Siderophores as \"Trojan Horses\": tackling multidrug resistance?

Article pubs.acs.org/jmc

Siderophore Receptor-Mediated Uptake of Lactivicin Analogues in Gram-Negative Bacteria Jeremy Starr,#,† Matthew F. Brown,#,† Lisa Aschenbrenner,§ Nicole Caspers,¶ Ye Che,⧧ Brian S. Gerstenberger,† Michael Huband,§ John D. Knafels,¶ M. Megan Lemmon,§ Chao Li,† Sandra P. McCurdy,§ Eric McElroy,† Mark R. Rauckhorst,† Andrew P. Tomaras,§ Jennifer A. Young,† Richard P. Zaniewski,§ Veerabahu Shanmugasundaram,⧧ and Seungil Han*,#,¶ †

Medicinal Chemistry, ⧧Computational Chemistry, §Antibacterials Research Unit, and ¶Structural Biology, Pfizer Global Research and Development, Eastern Point Road, Groton, Connecticut 06340, United States S Supporting Information *

ABSTRACT: Multidrug-resistant Gram-negative pathogens are an emerging threat to human health, and addressing this challenge will require development of new antibacterial agents. This can be achieved through an improved molecular understanding of drug−target interactions combined with enhanced delivery of these agents to the site of action. Herein we describe the first application of siderophore receptormediated drug uptake of lactivicin analogues as a strategy that enables the development of novel antibacterial agents against clinically relevant Gram-negative bacteria. We report the first crystal structures of several sideromimic conjugated compounds bound to penicillin binding proteins PBP3 and PBP1a from Pseudomonas aeruginosa and characterize the reactivity of lactivicin and β-lactam core structures. Results from drug sensitivity studies with β-lactamase enzymes are presented, as well as a structure-based hypothesis to reduce susceptibility to this enzyme class. Finally, mechanistic studies demonstrating that sideromimic modification alters the drug uptake process are discussed.

■

INTRODUCTION Since Fleming’s landmark publication of the discovery of penicillin in 1929,1 the β-lactam antibiotic class has produced more drugs than any other chemotype and has had a profound impact on extending human life. Despite the development of new anti-infective drugs, bacterial infections continue to represent a major global health threat, especially infections with serious Gram-negative pathogens such as Pseudomonas aeruginosa, Acinetobacter baumannii, Klebsiella pneumoniae, and Escherichia coli.2 Particularly vulnerable are hospitalized patients with comorbidities or weakened immune systems who are at high risk of infection with multidrug-resistant variants. βLactams continue to be utilized as the foundation of treatment for multiple conditions arising from these infections, such as ventilator-associated pneumonia, intra-abdominal infections, and bacteremia,3−5 but their effectiveness continues to diminish due to ever increasing prevalence of drug-resistant strains.6,7 There is a critical need for new therapeutic options for these difficult to treat infections.8 Penicillin binding proteins (PBPs) are a large class of enzymes ubiquitously expressed across the spectrum of prokaryotic pathogens and are the biochemical targets of βlactam drugs.9 Most PBPs are bound in the peptidoglycan layer and, in the case of Gram-positive bacteria, are readily accessible to drugs capable of transiting the relatively permeable cell wall. By contrast, an inhibitor must pass through the formidable © 2014 American Chemical Society

outer membrane of Gram-negative bacteria to gain access to its target in the periplasmic space. Classically, a β-lactam with Gram-negative activity exhibits specific structural and physicochemical properties carefully tuned to enable passage through porin channels while not engaging efflux transporters.10 Inhibition of PBPs then leads to the introduction of flaws in the peptidoglycan layer, resulting in loss of structural integrity, osmotic control, morphological changes, and eventual cell lysis. In response to drug pressure, bacteria utilize a variety of drug-neutralizing mechanisms, including the production of β-lactamase (BLA) enzymes. To combat this resistance mechanism clinically, β-lactams are often administered with BLA inhibitors. However, inhibitor-resistant BLA variants are prevalent and contribute to significant treatment failure. Therefore, novel PBP inhibitors with reduced susceptibility to BLAs are critical to the development of next-generation antibacterial therapies. It is noteworthy that while decades of research across the pharmaceutical industry has generated dozens of marketed β-lactam drugs, no other PBP-inhibiting chemotype has thus far reached the market. Bacteria have developed a highly effective mechanism to acquire iron, a critical element, from their host organism. The process involves the release of siderophores with extremely Received: February 10, 2014 Published: April 2, 2014 3845

dx.doi.org/10.1021/jm500219c | J. Med. Chem. 2014, 57, 3845−3855

Journal of Medicinal Chemistry

Article

Figure 1. Lactivicin (1) and comparator agents.

could positively influence the discovery of novel, nextgeneration Gram-negative antibacterial agents.

high affinity for iron, which can effectively compete with the host’s endogenous iron binding proteins. Following secretion and chelation of iron, the siderophore complexes are actively retrieved through binding to specialized bacterial receptors. βLactams conjugated to structurally simple, small molecule siderophore mimics (hereafter referred to as “sideromimics”) such as catechols and hydroxypyridones have been described which utilize this internalization process, leading to active cellular uptake. Despite extensive industrial and academic research efforts utilizing this “Trojan horse” approach to enhance intracellular delivery of β-lactams,11−20 a similar approach has not been reported for non-β-lactam PBP inhibitors. Here we describe the application of the sideromimic conjugation strategy to lactivicins, a non-β-lactam natural product class of PBP inhibitors of microbial origin (Figure 1).21−23 We report that lactivicin−sideromimic conjugates provide an improved PBP profile relative to aztreonam, with potent inhibition of PBP3 and 1a as well as enhanced potency vs PBP1b. This expanded PBP inhibitory profile may eventually prove to be an advantage for the lactivicin drug class. We report the first X-ray crystal structures of sideromimic-conjugated lactivicin analogues bound to Pseudomonas aeruginosa PBP3 (Pae PBP3) and PBP1a (Pae PBP1a), as well as the related βlactam sideromimic-conjugate 4 (BAL30072) (Figure 1)16,17 bound to Pae PBP3 and discuss novel insights gained from the effectiveness of these compounds. We report the sensitivity of the lactivicin sideromimic conjugates to β-lactamase enzymes, and based on structure-derived knowledge, we propose that optimal attachment and positioning of the sideromimic moiety enhances intrinsic activity and could lead to reduced susceptibility to drug-degrading BLA enzymes. In addition, to aid analogue design and improve the molecular understanding of covalent bond formation, relative reactivities of lactivicin and monobactam analogues are characterized using a computational method which defines the activation barrier of a transition-state structure derived from the reaction of the electrophilic carbonyl with a simple model nucleophile, CH3O−. We have also characterized the role of bacterial siderophore receptors in the uptake of the new phthalimide-conjugated lactivicin analogue 17 and have discovered that it utilizes a broader set of Ton-Bdependent siderophore receptors as compared to related βlactam sideromimic conjugates. Taken together, these results

■

RESULTS AND DISCUSSION Relative Reactivity of Lactivicin and Monobactam Analogues. Critical to the mechanism of β-lactam drugs is the presence of an electrophilic carbonyl that covalently modifies a highly conserved PBP active site serine hydroxyl group leading to the formation of an acyl enzyme complex that is significantly lower in energy due to relief of ring strain. By comparison, a monocyclic five-membered lactam or larger homologue would not be expected to exhibit a strain-induced irreversibility due to the exceptional stability of such lactams, while a highly strained three-membered ring lactam would be too reactive in water to survive en route to its bacterial target. However, the primacy of four-membered ring lactams as PBP inhibitors has been challenged.24−26 For example, the lactivicin class is unusual because it is the only known naturally occurring PBP inhibitor class that does not contain a β-lactam. Instead of a β-lactam ring, lactivicin (1, Figure 1) and its derivatives contain cycloserine (a five-membered lactam) and γ-lactone (a fivemembered cyclic ester) rings. Recently published X-ray crystallographic data supports the covalent mechanism of action of lactivicins wherein the carbonyl of the cycloserine ring reacts with the Gram-positive Streptococcus pneumoniae active site serine hydroxyl to form a covalent bond analogous to that formed by a typical β-lactam drug (Figure 2A).27 Nucleophilic substitution at the carbonyl group of an amide usually occurs in a stepwise manner initiated by the formation of a tetrahedral intermediate as the rate-determining step (Figure 2A). Here the relative reactivity of covalent inhibitors such as β-lactams and lactivicins was theoretically characterized by molecular orbital calculations as the activation barrier (ΔG⧧) of the transition state structure in an aqueous solution with a simple nucleophile (CH3O−) representing the active site serine. While this approach does not factor in the multiple hydrogenbonding interactions which stabilize the transition state in the oxyanion hole of the protein, it can be useful in the design of covalent inhibitors. For example, previously reported semiempirical calculations (CNDO/2) by Boyd et al. described an optimal reactivity range for β-lactams which provides effective covalent inhibition of PBPs combined with the hydrolytic stability needed to provide antibacterial activity in an aqueous 3846

dx.doi.org/10.1021/jm500219c | J. Med. Chem. 2014, 57, 3845−3855

Journal of Medicinal Chemistry

Article

environment.28 Here, we compare ΔG⧧ values, derived from high-level DFT calculations, for a number of reported lactivicin analogues to evaluate the influence of ring strain and electronwithdrawing substituents on reactivity and subsequent antibacterial activity (Figure 2B). While the expanded lactam ring size of lactivicins relative to β-lactams such as aztreonam (Figure 1) may be expected to lead to reduced reactivity, the calculated activation barriers for each are quite similar (ΔG⧧ for 1 and monobactam 9 = 31.3 and 30.7 kcal/mol, Figure 2B). Our assumption is that the reduced ring strain of lactivicins relative to monobactams is overcome by the electronic activation provided by the appended lactone moiety and the cycloserine ring oxygen, thus providing similar ΔG⧧ values, sufficient PBP reactivity, and the resulting antibacterial activity. In addition, lactivicins are known to undergo a ring-chain tautomerization (i.e., 5 to 6) which leads to equilibration of the γ-lactone chiral center (Figure 2A).29 The ring-opened tautomer 6 may be expected to be a highly reactive species, and calculations support this statement as virtually no barrier exists, suggesting that a spontaneous reaction would occur with the nucleophile CH3O−. Nevertheless, the fact that lactivicin analogues exhibit sufficient chemical stability to provide antibacterial activity in an aqueous environment suggests this may not be the case. Harada et al. provide an explanation suggesting that the imminium ion 6 may be stabilized by the tethered carboxylate(s).29 So while the existence of tautomer 6 is necessary to enable the reported lactone equilibration process, its impact on overall chemical stability and reactivity with the PBP active site serine remains unknown. The potential of tautomer 6 to react with PBP is captured in Figure 2a; however, the calculations described above are based solely on the transition of the ring-closed tautomer 5 to 7. A number of research groups have reported efforts to modify the lactivicin core, most of which lead to a significant or complete loss of antibacterial activity. These results can be rationalized with calculated ΔG⧧ values. For example, the expanded lactam ring analogue 10 was reported to be inactive

Figure 2. (A) Proposed mechanisms of acylation of lactivicin analogues. (B) The activation barrier (ΔG⧧) of forming the transition-state structure derived from the reaction of the electrophilic carbonyl with a simple nucleophile CH3O− (DFT calculations).

Table 1. MICs and Pae PBP IC50 Data for Lactivicin Analogues

MIC (μg/mL)a a

Pae PBP IC50 (μM)

compound

Eco EC-28

Pae PA01

Pae 1091-05

Kpn KP-3700

Aba AB-3167

1a

1b

2

3

nb

aztreonam 13 14 15R 15S 16R 16S 17

0.125 NTc 2 >64 >64 1 1 0.25

4 32 8 >64 >64 16 2 0.5

4 32 4 >64 >64 2 64 0.5

>64 4 32 >64 >64 16 8 0.5

32 32 0.5 >64 >64 0.5 0.5 0.06

3.34 ± 0.57 0.046 0.046 0.11 ± 0.032 0.065 ± 0.033 0.046 0.046 0.030

2.87 ± 0.64 1.23 2.46 2.44 ± 0.03 1.74 ± 0.87 0.82 0.82 0.27 ± 0.39

>300 33.3 22.2 47.1 ± 23.5 33.3 6.41 ± 5.23 3.7 3.7 ± 1.8

0.008 ± 0.004 0.092 0.14 0.11 ± 0.03 0.065 ± 0.033 0.065 ± 0.033 0.092 0.046

8 1 1 2 2 2 2 2

a c

Eco: E. coli, Pae: Pseudomonas, Kpn: Klebsiella pneumoniae, Aba: Acinetobacter baumannii. bn = number of replicates for PBP IC50 determination. NT: not tested. 3847

dx.doi.org/10.1021/jm500219c | J. Med. Chem. 2014, 57, 3845−3855

Journal of Medicinal Chemistry

Article

Figure 3. Interactions of compound 14 and 4 in the active site of the Pae PBP3. (A) Active site of Pae PBP3 bound to compound 14 (navy). (B) Active site of Pae PBP3 bound to 4 colored in green. (C) Molecular surface of the Pae PBP3 in the active site region in complex with compound 4 (green). The residues of hydrophobic aromatic wall (Tyr-Tyr-Phe) and Val333 are shown in orange. (D) Molecular surface of the Pae PBP3 active site in complex with aztreonam in cyan (PDB code: 3PBS). (E) Molecular surface of the Pae PBP3 active site in complex with 3 in pink (PDB code: 3PBT).

vs a panel for Gram-positive and negative bacteria.30 Ring expansion would be expected to reduce reactivity, and this is reflected in the 2.7 kcal/mol increase in the ΔG⧧ value for 10 vs 1 (Figure 2B). In addition, a number of reports describe efforts to replace the lactone N-activating group. For example, 11 contains the N-sulfate activating group common to the monobactam drug class. The significantly lower ΔG⧧ of this analogue (24.6 kcal/mol) as compared to 1 (31.3 kcal/mol) or monobactam 9 (30.7 kcal/mol) is suggestive of high level reactivity and poor hydrolytic stability, which may explain the reported lack of antibacterial activity.31 Similarly, the low calculated ΔG⧧ (19.3 kcal/mol) for compound 12 explains the reported lack of antibacterial activity and poor aqueous stability.32 While the utilization of calculated ΔG⧧ values has great potential to enable the rational design of next-generation lactivicin (or β-lactam) core structures, our initial efforts maintained the cycloserine-lactone core structure found in the lactivicin natural product (vide infra). Initial Application of Sideromimic-Conjugated Lactivicins: Oxime-Linked Dihydroxypyridone 14. Lactivicin lead optimization efforts reported by Takeda describe a number of potent analogues, including 13 (Table 1), possessing a fairly balanced spectrum of activity toward Gram-negative pathogens and in vivo efficacy.33−35 However, when evaluated against a contemporary panel of clinically relevant Gram-negative bacteria, weak whole cell activity (MIC) was observed for 13 vs P. aeruginosa and A. baumannii strains (Table 1). As discussed earlier, conjugation of sideromimics such as catechols and hydroxypyridones to β-lactams has been shown to impart improved cellular potency via utilization of the bacterial iron uptake process. Recent examples include 3 (MC-1) and 4, both of which employ hydroxypyridones as the sideromimics (Figure 1).16,18 Incorporation of a hydroxypyridone moiety in lactivicin, as in compound 14, quickly demonstrated the potential of the siderophore receptor uptake strategy, providing improved MICs in P. aeruginosa and A. baumannii strains while retaining PBP inhibitory potency similar to comparator analogue 13 which lacks a sideromimic (Table 1). Furthermore, the

sideromimic-containing C4 side chain of compound 14 is identical to that found in the monocyclic β-lactam analogue 4, which is currently in clinical evaluation for the treatment of Gram-negative infections. Compound 14 provided our first opportunity to examine the covalently bound conformation of a lactivicin−sideromimic analogue with its biological target and enable comparison to the related monocarbam analogue 4. Complexes of Pae PBP3 with 14 and 4. The 2.3 Å resolution crystal structure of Pae PBP3 with 14 revealed that both the L-cycloserine and γ-lactone rings are open and the antibiotic is covalently linked to the nucleophilic hydroxyl group of Ser294 (Figure 3A). The carboxylic acid group of 14 resulting from the opening of the lactone ring is stabilized by a H-bonding interaction with conserved residues Ser349, Ser485, and Thr487. The interaction of the core of 14 with Pae PBP3 is similar to the published S. pneumoniae PBP1b−lactivicin complex which involves all three signature motifs (SXXK, SXN, and KSGT).27 Importantly, in the structure of the complex of 14, the electron densities corresponding to the ethylene carboxylate moiety resulting from the opening of the lactone ring and that of the dihydroxypyridone sideromimic moiety are weak and discontinuous, suggesting that these side chains are not only flexible but also do not form strong interactions with Pae PBP3 (Figure S1, Supporting Information). These flexible side-chain conformations suggested that the potency of 14 could be further improved by structural modifications to enable interaction with the unique Tyr-TyrPhe aromatic wall present in Pae PBP3 (vide infra). Given the structural similarities between 14 and 4, a 2.0 Å resolution crystal structure of the Pae PBP3 in complex with 4 was prepared to enable direct comparison of the two series (Figure 3B). While not well-resolved, the 4-dihydroxypyridone moiety is positioned near a hydrophobic pocket formed by Tyr503, Tyr532, and Phe533 and the N-hydroxy group appears to be within hydrogen bonding distance of the phenolic oxygens in both Tyr532 and Tyr503. The low resolution in this region of the structure limits the ability to confidently define meaningful drug−protein interactions. This is different from 3848

dx.doi.org/10.1021/jm500219c | J. Med. Chem. 2014, 57, 3845−3855

Journal of Medicinal Chemistry

Article

Figure 4. Crystal structure of Pae PBP1a in complex with compound 17 and comparison of the structures with Pae PBP3 and Aba PBP3. (A) Overall structure of Pae PBP1a complexed with compound 17. The bound compound 17 is shown as spheres. (B) Active site of Pae PBP1a. The catalytic Ser461 is shown in red. The loop connecting β3 and β4 is shown in cyan. (C) Overlap of proteins from 3−Pae PBP3 (orange) and 17− PBP1a (cyan) structures, illustrating that PBP3 Phe533 occupies the same general space as PBP1a Tyr733. (D) Comparison of the extended loop connecting β3 and β4 in apo-Pae PBP3 (green) and apo-Aba PBP3 (pink).

no meaningful whole cell activity against any of the Gramnegative pathogens of interest despite both having PBP inhibitory potency comparable to the Takeda prototype compound 13 as well as the hydroxypyridone-conjugate 14 (Table 1). The increased size and lipophilicity of the Cbz group was likely deleterious to passive transit through porins and suggested that structurally similar compounds would gain little exposure to the periplasm without facilitated transport. Furthermore, on the basis of the similar PBP activity, there was little apparent preference for one configuration of the amido substituent over the other. Replacement of the Cbz groups in 15R and 15S with a hydroxypyridone sideromimic provided the diastereomeric compounds 16R and 16S. With siderophore receptor facilitated transport now enabled, both analogues exhibited improved MICs relative to 15R and 15S, despite only limited differences in PBP inhibition relative to the comparators lacking a sideromimic, providing additional evidence that sideromimics play a critical role in cell entry in this series. Again, despite a difference in stereochemistry of sideromimic attachment, the PBP inhibitory activity exhibited little differentiation between the diastereomers. However, fairly significant differences in P. aeruginosa cellular potency were observed for the 16R/S isomer pair. The reason for this is currently unknown. In an effort to further optimize drug PBP3 enzyme interactions, a design strategy to exploit potential aryl−aryl interactions between the sideromimic and the Tyr503, Tyr532, and Phe533 aromatic residues led to analogues bearing phenyl rings. From this effort, compound 17 was found to exhibit both

the published complexes of aztreonam and monocarbam 3 with Pae PBP3 in which the C4-linked gem-dimethyl group common to these compounds clearly interacts favorably with the hydrophobic pocket composed of Tyr503-Tyr532-Phe533 and the oxime-linked carboxylate forms a salt bridge with Arg489 (Figure 3C−E).36 Based on these favorable interactions, a design strategy was developed to incorporate this isobutyric acid side chain into the lactivicin analogues while exploring an alternate attachment point of the sideromimic group. Linkage of the Sideromimic to the α-Lactone Position. Of particular interest were α-amido-γ-lactones such as 15−17 (Table 1). A sideromimic attachment at the lactone α-position would potentially position this substituent in the semiopen channel distal from the active site serine where additional ligand−protein interactions might be possible. This channel has been exploited previously in the design of monocarbams such as 336 (Figure 3E) and related monobactams.19,20 A priori, predicting the optimal stereochemistry of the lactone α-amido connection was challenging. The available data from the crystal structure of P. aeruginosa PBP3 covalently bound to 14 showed only the ring-opened adduct, as expected, and could not inform about a preferred binding orientation of the ring-closed species. This analysis was further complicated by the dynamic stereochemistry of the tertiary γ-carbon of the lactone resulting from the ring-opening equilibration process described in Figure 2A. Therefore, exemplars of both configurations of the α-lactone substituent were prepared and evaluated. The resulting compounds (15R and 15S) bearing a benzyloxycarbonyl (Cbz) group (not a sideromimic) exhibited 3849

dx.doi.org/10.1021/jm500219c | J. Med. Chem. 2014, 57, 3845−3855

Journal of Medicinal Chemistry

Article

of 17 bound to Pae PBP3 was not prepared, comparison of the previously reported 3−Pae PBP3 structure36 with the 17−Pae PBP1a structure demonstrates a similar positioning of the PBP3 Phe533 and PBP1a Tyr733 side chains, suggesting that analogous π-stacking interactions with the phthalimido group of 17 may occur in both PBP1a and PBP3 (Figure 4C). Evidence Supporting Siderophore Receptor-Mediated Drug Uptake. P. aeruginosa strains are able to express an impressive range of Ton-B-dependent siderophore receptors.39 This high level of functional redundancy is critical to ensure that adequate levels of iron can be obtained in a variety of environments. Previous work from our group has suggested a role for the siderophore receptors PiuA and PirA for the uptake of a variety of hydroxypyridone-conjugated β-lactams.20,40 Because analogues 14 and 16R/S incorporate the same hydroxypyridone sideromimics, our assumption is that they utilize a similar uptake process; however, this remains to be determined. Because lactivicin analogue 17 incorporates a structurally unique phthalimide sideromimic, a study was conducted to determine which receptors are involved in cellular drug uptake. Whole cell activity was recorded for 17 vs the wild-type P. aeruginosa strain PAO1 as well as an isogenic panel of strains lacking one or more siderophore receptors (Table 2). The study was conducted with both standard

the lowest MICs and PBP IC50s to date for the series, perhaps due to an improved interaction between the phthalimide phenyl ring and the PBP active sites and to the competence of the phthalimide as a substrate for active transport via siderophore receptors (vide infra). Interestingly, attachment of a 4,5-dihydroxyphthalimide to a cephalosporin core has been reported to provide potent antibacterial analogues as well, suggesting that this sideromimic may have general utility for drug transport.37 Monobactam antiobiotics such as aztreonam typically demonstrate potent inhibition of PBP3, with lesser activity vs PBP1a and 1b and no relevant activity vs PBP2 (Table 1). While selective PBP3 inhibitors like aztreonam can successfully treat susceptible infections, enhanced potency vs other PBPs could lead to improved antibacterial performance. On the basis of initial data, it would appear that the lactivicin drug class provides an improved PBP profile relative to aztreonam, providing potent inhibition of PBP3 and 1a as well as enhanced potency vs PBP1b and 2 (Table 1). While the impact of this expanded PBP inhibitory profile remains to be determined, it may eventually prove to be an advantage for the lactivicin drug class. Class A Pae PBP1a Structure and Its Complex with a Phthalimide Sideromimic, Compound 17. The potent PBP1a inhibitory activity of phthalimide 17 catalyzed crystallographic studies to investigate its binding pose with this class A PBP. Following an extensive effort with multiple constructs, Pae PBP1a crystals were obtained after controlled proteolysis with trypsin (see Experimental Section for details). The crystallized Pae PBP1a is composed of four major polypeptides, including: Pro47-Met64, Thr256-Leu494, Thr506-Val609, Glu652-Glu792. In Pae PBP1a, the initial transmembrane helix is followed by an N-terminal transglycosylase (TG) domain connected through a β-rich linker to a C-terminal transpeptidase (TP) domain. The entire TG domain is absent in the Pae PBP1a structure due to intrinsic flexibility and tryptic cleavage. The interdomain linker contains a six-stranded βsheet where four strands are from TG and TP domains. The TP domain in the center of Pae PBP1a (residues 292−762) shares a similar overall fold with other transpeptidases and serine β-lactamases. In addition to TP and TG domains that are commonly found in bifunctional PBPs, Pae PBP1a contains an OB (oligonucleotide/oligosaccharide binding)-fold domain inserted in the TP domain between the first α-helix and the first β-strand, also seen in the A. baumannii PBP1a (Aba PBP1a) structure (Figure 4A).38 The overall Pae PBP1a structure can be superimposed onto the Aba PBP1a structure with an rmsd of 1.5 Å for 384 Cα atoms. In complex with 17, the electron density of the P. aeruginosa PBP1a active site unambiguously revealed a covalent acylenzyme interaction (Figure S1, Supporting Information). The phthalimide ring and the isobutyrate groups of 17 are well defined, and the electron density is continuous between the active site Ser461 hydroxyl group and the connecting lactivicin acyl carbon, with the acyl carbonyl oxygen lying in the oxyanion hole defined by the Ser461 and Thr698 main chain amides. The carboxylate at the γ-lactone position in 17 is anchored by Hbonds with the side chains of Thr696 and Thr698, and a watermediated H-bond with the Gly735 backbone. Though the ethylene carboxylate resulting from lactone ring opening does not show a clear interaction with PBP1a, the group does provide rigidity to the phthalimide ring, facilitating a π-stacking interaction with Tyr733 (Figure 4B). While a crystal structure

Table 2. Pae PAO1 Isogenic Siderophore Receptor Mutant Panel MICs (μg/mL)a 17 strains

MHB

low-Fe

MHB/lowFe

aztreonam lowFe

PAO1 piuA pirA f pvA f ptA pfeA piuA fpvA piuA fptA piuA pirA piuA pfeA pirA pfeA piuA pfeA pirA

0.5(1) 8(16) 0.5(1) 16(32) 4(8) 0.5(1) 16(32) 8(16) 16(32) 8(16) 0.5(1) 16(32)

≤0.03(1) 0.125(≥4) ≤0.03(1) ≤0.03(1) 0.06(≥2) ≤0.03(1) 0.25(≥8) 0.5(≥16) 0.25(≥8) 0.125(≥4) ≤0.03(1) 0.25(≥8)

16 64 16 512 64 16 64 16 64 64 16 64

4(1) 4(1) 4(1) 2(0.5) 4(1) 4(1) 4(1) 4(1) 4(1) 4(1) 4(1) 4(1)

a

Data in parentheses = strain MIC/PAO1 MIC.

Mueller Hinton Broth (MHB) as well as a modified broth depleted of iron to mimic the low iron in vivo environment to determine if the drug uptake mechanism varies based on availability of iron. A significant shift (>4×) in MIC for a receptor deficient strain vs the parent strain PAO1 is indicative of the involvement of the deleted receptor(s) in drug uptake. For example, under standard conditions (MHB), PiuA, FpvA, and FptA single mutants all provided an 8-fold or greater shift in MIC relative to PAO1 when treated with 17, as did a number of double mutant strains. However, under low iron conditions, only the PiuA single mutant showed a modest MIC shift (4fold), with no relevant shift observed for the FpvA or FptA single mutants, suggesting an adaption in receptor expression and/or function has occurred with differing iron levels, which is a known phenomenon.41 More robust MIC shifts, especially in the low iron media, were observed for a number of the double mutant strains, suggesting that multiple receptors including 3850

dx.doi.org/10.1021/jm500219c | J. Med. Chem. 2014, 57, 3845−3855

Journal of Medicinal Chemistry

Article

Table 3. Isogenic Library of E. coli Expressing β-Lactamases MIC (μg/mL) BLA-expressed DH5α (parent strain) pUCP26 (empty vector) SHV-5 SHV-12 TEM-1 TEM-24 CTX-M-15 VIM-2 NDM-1 OXA-24/40 OXA-58 KPC-3 a

ampicillin 2 2 >64 >64 >64 >64 >64 >64 >64 32 >64 64

cefepime

3

0.03 0.03 4 0.5 0.125 0.5 8 0.06 8 0.06 0.03 0.125

0.125 0.125 0.5 0.25 0.03 0.125 0.25 0.015 0.03 0.25 0.125 0.06

4

13

0.125 0.125 >64 >64 0.03 >64 1 0.03 0.06 0.125 0.125 0.125

0.25 0.25 0.5 0.25 0.25 0.5 0.5 0.25 0.5 1 0.25 0.5

14 1 0.5 32 16 1 4 4 0.5 1 1 1 1

17 0.03 0.03 0.03 0.015 0.06 0.06 0.015 0.015 0.03 0.03 NTa 0.03

NT: not tested.

any β-lactam or related drug. Table 3 reports the data generated for 3, 4, cefepime, ampicillin, and lactivicin analogues 13, 14, and 17. Results are reported as MICs, and susceptibility to a particular BLA can be inferred from a >4-fold MIC shift of the corresponding mutant E. coli strain relative to the empty vector control strain. As expected, the positive control drug, ampicillin, was hydrolyzed and thus inactivated by the full range of BLAs, and the more robust cephalosporin control, cefepime, displayed susceptibility to SHV-5, CTX-M-15, and NDM-1 while resisting hydrolysis by the other BLAs in the panel. As previously reported, the monocarbam 3 exhibited significant stability toward inactivation by a broad range of BLAs with only borderline susceptibility to SHV-5.40 In general, the lactivicin analogues also provide broad stability to BLAs similarly to 3, including the strain expressing the New Delhi metallo-βlactamase (NDM-1). The original Takeda compound 13 lacking a sideromimic moiety exhibited good stability across the entire panel, with only the MIC vs the OXA-24/40 expressing strain being 4-fold higher than the empty vector control strain. The phthalimide-conjugated lactivicin 17 also performed well, consistently inhibiting the BLA mutant strains at MICs within 2-fold of the empty vector control including the OXA-24/40 BLA-expressed E.coli strain. In contrast, both 4 and the related lactivicin derivative 14 displayed liabilities, notably with respect to SHV-5 and SHV-12, and to a lesser extent, CTX-M-15 and TEM-24. Given the significantly different core structure of 14 compared to 4, the similar BLA liabilities may reside with the common placement of the dihydroxypyridone moiety shared by these compounds. When considered alongside the original Takeda lead 13, the dramatic increase in susceptibility of 14 to SHV-5 and SHV-12 further suggests that this placement of the sideromimic enhances susceptibility to some BLAs and may potentially limit utility against Enterobacteriaceae carrying these common extended spectrum β-lactamases (ESBLs). By installing the known isobutryic acid C4 group found in 3, 13, and aztreonam, and by modifying the sideromimic point of attachment, compound 17 demonstrates a significant advantage over 4 and 14 with regard to BLA susceptibility in this panel. In addition, the 64-fold improvement in potency for 17 vs 14 when tested against the K. pneumoniae strain KP-3700 may provide additional evidence for the BLA susceptibility discussion above as this strain is known to express SHV-5 (Table 1). Bonomo et al. recently published a crystal structure of a preacylation complex of the BLA inhibitor sulbactam bound to a S70C variant of the SHV-1 enzyme.44

PiuA, PirA, FpvA, FptA, and perhaps others are involved in the uptake of 17. The PiuA and PirA receptors are known to recognize catechol-containing siderophores, such as enterobactin, while FpvA and FptA are involved in the uptake of the major P. aeruginosa siderophores pyoverdine and pyochelin, respectively. In comparison to the previously reported results for hydroxypyridone-containing β-lactams,20,40 compound 17 appears to utilize a broader set of siderophore receptors, which could translate to a therapeutic advantage in vivo where siderophore receptor expression and function is expected to vary depending on the availability of iron. In addition, 17 appears to provide a relative potency advantage in the low iron media, and the magnitude of the activity difference observed in normal vs low-iron conditions for 17 (16−512-fold) is significantly greater than that reported for the pyridoneconjugated β-lactams. (0.5−32-fold).20,40 While these preliminary in vitro results with phthalimideconjugated lactivicin analogue 17 may appear promising, a recent report demonstrates that the gold standard in vitro potency (i.e., MIC) and frequency of resistance (FOR) methodologies utilized ubiquitously in antibacterial drug research fail to predict in vivo efficacy for the related monobactam−sideromimic conjugate, 2 (MB-1) (Figure 1).42 The lack of efficacy appears to involve an adaptive response of bacteria when exposed to 2, leading to modified expression and utilization of siderophores and siderophore receptors, thus reducing the effectiveness of 2 in murine P. aeruginosa infection model studies. The authors describe modified in vitro methodologies which correlate nicely with 2 in vivo outcomes, and it remains to be determined if the modified siderophore receptor repertoire of phthalimide 17 conveys any advantage over 2 in these assays. Profiling β-Lactamase Susceptibility. Given that lactivicins inactivate PBPs in a fashion analogous to β-lactams, it is not surprising that they have been reported to be ineffective in killing bacteria harboring certain BLA enzymes.23,43 Therefore, to assess the relative BLA susceptibility for new lactivicin analogues, an isogenic panel of E. coli strains was assembled wherein each strain was engineered to express a single, specific, commonly occurring BLA.40 While this panel was not developed to match wild-type expression levels of BLAs or assess the impact of BLAs working in concert, evaluation of whole cell activity (i.e., MICs) against this panel as compared to the parent E. coli strain harboring the empty vector (pUCP26) provides a snapshot of the relative risk of BLA susceptibility of 3851

dx.doi.org/10.1021/jm500219c | J. Med. Chem. 2014, 57, 3845−3855

Journal of Medicinal Chemistry

Article

bound to Pae PBP3 and Pae PBP1a. In the course of this work, we identified compound 17, a novel phthalimide-conjugated lactivicin analogue. Efforts to understand the Ton-B-dependent siderophore receptors involved in the uptake of compound 17 suggest that it may utilize a broader set of receptors than related hydroxypyridone-conjugated β-lactams, potentially providing a unique advantage in uptake and Gram-negative effectiveness. Further profiling of 17 for susceptibility against BLAs indicates the potential for generally low susceptibility, and evaluation vs PBPs suggests the potential for a broader inhibitory profile relative to monobactams. Overall, the results demonstrate that the lactivicin−sideromimic conjugate class merits additional research in the quest to find new effective antibacterial agents.

Modeling of the ring-closed analogues 14 and 17 in this crystal structure lends further support to the hypothesis that sideromimic placement could impact BLA susceptibility, as 14 can be accommodated in the active site with the sideromimic moiety pointing toward solvent, while preserving key interactions such as proximity to nucleophile and positioning in the oxyanion hole. However, the phthalimide group of compound 17 appears to clash with the protein due to the limited space available in this region, while maintaining other key interactions and thereby leading to poor binding (Figure 5). While a more definitive BLA susceptibility

■

EXPERIMENTAL SECTION

Cloning, Expression, and Purification of Pae PBP3 and PBP1a. Pae PBP3 was prepared as discussed previously.36 PBP1a (residues 36−822) from PAO1 genomic DNA was cloned into the pET28a vector (Novagen) with an N-terminal His6 fusion tag. Protein was expressed in E. coli BL21 (Gold) cells grown in autoinduction media (Novagen) overnight at 25 °C. Cell pellets were lysed in 5 volumes (w/v) of B-PER protein extraction reagent (Thermo Scientific Pierce) supplemented with 0.4 M NaCl, 10 mM MgCl2, Complete EDTA-free protease inhibitor tablets (Roche), and Benzonase (Novagen). The lysate was clarified by centrifuging at 4 °C for 1 h in a Sorvall SS-34 rotor at 30 000g and loaded onto two tandem 5 mL HisTrap FF crude columns (GE Healthcare) equilibrated in buffer A (25 mM TrisHCl, 0.4 M NaCl, pH 8.0) with 20 mM imidazole pH 8.0 (Emerald Bio) added. After the columns were washed with 20 column volumes of buffer, PBP1a was eluted with a linear gradient to 250 mM imidazole over 20 column volumes. The eluted peak was concentrated and loaded onto a HiLoad Superdex 200 16/60 gel filtration column (GE Healthcare) equilibrated in buffer A. The peak corresponding to monomeric PBP1a (data not shown) was pooled and concentrated to 20 mg/mL using an Amicon Ultra 30 kDa MWCO centrifugal concentrator. Aliquots were flash-frozen in liquid nitrogen and stored at −80 °C. PBP1a was digested in a series of 84 μL reactions at a concentration of 2.67 mg/mL in 100 mM Tris-HCl, 100 mM NaCl, 10 mM CaCl2, and 1.1 μg of TPCK trypsin (Thermo Scientific Pierce). After 2 h at 4 °C, the digestions were stopped by adding PMSF to 10 mM. Ten digestion reactions were pooled and loaded onto a Superdex 200 HR 10/30 column equilibrated in buffer A. Fractions with digested PBP1a were pooled, concentrated to 14 mg/mL, flash-frozen in liquid nitrogen, and stored at −80 °C. Crystallization and Data Collection of Pae PBP3 and PBP1a. Crystals of apo-PBP3 were obtained with a reservoir solution containing 30% PEG 4000, 0.2 M MgCl2, and 0.1 M Tris pH 8.5. The compound was soaked into the crystals. Soaks were performed with 1 mM compound in reservoir solution for 1 day. Crystals were cryoprotected by dragging the crystals through MiTeGen’s LV CryoOil (MiTeGen, LLC) and flash frozen in liquid nitrogen. PBP1a was crystallized by sitting drop vapor diffusion at room temperature using equal volumes of 10 mg/mL protein and well solution. Crystallization conditions were screened using Crystal Screen HT (Hampton Research). Crystals grew from condition 11 (0.1 M sodium citrate tribasic dihydrate pH 5.6, 1.0 M ammonium phosphate monobasic) and reached their full size (∼100 μm) in 2−3 weeks. Crystals were soaked for 4 h in a solution containing 0.1 M sodium citrate tribasic dihydrate pH 5.6, 1.2 M ammonium phosphate monobasic, and 2 mM compound 17. The soaked crystals were cryoprotected in 0.07 M sodium citrate tribasic dihydrate pH 5.6, 0.7 M ammonium phosphate monobasic, and 35% glycerol and flashfrozen in liquid nitrogen. Structure Determination. Data were processed using the HKL2000 software suite.45 The structures of PBP3− and PBP1a− inhibitor complexes were solved by molecular replacement methods with the CCP4 version of PHASER46 using Acinetobacter baumannii

Figure 5. Proposed models of compound 14 and 17 in the active site of crystal structure of S70C SHV-1 β-lactamase. The S70C mutation was designed to affect the reactivity of the catalytic residue to allow for capture of the preacylation complex. (A) Model of S70C SHV-1 βlactamase with compound 14. The volume occupied by the ring-closed form of the analogue is shown as a blue wire surface and the active site of SHV-1 as an indigo solid surface. (B) Model of compound 17. The volume occupied by the ring-closed form of the analogue is shown as a pink wire surface and the active site of SHV-1 as an indigo solid surface. The arrow points to a region in the active site where there is significant steric clash with compound 17.

assessment would require evaluation against an expanded panel of clinical isolates known to express specific BLAs, these initial results may provide useful guidance for future drug optimization efforts.

■

CONCLUSION We have successfully applied the sideromimic-conjugation strategy to lactivicins, providing new analogues with enhanced in vitro Gram-negative activity. Analogue design was significantly aided by learnings derived from structures of inhibitors 3852

dx.doi.org/10.1021/jm500219c | J. Med. Chem. 2014, 57, 3845−3855

Journal of Medicinal Chemistry

■

PBP1a (PDB code: 3UDF) and Pae PBP3 (PDB code: 3PBN) as a search model. After molecular replacement, maximum likelihood-based refinement of the atomic position and temperature factors were performed with autoBUSTER47 and the atomic model was built with the program COOT.48 The stereochemical quality of the final model was assessed with PROCHECK.49 Crystallographic statistics for the final models are shown in Table S1, Supporting Information. Figures were prepared with PYMOL (www.pymol.org). Reactivity Modeling. Density functional theory (DFT) calculations, full geometry optimizations, and frequency analyses were carried out with B3LYP hybrid functional and triple-ζ 6-311+G(d,p) basis set using the Gaussian 09 package.50 A scaling factor of 0.9877 was used to correct zero-point vibrational energies.51 Implicit solvation using SMD polarizable continuum model of Cramer and Thrular was utilized.52 Transition state optimizations were conducted for additions of a simple model nucleophile (CH3O−), as a computational surrogate for the catalytic serine residue, to the electrophilic carbonyl. The nature of stationary points was checked by means of frequency calculations, and transition states were further verified by IRC calculations.53 The activation barrier (ΔG⧧) determined at room temperature (298.15 K) was used for ranking relative reactivity. MICs, PBP IC50s, PAO1 Isogenic Siderophore Receptor Mutant Panel, and E. coli Isogenic BLA Panel. The minimum inhibitory concentration (MIC) values were determined using the broth microdilution protocol according to the methods of the Clinical and Laboratory Standards Institute (CLSI).54 The experimental methods utilized to generate the P. aeruginosa siderophore receptor KO data as well as the E. coli isogenic BLA panel have been described in detail,40 as has the method utilized to generate PBP IC50 values.20 Chemistry. Detailed synthetic procedures and spectral characterization for all compounds are provided in the Supporting Information. The methods of purification and purity determination of compounds 13, 14, 15R/S, 16R/S, and 17 are as follows: Method A. Phenomenex Max-RP, C18, 150 × 21.2 mm, 5 μm; mobile phase A: 0.1% formic acid in water (v/v); mobile phase B: 0.1% formic acid in methanol (v/v); gradient: 5% B for 1.50 min, 5% to 100% B over 8.5 min 100% B for 1 min; flow rate: 28 mL/minute. Temperature: not controlled; detection: DAD 210−360 nm; MS (+) range 150−700 m/z; injection volume: 10 μL. Method B. Phenomenex Gemini-NX, 4.6 mm × 50 mm, C18, 3 μm, 110A; mobile phase A: 0.1% formic acid in water (v/v); mobile phase B: 0.1% formic acid in acetonitrile (v/v); gradient: 0% to 100% B over 4.1 min, 100% B 0.4 min; flow rate: 1.5 mL/min. Temperature: 60 °C.; detection: 200−450 nm; MS (+) range 100−1200 m/z; injection volume: 5 μL; instrument: column oven from Agilent Technologies, Wilmington, DE; autosampler and MS detector from Waters Corporation, Milford, MA; ELS detector from Varian medical devices, Palo Alto, CA. Method C. Waters Acqity HSS T3, C18, 2.1 × 50 mm, 1.7 μm; mobile phase A: 0.1% formic acid in water (v/v); mobile phase B: 0.1% formic acid in acetonitrile (v/v); gradient: 5% B over 0.1 min, 5% to 95% B over 2.5 min, 95% B 0.35 min; flow rate: 1.25 mL/min. Temperature: 60 °C.; detection: 200−450 nm; MS (+) range 100− 2000 m/z; injection volume: 5 μL; instrument: Waters Acquity UPLC. Compound 13. Purification method A, purity determination method B. The compound was found to have a purity > 95%. Compound 14. Crude product was triturated with diethyl ether followed by 4:1 dichloromethane: diethyl ether. Purity determination method B. The compound was found to have a purity > 95%. Compound 15R. Purification method A, purity determination method B. The compound was found to have a purity = 87%. Compound 15S. Purification method A, purity determination method B. The compound was found to have a purity = 94%. Compound 16R. Purification method A, purity determination method B. The compound was found to have a purity > 95%. Compound 16S. Purification method A, purity determination method B. The compound was found to have a purity = 84%. Compound 17. Purification method A, purity determination method C. The compound was found to have a purity > 95%.

Article

ASSOCIATED CONTENT

S Supporting Information *

The experimental details regarding the synthesis of PBP inhibitors and reactivity modeling. This material is available free of charge via the Internet at http://pubs.acs.org. Accession Codes

The atomic coordinates and structure factors have been deposited in the Protein Data Bank with entry codes 4OOL, 4OOM, and 4OON.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: (860)-686-1788. E-mail: seungil.han@pfizer.com. Author Contributions #

J.S, M.F.B., and S.H. contributed equally.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS

We thank Mark Plummer, Mark Mitton-Fry, John Mueller, Paul Miller, and Mark Noe for helpful discussions. Our appreciation is also extended to Jinshan (Mike) Chen for managing external chemistry resources. Finally, we thank Juergen Bulitta for providing helpful comments regarding this manuscript.

■

ABBREVIATIONS USED MIC, minimum inhibitory concentration; PBP, penicillin binding protein; MHB, Mueller−Hinton broth; Pae, Pseudomonas aeruginosa; Kpn, Klebsiella pneumoniae; Eco, Escherichia coli; Aba, Acinetobacter baumannii; BLA, β-lactamase; FOR, frequency of resistance; TG, transglycosylase; TP, transpeptidase

■

REFERENCES

(1) Fleming, A. The antibacterial action of cultures of a Penicillium, with special reference to their use in the isolation of B. influenzae. Br. J. Exp. Pathol. 1929, 10, 226−236. (2) Boucher, H. W.; Talbot, G. H.; Bradley, J. S.; Edwards, J. E.; Gilbert, D.; Rice, L. B.; Scheld, M.; Spellberg, B.; Bartlett, J. Bad bugs, no drugs: no ESKAPE! An update from the Infectious Diseases Society of America. Clin. Infect. Dis. 2009, 48, 1−12. (3) Solomkin, J. S.; Mazuski, J. E.; Bradley, J. S.; Rodvold, K. A.; Goldstein, E. J. C.; Baron, E. J.; O’Neill, P. J.; Chow, A. W.; Dellinger, E. P.; Eachempati, S. R.; Gorbach, S.; Hilfiker, M.; May, A. K.; Nathens, A. B.; Sawyer, R. G.; Bartlett, J. G. Diagnosis and management of complicated intra-abdominal infection in adults and children: guidelines by the Surgical Infection Society and the Infectious Diseases Society of America. Clin. Infect. Dis. 2009, 50, 133−164. (4) Anonymous. Guidelines for the management of adults with hospital-acquired, ventilator-associated, and healthcare-associated pneumonia. Am. J. Respir. Crit. Care Med. 2005, 171, 388-416. (5) Dellinger, R. P.; Levy, M. M.; Rhodes, A.; Annane, D.; Gerlach, H.; Opal, S. M.; Sevransky, J. E.; Sprung, C. L.; Douglas, I. S.; Jaeschke, R.; Osborn, T. M.; Nunnally, M. E.; Townsend, S. R.; Reinhart, K.; Kleinpell, R. M.; Angus, D. C.; Deutschman, C. S.; Machado, F. R.; Rubenfeld, G. D.; Webb, S. A.; Beale, R. J.; Vincent, J. L.; Moreno, R. Surviving sepsis campaign: International guidelines for management of severe sepsis and septic shock: 2012. Crit. Care Med. 2013, 41, 580− 637. (6) Livermore, D. M. Current epidemiology and growing resistance of Gram-negative pathogens. Korean J. Intern. Med. 2012, 27, 128−142.

3853

dx.doi.org/10.1021/jm500219c | J. Med. Chem. 2014, 57, 3845−3855

Journal of Medicinal Chemistry

Article

(7) Patel, G.; Bonomo, R. A. Status report on carbapenemases: challenges and prospects. Expert Rev. Anti-Infect. Ther. 2011, 9, 555− 570. (8) Obrecht, D.; Bernardini, F.; Dale, G.; Dembowsky, K. Emerging new therapeutics against key Gram-negative pathogens. Annu. Rep. Med. Chem. 2011, 46, 245−262. (9) Waxman, D. J.; Strominger, J. L. Penicillin-binding proteins and the mechanism of action of β-lactam antibiotics. Annu. Rev. Biochem. 1983, 52, 825−869. (10) O’Shea, R.; Moser, H. E. Physicochemical properties of antibacterial compounds: implications for drug discovery. J. Med. Chem. 2008, 51, 2871−2878. (11) Wencewicz, T. A.; Moellmann, U.; Long, T. E.; Miller, M. J. Is drug release necessary for antimicrobial activity of siderophore-drug conjugates? Syntheses and biological studies of the naturally occurring salmycin “Trojan Horse” antibiotics and synthetic desferridanoxamineantibiotic conjugates. BioMetals 2009, 22, 633−648. (12) Ji, C.; Juarez-Hernandez, R. E.; Miller, M. J. Exploiting bacterial iron acquisition: siderophore conjugates. Future Med. Chem. 2012, 4, 297−313. (13) Sykes, R. B.; Koster, W. H.; Bonner, D. P. The new monobactams: Chemistry and biology. J. Clin. Pharmacol. 1988, 28, 113−119. (14) Barbachyn, M. R.; Tuominen, T. C. Synthesis and structureactivity relationships of monocarbams leading to U-78608. J. Antibiot. 1990, 43, 1199−1203. (15) Arnould, J. C.; Boutron, P.; Pasquet, M. J. Synthesis and antibacterial activity of C-4 substituted monobactams. Eur. J. Med. Chem. 1992, 27, 131−140. (16) Page, M. G. P.; Dantier, C.; Desarbre, E. In vitro properties of BAL30072, a novel siderophore sulfactam with activity against multiresistant Gram-negative bacilli. Antimicrob. Agents Chemother. 2010, 54, 2291−2302. (17) Mushtaq, S.; Warner, M.; Livermore, D. Activity of the siderophore monobactam BAL30072 against multiresistant nonfermenters. J. Antimicrob. Chemother. 2010, 65, 266−270. (18) Flanagan, M. E.; Brickner, S. J.; Lall, M.; Casavant, J.; Deschenes, L.; Finegan, S. M.; George, D. M.; Granskog, K.; Hardink, J. R.; Huband, M. D.; Hoang, T.; Lamb, L.; Marra, A.; Mitton-Fry, M.; Mueller, J. P.; Mullins, L. M.; Noe, M. C.; O’Donnell, J. P.; Pattavina, D.; Penzien, J. B.; Schuff, B. P.; Sun, J.; Whipple, D. A.; Young, J.; Gootz, T. D. Preparation, Gram-negative antibacterial activity, and hydrolytic stability of novel siderophore-conjugated monocarbam diols. ACS Med. Chem. Lett. 2011, 2, 385−390. (19) Mitton-Fry, M. J.; Arcari, J. T.; Brown, M. F.; Casavant, J. M.; Finegan, S. M.; Flanagan, M. E.; Gao, H.; George, D. M.; Gerstenberger, B. S.; Han, S.; Hardink, J. R.; Harris, T. M.; Hoang, T.; Huband, M. D.; Irvine, R.; Lall, M. S.; Megan Lemmon, M.; Li, C.; Lin, J.; McCurdy, S. P.; Mueller, J. P.; Mullins, L.; Niosi, M.; Noe, M. C.; Pattavina, D.; Penzien, J.; Plummer, M. S.; Risley, H.; Schuff, B. P.; Shanmugasundaram, V.; Starr, J. T.; Sun, J.; Winton, J.; Young, J. A. Novel monobactams utilizing a siderophore uptake mechanism for the treatment of gram-negative infections. Bioorg. Med. Chem. Lett. 2012, 22, 5989−5994. (20) Brown, M. F.; Mitton-Fry, M. J.; Arcari, J. T.; Barham, R.; Casavant, J.; Gerstenberger, B. S.; Han, S.; Hardink, J. R.; Harris, T. M.; Hoang, T.; Huband, M. D.; Lall, M. S.; Lemmon, M. M.; Li, C.; Lin, J.; McCurdy, S. P.; McElroy, E.; McPherson, C.; Marr, E. S.; Mueller, J. P.; Mullins, L.; Nikitenko, A. A.; Noe, M. C.; Penzien, J.; Plummer, M. S.; Schuff, B. P.; Shanmugasundaram, V.; Starr, J. T.; Sun, J.; Tomaras, A.; Young, J. A.; Zaniewski, R. P. Pyridoneconjugated monobactam antibiotics with Gram-negative activity. J. Med. Chem. 2013, 56, 5541−5552. (21) Harada, S.; Tsubotani, S.; Hida, T.; Ono, H.; Okazaki, H. Structure of lactivicin, an antibiotic having a new nucleus and similar biological activities to β-lactam antibiotics. Tetrahedron Lett. 1986, 27, 6229−6232.

(22) Nozaki, Y.; Katayama, N.; Ono, H.; Tsubotani, S.; Harada, S.; Okazaki, H.; Nakao, Y. Binding of a non-beta-lactam antibiotic to penicillin-binding proteins. Nature 1987, 325, 179−180. (23) Nozaki, Y.; Katayama, N.; Harada, S.; Ono, H.; Okazaki, H. Lactivicin, a naturally occurring non-β-lactam antibiotic having βlactam-like action: biological activities and mode of action. J. Antibiot. 1989, 42, 84−93. (24) Aszodi, J.; Rowlands, D. A.; Mauvais, P.; Collette, P.; Bonnefoy, A.; Lampilas, M. Design and synthesis of bridged γ-lactams as analogues of β-lactam antibiotics. Bioorg. Med. Chem. Lett. 2004, 14, 2489−2492. (25) Coleman, K. Diazabicyclooctanes (DBOs): A potent new class of non-β-lactam β-lactamase inhibitors. Curr. Opin. Microbiol. 2011, 14, 550−555. (26) Zervosen, A.; Sauvage, E.; Frere, J.-M.; Charlier, P.; Luxen, A. Development of new drugs for an old target - the penicillin binding proteins. Molecules 2012, 17, 12478−12505. (27) Macheboeuf, P.; Fischer, D. S.; Brown, T., Jr.; Zervosen, A.; Luxen, A.; Joris, B.; Dessen, A.; Schofield, C. J. Structural and mechanistic basis of penicillin-binding protein inhibition by lactivicins. Nat. Chem. Biol. 2007, 3, 565−569. (28) Boyd, D. B.; Herron, D. K.; Lunn, W. H. W.; Spitzer, W. A. Electronic structures of cephalosporins and penicillins. 11. Parabolic relationships between antibacterial activity of cephalosporins and βlactam reactivity predicted from molecular orbital calculations. J. Am. Chem. Soc. 1980, 102, 1812−1814. (29) Harada, S.; Tsubotani, S.; Hida, T.; Koyama, K.; Kondo, M.; Ono, H. Chemistry of a new antibiotic: Lactivicin. Tetrahedron 1988, 44, 6589−6606. (30) Wolfe, S.; Wilson, M.-C.; Cheng, M.-H.; Shustov, G. V.; Akuche, C. I. Cyclic hydroxamates, especially multiply substituted [1,2]oxazinan-3-ones. Can. J. Chem. 2003, 81, 937−960. (31) Baldwin, J. E.; Ng, S. C.; Pratt, A. J. Synthesis of phenoxyacetylN-sulphonyl cycloserine. Tetrahedron Lett. 1987, 28, 4319−4320. (32) Ueda, Y.; Crast, L. B., Jr.; Mikkilineni, A. B.; Partyka, R. A. Synthesis of phenoxyacetyl-N-(hydroxydioxocyclobutenyl)cycloserines. Tetrahedron Lett. 1991, 32, 3767−3770. (33) Natsugari, H.; Kawano, Y.; Morimoto, A.; Yoshioka, K.; Ochiai, M. Synthesis of lactivicin and its derivatives. J. Chem. Soc., Chem. Commun. 1987, 62−63. (34) Tamura, N.; Matsushita, Y.; Yoshioka, K.; Ochiai, M. Synthesis of lactivicin analogs. Tetrahedron 1988, 44, 3231−3240. (35) Tamura, N.; Matsushita, Y.; Kawano, Y.; Yoshioka, K. Synthesis and antibacterial activity of lactivicin derivatives. Chem. Pharm. Bull. 1990, 38, 116−122. (36) Han, S.; Zaniewski, R. P.; Marr, E. S.; Lacey, B. M.; Tomaras, A. P.; Evdokimov, A.; Miller, J. R.; Shanmugasundaram, V. Structural basis for effectiveness of siderophore-conjugated monocarbams against clinically relevant strains of Pseudomonas aeruginosa. Proc. Natl. Acad. Sci. U. S.A. 2010, 107, 22002−22007. (37) Baudart, M. G.; Hennequin, L. F. Synthesis and biological activity of C-3′ ortho-dihydroxyphthalimido cephalosporins. J. Antibiot. 1993, 46, 1458−1470. (38) Han, S.; Caspers, N.; Zaniewski, R. P.; Lacey, B. M.; Tomaras, A. P.; Feng, X.; Geoghegan, K. F.; Shanmugasundaram, V. Distinctive attributes of β-lactam target proteins in Acinetobacter baumannii relevant to development of new antibiotics. J. Am. Chem. Soc. 2011, 133, 20536−20545. (39) Winsor, G. L.; Lam, D. K. W.; Fleming, L.; Lo, R.; Whiteside, M. D.; Yu, N. Y.; Hancock, R. E. W.; Brinkman, F. S. L. Pseudomonas genome database: Improved comparative analysis and population genomics capability for Pseudomonas genomes. Nucleic Acids Res. 2011, 39, D596−D600. (40) McPherson, C. J.; Aschenbrenner, L. M.; Lacey, B. M.; Fahnoe, K. C.; Lemmon, M. M.; Finegan, S. M.; Tadakamalla, B.; O’Donnell, J. P.; Mueller, J. P.; Tomaras, A. P. Clinically relevant Gram-negative resistance mechanisms have no effect on the efficacy of MC-1, a novel siderophore-conjugated monocarbam. Antimicrob. Agents Chemother. 2012, 56, 6334−6342. 3854

dx.doi.org/10.1021/jm500219c | J. Med. Chem. 2014, 57, 3845−3855

Journal of Medicinal Chemistry

Article

(41) Carpenter, B. M.; Whitmire, J. M.; Merrell, D. S. This is not your mother’s repressor: The complex role of Fur in pathogenesis. Infect. Immun. 2009, 77, 2590−2601. (42) Tomaras, A. P.; Crandon, J. L.; McPherson, C. J.; Banevicius, M. A.; Finegan, S. M.; Irvine, R. L.; Brown, M. F.; O’Donnell, J. P.; Nicolau, D. P. Adaptation-based resistance to siderophore-conjugated antibacterial agents by Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 2013, 57, 4197−4207. (43) Brown, T., Jr.; Charlier, P.; Herman, R.; Schofield, C. J.; Sauvage, E. Structural basis for the interaction of lactivicins with serine β-lactamases. J. Med. Chem. 2010, 53, 5890−5894. (44) Rodkey, E. A.; Drawz, S. M.; Sampson, J. M.; Bethel, C. R.; Bonomo, R. A.; van den Akker, F. Crystal structure of a preacylation complex of the β-lactamase inhibitor sulbactam bound to a sulfenamide bond-containing thiol-β-lactamase. J. Am. Chem. Soc. 2012, 134, 16798−16804. (45) Otwinowski, Z.; Minor, W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 1997, 276, 307−326. (46) McCoy, A. J.; Grosse-Kunstleve, R. W.; Adams, P. D.; Winn, M. D.; Storoni, L. C.; Read, R. J. Phaser crystallographic software. J. Appl. Crystallogr. 2007, 40, 658−674. (47) Blanc, E.; Roversi, P.; Vonrhein, C.; Flensburg, C.; Lea, S. M.; Bricogne, G. Refinement of severely incomplete structures with maximum likelihood in BUSTER-TNT. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004, 60, 2210−2221. (48) Emsley, P.; Cowtan, K. Coot: model-building tools for molecular graphics. Acta Crystallogr., Sect. D.: Biol. Crystallogr. 2004, 60, 2126−2132. (49) Laskowski, R. A.; Moss, D. S.; Thornton, J. M. Main-chain bond lengths and bond angles in protein structures. J. Mol. Biol. 1993, 231, 1049−1067. (50) Frisch, M. J.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; J. A. Montgomery, J.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö .; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Revision C.1; Gaussian, Inc., Wallingford, CT, 2009. (51) Andersson, M. P.; Uvdal, P. New scale factors for harmonic vibrational frequencies using the B3LYP density functional method with the triple-ζ basis set 6−311+G(d,p). J. Phys. Chem. A 2005, 109, 2937−2941. (52) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. Universal solvation model based on solute electron density and on a continuum model of the Solvent defined by the bulk dielectric constant and atomic surface tensions. J. Phys. Chem. B 2009, 113, 6378−6396. (53) Gonzalez, C.; Schlegel, H. B. Reaction path following in massweighted internal coordinates. J. Phys. Chem. 1990, 94, 5523−5527. (54) Clinical Laboratory Standards Institute (CLSI). Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically: approved standard. CLSI, M07, 2009a. CLSI. Performance standards for antimicrobial susceptibility testing: 19th informational supplement. CLSI, M100, 2009b.

3855

dx.doi.org/10.1021/jm500219c | J. Med. Chem. 2014, 57, 3845−3855