The most notable feature of such compounds is the introduction of a difluoromethyl fragment in the P2 region. As shown in Figures 4A,B, a compound with broad-spectrum antimicrobial activity, LPC-058 (Compound 7), was reported in 2016, which was effective against E. coli, P. aeruginosa, and against A. baumannii. It is one of the few compounds with activity against A. baumannii (the difluoromethyl fragment may have increased the ability to cross the lipid barrier), but there were significant toxic side effects of this compound at high doses (Lemaître et al., 2017; Titecat et al., 2016). Notably, LPC-058 was very active against Y. pestis, with a MIC similar to that of ciprofloxacin (0.03 μg/mL). Another compound with broad-spectrum antimicrobial activity, LPC-069 (Compound 8), was reported in 2017. It was effective against E. coli, P. aeruginosa, and against A. baumannii, and the researchers found no significant toxicities under high dose conditions. In contrast to LPC-058, LPC-069 was not toxic in mice and was capable of curing plague.

This compound is slightly different from LPC-058 at the hydrophobic end, Jinshi Zhao et al. replaced the terminal aniline structure of LPC-058 with a cyclopropargyl structure, which may effectively increase the water solubility of the compound and may reduce its toxicity. The study demonstrated that when LPC-233 (Compound 9) was administered via the intravenously (IV), intraperitoneally (IP), or orally (PO) routes, it effectively eliminated bacterial infections in a variety of murine models (Zhao et al., 2023).

At the highest dose (40 mg/kg per 12 h), LPC-058 exhibited side effects such as diarrhea, leukocytosis, and hepatotoxicity (Lemaître et al., 2017). The toxicity of LPC-058 may be related to the aniline moiety. Compared to LPC-058, in a 7-day repeated-dose toxicity study in rats, LPC-233 was safely administered for seven consecutive days at a maximum dose of 125 mg/kg every 12 h (250 mg/kg per day). Most notably, it had high oral bioavailability (73% under fasting conditions) and an oral dose of 40 mg/kg per 12 h completely cleared bacterial infection in a murine UTI model (Zhao et al., 2023). These two factors may contribute to the unique safety profile of LPC-233. (1) LPC-233 contains a hydroxamic acid moiety but no amino group, thus attenuating the amino-associated toxicity observed in ACHN-975. (2) LPC-233 is highly bound to plasma, with binding fractions (fb) ranging from 96.4% in mice to 98.8% in humans. Limited amounts of unbound LPC-233 in plasma may also contribute to the lack of cardiovascular toxicity. The lack of plasma effect may be explained by the extremely tight binding affinity of LpxC for LPC-233 (Ki = 8.9 pM), as demonstrated by plasma protein binding experiments in the presence of LpxC, which competes with plasma proteins for binding of LPC-233, thereby effectively stripping bound LPC-233 from plasma proteins. Thus, the high percent plasma binding properties may provide a safety net to mitigate potential side effects of LPC-233.

It is worth noting that preclinical studies have shown a promising safety profile, with no detectable adverse cardiovascular toxicity in dogs at a dose of 100 mg/kg. These promising properties make LPC-233 an attractive candidate for further development in clinical trials. LPC-233 is being considered for Phase I clinical trials to investigate its safety and efficacy in humans.

Since the first alkylsulfone hydroxamic acid inhibitor was reported and development, medicinal chemists have become increasingly interested in a variety of similar compounds. In Figure 5A, Compound 10 inhibitory activity against P. aeruginosa (PA01) was better than that of CHIR-090. The IC₅₀ of compound 10 against the LpxC enzyme of P. aeruginosa was 3.6 nM, and further optimization produced compounds 11 and 12. The addition of appropriately positioned substituents such as chlorine, fluorine and methoxy to the terminal substituent of the phenyl group of 10 was found to appropriately increase the potency of the compounds. As indicated, representative compound 11 exhibits broad-spectrum antimicrobial activity against Gram-negative bacteria, including E. coli, P. aeruginosa and K. pneumoniae (Tomaras et al., 2014).

As shown in Figure 5B, the intermolecular interactions associated with the Zn ion binding site are similar to those of CHIR-090. The alkyl sulfone structure in the P2 region extends into the UDP pocket and forms a hydrogen bonding force with Phe191. In contrast to previous studies, the oxygen atom in the amide structure within the P4 region forms a significant hydrogen bonding force with His19. This indicates that the structure may also function as a linker. It can be seen that the shorter hydrophobic structure as well as the absence of hydrophilic ends does not affect the antimicrobial activity of the inhibitors.

In 2020, Melanie Rodríguez-Alvarado et al. substituted F for Me to obtain compound 13 (Rodríguez-Alvarado et al., 2020). Theoretically, fluorine is closer to smaller hydrogen atoms than to larger methyl groups, and the substitution of fluorine for acidic H in the stereocenter to stabilize its conformation is expected to further enhance the activity of the compound. Unfortunately, it turned out in the end that the substitution of Me (or H) with F in the α-stereocenter of these LpxC inhibitors may not be very advantageous, and their activity is weaker than that of the lead compounds.

With the development of LpxC, such compounds are slowly fading into obscurity. The P2 region of such compounds is an alkyl sulfone structure, which has a close interaction force with LpxC. The most important feature of this class of compounds is that their P3 region is a non-alkyne structure, which provides a favorable reference for the subsequent development of LpxC inhibitors.

As shown in Figure 6A, an isoxazole analog compound 14 was reported in Novartis’ patent, while compound 15 was also included in the patent with superior anti-bacterial activities against E. coli (MIC = 0.063 μg/mL) and P. aeruginosa (MIC = 0.5 μg/mL) (Abdel-Magid, 2015).

In 2018, Patrick S. Lee et al. designed and synthesized compound 16 by computational scaffold jumping method (Lee et al., 2018). Due to the hydrogen bonding forces between His19 and the oxazolidinone structure, as well as the insertion of the alkyl sulfone into the UDP cavity. This may result in the loss of important hydrogen bonding interactions between the alkyl sulfone and the lysine residue (Lys 238), which is located near the catalytic site (Figure 6B). Compound 16 was potent against P. aeruginosa in a mouse model of neutropenic thigh infection and it showed strong antibacterial activity against Gram-negative bacteria such as E. coli and K. pneumoniae.

In 2019, Frederick Cohen et al. designed and developed the LpxC-516 (compound 17) (Figure 7A), which has favorable activity against P. aeruginosa (Krause et al., 2019). As shown in Figure 7B, similar to previous findings, there is a significant hydrogen bonding interaction between the alkyl sulfone headgroup and the lysine residue (Lys 238), which is located near the catalytic site adjacent to the methyl ether extending into the UDP pocket. The difference is that there is also a Thr190 at the hydroxamic acid group position that forms an important hydrogen bonding force with the hydroxyl group. The hydroxyl group at the hydrophobic terminus does not fully extend into the solvent-exposed region, but rather there is a hydrogen bonding force with the Val211 of the LpxC protein, which binds it tightly to the protein. LpxC-516 hydroxamic acid salt is unstable at elevated pH and does not increase in solubility at lower pH, Frederick Cohen et al. made it into a phosphate precursor (Compound 18), which was successful in increasing the drug solubility of the precursor drug. Its phosphate will be rapidly converted to the active drug, with a half-life of 2.4 min and 100% bioavailability of the precursor drug. Unfortunately when tested in a rat CV safety assay, the final lead prodrug compound showed unacceptable CV toxicity, which contrasts with the results obtained when the parent compound was given in the HPCD formulation, the reason for this discrepancy remains unclear (Krause et al., 2019).

Surivet et al. (2020), obtained compound 19 by structural optimization on the basis of PF-5081090 (Panchaud et al., 2019; Surivet et al., 2020). The UDP pocket alkyl sulfone structure and the hydroxamic acid structure were retained, and a hydrophobic channel for imidazolone was introduced in the P3 region. In previous SAR studies on CHIR-090 analogs, inhibitors with small alkyne units in the threonyl hydroxamic acid series achieved broad and potent antibiotic activity. As shown in Figure 7C, Compound 19 showed good activity against a wide range of Gram-negative bacteria.

SAR study of compound 19 indicated that when A group is a benzene ring, all these derivatives were potent LpxC inhibitors, as evidenced by their strong binding affinity for the LpxC enzymes of Ec (ΔTm = 11–12 K) and Pa (ΔTm = 13–14 K). The R group has a great influence on the antimicrobial activity of the bis-alkynyl (A group is alkynyl) compounds compared to the phenylethynyl (A group is a benzene ring) compounds. The compounds of this series (A group is alkynyl) exhibited good solubility (>350 μg/mL) at pH = 7, whereas their respective phenyl analogs (A group is a benzene ring) exhibited limited solubility (ranging between 7 and 112 μg/mL) at the same pH (Surivet et al., 2020).

As shown in Figure 7D, the overall conformation of the compound is altered due to the introduction of the five-membered heterocycle, the binding mode for the binding site of its hydroxamic acid to zinc ions is slightly different from that of the methylsulfone-based hydroxamic acid LpxC inhibitors. The binding mode here is identical to that of CHIR-090, and in addition to the three amino acids that immobilize the zinc ion, there are also two amino acids, Met62 and Glu77, as in CHIR-090, that form important hydrogen bonding interactions with the hydroxamic acid. In addition to this, unlike other methylsulfone hydroxamic acid LpxC inhibitors, there is also a hydrogen bonding force that Phe191 shares with methylsulfone and pyrrolobenzimidazolone. The cyclopropyl ring occupies a shallow hydrophobic side pocket in both fine conformations, with the hydroxymethyl group pointing toward the solvent in the opposite direction of the cyclopropyl unit (in both conformations). The hydrophobic terminus extends into the solvent-exposed region, and due to the bending of the terminal hydroxyl group, compound 19 has hydrogen bonding forces with Gly209 of this protein. Similar to all related derivatives, the methylsulfone hydroxamic acid ester head and the pyrroloimidazolone core bind to Pa LpxC, and the bis-alkynyl metal needle fills the substrate channel, projecting the moiety R onto the edge of the protein.

In 2016, Tetsuya Kawai et al. focused on the improvement of enzyme inhibition and antimicrobial activity to optimize sulfonamide-based non-alkyne LpxC inhibitors (Kawai et al., 2017). Compound 20 showed moderate enzyme inhibitory activity against P. aeruginosa and in vitro antimicrobial activity against a wide range of Gram-negative bacteria (Figure 8A). As shown in Figure 8B, the nitrogen of sulfonamide appears to have the potential to utilize the critical Lys 238 hydrogen bond in P. aeruginosa LpxC. In Figure 8A, compound 21 with 2-arylbenzofuran as a hydrophobic moiety was found to have very excellent in vitro antimicrobial activity against a wide range of Gram-negative bacteria (Ec, Pa and Kp) and it has a good slow release activity. The reason for this may be that the introduction of 2-arylbenzofuran in the P4 region led to a change in the overall conformation of compound 21 (Figure 8C). The binding pattern of its hydroxamic acid to zinc ions is similar to that of CHIR-090, and the tailed furan structure forms a hydrogen bonding interaction with Arg201. Compound 21 showed impressive antimicrobial activity and was a precursor for further exploration as an antimicrobial agent for non-alkyne LpxC inhibitors.

In 2019, Magdalena Galster et al. performed a divergent synthesis of the hydroxamic acid portion of benzyloxyacetyl hydroxamic acid to obtain phenylglycol derivatives exhibiting a variety of Zn²⁺ binding groups (Galster et al., 2019). For example, hydrazides, amides, sulfonamides, o-diols, thiols, thioesters, and hydroxypyridone derivatives were synthesized and tested for antimicrobial activity, and ultimately no compounds were found to be more prominently active.

In 2020, Katharina Hof et al. performed a divergent synthesis of the hydroxyl portion of benzyloxyacetyl hydroxamic acid to obtain a series of triazole-based LpxC inhibitors, and ultimately found that the triazole portion was a suitable functional group (Figure 9, compounds 29–31) (Hoff et al., 2020). This could address additional binding sites for LpxC, but in in vitro antimicrobial assays, the ability of its series of compounds to resist E. coli was not quite as prominent.

In 2024, Sebastian Mielniczuk et al. developed a benzyloxyacetohydroxamic acid LpxC-based inhibitor, it specifically targets the UDP binding site of the enzyme (Mielniczuk et al., 2024). Compound 33 showed high affinity for the LpxC enzymes of E. coli and P. aeruginosa, and it exhibited low MIC values for E. coli (E.coli BL21 (DE3) (1–4 μg/mL) and E. coli D22 (0.031 μg/mL)). Compound 33 has potential off-target effects on several mammalian zinc-dependent enzymes (MMP 1–3 and TACE), suggesting good selectivity for LpxC. Compound 33 has low plasma stability in mouse plasma and thus needs to be improved in further optimization steps. Structural optimization of the P2 region alone may achieve an improvement over LpxC inhibitors.

In 2021, Alexander Dreger et al. obtained a series of C-furanoside LpxC inhibitors in a chiral pool synthesis using D- and L-xylose as starting materials (Dreger et al., 2021). This is similar to a class of glycoside-containing LpxC inhibitors synthesized by Oddo et al. in 2012, with R and S conformational differences only at the hydroxyl carbon atom of the five-membered sugar ring. Compound 34 exhibited 2,3-trans- and 4,5-trans-configurations and was the most active stereoisomer of the furanoside series studied (Figure 10A). As in Figure 10B, the hydroxyl group on the furan of compound 34 extends into the cavity and forms a hydrogen bond with water. At the site where the hydroxamic acid binds to the zinc ion, in addition to the three amino acids that immobilize the zinc ion, four amino acids, Glu77, His264, Met62, and Thr190, form strong hydrogen bonds with the hydrogen atom of the hydroxyl group of the hydroxamic acid, the hydrogen atom of the hydroxyl group, the hydrogen atom of the amino group, and the oxygen atom of the carbonyl group, respectively. This allows the protein to strongly immobilize the compound 34 in the protein. In the same year, Alexander Dreger et al. designed a series of monohydroxy-substituted tetrahydrofuran derivatives LpxC inhibitors based on compound 34 (Dreger et al., 2021). The most promising LpxC inhibitor among them is 3-hydroxytetrahydrofuran derivative 35, but its inhibitory activity is inferior to compound 34.

Notably, in experiments with and without the efflux pump inhibitor PAßN, compound 34 was tested against P. aeruginosa ATCC-27853 and a significant 8-fold reduction of the MIC was observed in the presence of PAßN (three serial dilution steps). In a further optimization step, the bacterial permeability of the C-glycoside LpxC inhibitor needs to be increased.

In 2014, AstraZeneca Murphy Benenato et al. screened the library of hydroxamic acid compounds and obtained compound 36 (Figure 11B), which is a derivative of the matrix metalloproteinase (MMP-2, −9, −8, −13) inhibitor JR-130830 (Murphy-Benenato et al., 2014). Compound 36 has activity against P. aeruginosa, but its activity against E. coli ARC523 is weak (MIC>200 μmol/L). Using compound 36 as the lead compound and retaining the tetrahydropyran ring and optimize the hydrophobic region. As shown in Figure 11B, the sulfoxide structure of compound 37 formed a hydrogen bond with His19 leading to a higher affinity of compound 37 for the protein. The hydrophobic end of compound 37 extends further into the hydrophobic channel, and shows increased activity against E. coli compared to compound 36. Finally although it exhibited low molar potency against P. aeruginosa LpxC, its effective MMP inhibition was retained, which increased off-target effects. Since then, such compounds have not been mentioned.

Based on ACHN-975, compounds 38 and 39 (Figure 12A) were optimized and synthesized, respectively. In 2017, Jing et al. reported that the linker is a dicyclic analog, representing compound 38 with good broad-spectrum antimicrobial activity (Zhang et al., 2017). As in Figure 12B, Phe191 is 2.3 Å away from the amino group of compound 38 and forms a hydrogen bond, and the terminal hydroxyl group bends inward and forms a hydrogen bond with Gly209. Acute tolerability problems observed in phase IV rats prevented its entry into clinical trials due to its intravenous acute toxicity and inhibition of sodium channel site II (97%, 25 μmol/L, IC₅₀ = 400 nmol/L).

Based on ACHN-975, in order to improve the water solubility and reduce the toxicity, Piizzi et al. of Novartis optimized the bis-alkynyl structure of the P4 region of ACHN-975. Compounds 39 was effective against MDR P. aeruginosa, however, its antimicrobial activity is no more desirable than ACHN-975 (Piizzi et al., 2017). As shown in Figure 12C, in the important Zn ion binding site and UDP pocket region, compound 39 interacts with the LpxC protein of P. aeruginosa similarly to CHIR-090, and the alkynyl propoxy occupies the hydrophobic cavity and extends into the solvent-exposed region.

In 2016, Yucheng Wang et al. introduced the phosphate group into the P1 region of CHIR-090 to obtain compound 40, which showed comparable inhibition of E. coli and P. aeruginosa to CHIR-090 (Figure 13A). As in Figure 13B, Met62 and His264 are 1.8 Å and 2.7 Å away from the hydroxyl group of phosphoric acid, respectively, and form hydrogen bonds with it. Phe191 forms a hydrogen bond with the hydroxyl group and amide of threonine, and due to the introduction of the phosphate group, His19 is more distant from the amide of compound 40 (3.5 Å) relative to CHIR-090 (3.3 Å).

In 2020, Takeru Furuya et al. used CHIR-090 as a lead compound, focusing on the hypothesis that the N-hydroxycarboxamide warhead could overcome the previous toxicity problems of LpxC inhibitors by substituting the terminal hydroxamic acid (Furuya et al., 2020). As shown in Figure 14A, Compound 41, which was just successfully obtained, was inhibitory to the LpxC enzymes of E. coli and P. aeruginosa, and showed comparable viability to E. coli ATCC-25922. As in Figure 14B, the binding site of compound 41 to the zinc ion is the carbonyl structure of the N-hydroxyformamide, and then Phe191 forms a hydrogen bond with the hydroxyl group of compound 41. The use of benzotriazole structure to replace the benzene ring structure effectively enhanced the activity, the reason for this is that the benzotriazole forms a strong hydrogen bond with the imidazole group of His19 (Figure 14C). A potent and effective LpxC inhibitor against Gram-negative bacterial infections was obtained (compound 42), but it had a strong inhibition rate of AChE and nearly 2/3 of the mice died in the safety evaluation of the hemodynamic test in rats. To solve this problem, the morpholine tail was replaced with known morpholine bioelectronic equivalents and structurally related heterocycles. Compound 43 stood out in terms of lower rat plasma protein binding (9% unbound compared to 3% for compound 42) and weaker AChE inhibition (>100 μM compared to 3.5 μM for compound 42). In safety evaluations of hemodynamic tests, rats receiving compound 43 showed hypotension with plasma concentrations of 11–16 μg/mL and did not rebound during the recovery period. Despite acceptable PK and in vitro safety data, the compound 43 showed unacceptable swelling effects in rat studies.

In 2020, Yousuke Yamada et al. synthesized 2-(1-S-hydroxyethyl) imidazole derivatives exhibiting low nanomolar inhibition of LpxC, which are virtually unaffected by the presence of albumin (Yamada et al., 2020). In 2021, Fumihito Ushiyama et al. designed TP 0586532 to be effective against carbapenem-resistant K. pneumoniae without posing cardiovascular risks (Ushiyama et al., 2021). TP 0586532 shows potent activity against a broad range of carbapenemase gene-positive strains, including blaNDM, blaKPC, blaVIM, blaOXA and blaIMP (Fujita et al., 2022). TP 0586532 has excellent activity against mucin-resistant strains, ESBLs-producing E. coli, and carbapenem-resistant K. pneumoniae. As shown in Figure 15A, the IC₅₀ of TP 0586532 for human MMPs is more than 700 times higher than that for E.coli LpxC, so TP 0586532 has excellent selectivity for human MMPs.

In Figure 15B, the N and O of 2-(1-S-hydroxyethyl) imidazole chelate Zn ions by hydrogen bonding forces, which is the same as compound 44. The five-membered oxazole heterocycle of TP 0586532 then forms a hydrogen bond with His19, the hydrophobic structure is inserted into the hydrophobic cavity in a compliant manner, and the carboxylic acid structure at the hydrophobic terminus bends toward the side of the LpxC protein, but there is no associated hydrogen-bonding force with the amino acid residue in question.

In vitro LPS level assays demonstrated that TP 0586532 exerted antimicrobial activity by inhibiting LPS synthesis and that TP 0586532 reduced total LPS levels in a concentration-dependent manner (Fujita et al., 2022). TP 0586532 enhances the antimicrobial activity of various antibiotics (especially meropenem) against Enterobacteriaceae, including CRE (Yoshida et al., 2022). The mechanism of action is related to increasing bacterial cell membrane permeability and facilitating the entry of meropenem into the cell. This study demonstrates that meropenem in combination with TP 0586532 has therapeutic potential for the treatment of severe CRE infections, including strains with high levels of resistance to meropenem. This also suggests that combination therapy with antimicrobials of different antimicrobial mechanisms can compensate for the deficiency of LpxC inhibitors. TP 0586532 exerts potent efficacy in mouse models of systemic infection and pneumonia caused by gram-negative pathogens (Fujita et al., 2022; Ushiyama et al., 2021). TP 0586532 is expected to be the most promising LpxC inhibitor to enter clinical trials after ACHN-975.

Small molecule LpxC inhibitors with 2-(1-S-hydroxyethyl) imidazole as the active center have become a hotspot for optimization of LpxC inhibitors. As shown in Figure 15C, in 2022, Forge Therapeutics’ patent disclosed that compounds 46 and 47 could be used to treat bacterial infections, particularly pneumonia, both compounds had better antimicrobial activity than TP 0586532 (WO2022173756; WO2022173758, Min et al., 2022a,b). In 2024, Blacksmith Pharmaceuticals’ patent disclosed that compound 48 could be used to treat Pseudomonas aeruginosa infections, lung disease, and infectious pneumonia (WO2024036170, David et al., 2024). In 2024, researchers from Shanghai Angry Pharmaceuticals and Zhejiang Haizheng Pharmaceuticals introduced an aromatic acetylene group in the P4 region to obtain compound 49 for treating Gram negative bacterial infections, of which oral bioavailability is 89.4% (WO2024083177, Xiangui et al., 2024).