Synthesis, Antibacterial and Antiribosomal Activity of the 3C-Aminoalkyl Modification in the Ribofuranosyl Ring of Apralogs (5-O-Ribofuranosyl Apramycins)

The synthesis and antiribosomal and antibacterial activity of both anomers of a novel apralog, 5-O-(5-amino-3-C-dimethylaminopropyl-D-ribofuranosyl)apramycin, are reported. Both anomers show excellent activity for the inhibition of bacterial ribosomes and that of MRSA and various wild-type Gram negative pathogens. The new compounds retain activity in the presence of the aminoglycoside phosphoryltransferase aminoglycoside modifying enzymes that act on the primary hydroxy group of typical 4,5-(2-deoxystreptamine)-type aminoglycoside and related apramycin derivatives. Unexpectedly, the two anomers have comparable activity both for the inhibition of bacterial ribosomes and of the various bacterial strains tested.


Introduction
In our quest to improve the antibacterial activity of apramycin 1, an atypical 2deoxystreptamine-type aminoglycoside antibiotic with reduced toxicity, minimal susceptibility to aminoglycoside modifying enzymes [1][2][3][4] (AMEs) and ribosomal methyltransferases (RMTs), and strong activity against a broad spectrum of ESKAPE pathogens [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19], we have developed the 5-O-furanosyl apramycins, or apralogs [20][21][22][23][24]. The present optimal apralogs carry aminoalkyl substituents at the 3-position of the furanosyl ring, eg, 2, and/or aminodeoxy substitution at the 5-position of the furanose ring as in 3 and 4, and have increased levels of activity against ESKAPE pathogens while retaining the outstanding toxicity profile and minimal susceptibility to resistance mechanisms that characterize apramycin itself. In our continuing quest to further improve the apralogs we designed and report here on the synthesis and evaluation of the new apralogs 5 and 6. Like the previous apralogs 2 and 4, these novel derivatives carry activity-enhancing aminoalkyl substituents at the ribose 3-position, but now appended via a carbon-carbon bond as opposed to the previous ether linkages. This modification allows the retention of a hydroxy group at the ribose 3-position with the potential to engage in adventitious hydrogen bonding interactions in the hydrated binding site and the consequent potential to further increase activity and selectivity ( Figure 1). Ultimately, we find that 5 and 6 have essentially identical activity and ribosomal selectivity, indicating that the modifications introduced override the importance of anomeric configuration in the ribofuranosyl bond that characterized the early apralogs.
increase activity and selectivity ( Figure 1). Ultimately, we find that 5 and 6 have essentially identical activity and ribosomal selectivity, indicating that the modifications introduced override the importance of anomeric configuration in the ribofuranosyl bond that characterized the early apralogs.

Synthesis
Apralogs 5 and 6 were synthesized by glycosylation of selectively protected apramycin derivative 18 with glycosyl donor 17 as the key step (Scheme 1). Alcohol 18 was accessed from apramycin 1 in four steps as described previously [25,26], whereas 3C-aminoalkyl ribofuranose 17 was synthesized from protected xylofuranose 7 (Scheme 1). Thus, treatment of 7 with thionyl chloride in pyridine furnished cyclic sulfite [27] that was further reacted with NaN3 to afford 8 in 86% yield. Subsequent oxidation with Dess-Martin periodinane (DMP) delivered ketone 9 that underwent highly stereoselective addition of Grignard reagent 11, prepared from THP-protected bromopentanol 10 [28] and metallic magnesium. Subsequent benzoyl protection of the resulting tertiary alcohol was followed by THP cleavage with Bu4N-Br3 and oxidation of the so-formed primary alcohol to aldehyde 15 in 86% yield over three steps from 12. The desired N,N-dimethylamino moiety was installed by reductive amination of 15 in 72% yield, after which a swap of the acetonide protection for the corresponding 2,3-diacetate delivered glycosyl donor 17 in 83% yield as a 3:2 mixture of α:β-anomers.
Glycosidic bond formation between glycosyl donor 17 and alcohol 18 was a non-trivial task. Initial attempts using excess (6 equiv) of BF3OEt2, TMS-OTf and TES-OTf as acidic promoters in the presence of 3Å MS (type A zeolite, 400 mg per mmol of 17) as water scavenger only led to 5% conversion of alcohol 18, perhaps because of the alkaline nature of zeolite sieves [29]. While pretreatment of the molecular sieves with acid resulted in a slight improvement of the BF3OEt2-promoted glycosylation (10%), a preparative useful 30% yield of the desired 19 was eventually obtained in the absence of molecular sieves. Under these conditions glycoside 19 was formed as 1:2 mixture of α:β-anomers that were obtained as individual isomers after straightforward separation from unreacted 18 and hydrolyzed glycosyl donor 17 by preparative HPLC. Each of epimeric glycosides 19β and 19α was deprotected by a sequence of saponification, followed by hydrogenolysis of azides (Scheme 1), with final purification achieved by preparative HPLC, followed by treatment with acetic acid and trituration with MeCN to give apralogs 5 and 6 in the form of their peracetate salts.

Synthesis
Apralogs 5 and 6 were synthesized by glycosylation of selectively protected apramycin derivative 18 with glycosyl donor 17 as the key step (Scheme 1). Alcohol 18 was accessed from apramycin 1 in four steps as described previously [25,26], whereas 3C-aminoalkyl ribofuranose 17 was synthesized from protected xylofuranose 7 (Scheme 1). Thus, treatment of 7 with thionyl chloride in pyridine furnished cyclic sulfite [27] that was further reacted with NaN 3 to afford 8 in 86% yield. Subsequent oxidation with Dess-Martin periodinane (DMP) delivered ketone 9 that underwent highly stereoselective addition of Grignard reagent 11, prepared from THP-protected bromopentanol 10 [28] and metallic magnesium. Subsequent benzoyl protection of the resulting tertiary alcohol was followed by THP cleavage with Bu 4 N-Br 3 and oxidation of the so-formed primary alcohol to aldehyde 15 in 86% yield over three steps from 12. The desired N,N-dimethylamino moiety was installed by reductive amination of 15 in 72% yield, after which a swap of the acetonide protection for the corresponding 2,3-diacetate delivered glycosyl donor 17 in 83% yield as a 3:2 mixture of α:β-anomers.

Activity and Selectivity at the Drug Target
We first checked the activity of the new apralogs for activity at the target level, the ribosomal decoding A site [30][31][32][33][34][35][36], through their ability to disrupt bacterial protein synthesis in cell-free translation assays [37], with apramycin 1 and the apralogs 2, 3 and 4 as comparators (Table 1). We also screened for inhibition of protein synthesis by a set of humanized bacterial ribosomes in which the complete bacterial decoding A site has been Scheme 1. Synthesis of Apralogs 5 and 6. Glycosidic bond formation between glycosyl donor 17 and alcohol 18 was a non-trivial task. Initial attempts using excess (6 equiv) of BF 3 OEt 2 , TMS-OTf and TES-OTf as acidic promoters in the presence of 3Å MS (type A zeolite, 400 mg per mmol of 17) as water scavenger only led to 5% conversion of alcohol 18, perhaps because of the alkaline nature of zeolite sieves [29]. While pretreatment of the molecular sieves with acid resulted in a slight improvement of the BF 3 OEt 2 -promoted glycosylation (10%), a preparative useful 30% yield of the desired 19 was eventually obtained in the absence of molecular sieves. Under these conditions glycoside 19 was formed as 1:2 mixture of α:β-anomers that were obtained as individual isomers after straightforward separation from unreacted 18 and hydrolyzed glycosyl donor 17 by preparative HPLC. Each of epimeric glycosides 19β and 19α was deprotected by a sequence of saponification, followed by hydrogenolysis of azides (Scheme 1), with final purification achieved by preparative HPLC, followed by treatment with acetic acid and trituration with MeCN to give apralogs 5 and 6 in the form of their peracetate salts.

Activity and Selectivity at the Drug Target
We first checked the activity of the new apralogs for activity at the target level, the ribosomal decoding A site [30][31][32][33][34][35][36], through their ability to disrupt bacterial protein synthesis in cell-free translation assays [37], with apramycin 1 and the apralogs 2, 3 and 4 as comparators (Table 1). We also screened for inhibition of protein synthesis by a set of humanized bacterial ribosomes in which the complete bacterial decoding A site has been replaced by that of the human mitochondrial (Mit13) or A1555G mutant mitochondrial ribosome (A1555G) (Figure 2) [38], as AGA binding to the cognate decoding A sites of the human mitochondrial and especially the A1555G mutant mitochondrial ribosomes in the cochlea is considered to be one of the main causes of AGA-induced ototoxicity [30,[39][40][41][42][43][44][45]. Finally, we screened for inhibition of protein synthesis by similarly engineered bacterial ribosomes carrying the human cytosolic decoding A site (Cyt14) to assess the possibility of broader systemic toxicity ( Figure 2). Compounds 5 and 6 show very similar levels of activity for the inhibition of the wild-type bacterial ribosome and for that of the hybrid ribosomes carrying the eukaryotic decoding A sites, indicating that the anomeric configuration of the ribofuranosyl ring is of no consequence in this pair of isomers. The activity of 5 and 6 against the wild-type bacterial ribosome is comparable to that of 2, 2-fold better than of apramycin itself and the apralog 3 and 2-3-fold-less than that of apralog 4. In terms of selectivity for the bacterial ribosome over the three eukaryotic hybrid ribosomes, the two novel compounds retain the overall favorable profile of apramycin and the apralogs in general (Table 1).

Figure 2.
Decoding A sites of prokaryotic and eukaryotic ribosomes. The bacterial AGA b pocket is boxed. The bacterial numbering scheme is illustrated for the AGA binding pocket. C from the bacterial ribosome binding pocket are colored green. The A1555G mutant con hypersusceptibility to AGA ototoxicity is colored red.
Compounds 5 and 6 show very similar levels of activity for the inhibition of the type bacterial ribosome and for that of the hybrid ribosomes carrying the eukaryo coding A sites, indicating that the anomeric configuration of the ribofuranosyl rin no consequence in this pair of isomers. The activity of 5 and 6 against the wild-typ terial ribosome is comparable to that of 2, 2-fold better than of apramycin itself an apralog 3 and 2-3-fold-less than that of apralog 4. In terms of selectivity for the ba ribosome over the three eukaryotic hybrid ribosomes, the two novel compounds the overall favorable profile of apramycin and the apralogs in general (Table 1).

Antibacterial Activity against Wild-Type Bacterial Strains
All newly prepared compounds and the comparators were tested for activity a a series of ESKAPE pathogens made up of a Gram-positive methicillin-resistant Sta coccus aureus (MRSA) strain, and a panel of wild-type Gram negative pathogens ( richia coli, Klebsiella pneumoniae, Enterobacter cloacae, Acinetobacter baumannii, Pseudo aeruginosa) ( Table 2). Consonant with their inhibition of the wild-type bacterial ribosomes, compou and 6 have very similar antibacterial activity against MRSA and the wild-type Gram

Antibacterial Activity against Wild-Type Bacterial Strains
All newly prepared compounds and the comparators were tested for activity against a series of ESKAPE pathogens made up of a Gram-positive methicillin-resistant Staphylococcus aureus (MRSA) strain, and a panel of wild-type Gram negative pathogens (Escherichia coli, Klebsiella pneumoniae, Enterobacter cloacae, Acinetobacter baumannii, Pseudomonas aeruginosa) ( Table 2). Consonant with their inhibition of the wild-type bacterial ribosomes, compounds 5 and 6 have very similar antibacterial activity against MRSA and the wild-type Gram negative pathogens screened (Table 2). Again in agreement with the antiribosomal activities, the two compounds display comparable activity to apramycin itself and to the apralogs 2 and 3, and 2-fold less activity than 4 against all pathogens tested, with the exceptions of Acinetobacter baumannii and Pseudomonas aeruginosa, where they showed 4-8 fold less activity.

Antibacterial Activity against Resistant Bacterial Strains
To gauge the ability of the new apralogs to overcome resistance due to the presence of AMEs, they were screened against a panel of engineered E. coli each member of which carries a specific resistance determinant (Table 3). Four APH isoforms were included in this survey, together with one bearing the AAC(3)-IV AME known to be problematic in the apramycin series [20,21], and two carrying G1405-acting RMTs (ArmA and RmtB), which strongly mitigate the activity of all DOS-type AGAs currently used in the clinic (Table 3). Table 3. Activities against E. coli in the Presence of Specific Resistance Determinants (MIC, µg/mL) a .

APH(3 )-VI
All values were determined in duplicate using twofold dilution series.
As indicated in Table 3, 5 and 6 retained excellent activity against E. coli strains bearing four different APH(3 ) isoforms and in particular against the APH(3 ,5 )-Ia isoform [46], which has the ability to phosphorylate at the ribose 5-position and so abrogate the activity of the 4,5-DOS AGAs in general and of apralogs such as 2 that retain the hydroxy group in the ribose side chain. Notably, like other apralogs, 5 and 6 afford a significant measure of protection against the action of the AAC(3)-IV isozyme, the only AME with the ability to modify and reduce the activity of apramycin itself [47]. Finally, the novel modification in 5 and 6 does not lead to resistance arising from the presence of ribosomal methyltransferases acting on G1405 [48].

Discussion
Compounds 5 and 6 retain excellent levels of activity for inhibition of the bacterial ribosome and correspondingly strong levels of antibacterial activity against MRSA and wild-type Gram negative pathogens. Compounds 5 and 6 show comparable selectivity for inhibition of the bacterial ribosome over the eukaryotic ribosomes to other apralogs and a similar profile to other 5 -amino-5 -deoxy apralogs when challenged with E. coli carrying the APH(3 )-Ia and AAC(3)-IV AMEs. As such the novel 3-C-(aminoalkyl)-3-hydroxy modification in the ribose ring of the apralogs is a viable modification, but based on the present data does not offer any particular advantages over the existing series of compounds and in particular the advanced apralog 4. It is, however, noteworthy, that the antiribosomal and antibacterial activities of the two compounds are essentially identical, indicating that the anomeric configuration in the ribofuranosyl ring is of no consequence in this series. This observation differs significantly from that previously reported for 2 and its α-ribofuranosyl epimer 20 (Figure 3), where the β-isomer was some 400 times more active for inhibition of the bacterial ribosome, and between 2-and 8-fold more active in MIC assays against wild-type Gram negative organisms [20]. As the β-ribofuranosyl configuration is usually necessary to position the primary side chain hydroxyl group of the ribose moiety for a critical hydrogen bonding interaction with both N2 in ring I and with G1491 in the drug binding pocket, this result suggests that the N2 -OH/NH 2 5 '-G1491 hydrogen bond is not critical in the present molecules. This is presumably because of the presence of six basic amines, which we have previously shown surmounts the importance of this hydrogen bond [22].

Conclusions
The synthesis of the α-and β-anomers of a novel 5-O-(3C-aminoalkyl-5-am furanosyl)apramycin is described. The new modification affords strong activity inhibition of protein synthesis by the bacterial ribosome, and for the inhibition o

Conclusions
The synthesis of the αand β-anomers of a novel 5-O-(3C-aminoalkyl-5-aminoribofuranosyl)apramycin is described. The new modification affords strong activity for the inhibition of protein synthesis by the bacterial ribosome, and for the inhibition of MRSA and typical Gram-negative pathogens. Consistent with other apralogs carrying the 5-amino-5-deoxy modification in the ribofuranosyl ring, the new compounds are not susceptible to deactivation by the APH(3 ,5 )-Ia type AME. Unexpectedly, both anomers of the new compound show essentially identical activity.

General Experimental
All reagents and solvents were purchased from commercial suppliers and were used without further purification unless otherwise specified. All experiments were carried out under a dry argon atmosphere unless otherwise specified. Unless noted otherwise, progress of reactions was monitored by thin-layer chromatography on pre-coated aluminum-  The title compound was prepared according to literature procedure [49]. Accordingly, a stirred solution of 1,2-O-isopropylidine-α-D-xylofuranose 7 (5 g, 26.29 mmol, 1 equiv) in anhydrous dichloromethane (100 mL) was cooled to 0 • C (crushed ice bath) and treated with anhydrous pyridine (4.89 mL, 60.46 mmol, 2.3 equiv) under argon atmosphere. Then, a solution of SOCl 2 (2.19 mL, 30.23 mmol, 1.15 equiv) in anhydrous dichloromethane (20 mL) was added dropwise at 0 • C over a period of 20 min. The resulting yellowish solution was stirred at 0 • C for 2 h, and the reaction progress was monitored by GC-MS assay. Upon completion of the reaction, a solution was transferred to a separatory funnel and washed with water (3 × 50 mL). The DCM layer was dried over Na 2 SO 4 , filtered off concentrated under reduced pressure keeping the water bath temperature below 30 • C to avoid product decomposition. The yellow residue was dissolved in anhydrous DMF (50 mL) and NaN 3 (5.12 g, 78.9 mmol, 3 equiv) was added. The resulting brown suspension was heated at 110 • C with stirring for 18 h, then it was cooled to ambient temperature and all volatiles were removed in vacuo. The residue was dissolved in EtOAc (100 mL) and washed with water (100 mL). The water layer was back-extracted with Et 2 O (3 × 100 mL). The combined EtOAc and Et 2 O extracts were washed with water (100 mL) to remove residual DMF and inorganic salts, then dried over Na 2 SO 4 , filtered and concentrated under reduced pressure. The resulting yellow oily residue was purified on Biotage SNAP KP-Sil 50 g silica cartridge (gradient elution from 100% petroleum ether (PE) to 45% EtOAc/PE) to give 8 (4.85 g, 86%) as a colorless sticky mass.

(5-((Tetrahydro-2H-pyran-2-yl)oxy)pentyl)magnesium bromide (11)
An oven-dried round-bottom two neck flask equipped with magnetic stir-bar was cooled to ambient temperature under an argon atmosphere. Magnesium turnings (2.66 g, 101.13 mmol, 2 equiv) were placed in the flask and activated by intensive stirring for 12 h under argon atmosphere at ambient temperature. Anhydrous THF (5 mL) was then added, a reflux condenser was mounted and the slurry was heated at 60 • C (water bath) under an argon atmosphere. 1,2-Dibromomethane (435 µL, 0.1 equiv) was added dropwise under argon, and after gas evolution ceased, a solution of bromide 10 (12.74 g, 50.57 mmol, 1 equiv) in anhydrous THF (50 mL) was added dropwise at 60 • C over a period of 45 min. The resulting gray suspension was stirred at ambient temperature for additional 3 h, then stirring was turned off and the suspension was left undisturbed overnight under argon atmosphere. The supernatant was carefully transferred via cannula to an ovendried round-bottom flask and diluted with anhydrous THF (45 mL). Concentration of the Grignard reagent 11 was determined to be 0.38 M by titration with menthol and 1,10-phenanthroline [52]. was stirred at ambient temperature for 30 min, whereupon it was cooled to −78 • C (dry ice/acetone bath). A solution of ketone 9 (1.45 g, 6.80 mmol) in anhydrous THF (2.5 mL) was added rapidly at a rate to keep temperature below −60 • C. The resulting yellow suspension was stirred at −78 • C for 1 h, warmed to ambient temperature over a period of 30 min and quenched with saturated aqueous NH 4 Cl solution (25 mL). The yellow slurry was transferred to a separation funnel, diluted with water (100 mL), and the product was back-extracted with EtOAc (3 × 50 mL). The organic extracts were combined, dried over Na 2 SO 4 and filtered off. The solvent was evaporated under reduced pressure and the yellow residue was purified on a KP-Sil 50 g silica cartridge (gradient elution from 10% EtOAc/PE to 50% EtOAc/PE) to give 12 (1.