553,1000
Agar	PanReac Applichem	Cat# A3477,0500
Isopropil β-D-1-tiogalattopiranoside, IPTG, Isopropil β-D-tiogalattoside (IPTG)	Apollo Scientific	Cat# BIMB1008
Sodium chloride	PanReac Applichem	Cat# 131659.1211
5,5′-dithio-bis(2-nitrobenzoic acid) (DTNB)	Sigma	Cat# D-8130
Copper (II) sulfate pentahydrate	Sigma	Cat# C8027-500G
Sodium carbonate	Sigma-Aldrich	Cat# S7795
Sodium bicarbonate	Sigma	Cat# S-7277
Sodium L-tartrate dibasic dihydrate	Sigma-Aldrich	Cat# 228729-110G
Sodium hydroxide	PanReac Applichem	Cat# 141929.1211
1,2,4-triazole	Sigma-Aldrich	Cat# T46108-25G
4,4′-dicarboxy-2,2′-biquinoline (bicinchoninate) disodium salt	Apollo Scientific	Cat# B15837
Ammonium bicarbonate	Fluka Analytical	Cat# 40867-50G-F
Potassium chloride	ACEF	Cat# 001187-1
Potassium dihydrogen phosphate	ACEF	Cat# 001191-1
Sodium phosphate dibasic	Sigma-Aldrich	Cat# 102618152
Magnesium sulfate	Sigma	Cat# M2643-500G
Ammonium chloride	ACEF	Cat# 000216-1
D-(+)-Glucose	Sigma Life Science	Cat# 1002432167
Calcium chloride dihydrate	Sigma-Aldrich	Cat# C3881-500G
Tris(hydroxymethyl)aminomethane (TRIS)	VWR	Cat# 103154M
Sodium dodecyl sulfate (SDS)	VWR	Cat# 0227-100G
Bovine serum albumin	Sigma-Aldrich	Cat# A6003-1G
Lysozime from chicken egg white	Sigma-Aldrich	Cat# 62971-10G-F
Phenylmethanesulphonyl fluoride (PMSF)	Apollo Scientific	Cat# PC6222M
Pepstatin A	PanReac Applichem	Cat# A2205,0010
Tris(2-carboxyethyl)phosphine Hydrochloride (TCEP)	Apollo Scientific	Cat# BIT0122
Benzamidine	Fluka	Cat# 12072
Ethylenediaminetetraacetic acid (EDTA) disodium salt 2-hydrate	PanReac Applichem	Cat# 131669.1210
Deuterium oxide	Chembridge Isotope Lboratories, Inc.	Cat# DLM-4-25
O-acetyl L-serine hydrochloride	Sigma-Aldrich	Cat# A6262-1G
L-cysteine	Sigma	Cat# C-7755
Glutathione, reduced	Sigma-Aldrich	Cat# G4251-5G
colistin	Sigma-Aldrich	Cat# C4461-100 MG; Lot. # 049M-4836V
gentamicin	Sigma-Aldrich	Cat# G1264-250 MG; Lot. # SLBK9973V
ampicillin	Sigma-Aldrich	Cat# A1593-25G; Lot# 016M4816V
kanamicin	Sigma-Aldrich	Cat# K4000-25G; Lot# SLBR6873V
ciprofloxacin	Sigma-Aldrich	Cat# 17850-25G-F; Lot. # BCBX9934
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (Tetrazolium blue chloride)	Sigma-Aldrich	Cat# 1002110803; Lot. #MKBW0025V
Mueller-Hinton broth (MHB)	BD	Cat# 275730
Dulbecco’s Modified Eagle Medium (DMEM)	GIBCO	Cat# 11965092
pET19m:StOASS-A	Magalhães et al. (2018)	N/A
pET19m:StOASS-A	Magalhães et al. (2018)	N/A
StOASS-A	this study	N/A
StOASS-B	this study	N/A
Tobacco Etch Virus (TEV) Protease	this study	N/A
Experimental models: Cell lines
MDBK ATCC CRL-6071	ATCC	N/A
Software and algorithms
Graphpad Prism 10	DotMatics	https://www.graphpad.com/features
Sigmaplot 12.0	Grafiti	https://grafiti.com/sigmaplot-detail/
MATLAB 2024	MathWorks	https://it.mathworks.com/products/matlab.html?s_tid=hp_products_matlab
Mestrenova 15.0.1	Mestrelab Research	https://mestrelab.com/
Topspin 4.1.1	Bruker	https://www.bruker.com/en/products-and-solutions/mr/nmr-software/topspin.html
MassLynx	Waters	https://www.waters.com/nextgen/it/it/products/informatics-and-software/mass-spectrometry-software/masslynx-mass-spectrometry-software.html
Deposited data
A dataset of structural information of a series of benzotriazole-3-carboxyamide as OASS inhibitors	Science DataBank	CSTR: https://cstr.cn/31253.11.sciencedb.28774 DOI: https://doi.org/10.57760/sciencedb.28774
Other
Amicon Ultra 0.5 mL 10 kDa	Merck Millipore	Cat# UFC501024
Dialysis tubing cellulose membrane	Sigma-Aldrich	Cat# D9527-100FT
Talon Superflow metal affinity resin	GE Healthcare	N/A
Reverse-phase C18 XSelect® HSS T3 column 2.1 × 50 mm, 2.5 particle size	Waters	Cat# 186006149
2695 Alliance separation system equipped with a Quattro API tandem quadrupole mass spectrometer	Waters	N/A

Experimental model and study participant details

Bacterial strains

This study utilized the following bacterial strains: Escherichia coli ATCC 25922 (GenBank accession no. CP009072); Salmonella enterica subsp. enterica serovar Typhimurium ATCC 14028 (GenBank accession no. CP001363.1); Klebsiella pneumoniae ATCC 13883 (NCBI Reference Sequence: NZ_JSZI00000000.1); E. coli JW2407-1 (Genotype: F-, Δ(araD-araB)567, ΔlacZ4787(rrnB-3), λ-, ΔcysK743kan, rph-1, Δ(rhaD-rhaB)568, hsdR514); E. coli JW2414-3 (Genotype: F-, Δ(araD-araB)567, ΔlacZ4787(rrnB-3), λ-, ΔcysM750kan, rph-1, Δ(rhaD-rhaB)568, hsdR514); E. coli MG1655 (GenBank accession no. U00096.3). All the strains are stored at −80°C.

Madin-Darby bovine kidney cells (MDBK)

The study also utilized the Madin-Darby bovine kidney cells (MDBK) ATCC NBL-1 (CCL-22) cell line. Age: adult, gender: male; karyotype: chromosome frequency distribution 50 cells: 2n = 60. The stemline chromosome number is hypodiploid with 2S component occurring at 5%. There were a total of 11–14 marker chromosomes (4–5 metacentric, 3–4 submetacentric and 4–5 acro-telocentric) common to most hypodiploid metaphases. The X was monosomic. Neither HSR chromosomes nor DM’s were seen. The cell line is stored under liquid nitrogen vapor. Cells were grown in complete Eagle’s minimal essential medium (cEMEM, containing 2 mM of L-glutamine, 1 mM of sodium pyruvate, 100 μg/mL of streptomycin, 100 IU/mL of penicillin, and 0.25 μg/mL of amphotericin B (all from Gibco, Waltham, MA, USA), complemented with 10% FBS) at 37°C with 5% CO₂ in a humidified incubator.

Method details

Chemistry

Reactions were monitored by thin-layer chromatography on silica gel-coated aluminum foils at both 254 nm and 365 nm wavelengths. Where indicated, intermediates and final products were purified through silica gel flash chromatography, using appropriate solvent mixtures. ¹H NMR and ¹³C NMR spectra for the characterization of BT derivatives were recorded on a Bruker Advance spectrometer at 400 and 100 MHz, respectively, with TMS as an internal standard. 1H NMR spectra are reported in this order: multiplicity and number of protons. Standard abbreviation indicating the multiplicity was used as follows: s = singlet, d = doublet, dd = doublet of doublets, t = triplet, q = quadruplet, m = multiplet, and br = broad signal. All compounds were tested as 95−100% purity samples (by HPLC/MS).

General procedure for the synthesis of compounds 3a-v

Carbonyldiimidazole (CDI, 1.5 equiv) was added to a solution of carboxylic acid 1 (1 equiv) in anhydrous DMF (3 mL/mmol). The reaction mixture was stirred at room temperature for 1h. Subsequently, the corresponding amine 2a–v (1.1 equiv) was added, and the reaction was maintained under stirring for 18 h at 70°C. Completion of the reaction was monitored by TLC. The reaction mixture was quenched with water and extracted three times with ethyl acetate (3 × 20 mL). The combined organic layers were washed with brine, dried over anhydrous Na₂SO₄, and concentrated under reduced pressure. The crude product was purified by flash chromatography using a gradient of dichloromethane/methanol as the mobile phase, starting from 99:1 to 92:8. The desired products were obtained in yields ranging from 74% to 46%, depending on the specific amine utilized.

N-(5-methylpyridin-2-yl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3a).

¹H NMR (400 MHz, DMSO-d6) δ: 10.86 (s, 1H); 8.65 (s, 1H); 8.20 (d, J = 1.4, 1H); 8.07 (d, J = 8.4, 1H); 8.03 (dd, J₁ = 1.1, J₂ = 8.7, 1H); 7.93 (d, J = 8.7, 1H); 7.63 (dd, J₁ = 2.0, J₂ = 8.4, 1H); 2.25 (s, 3H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 166.04; 150.49; 148.29; 139.02; 129.28; 126.13; 112.39; 114.99; 17.87.

HRMS/ESI: Calcd for C₁₃H₁₁N₅O [M-H]^-: 251.7664; found: 251.7631.

N-(5-(trifluoromethyl)pyridin-2-yl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3b).

¹H NMR (400 MHz, DMSO-d6) δ: 11.41 (s, 1H); 8.77 (s, 1H); 8.69 (s, 1H); 8.39 (d, J = 8.8, 1H); 8.22 (dd, J₁ = 2.1, J₂ = 8.9, 1H); 8.03 (d, J = 8.7, 1H); 7.94 (d, J = 8.6, 1H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 166.86; 155.87; 145.82; 136.36; 130.71; 126.26; 114.63.

HRMS/ESI: Calcd for C₁₃H₈F₃N₅O [M-H]^-: 306.2362; found: 306.2317.

N-(thiophen-2-ylmethyl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3c).

¹H NMR (400 MHz, DMSO-d6) δ: 9.28 (s, 1H); 8.44 (s, 1H); 7.92 (q, J = 5.2, 2H); 7.35 (d, J = 3,6, 1H); 7.01 (d, J = 5.6, 1H); 6.93–6.92 (m, 1H); 4.64 (d, J = 5.8, 2H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 166.20; 143.09; 127.20; 127.02; 126.00; 125.55; 38.45.

HRMS/ESI: Calcd for C₁₃H₈F₃N₅O [M-H]^-: 257.2990; found: 257.2945.

N-(thiazol-2-yl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3days).

¹H NMR (400 MHz, DMSO-d6) δ: 16.06 (bs, 1H); 13.01 (bs, 1H); 8.81 (s, 1H); 8.14 (d, J = 8.5, 1H); 8.00 (d, J = 8.5, 1H); 7.59 (d, J = 3.5, 1H); 7.30 (d, J = 3.6, 1H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 165.41; 159.39; 140.69; 139.44; 137.94; 113.05; 129.33; 112.78; 125.77; 117.71; 114.47; 114.35.

HRMS/ESI: Calcd for C₁₀H₇N₅OS [M-H]^-: 244.2600; found: 244.2645.

N-(3-methylpyridin-2-yl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3e).

¹H NMR (400 MHz, DMSO-d6) δ 10.73 (s, 1H); 8.68 (s, 1H); 8.34 (d, J = 3.5, 1H); 8.08 (d, J = 8.6, 1H); 8.00 (d, J = 8.6, 1H); 7.75 (dd, J₁ = 0.9, J₂ = 7.5, 1H); 7.28 (dd, J₁ = 4.8, J₂ = 7.5, 1H); 2.24 (s, 3H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 165.59; 150.79; 146.40; 139.83; 131.45; 130.22; 126.08; 122.65; 116.88; 114.14; 18.07.

HRMS/ESI: Calcd for C₁₃H₁₁N₅O [M-H]^-: 252.0964; found: 252.0987.

N-(2-(pyridin-2-ylamino)ethyl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3f).

¹H NMR (400 MHz, DMSO-d6) δ: 8.83 (s, 1H); 8.46 (bs, 1H); 7.99–7.95 (m, 3H); 7.49–7.45 (m, 1H); 7.05 (s, 1H); 6.62–6.53 (m 2H); 3.49 (s, 4H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 166.56; 139.72; 158.22; 158.36; 145.94; 112.18; 134.27; 109.50; 40.86; 39.11.

HRMS/ESI: Calcd for C₁₄H₁₄N₆O [M-H]^-: 281.3070; found: 281.3065.

N-cyclopentyl-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3g).

¹H NMR (400 MHz, DMSO-d6) δ: 15.90 (bs, 1H); 8.47 (s, 1H); 7.96–7.90 (m, 2H); 4.31–4.23 (m, 1H); 1.95–1.52 M, 8H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 165.90; 132.22; 125.76; 115.66; 114.17; 51.59; 32.61; 24.13.

HRMS/ESI: Calcd for C₁₂H₁₄N₄O [M-H]^-: 229.2710; found: 229.2756.

N-(pyridin-3-yl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3h).

¹H NMR (400 MHz, DMSO-d6) δ: 10.58 (s, 1H); 8.93 (s, 1H); 8.62 (s, 1H); 8.29 (s, 1H); 8.19 (d, J = 5.6, 1H); 8.01–7.97 (m, 2H); 8.07–8.01 (m, 2H); 7.39–7.36 (m, 1H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 166.00; 145.27; 142.42; 136.42; 138.24; 127.92; 124.06.

HRMS/ESI: Calcd for C₁₂H₉N₅O [M-H]^-: 238.1824; found: 228.1817.

N-cycloheptyl-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3i).

¹H NMR (400 MHz, DMSO-d6) δ: 15.91 (bs, 1H); 8.44 (d, J = 9.4, 1H); 1.88 (s, 2H); 1.67–1.46 (m, 13H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 167.30;145.24; 139.11; 134.60; 127.29; 113.89; 112.67; 51.38; 31.28; 29.17; 25.54.

HRMS/ESI: Calcd for C₁₄H₁₈N₄O [M-H]^-: 257.2259; found: 257.2239.

N-(2-phenoxyethyl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3j).

¹H NMR (400 MHz, DMSO-d6) δ: 15.94 (bs, 1H); 8.91 (t, J = 5.3, 1H); 8.49 (s, 1H); 8.00–7.94 (m, 2H); 7.32–7.26 (m, 2H); 6.99–6.91 (m, 3H); 4.16 (t, J = 5.8, 2H); 3.72–3.67 (m, 2H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 166.61; 156.92; 131.65; 129,83; 127.91; 125.57; 121.12; 114.82; 66.16; 39.99.

HRMS/ESI: Calcd for C₁₅H₁₄N₄O₂ [M-H]^-: 281.3017; found: 281.3044.

N-(2-(thiophen-2-yl)ethyl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3k).

¹H NMR (400 MHz, DMSO-d6) δ: 8.78 (t, J = 5.4, 1H); 8.39 (s, 1H); 7.90 (s, 2H); 7.29 (d, J = 3.2, 1H); 6.92–6.89 (m, 2H); 3.53–3.49 (m, 2H); 3.06 (t, J = 7.1, 2H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 166.15; 141.98; 127.48; 127.65; 125.72; 124.49; 41.68; 29.69.

HRMS/ESI: Calcd for C₁₃H₁₂N₄OS [M-H]^-: 271.5471; found: 271.5489.

N-(pyridin-2-yl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3L).

¹H NMR (400 MHz, DMSO-d6) δ: 10.94 (s, 1H); 8.66 (s, 1H); 8.38–8.37 (m, 1H); 8.03 (dd, J₁ = 1.3, J₂ = 4.6, 1H); 7.93 (d, J = 5.2, 1H); 7.84–7.81 (m, 1H); 7.14 (dd, J₁ = 4.7, J₂ = 8.3, 1H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 166.19; 152.66; 148.54; 145.67; 138.57; 120.46; 117.86; 115.37.

HRMS/ESI: Calcd for C₁₂H₉N₅O [M-H]^-: 238.2319; found: 238.2341.

N-cyclohexyl-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3m).

¹H NMR (400 MHz, DMSO-d6) δ: 10.94 (s, 1H); 8.66 (s, 1H); 8.38–8.37 (m, 1H); 8.03 (dd, J₁ = 1.3, J₂ = 4.6, 1H); 7.93 (d, J = 5.2, 1H); 7.84–7.81 (m, 1H); 7.14 (dd, J₁ = 4.7, J₂ = 8.3, 1H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 166.19; 152.66; 148.54; 145.67; 138.57; 120.46; 117.86; 115.37.

HRMS/ESI: Calcd for C₁₃H₁₆N₄O [M-H]^-: 243.5411; found: 243.5437.

N-(thiophen-3-yl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3n).

¹H NMR (400 MHz, DMSO-d6) δ: 10.83 (s, 1H); 8.58 (s, 1H); 8.00 (d, J = 6, 1H); 7.96 (d, J = 6, 1H); 7.74 (d, J = 2, 1H); 7.47–7.46 (m, 1H); 7.32 (d, J = 3.2, 1H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 164.49; 137.45; 131.71; 125.95; 125.08; 122.67; 110.20.

HRMS/ESI: Calcd for C₁₁H₈N₄OS [M-H]^-: 243.2720; found: 243.2749.

N-pentyl-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3o).

¹H NMR (400 MHz, DMSO-d6) δ: 8.63 (bs, 1H); 8.45 (s, 1H); 7.94 (s, 2H); 1.58–1.54 (m, 2H); 1.33–1.31 (m, 5H); 0.90–0.87 (m, 3H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 166.17; 132.17; 125.57; 115.33; 114.30; 29.19; 22.36; 14.40.

HRMS/ESI: Calcd for C₁₂H₁₆N₄O [M-H]^-: 231.7142; found: 231.7151.

N-(2-(pyridin-2-yloxy)ethyl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3p).

¹H NMR (400 MHz, DMSO-d6) δ: 8.73 (s, 1H) 8.33 (s, 1H); 7.89 (d, J = 6, 1H); 7.83 (d, J = 6, 1H); 7.51 (d, J = 4.8, 1H); 7.37–7.34 (m, 1H); 4.05 (t, J = 4, 2H); 3.58–3.55 (m, 2H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 166.73; 162.39; 140.51; 140.01; 120.11; 105.56; 48.87.

HRMS/ESI: Calcd for C₁₄H₁₃N₅O₂ [M-H]^-: 282.0736; found: 282.0714.

N-(2-(pyridin-4-yloxy)ethyl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3q).

¹H NMR (400 MHz, DMSO-d6) δ: 15.72 (bs, 1H); 8.79 (t, J = 4.8, 1H); 8.37 (s, 1H); 7.95–7.86 (m, 2H); 7.63 (d, J = 7.2, 2H); 6.05 (d, J = 7.2, 2H); 4.05 (t, J = 5.2, 2H); 3.62 (d, J = 5.2, 2H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 141.96; 117.94; 55.05; 46.21; 91.79.

HRMS/ESI: Calcd for C₁₄H₁₃N₅O₂ [M-H]^-: 282.6719; found: 282.6741.

N-(3-(trifluoromethyl)phenyl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3r).

¹H NMR (400 MHz, DMSO-d6) δ: 16.00 (bs, 1H); 10.73 (s, 1H); 8.68 (s, 1H); 8.30 (s, 1H); 8.11–8.04 (m, 3H); 7.66–7.62 (m, 1H); 7.50–7.48 (d, J = 8, 1H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 166.86; 155.87; 145.82; 136.36; 130.71; 126.26; 114.63.

HRMS/ESI: Calcd for C₁₄H₉F₃N₄O [M-H]^-: 305.7256; found: 305.7227.

N-(5-methylthiazol-2-yl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3s).

¹H NMR (400 MHz, DMSO-d6) δ: 8.77 (s, 1H); 8.13 (d, J = 9.2.1H); 7.99 (d, J = 20.4, 1H); 7.24 (s, 1H); 7.04 (s, 1H); 2.39 (s, 3H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 165.25; 157.80; 134.70; 129.38; 126.90; 126.18; 11.61.

HRMS/ESI: Calcd for C₁₁H₉N₅OS [M-H]^-: 258.9006; found: 258.9021.

N-phenyl-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3t).

¹H NMR (400 MHz, DMSO-d6) δ: 10.46 (s, 1H); 8.69 (s, 1H); 8.08 (d, J = 8, 1H); 7.99 (d, J = 8, 1H); 7.87 (d, J = 8, 2H); 7.36 (t, J = 4, 2H); 7.09 (t, J = 7.6, 1H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 165.70; 139.64; 132.27; 129.05; 126.07; 124.19; 120.98; 116.57; 114.32.

HRMS/ESI: Calcd for C₁₃H₁₀N₄O [M-H]^-: 237.7456; found: 237.5468.

N-(3-methoxyphenyl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3u).

¹H NMR (400 MHz, DMSO-d6) δ: 10.34 (s, 1H); 8.58 (s, 1H); 7.99–7.95 (m, 2H); 7.47 (s, 1H); 7.36 (d, J = 5.2, 1H); 7.23 (t, J = 5.2, 1H); 6.66 (d, J = 4, 1H); 3.72 (s, 3H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 165.73; 160.00; 140.90; 129.95; 126.07; 113.14; 109.80; 106.62; 55.56.

HRMS/ESI: Calcd for C₁₄H₁₂N₄O₂ [M-H]^-: 267.2760; found: 267.2784.

N-((3s,5s,7s)-adamantan-1-yl)-1H-benzo[d][1,2,3]triazole-5-carboxamide (compound 3v).

¹H NMR (400 MHz, DMSO-d6) δ: 8.41 (bs, 1H); 7.88 (m, 2H); 7.82 (s, 1H); 2.12–2.07 (m, 11H); 1.67 (s, 7 H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 166.21; 133.36; 125.97; 115.83; 113.89; 52.19; 41.33; 36.59; 29.39.

HRMS/ESI: Calcd for C₁₇H₂₀N₄O [M-H]^-: 295.4972; found: 295.4991.

General procedure for Boc protection (compounds 5, 9)

To a solution of the appropriate ethyleneamine (4 or 8, 8 equiv) in anhydrous 1,4-dioxane (3 mL/mmol), tert-butyl dicarbonate (1 equiv) was added dropwise over 1h at room temperature. Then the reaction mixture was stirred for 18 h at room temperature, and the completion of the reaction was monitored by TLC. The reaction was quenched with water and extracted three times with dichloromethane (3 × 20 mL). The combined organic layers were washed with brine, dried over anhydrous Na₂SO₄, and concentrated under reduced pressure. The crude product was obtained in 90% yield and was utilized directly in subsequent reactions without further purification.

tert-butyl (2-aminoethyl)carbamate (5).

¹H NMR (400 MHz, DMSO-d6) δ: 6.76 (bs, 1H); 4.13 (s, 2H); 3.42–2.76 (m, 4H); 1.42 (s, 9H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 159.23; 81.54; 50.66; 44.23; 27.13.

HRMS/ESI: Calcd for C₇H₁₆N₂O₂ [M + H]⁺: 161.4210; found: 161.4223.

tert-butyl (2-hydroxyethyl)carbamate (9).

¹H NMR (400 MHz, DMSO-d6) δ: 4.99 (bs, 1H); 3.72 (d, J = 4, 2H); 3.32–3.28 (m, 2H); 2.54 (s, 1H); 1.31 (s, 9H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 159.45; 80.87; 61.14; 45.26; 28.33.

HRMS/ESI: Calcd for C₇H₁₅NO₃ [M + H]⁺: 162.2010; found: 162.2049.

tert-butyl (2-(pyridin-2-ylamino)ethyl)carbamate (7)

A solution of 5 (1 equiv), 6 (1.5 equiv), L-proline (0.3 equiv) and K₂CO₃ (2 equiv) in anhydrous DMSO (3 mL/mmol) was stirred for 18 h at 90°C. The reaction progress was monitored by TLC. Upon completion, the reaction mixture was filtered through a Celite pad to remove insoluble materials. The organic layer was washed sequentially with water and brine and then concentrated under reduced pressure. The crude product was purified by flash chromatography using dichloromethane/methanol (98:2) as the eluent, affording the desired product as a yellow oil in 43% yield.

tert-butyl (2-(pyridin-2-ylamino)ethyl)carbamate (7).

¹H NMR (400 MHz, CDCl₃) δ: 8.09 (d, J = 4, 1H); 7.43–7.39 (m, 1H); 6.60–6.57 (m, 1H); 6.43 (d, J = 8, 1H); 5.23 (bs, 1H); 4.73 (bs, 1H); 3.51–3.37 (m, 4H); 1.46 (s, 9H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 158.49; 155.89; 148.23; 138.44; 117.17; 106.25; 79.26; 48.44; 40.12; 28.71.

HRMS/ESI: Calcd for C₁₂H₁₉N₃O₂ [M-H]^-: 236.1459; found: 236.1423.

General procedure for the synthesis of compounds 11w-y

A solution of the appropriately substituted iodide (10w-y, 1.5 equiv) in anhydrous DMF (3 mL/mmol) was added to a solution of compound 9 (1 equiv) in anhydrous DMF (3 mL/mmol) at room temperature. Subsequently, K₂CO₃ (1.5 equiv) was added. The reaction mixture was stirred at room temperature for 18 h, with the progress monitored by TLC. Upon completion, the reaction mixture was washed with 1 M NaOH and extracted with chloroform (3 × 20 mL). The combined organic layers were washed with brine, dried over anhydrous Na₂SO₄, and concentrated under reduced pressure. The resulting crude product was purified by flash chromatography using a dichloromethane/methanol gradient as the mobile phase, starting from 99:1 and gradually changing to 9:1. The desired products were obtained in yields ranging from 63% to 51%, depending on the specific iodide utilized.

tert-butyl (2-phenoxyethyl)carbamate (11w).

¹H NMR (400 MHz, DMSO-d6) δ: 7.29 (m, 2H); 6.94 (d, J = 7, 3H); 6.15 (s, 1H); 3.93–2.87 (m, 4H); 1.27 (s, 9H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 159.31; 155.96; 129.54; 120.35; 114.12; 79.48; 68.71; 40.55; 28.12.

HRMS/ESI: Calcd for C₁₃H₁₉NO₃ [M-H]^-: 236.2990; found: 236.2981.

tert-butyl (2-(pyridin-2-yloxy)ethyl)carbamate (11x).

¹H NMR (400 MHz, DMSO-d6) δ: 8.14 (q, J₁ = 5, J₂ = 5, 1H); 7.72 (t, J₁ = 2, J₂ = 5, 1H); 6.97 (m, 2H); 6.79 (d, J = 8, 1H); 4.23 (t, J₁ = 6, J₂ = 6, 2H); 3.28 (t, J₁ = 6, J₂ = 6, 2H); 1.38 (s, 9H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 163.45; 155.49; 146.98; 138.17; 116.17; 109.23; 79.21; 68.43; 40.25; 28.34.

HRMS/ESI: Calcd for C₁₂H₁₈N₂O₃ [M-H]^-: 237.2870; found: 237.2855.

tert-butyl (2-(pyridin-4-yloxy)ethyl)carbamate (11yearears).

¹H NMR (400 MHz, DMSO-d6) δ: 7.53 (d, J = 8, 2H); 6.97 (t, J = 4, 1H); 6.04 (d, J = 8, 2H); 3.86–3.83 (m, 2H); 3.25–3.21 (m, 2H); 1.34 (s, 9H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 161.02; 155.47; 150.42; 109.78; 79.58; 68.49; 40.22; 28.13.

HRMS/ESI: Calcd for C₁₂H₁₈N₂O₃ [M-H]^-: 237.1317; found: 237.1349.

General procedure for Boc deprotection (compounds 2f, 2j, 2p, 2q)

To a solution of 7 (1 equiv) in anhydrous dichloromethane (3 mL/mmol), trifluoroacetic acid (5 equiv) was added at room temperature The reaction mixture was stirred for 18h at room temperature, and completion of the reaction was monitored by TLC. After completion, the solvent and excess TFA were removed under reduced pressure. The product was obtained in quantitative yield and was directly utilized in the subsequent reaction without further purification.

N1-(pyridin-2-yl)ethane-1,2-diamine (2f).

¹H NMR (400 MHz, DMSO) δ: 7.93 (d, J = 4, 1H); 7.33 (t, J = 4, 1H); 6.45–6.42 (m, 3H); 3.22–3.18 (m, 4H); 2.67 (t, J = 4, 2H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 158.77; 148.65; 138.48; 117.19; 106.25; 51.06; 37.00.

HRMS/ESI: Calcd for C₇H₁₁N₃ [M + H]⁺: 137.1860; found: 137.1844.

2-phenoxyethan-1-amine (2j).

¹H NMR (400 MHz, DMSO-d6) δ: 7.29–7.18 (m, 3H); 6.94 (d, J = 7, 3H); 6.15 (s, 1H); 3.93–2.78 (m, 4H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 159.42; 129.33; 120.21; 114.67; 71.29; 41.14.

HRMS/ESI: Calcd for C₈H₁₁NO [M + H]⁺: 138.0841; found: 138.0844.

2-(pyridin-2-yloxy)ethan-1-amine (2p).

¹H NMR (400 MHz, DMSO-d6) δ: 7.94 (s, 3H); 7.63 (m, 1H); 7.46 (m, 1H); 6.44 (d, J = 9, 1H); 6.29 (m, 1H); 4.11 (d, J = 12, 2H); 3.14 (t, J1 = 6, J2 = 6, 1H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 163.24; 146.48; 138.18; 116.71; 109.23; 71.66; 41.23.

HRMS/ESI: Calcd for C₇H₁₀N₂O [M + H]⁺: 139.0702; found: 139.0723.

2-(pyridin-4-yloxy)ethan-1-amine (2q).

¹H NMR (400 MHz, DMSO-d6) δ: 8.44 (d, J = 8, 2H); 8.09 (bs, 2H); 7.17 (d, J = 8, 2H); 4.49 (t, J = 4, 2H); 3.38 (bs, 2H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 161.38; 150.52; 109.41; 71.24; 41.18.

HRMS/ESI: Calcd for C₈H₁₁NO [M + H]⁺: 139.1743; found: 139.1703.

General procedure for the synthesis of compounds 13aa-ae

A solution of compound 1 (1 equiv) in anhydrous THF (3 mL/mmol) was treated with CDI (1 equiv). The reaction mixture was stirred at room temperature for 1 h, after which the appropriate amine (12a-ae, 1 equiv) was added. The reaction was stirred at room temperature for an additional 2 h. The progress of the reaction was monitored by thin-layer chromatography (TLC). Upon completion, the solvent was removed under reduced pressure. The crude product was purified by flash chromatography using a mobile phase gradient of dichloromethane/methanol, ranging from 99:1 to 92:8. The desired products were obtained in yields ranging from 74% to 46%, depending on the amine employed.

(1H-benzo[d][1,2,3]triazol-5-yl)(morpholino)methanone (13aa).

¹H NMR (400 MHz, DMSO-d6) δ: 7.95 (s, 1H); 7.88–7.82 (m, 2H); 3.98–3.71 (m, 8H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 168.22; 134.71; 125.50; 114.84; 49.49; 46.50; 26.43; 24.42.

HRMS/ESI: Calcd for C₁₁H₁₂N₄O₂ [M-H]^-: 233.2430; found: 233.2471.

N-cyclohexyl-N-methyl-1H-benzo[d][1,2,3]triazole-5-carboxamide (13 ab).

¹H NMR (400 MHz, DMSO-d6) δ: 16.00 (bs, 1H); 8.57 (s, 1H); 8.00 (q, J = 8.8, 2H); 3.92 (s, 3H); 2.25 (s, 11H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 169.09; 133.53; 125.19; 115.43; 114.61; 54.90; 46.09.

HRMS/ESI: Calcd for C₁₄H₁₈N₄O [M-H]^-: 257.1849; found: 257.1856.

(1H-benzo[d][1,2,3]triazol-5-yl)(pyrrolidin-1-yl)methanone (13ac).

¹H NMR (400 MHz, DMSO-d6): 15.84 (bs, 1H); 8.09 (s, 1H); 7.93 (d, J = 8.4, 1H); 7.58 (dd, J₁ = 1.1, J₂ = 8.6, 1H); 3.50 (t, J = 6.9, 2H); 3.42 (t, J = 6.4, 2H); 1.92–1.77 (m, 4H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 168.15; 134.79; 125.52; 114.92; 49.49; 46.48; 26.44; 24.33.

HRMS/ESI: Calcd for C₁₁H₁₂N₄O [M-H]^-: 215.8592; found: 215.8574.

(1H-benzo[d][1,2,3]triazol-5-yl)(4-methylpiperazin-1-yl)methanone (13ad).

¹H NMR (400 MHz, DMSO-d6) δ: 7.92–7.90 (m, 2H), 7.39 (d, J = 5.6, 1H), 3.44 (m, 4H); 2.29 (m, 4H); 2.16 (s, 3H).

¹³C NMR (400 MHz, DMSO-d6) δ: 169.15; 133.71; 125.33; 115.43; 114.59; 46.14.

HRMS/ESI: Calcd for C₁₂H₁₅N₅O [M-H]^-: 244.1233; found: 244.1247.

(1H-benzo[d][1,2,3]triazol-5-yl)(piperidin-1-yl)methanone (13ae).

¹H NMR (400 MHz, DMSO-d6) δ: 7.92–7.87 (m, 2H), 7.38 (d, J = 5.6, 1H) 3.58–3.23 (m, 4H); 1.58–1.45 (m, 6H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 165.43; 132.33; 125.74; 115.72; 114.20; 49.02; 32.89; 25.74; 25.41.

HRMS/ESI: Calcd for C₁₂H₁₄N₄O [M-H]^-: 229.2710; found: 229.2728.

General procedure for the synthesis of compounds 16af-ai

Carbonyldiimidazole (CDI, 1.5 equiv) was added to a solution of the appropriate carboxylic acids (15af-ai, 1 equiv) in anhydrous DMF (3 mL/mmol). The reaction mixture was stirred at room temperature for 1h. Subsequently, compound 14 (1.1 equiv) was added, and the reaction was maintained under stirring for 18 h at 70°C. Completion of the reaction was monitored by TLC. The reaction mixture was quenched with water and extracted three times with ethyl acetate (3 × 20 mL). The combined organic layers were washed with brine, dried over anhydrous Na₂SO₄, and concentrated under reduced pressure. The crude product was purified by flash chromatography using a gradient of dichloromethane/methanol as the mobile phase, starting from 99:1 to 92:8. The desired products were obtained in yields ranging from 74% to 46%, depending on the specific carboxylic acid utilized.

N-(1H-benzo[d][1,2,3]triazol-5-yl)picolinamide (16af).

¹H NMR (400 MHz, DMSO-d6) δ: 15.84 (bs, 1H); 10.94 (s, 1H); 8.78 (s, 1H); 8.77 (s, 1H); 8.20 (d, J = 8, 1H); 8.12 (t, J = 1.6, 1H); 8.12–7.71 (m, 2H); 7.70–7.96 (m, 1H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 166.19; 152.66; 148.54; 145.67; 138.57; 120.46; 117.86; 115.37.

HRMS/ESI: Calcd for C₁₂H₉N₅O [M-H]^-: 238.2327; found: 238.2344.

N-(1H-benzo[d][1,2,3]triazol-5-yl)thiophene-3-carboxamide (16 ag).

¹H NMR (400 MHz, DMSO-d6) δ: 10.83 (s, 1H); 8.58 (s, 1H); 8.00 (d, J = 6, 1H); 7.96 (d, J = 6, 1H); 7.74 (d, J = 2, 1H); 7.47–7.46 (m, 1H); 7.32 (d, J = 3.2, 1H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 164.49; 137.45; 131.71; 125.95; 125.08; 122.67; 110.20.

HRMS/ESI: Calcd for C₁₁H₈N₄OS [M-H]^-: 243.0419; found: 243.0431.

N-(1H-benzo[d][1,2,3]triazol-5-yl)thiazole-2-carboxamide (16ah).

¹H NMR (400 MHz, DMSO) δ: 15.63 (bs, 1H); 11.08 (s, 1H); 8.51 (s, 1H); 8.17 (d, J = 4, 1H); 8.14 (d, J = 4, 1H); 7.95 (d, J = 8, 1H); 7.86 (dd, J = 9, 1H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 164.10; 158.61; 144.50; 136.62; 127.27; 120.24.

HRMS/ESI: Calcd for C₁₀H₇N₅OS [M-H]^-: 244.2619; found: 244.2643.

N-(1H-benzo[d][1,2,3]triazol-5-yl)benzamide (16ai).

¹H NMR (400 MHz, DMSO-d6) δ: 10.46 (s, 1H); 8.69 (s, 1H); 8.08 (d, J = 8, 1H); 7.99 (d, J = 8, 1H); 7.87 (d, J = 8, 2H); 7.36 (t, J = 4, 2H); 7.09 (t, J = 7.6, 1H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 165.70; 139.64; 132.27; 129.05; 126.07; 124.19; 120.98; 116.57; 114.32.

HRMS/ESI: Calcd for C₁₃H₁₀N₄O [M-H]^-: 237.2540; found: 237.2589.

General procedure for the synthesis of compounds 17L and 17n

A solution of the benzotriazole derivative (3L-n, 1 equiv) in anhydrous THF (3 mL/mmol) was added to NaH (60% dispersion in mineral oil, 1 equiv) at 0°C. The reaction was stirred for 30 min at 0°C, and the iodomethane (1 equiv) was added. the reaction was stirred at room temperature for 3 h. Completion of the reaction was monitored by TLC. The reaction mixture was quenched with water and extracted three times with ethyl acetate (3 × 20 mL). The combined organic layers were washed with brine, dried over anhydrous Na₂SO₄, and concentrated under reduced pressure. The crude product was purified by flash chromatography using a gradient of dichloromethane/methanol as the mobile phase, starting from 99:1 to 92:8. The desired products were obtained in yields ranging from 74% to 46%.

2-methyl-N-(pyridin-2-yl)-2H-benzo[d][1,2,3]triazole-5-carboxamide (17L).

¹H NMR (400 MHz, DMSO-d6) δ: 10.99 (s, 1H); 8.71 (s, 1H); 8.42 (d, J = 1.2, 1H); 8.41 (d, J = 1.2, 1H); 8.23 (s, 2H); 8.20–7.87 (m, 1H); 7.21–7.18 (m, 1H); 4.57 (s, 3H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 166.23; 152.65; 148.44; 145.67; 143.68; 138.63; 132.47; 120.38; 119.19; 118.09; 115.27; 44.06.

HRMS/ESI: Calcd for C₁₃H₁₁N₅O [M-H]^-: 252.0963; found: 252.0971.

2-methyl-N-(thiophen-3-yl)-2H-benzo[d][1,2,3]triazole-5-carboxamide (17n).

¹H NMR (400 MHz, DMSO-d6) δ: 10.87 (s, 1H); 8.50 (s, 1H); 8.06–7.97 (m, 1H); 7.76 (s, 1H); 7.58 (s, 1H); 7.35 (s, 1H); 4.57 (s, 3H).

¹³C NMR (100.6 MHz, DMSO-d6) δ: 164.60; 148.53; 146.77; 136.40; 132.82; 126.95; 125.07; 122.59; 118.28;118.22; 110.18; 44.09.

HRMS/ESI: Calcd for C₁₂H₁₀N₄OS [M-H]^-: 257.2948; found: 257.2954.

Compounds solubilization

Compounds stock solutions were prepared at 100 mM in 100% DMSO. Most compounds are easily dissolved. When insoluble parts were present, compounds were dissolved by sonication in a sonicator bath for 8 min at room temperature.

Bacterial strains

E. coli ATCC 25922; S. enterica subsp. enterica serovar Typhimurium ATCC 14028; K. pneumoniae ATCC 13883 were purchased from American Type Culture Collection (ATCC, USA). E. coli strains JW2407-1 (ΔcysK743:kan) and JW2414-3 (ΔcysM750:kan), derived from E. coli K12, were purchased from the Coli Genetic Stock Center (Yale University, New Haven, CT, USA).²⁹ E. coli MG1655 strain was a generous gift of Prof. Emanuela Frangipani, University of Urbino, Italy.

Evaluation of minimal inhibitory concentration (MIC) by broth microdilution assays

MIC values were evaluated for each compound considered following CLSI guidelines with some modifications in two different broth media: M9 minimal medium and Mueller Hinton Broth (MHB, Difco, Sparks, USA, lot. n. 91568).⁴² M9 Minimal Medium was prepared following the recipe provided by Sambrook and Russell.⁴³ 2-fold dilutions 6.4–0.05 mg/mL of the tested compounds were performed in DMSO in separate 96-well microtiter U-plates (Greiner, Milan, Italy). Tested strains (E. coli ATCC 25922; S. enterica subsp. enterica serovar Typhimurium ATCC 14028; K. pneumoniae ATCC 13883) were brought to the logarithmic phase of growth in Mueller Hinton medium (MHB) by incubation at 37°C for 24 h. Bacterial suspensions were then centrifuged (2000 rpm, 4°C for 20 min), and the pellets were resuspended in phosphate buffer (pH 7, 0.1 M) to reach a final bacterial concentration of 10⁸ CFU/mL, adjusted by spectrophotometry (OD ranged between 0.08 and 0.13 at 620 nm). The suspensions obtained were further diluted 1:100 in broth medium to reach a bacterial concentration of 10⁶ CFU/mL. In a 96-well microtiter U-plate, for each well of the replicates 49 μL of broth medium were added and 50 μL were inoculated within 30 min, to obtain a final concentration of 5x10⁵ CFU/mL in a total volume of 99 μL. In each well of the plate, 1 μL of each compound at variable dilution in DMSO was added. A final dilution range of 256–0.5 μg/mL was tested for each compound, in the presence of 1% DMSO. For each test, three independent experiments with three replicates were performed. Growth and sterility controls were performed for each strain and for each tested compound. Plates were then incubated for 24 h at 37°C under aerobic conditions. After incubation, plates were read by unaided eye with a microtiter reading mirror and then the inhibition percentages for each tested concentration were calculated through the spectrophotometric measurement of optical density (OD) at 620 nm. MIC values were calculated as the arithmetic mean ± standard deviation (SD) of unaided eye reading. A quality control microorganism (E. coli ATCC 25922) was tested periodically to validate the accuracy of the procedure.

Protein expression and purification

Recombinant proteins were expressed starting from the S. Typhimurium gene sequences (optimized for E. coli codon usage, UniProtKB sequences P0A1E3 for OASS-A and P29848 for OASS-B) cloned between NdeI and BamHI restriction sites in a pET19m vector, a pET16b-derived plasmid in which factor X_A recognition site is substituted with Tobacco Etch Virus (TEV) protease recognition site.⁴⁴ The proteins, presenting an N-terminal hexahistidine-tag, were overexpressed in E. coli BL21 Tuner host in LB medium in the presence of 1 mM IPTG. After a 4-h induction at 37°C, cells were harvested by centrifugation. The pellets were resuspended in lysis buffer (50 mM sodium phosphate, 300 mM NaCl, pH 8.0) in the presence of protease inhibitors (0.2 mM PMSF, 0.2 mM benzamidine, 1.5 μM pepstatin A), 1 mM TCEP, 1 mg/mL lysozyme, and 0.2 mM PLP. After a 45-min incubation under agitation, the suspensions were sonicated and then centrifugated to separate the soluble fraction from the debris. Surnatants were then loaded on a pre-equilibrated cobalt column, connected to an ÄKTA FPLC system (GE Healthcare); the resin was washed with buffer containing 20 mM imidazole and the proteins were eluted using a buffer containing 300 mM imidazole. The protein-containing fractions were pooled, 1 mM EDTA added to the solution which was then dialyzed o/n in 10 mM HEPES, pH 8.0. Aliquots were flash-frozen in liquid nitrogen and stored at −80°C until further use.

Spectroscopic assays

The formation and disappearance of α-aminoacrylate intermediate was followed by recording absorbance spectra on a Cary 4000 UV-vis spectrophotometer (Agilent Technologies), equipped with a thermostatic bath. OASS catalyzes also a deacetylase side-reaction, by which in the presence of water α-aminoacrylate is degraded into pyruvate and ammonia, restoring the internal aldimine signal.⁴⁵ Noteworthy, the degradation of α-aminoacrylate intermediate in OASS-B is faster than in the isoform A, in which the intermediate is instead relatively stable^45,46; therefore, OAS was used at 750 μM for OASS-B and 75 μM for OASS-A, as the best assay conditions individuated for the qualitative evaluation of the compounds during a window time of 20 min, suitable for the activity test (Figure S2). The assay solutions contained 30 μM of either StOASS-A or StOASS-B and were added OAS and BT derivatives (or 1,2,4-triazole in control reactions) at 1 mM, in the presence of 10% DMSO. Reactions were carried out at 25°C in 100 mM HEPES, pH 7, and spectra were recorded after the addition of OAS and after either 1- or 10-min incubation with BT derivative or 1,2,4-triazole. Spectra were corrected for buffer contribution.

NMR

False product preparation – 0.9 mg of compound 3L (4.6 mM) were incubated with an equimolar amount of OAS in the presence of either 4 μM StCysK or StCysM, in 100 mM ammonium bicarbonate, 5% DMSO, pH 7. The reactions were incubated at RT for 60 min in the dark to avoid bleaching of the enzyme PLP cofactor. At the end of the enzymatic reaction, the solution was basified using sodium hydroxide (final pH 10) to promote the complete solubilization of the product, which is partially insoluble in a neutral environment. The false product was purified from the enzyme via diafiltration by 10 kDa-centrifugal devices (Merck-Millipore) and the collected flow through was lyophilized and stored at −20°C until further use. The powder was then resuspended in 600 μL of deuterated water (D₂O), and the solution was transferred to a 5 mm NMR tube. Spectra were recorded at 25°C on a JEOL ECZ 600R Spectrometer equipped with broadband ROYAL probe and on a 400 MHz Bruker Advance spectrometer equipped with Selective Inverse (SEI) probe. The determination of OAS peaks was performed by adding 2 μL of 100 mM OAS stock solution in D₂O (final concentration 300 μM) directly to the tube containing the solubilized false product. 1H-1D, 2D-1H COSY and 1H-13C HMQC spectra were acquired by applying presaturation during relaxation delay to suppress residual water signal. The 2D DOSY experiments were conducted with ledbpgppr2s pulse sequence.⁴⁷ Values for the diffusion time, d20 (Δ) and the gradient pulse length, p30 (δ∗0.5) were 80 and 1.0 ms, respectively. Spectra were processed and analyzed using Mestrenova 15.0.1 and Topspin 4.1.1, DOSY analysis was performed via dynamic center software.

Thiol assay

Intracellular reduced thiols were measured in E. coli ATCC 25922 bacterial cells following the procedure described by Turner et al.⁴⁸ with modifications. Briefly, a 1:1000 dilution of an overnight culture in LB (around OD₆₀₀ = 3.5) was grown at 37°C for 3 h in M9 medium, under shaking at 160 RPM in a flask (medium:air ratio 1:10). After 3 h the culture reached OD₆₀₀ = 0.25, and it was split into smaller flasks (medium:air ratio of 1:10), added with the compounds under study (3days, 3L, 3n), DMSO or 1,2,4-triazole,²¹ and let grow at 37°C under shaking at 160 RPM for 20 h. The DMSO-treated subculture was chosen as a negative control, while 1,2,4-triazole was used as a positive control. All compounds were tested at two concentrations equal to or 3-fold the compounds MIC, while 1,2,4-triazole was tested at a concentration of 5 mM (345 μg/mL). For all conditions, DMSO was maintained at 1%. Cultures were sampled at 0, 1 h, 2 h, 4 h, 6 h and 20 h time points after treatment withdrawing 2 mL (0, 1 h, 2 h) or 1 mL (4 h, 6 h, 20 h) from each condition. Pellets were obtained from each sample by centrifuging at 16,000 xg for 10 min at 4°C and then frozen at −20°C after removal of the supernatant, until further use. Pellets were resuspended in 1 mL of deoxygenated experimental buffer (50 mM Tris, 5 mM EDTA, 0.1% SDS, 0.1 mM DTNB, pH 8) by vortexing, and statically incubated for 30 min at 37°C. The samples were centrifuged for 20 min at 16,000 xg at RT. 200 μL of each sample was transferred to a 96-well plate and the absorbance was read at 405 nm. The molar concentration of RSH was calculated from the slope of the calibration curve, obtained using L-cysteine, freshly solubilized in the same buffer. Protein concentration in the supernatant was determined by bicinchoninic acid (BCA) assay,⁴⁹ using bovine serum albumin as a calibration curve standard. Each condition was replicated as a biological triplicate and thiol measurements were conducted in duplicate. Data significance was calculated with GraphPad Prism 10 applying a two-way ANOVA analysis (multiple comparisons correction by Dunnett test, alpha threshold 0.05) comparing each treatments time-point to those relative to DMSO-treated cultures.

Phenotype recovery assay of E. coli

Phenotype recovery was evaluated in the presence of 10 μg/mL or 100 μg/mL glutathione following the same procedure described for MIC determination in M9 medium, with the culture diluted to 5x10⁵ CFU/mL in the presence of glutathione. Growth and sterility controls in the presence and in the absence of glutathione were added to each plate.

Cytotoxicity assay on MDBK cells

The cytotoxicity assay was performed using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) test, following the method proposed by Donofrio et al.⁵⁰ The MDBK cell line used in this study was purchased from ATCC (ATCC CRL-6071), a reputable cell bank that provides authenticated and mycoplasma-free cell lines. As such, the cells were considered authenticated by the provider and not further tested in our laboratory. Mycoplasma testing was not additionally performed, as the ATCC certificate guarantees their absence at the time of purchase. MDBK cells were cultured on microtiter tissue culture plates in DMEM medium for 24 h at 37°C in the presence of 5% CO₂. After incubation, when the cell monolayer was at confluence, an aliquot of 1 μL of each compound in DMSO at different concentrations was added in each well, containing a volume of 100 μL of DMEM medium and MDBK cells, then the plates were reincubated at the same conditions. After incubation, 10 μL of MTT at 200 mg/mL was added to each well and incubated at 37°C for 6 h. At the end of the incubation, 100 μL of the solubilization solution (10% SDS in 0.01 M HCl) was added to each well and then incubated o/n. The MTT compound, a yellow tetrazolium salt, is reduced by mitochondrial enzymes (succinate dehydrogenase) of metabolically active eukaryotic cells to insoluble formazan crystals. In the presence of metabolically active cells, after the addition of a detergent solution (10% SDS in sterile PBS) that allows the formazan to be released from the cells, a violet color is seen in the medium. Instead, in the presence of non-viable cells, MTT is not reduced to formazan and therefore the solution will remain yellow. After incubation, plates were read with a spectrophotometer at λ = 620 nm. Positive controls - without any compounds – and with 1% DMSO were performed for each plate and three replicates for two independent experiments were performed for each compound.

Checkerboard assays

Antimicrobial activities of associations of 1,2,4 triazole, or compounds 3days, 3L and 3n with conventional antimicrobials were evaluated through checkerboard assay with minor modifications in M9 Minimal Medium.²⁵ Antimicrobials were all purchased from Sigma-Aldrich, MO, USA, namely colistin (batch. 049M-4836V), gentamicin (batch. SLBK9973V), ampicillin (batch. 016M4816V) and ciprofloxacin (batch. BCBX9934). For each assay, three experiments with three replicates were assessed. For each tested concentration of the different compounds, 96-wells microtiter U-plates were prepared with 2-fold serial dilutions of antimicrobial starting from the MIC value (μg/mL) for ten consecutive dilutions in 50 μL of M9 broth. In each well of the same replicate, 1 μL of each compound in DMSO was added at a fixed concentration, 100 times higher than the final desired concentration based on the MIC value obtained for the compound alone. Subsequently, 49 μL of the bacterial suspension at a concentration of 10⁶ CFU/mL, adjusted spectrophotometrically as reported above, were added to each well, reaching the final bacterial concentration of 5x10⁵ CFU/mL. Growth and sterility controls were performed for each experiment and for each replicate. Finally, the plates were incubated at 37°C in aerobic atmosphere for 24 h. After incubation, plates were read as in MIC assays. To evaluate the antimicrobial effect of the two molecules in association, the FIC Index was calculated as follows. The MICs of each of the two molecules tested individually and in combination with each other were evaluated and the results have been included in the following formula as reported by Meletiadis et al.⁵¹:

$$FIC=\left[\frac{{MIC}_{AB}}{{MIC}_A}\right]+\left[\frac{{MIC}_{BA}}{{MIC}_B}\right]$$

Where MIC_A is the MIC of the compound under evaluation and MIC_AB in combination is the MIC of compound under evaluation in combination with the conventional antimicrobial. MIC_B is the MIC of conventional antimicrobial and MIC_BA is the MIC of the conventional antimicrobial in combination with the compound under evaluation. From the results of the FIC index formula, the antimicrobial activity in combination of the two molecules can be considered: synergistic, additive, indifferent of antagonistic. If the FIC index is ≤0.5 the association is synergic, additive if FIC is between 0.5 and 1, indifferent if FIC is between 1 and 4 and antagonistic if FIC is ≥ 4.⁵¹

Quantification and statistical analysis

Statistical analysis was applied to calculate the significance of intracellular thiol concentration after the treatment of E. coli ATCC 25922 cultures with benzotriazole derivatives. Experimental details about growth and sample handling and treatment are reported in “Thiol assay” paragraph, in the STAR Methods section. For statistical analysis, each growth condition was replicated as a biological triplicate and thiol measurements were conducted in duplicate. Data significance was calculated with GraphPad Prism 10 (DotMatics) applying a two-way ANOVA analysis (multiple comparisons correction by Dunnett test, alpha threshold 0.05) comparing each treatment time-point to the corresponding time-point of DMSO-treated cultures. Results are reported in Paragraph 2.7 “evaluation of compounds effect on intracellular thiols production” and Figure 6, panels B and D. The bars represent the average of at least 3 biological replicates, and the error bars represent the standard deviation. The statistical significance is indicated as follow: ∗, p < 0.1; ∗∗, p < 0.01; ∗∗∗, p < 0.001.

Published: October 21, 2025