Exploration of Spirocyclic Derivatives of Ciprofloxacin as Antibacterial Agents

The previously reported as well as newly synthesized derivatives of the 1-oxa-9-azaspiro[5.5]undecane were employed in the synthesis of thirty-six derivatives of ciprofloxacin using commercially available 7-chloro-1-cyclopropyl-6-fluoro-4-oxo-1,4-dihydroquinoline-3-carboxylic acid and the literature protocol involving the preparation of boron chelate complex to facilitate nucleophilic aromatic substitution. All new fluoroquinolone derivatives were tested against two gram-positive as well as three gram-negative strains of bacteria. With the activity spectrum of the new derivatives being substantially narrower than that of ciprofloxacin, compounds were distinctly active against two of the five strains: gram-negative Acinetobacter baumannii 987® and gram-positive Bacillus cereus 138®. Towards these two strains, a large group of compounds displayed equal or higher potency than ciprofloxacin.


Introduction
Spirocycles represent an emerging privileged structural class for drug design [1]. Not only are spirocyclic motifs omnipresent in the natural product realm [2]; they are also employed, with increasing frequency, in drug candidate development [3]. The latter observation most likely attests to the widespread recognition of the unique structural properties offered by spirocyclic frameworks. These include, though are not limited to, the pronounced three-dimensional character of spirocycles [4], the well-defined spatial projection of peripheral appendages off the spirocyclic scaffold [5], the inherent high degree of saturation (defined as F sp3 or fraction of sp 3 -hybridized heavy atoms) [6] and the presence of multiple stereocenters. Certainly, all these aspects do not necessarily facilitate the still daunting task of the finding of a new, biologically active lead molecule acting via a specific biological target. However, if such a molecule is identified around a spirocyclic scaffold or with the use of spirocyclic periphery motifs, optimizing it into potent, selective and overall developable [7] drug candidate is an undertaking less prone to failure due to ligand promiscuity, undesired metabolism and pharmacokinetics or overall unfavorable physicochemical profile [8]. In other words, the use of spirocycles in drug design frontloads many important aspects of drug discovery and development which are traditionally worried about at more advanced stages of the process [9].
In 2016, we developed a facile synthetic entry into spirocyclic building blocks 1 via the sulfuric-acid-promoted Prins cyclization of cyclic ketones with homoallylic alcohol [10].
Besides the general utility of these building blocks for drug discovery, one particular compound, N-Boc-protected spirocyclic piperidine 1a, was synthesized on a multigram scale and was envisioned as a starting material for a larger cluster of diversely substituted spirocyclic piperidines 2 for various medicinal applications (Scheme 1). In particular, many of these spirocyclic building blocks were successfully employed in the design of free fatty acid receptor 1 (GPR40) agonists for the treatment of type 2 diabetes mellitus. The synthesis was based on the conjugation of these building blocks to a fatty-acid-mimicking carboxylic acid warhead [11][12][13]. More recently, we followed a similar strategy in antibacterial area and attached these spirocycles to the pharmacophoric 5-nitrofuroyl moiety, obtaining a series of non-toxic nitrofurans that turned out to be efficacious in vitro against multidrug-resistant Mycobacterium tuberculosis [14]. Inspired by the propensity of spirocycles 2 to deliver the chemotypes of desired biological profiles depending on the nature of the pharmacophoric element, we continued to further exploit the privileged [15] character of this versatile medicinal chemistry toolkit. Employing compounds 2 in the design and synthesis of fluoroquinolone antibacterials such as ciprofloxacin became our next focal point. It is well known that variation of cyclic secondary amine appendages at position 7 of the quinolone ring (alone or in combination with other alterations of the core as well as the periphery) of this potent class of antibiotics has a strong bearing on the antimicrobial profile and has delivered numerous efficacious, broad-spectrum antibiotics [16]. Hence, using the alreadyamassed arsenal of building blocks 2 as well as several newly prepared derivatives, we aspired to synthesize and evaluate the antimicrobial profile of fluoroquinolones 3 ( Figure 1). Herein, we report on the results obtained in the course of realizing this strategy. scale and was envisioned as a starting material for a larger cluster of diversely substituted spirocyclic piperidines 2 for various medicinal applications (Scheme 1). In particular, many of these spirocyclic building blocks were successfully employed in the design of free fatty acid receptor 1 (GPR40) agonists for the treatment of type 2 diabetes mellitus. The synthesis was based on the conjugation of these building blocks to a fatty-acid-mimicking carboxylic acid warhead [11][12][13]. More recently, we followed a similar strategy in antibacterial area and attached these spirocycles to the pharmacophoric 5-nitrofuroyl moiety, obtaining a series of non-toxic nitrofurans that turned out to be efficacious in vitro against multidrug-resistant Mycobacterium tuberculosis [14]. Inspired by the propensity of spirocycles 2 to deliver the chemotypes of desired biological profiles depending on the nature of the pharmacophoric element, we continued to further exploit the privileged [15] character of this versatile medicinal chemistry toolkit. Employing compounds 2 in the design and synthesis of fluoroquinolone antibacterials such as ciprofloxacin became our next focal point. It is well known that variation of cyclic secondary amine appendages at position 7 of the quinolone ring (alone or in combination with other alterations of the core as well as the periphery) of this potent class of antibiotics has a strong bearing on the antimicrobial profile and has delivered numerous efficacious, broad-spectrum antibiotics [16]. Hence, using the already-amassed arsenal of building blocks 2 as well as several newly prepared derivatives, we aspired to synthesize and evaluate the antimicrobial profile of fluoroquinolones 3 ( Figure 1). Herein, we report on the results obtained in the course of realizing this strategy.

Results
For the synthesis of the library of novel fluoroquinolone analogs of ciprofloxacin, 36 spirocyclic piperidines 2a-aj were selected ( Figure 2). The synthesis of the majority of these compounds had been reported previously [11][12][13][14] while seven building blocks (2a, 2d, 2ab, 2af, 2ah, 2ai and 2aj) were synthesized from the earlier reported starting materials 4 [13], 5 [11] and 6 [11] as shown in Scheme 2. The synthesis of target fluoroquinolones 3a-aj commenced with commercially available 7-chloro-1-cyclopropyl-6-fluoro-4-oxo-1,4-dihydroquinoline-3-carboxylic acid (7) which was esterified to give ester 8. The latter was converted to boron chelation complex 9 using the published protocols [17][18][19]. In the latter, the chlorine in position 7 is particularly activated towards the nucleophilic aromatic substitution. The latter was brought about by heating compound 9 with spirocyclic piperidines 2a-aj at 60 °C in the presence of triethylamine. The boron chelation complex [20] was removed by exposing intermediates 10a-aj to a 2% aqueous sodium hydroxide solution. As a result, fluoroquinolones 3aaj were obtained in yields from moderate to nearly quantitative (Scheme 3).  [11][12][13][14]. The synthesis of target fluoroquinolones 3a-aj commenced with commercially av able 7-chloro-1-cyclopropyl-6-fluoro-4-oxo-1,4-dihydroquinoline-3-carboxylic acid which was esterified to give ester 8. The latter was converted to boron chelation comp 9 using the published protocols [17][18][19]. In the latter, the chlorine in position 7 is parti larly activated towards the nucleophilic aromatic substitution. The latter was broug about by heating compound 9 with spirocyclic piperidines 2a-aj at 60 °C in the presen of triethylamine. The boron chelation complex [20] was removed by exposing interme ates 10a-aj to a 2% aqueous sodium hydroxide solution. As a result, fluoroquinolones 3 aj were obtained in yields from moderate to nearly quantitative (Scheme 3).  Having synthesized spirocyclic derivatives of fluoroquinolone ciprofloxacin 3a-aj, we proceeded to evaluate their antibacterial profile against two gram-positive (Staphylococcus aureus ATCC 25923 and Bacillus cereus 138 ® ) as well as three gram-negative (Klebsiella pneumoniae 1062 ® , Acinetobacter baumannii 987 ® and Pseudomonas aeruginosa 7292/5 ® ) strains of bacteria ( ® indicates clinical isolate strains from the Pasteur Institute's own collection of bacterial strains). Ciprofloxacin was used in these assays as a positive control.
As is evident from the data presented in Table 1, the 36 compounds displayed varying degrees of activity against different strains, in contrast to ciprofloxacin which acted as a broad-spectrum antibiotic across the panel of all five strains. Clearly, the new set of fluoroquinolones possesses a narrower spectrum of activity compared to ciprofloxacin. However, the propensity of the newly synthesized set of fluoroquinolones to kill gram-negative Acinetobacter baumannii 987 ® bacteria is evident. Indeed, eight compounds (3d, 3f, 3j-k, 3q, 3r, 3u, 3ae) were equipotent to ciprofloxacin against this strain while fourteen compounds (3b, 3e, 3g, 3l, 3n-p, 3v-w, 3z, 3aa-ab, 3ad, 3af) had an even lower minimum inhibitory concentration (MIC) than ciprofloxacin. The other bacterium that was significantly affected by the best compounds in the set was gram-positive Bacillus cereus 138 ® . Out of 36 compounds tested, nine (3f-g, 3l, 3n, 3r, 3u, 3ac-ad, 3ag) had the same activity towards this strain as ciprofloxacin. In contrast to the latter, none of the compounds (except for weakly active 3d) displayed activity towards gram-negative Pseudomonas aeruginosa 7292/5 ® . Additionally, the new spirocyclic derivatives were only weakly active towards gram-positive Staphylococcus aureus ATCC 25923 and gram-negative Klebsiella pneumoniae 1062 ® . Table 1. Antibacterial activity (MIC, mg/mL) of compounds 3a-aj against five bacterial strains in comparison with ciprofloxacin (-: no activity; +: active but MIC is higher than that of ciprofloxacin; ++: MIC same as that of ciprofloxacin; +++: MIC lower than that of ciprofloxacin; NT: not tested).

Conclusions
In summary, we explored the possibility of using spirocyclic piperidines (amenable to the Prins cyclization of protected 4-piperidone and homoallylic alcohol in aqueous mineral acid and subsequent functional group interconversions) in the design of ciprofloxacin analogs. Using the literature-established procedure of activating the halogen-substituted fluoroquinolone core by boron complexation, 36 new ciprofloxacin analogs were synthesized and tested against two gram-positive and three gram-negative bacterial strains. The activity profile of the new spirocyclic compounds displayed significant sensitivity to the peripheral groups in the 1-oxa-9-azaspiro[5.5]undecane moiety. Overall, the new set of derivatives was distinctly active against two of the five strains: gram-negative Acinetobacter baumannii 987 ® and gram-positive Bacillus cereus 138 ® . Towards these two strains, a large group of compounds displayed equal or higher potency than ciprofloxacin. These findings substantially expand the utility of spirocyclic motifs in medicinal chemistry design and further attest to the privileged character of spirocycles.