Marinopyrrole Derivatives as Potential Antibiotic Agents against Methicillin-Resistant Staphylococcus aureus (I)

Infections caused by drug-resistant pathogens are on the rise. The ongoing spread of methicillin-resistant Staphylococcus aureus (MRSA) strains exemplifies the urgent need for new antibiotics. The marine natural product, marinopyrrole A, was previously shown to have potent antibiotic activity against Gram-positive pathogens, including MRSA. However, its minimum inhibitory concentration (MIC) against MRSA was increased by >500 fold in the presence of 20% human serum, thus greatly limiting therapeutic potential. Here we report our discovery of a novel derivative of marinopyrrole A, designated 1a, featuring a 2–4 fold improved MIC against MRSA and significantly less susceptibility to serum inhibition. Importantly, compound 1a displayed rapid and concentration-dependent killing of MRSA. Compared to the natural product counterpart, compound 1a provides an important natural product based scaffold for further Structure Activity Relationship (SAR) and optimization.


Introduction
Marinopyrroles were first reported to show antibiotic activity against methicillin-resistant Staphylococcus aureus (MRSA) several years ago [1]. Structural elucidation of the isolated marinopyrroles from an obligate marine Streptomyces (strain CNQ-418, collected from the sea floor near La Jolla, CA) by 2D NMR and X-ray crystallography revealed an unprecedented structure class. The densely halogenated 1,3-bipyrroles represented by marinopyrrole A (±)-1 exist as enantiopure M-(−)-atropisomer, (−)-1, at ambient temperature (Chart 1). The enantiomer (+)-1 was obtained after a chiral separation of synthetic (±)-1 using chiral HPLC [2]. Due to their novel class of molecular structures and promising biological properties, the marinopyrroles have attracted considerable attention. In the last few years since the discovery of this uncommon bispyrrole structure of marine natural products, several synthetic and semi-synthetic efforts have successfully yielded a few series of novel marinopyrrole derivatives [2][3][4][5][6].

Chart 1. Chemical structures of marinopyrrole A.
The global crisis of antibiotic resistance has spread rapidly over the last several decades. MRSA infections have reached epidemic proportions in many countries [7] and represent the most common cause of skin and soft tissue infections in the United States [8]. Both hospital-associated and community-associated MRSA can exhibit broader resistance to multiple classes of antibiotics [7,[9][10][11]. Hospital-associated MRSA (HA-MRSA) infections are common among healthcare facilities and are resistant to many antibiotics. However, community-associated MRSA (CA-MRSA) strains are highly virulent and infect even healthy individuals; the incidence of these infections has skyrocketed in the last decade. The relative abandonment of antibiotic discovery and development by the pharmaceutical industry has created a vacuum in the introduction of new antibiotics to treat these increasingly problematic infections. Except for the addition of the oxazolidinone linezolid [12] in 2000, the lipopeptide daptomycin [13] in 2003, and FDA's most recent approval of ceftaroline (Teflaro) [14] as an injectable antibiotic to treat adults with community acquired bacterial pneumonia (CABP) and acute bacterial skin and skin structure infections (ABSSSI) including MRSA, a very limited number of new antibiotics have been discovered over the last half a century, with only two new classes of anti-MRSA drugs having been approved in the last 40 years. Clearly, novel agents for the treatment of MRSA infections are urgently needed [15].
Recently, evaluation of the pharmacological properties has revealed that marinopyrrole A has potent concentration-dependent antibacterial activity against clinically-relevant hospital-and community-acquired MRSA strains, a prolonged post-antibiotic effect superior to the first-line current agents vancomycin and linezolid, and a favorable resistance profile [16]. Although marinopyrrole A (±)-1, (−)-1 and (+)-1 showed potent antibiotic activity [2] with limited toxicity to mammalian cell lines at >20× MIC, their antibiotic activity against MRSA were inhibited by 20% human serum [16]. Previous studies of synthetic or semi-synthetic marinopyrrole derivatives to date have yielded no new analogs that retained antibacterial activity in the presence of human serum [16], constituting a major challenge for therapeutic development. Here, we report our design and synthesis of novel marinopyrrole derivatives with excellent antibiotic activity against MRSA but only limited serum inactivation.
The anti-MRSA activity of the novel asymmetrical marinopyrroles shown in Scheme 1 was evaluated against a USA300 strain of community-associated MRSA. One of the derivatives, (4-chloro-2-hydroxyphenyl) (4,4′,5,5′-tetrachloro-2′-(2-hydroxybenzoyl)-1′H-1,3′-bipyrrol-2-yl)methanone (1a), exhibited potent antibacterial activity. The 2-4 fold improvement in antibacterial activity from the parent compound 1 was observed for this novel marinopyrrole derivative 1a as shown in Table 1. The corresponding dimethoxy-congener, (4-chloro-2-methoxyphenyl)(4,4′,5,5′-tetrachloro-2′-(2-methoxy benzoyl)-1′H-1,3′-bipyrrol-2-yl)methanone, 10a, was found to be inactive against MRSA at >200 μM. This observation is consistent with the result reported previously for compound 1 and its corresponding methoxy analog [2] as compound 1a showed >1000 fold more potency than 10a. This observed SAR was also supported by the current results from compounds 1b and 1c. Compound 1b, with a chloro-substituent ortho to the hydroxyl group, was >128-fold more potent than 10a, whereas derivative 1c with chloro-substituent para to the hydroxyl group showed >64-fold more anti-MRSA activity than 10b. Compound 1a has the chloro-substituent meta to the hydroxyl group exhibiting 8-16-fold and 4-8-fold more anti-MRSA activity than 1c and 1b, respectively. The fact that the free hydroxyl groups are essential for the antibacterial activity suggested that the binding sites may contain both hydrogen bond donors and hydrogen bond receptors. Although the internal hydrogen bonding between the phenyl hydroxyl and adjacent carbonyl group was observed in an X-ray structure of (−)-1 [1] and compounds 1a-1c may behave similarly, this interaction can be weakened or interrupted if appropriate amino acid residues, such as Arg, His, Asn, Asp, Glu or Gln, are close by for interactions. Although thorough investigation is required to understand the observed structure activity relationships generated by compounds 1a-1c, their physicochemical properties on the OR 5 group in B ring due to the difference in pK a values of their hydroxyl groups might, at least in part, play an essential role in addition to the different substitution pattern. Most significantly, not only did compound 1a show increased antibacterial activity compared to the parent compound, marinopyrrole A (1), but its activity was also less inhibited upon the addition of 20% human serum (MIC 12.5-25 µM vs. 100-200 µM). In comparison to contemporary MRSA agents, compound 1a is more potent than vancomycin and linezolid against USA300 MRSA strain TCH1516 [16].

In Vitro Time-Kill of Marinopyrrole Derivative 1a
In vitro time-kill studies were performed to characterize killing kinetics of derivative 1a at 20×, 10× and 1× MIC and to compare them with that of the parent natural product marinopyrrole A (−)-1 [16]. Derivative 1a displayed strong concentration-dependent MRSA killing that was much more rapid than vancomycin or linezolid and was similar in profile to the parent compound [16]. The potency of derivative 1a was especially evident at 20× the MIC (3.8 μM), yielding greater than a 4-log kill of MRSA at 4 h. This feature parallels the effects seen with the natural product parent compound (−)-1.
Compound (−)-1 was also shown to have a favorable therapeutic index at 24 h in treated mammalian cells [16]. The natural product marinopyrrole A had a low propensity for toxicity at concentrations below 20× MIC. Upon closer examination, derivative 1a displays more potent MRSA killing effects. For example, the tested concentration of 1a (MIC 0.39 μM) was half of the concentration of the parent natural product tested in time-kill analyses (MIC 0.75 μM or 0.375 μg/mL) [16]. Secondly, at 3.9 μM (10× MIC) and 6 h incubation, 1a causes 3-log decrease in bacterial survival ( Figure 1). This improved upon the kinetics of the natural product, (−)-1, which generated only a two-log decrease in bacterial survival [16]. Furthermore, the actual tested compound concentration of (−)-1 was 7.5 μM, a fold higher than that of derivative 1a (3.9 μM) (Figure 1). Similar to (−)-1, this activity is much more rapid than vancomycin or linezolid [16]. In summary, we have discovered a novel natural product based derivative that retains the favorable and potent concentration-dependent bacterial killing characteristic of the natural product but at a lower concentration.

Synthesis of Compounds 4-10c
All chemicals and solvents were purchased from commercial suppliers and used without further purification. Preparative flash column chromatography was performed on silica gel 60, 0.040-0.063 mm (EMD Chemicals). 1 H NMR (400 MHz) spectra were recorded on a Varian AS400 with a 60-place automated sample changer. High resolution ESI-MS spectra were recorded on an Agilent ESI-TOF LC-MS 6200 system. Preparative HPLC was performed on a Gilson HPLC system with UV detectors and Gilson 215 liquid handler for auto injection and fraction collections (customized by HT Labs, San Diego, CA, USA). Analytical HPLC was performed on an Agilent 1100 series with diode array detectors and auto samplers. The specifications of HPLC analysis are as follows: flow rate, 1 mL/min; column, Intertsil, 2.5 μm, 4.6 × 150 mm; wavelength, 254 and 280 nm; mobile phase, A: H 2 O with 0.1% HCO 2 H, B: MeOH, gradient of 50-95% B in 25 min. All tested compounds possessed a purity of not less than 95%.

In Vitro Antibacterial Assays
TCH1516, a USA300 strain of community-associated MRSA, was obtained from the American Type Culture Collection and used for biological assays. MICs were determined by broth microdilution according to Clinical and Laboratory Standards Institute guidelines except that Todd-Hewitt broth (THB) was used in place of Mueller-Hinton broth. Assays were done in 96-well tissue-culture treated plates. Vancomycin served as a control antibiotic. MIC assays in 20% human serum were assessed by bacterial metabolic activity in resazurin as described [16].

In Vitro Time-Kill Analysis
The bactericidal activity of derivative 1a against the CA-MRSA isolate TCH1516 were assessed by time-kill analysis essentially as described previously [16,[19][20][21]. Briefly, MRSA was grown overnight in Todd-Hewitt Broth (THB) at 37 °C with shaking. Following overnight growth, MRSA was inoculated in fresh media for growth to mid-logarithmic phase. At the start of the time-kill assay, bacteria (starting inoculum ~5 × 10 5 cfu/mL) were added to duplicate 5 mL polystyrene round-bottom tubes (Falcon, Bedford, MA, USA) containing 20×, 10×, 1×, none of the MIC of derivative 1a (0.39 μM) or an equivalent amount of DMSO vehicle control (Figure 1, none). These cultures were incubated in a shaking 37 °C incubator for 24 h. To determine the rate of antibiotic killing, small aliquots were removed from tubes at 0, 3, 6 and 24 h and serially diluted for cfu enumeration on THA plates. The limit of detection for the time-kill assay was 1.6 (log 10 cfu/mL).

Conclusions
In summary, we designed and synthesized a novel series of asymmetrical derivatives of marine natural product, marinopyrrole A, and discovered that a novel derivative, 1a (MIC = 0.19-0.39 μM), exhibits significantly greater activity than linezolid and vancomycin against MRSA. Moreover, in vitro time-kill kinetics of 1a demonstrated potent concentration-dependent killing. These qualities were similar to those displayed by the parent natural product in time-kill studies, but 1a was 2-4-fold more potent at lower compound concentrations. One of the main drawbacks of the natural product has been its significant inhibition in the presence of human serum. For example, the antibiotic activity of (−)-1 was inactivated (MIC increased 256-522×) in the presence of serum [16]. Here, 1a again exhibits an advantage, with only 32-64 fold increase in MIC in 20% human serum (Table 1). Further derivatization and SAR optimization will aim to engineer out the propensity for serum inhibition. Design and synthesis of both symmetrical and asymmetrical marinopyrrole derivatives are actively ongoing. The optimization of their structure and activity for use as potential antibiotic agents against MRSA will be reported in due course.