Chemoenzymatic Modification of Daptomycin: Aromatic Group Installation on Trp1

Abstract

Daptomycin is a cyclic lipodepsipeptide antibiotic used to treat infections caused by Gram-positive pathogens, including multi-drug resistant strains such as methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant enterococci (VRE). The emergence of daptomycin-resistant bacterial strains has renewed interest in generating daptomycin analogs. Previous studies have shown that replacing the tryptophan of daptomycin with aromatic groups can generate analogs with enhanced potency. Additionally, we have demonstrated that aromatic prenyltransferases can attach diverse groups to the tryptophan of daptomycin. Here, we report the use of the prenyltransferase CdpNPT to derivatize the tryptophan of daptomycin with a library of benzylic and heterocyclic pyrophosphates. An analytical-scale study revealed that CdpNPT can transfer various aromatic groups onto daptomycin. Subsequent scaled-up and purified reactions indicated that the enzyme can attach aromatic groups to N1, C2, C5 and C6 positions of Trp1 of daptomycin. In vitro antibacterial activity assays using six of these purified compounds identified aromatic substituted daptomycin analogs show potency against both daptomycin-susceptible and resistant strains of Gram-positive bacteria. These findings suggest that installing aromatic groups on the Trp1 of daptomycin can lead to the generation of potent daptomycin analogs.

Graphical Abstract

The aromatic prenyltransferase CdpNPT can modify Trp1 of daptomycin using various alkyl-pyrophosphate analogs. This study evaluated CdpNPT's ability to generate daptomycin analogs with 21 different benzylic and heterocyclic alkyl-PP analogs. All six structurally characterized Trp1-substituted daptomycin analogs demonstrated potent activity against DapS and DapR Gram-positive bacterial strains.

Introduction

Daptomycin is a last-resort antibiotic for treating infections caused by many Gram-positive strains, including vancomycin-resistant Enterococcus (VRE) and methicillin- and vancomycin-resistant Staphylococcus aureus (MRSA and VRSA).^[1] It is a cyclic lipodepsipeptide secondary metabolite isolated from Streptomyces roseosporus.^[2] Structurally, daptomycin is made up of 13 amino acid residues, consisting of a macrocycle composed of ten amino acids, and an exocyclic lipidated tripeptide (Figure 1A). This macrocycle is closed by an ester bond between the side chain of Thr-4 and the α-COOH group of kynurenine (Kyn-13). Notably, daptomycin includes six nonproteinogenic amino acids: D-Asn-2, Orn-6, D-Ala-8, D-Ser-11, (2S,3R)-3-methylglutamate (MeGlu-12), and Kyn-13 (Figure 1A).^[2–3] The residues 7–10 form the DXDG motif, crucial for binding to Ca²⁺, which is essential for its bactericidal activity.^[2–3] Its unique mechanism of action involves preferentially inserting into fluid membrane microdomains, inducing membrane rigidity that disrupts the interaction of peripheral membrane proteins, such as the essential lipid II synthase MurG, with these domains. This disruption impairs cell wall synthesis, ultimately leading to bacterial death and making it effective against a wide range of Gram-positive pathogens.^[4] Despite its efficacy, the clinical utility of daptomycin is limited by the emergence of resistant strains (DapR).^[5] This has spurred research into developing daptomycin analogs with improved activity against resistant strains. Generating daptomycin analogs represents a promising avenue for creating more effective and safer therapeutic options against resistant bacterial infections. Several strategies for generating daptomycin analogs have emerged, including chemoenzymatic,^{[3a, 6]} biosynthetic,^[7] and solid- and solution-phase methods.^[8] One promising approach is late-stage functionalization using biocatalysis, which allows for targeted, efficient, and cost-effective molecular modifications. We have established that the aromatic prenyltransferase CdpNPT from Aspergillus fumigatus can perform late-stage modification of the Trp1-position of daptomycin (Figure 1B).^[9] Alkyl-modifications at Trp1 have been shown to produce daptomycin analogs with enhanced activities.^[9] Additionally, studies on the structure-activity relationship of Trp1 in daptomycin have demonstrated that replacing Trp1 with an aromatic moiety enhances the antibiotic's activity against Gram-positive bacteria (Figure 1C).^{[8h, 8k]} In our previous work, we explored the feasibility of attaching diverse allylic alkyl groups to the Trp1 of daptomycin using CdpNPT.^[10] However, apart from attaching a para-methoxybenzyl moiety to Trp1, we did not explore a wider range of substitutions due to the limited availability of aromatic alkyl-donors for the CdpNPT reaction. Therefore, we set out to explore CdpNPT’s ability to generate daptomycin analogs with diverse aromatic moieties using synthetic benzylic and heterocyclic alkyl-donors, which would otherwise be tedious to synthesize via classical synthetic routes. Herein, we report the synthesis and biological activity assessment of a new series of daptomycin analogs with aromatic substitutions at Trp1.

Figure 1.

A) Daptomycin structure. B) Daptomycin analogs with Trp1-alkyl substitution.⁹ C) Daptomycin analogs with replacement of Trp1 by aromatic amino acids.^8h,8k

Results and Discussion

Donor Substrate Scope of CdpNPT with Daptomycin

Our goal was to explore the ability of CdpNPT to attach diverse aromatic groups to daptomycin using a library of alkyl pyrophosphate (alkyl-PP) analogs. We used a library of 21 alkyl-PP analogs (1–21, Figure 2A), that included both benzylic (1–13) and heterocyclic compounds (14–21) with varying electron-donating and electron-withdrawing substituents.^[10] The alkylation reactions were performed using recombinant CdpNPT, (overexpressed and purified as described previously)^[9b] and standard uniform assay conditions (1.2 mM alkyl-PP, 1 mM daptomycin, 5 μM CdpNPT, 25 mM Tris, 5 mM CaCl₂, 50 mM KCl, pH 7.5, 16 h at 35 °C). Production of daptomycin derivatives was determined by differences in the retention time using reverse-phase high-pressure liquid chromatography (RP-HPLC, see Supporting Information, Table S1 and Figure S1). Confirmation of positive reactions was conducted using LC coupled with mass spectrometric (LCMS) analysis (see Supporting Information Figures S2–S14).

Figure 2.

A) Library of alkyl-PP analogues used in this study. Analogs colored red were accepted as substrates in the CdpNPT catalyzed reaction with daptomycin. B) Analytical scale reaction of CdpNPT with daptomycin and alkyl-PP analogs, presented as average percent conversion with associated standard deviation (n=3, see Supporting Information, Table S1 and Figure S1) as determined by RP-HPLC. Areas of individual product peaks are shown using different colors. Each reaction was carried out in a 20-μL volume and contained 1.2 mM alkyl-PP analog, 1 mM daptomycin, and 5 μM purified CdpNPT in a reaction buffer (25 mM Tris pH 8.0, 5 mM CaCl₂, 50 mM KCl) incubated at 35 °C for 16 h. No product formation was observed in the absence of CdpNPT or alkyl-PP. All products were subsequently confirmed by LCMS (see Supporting Information, Figures S2–S14).

A detailed analysis of the turnover ratios observed in the analytical scale reactions revealed noteworthy patterns (Figure 2B). CdpNPT did not accept benzyl pyrophosphates with 3-methoxy, 4-nitro, 2,4-dimethoxy, 3,4,5-trimethoxy, or 3,5-difluoro substitutions (4, 9, 10, 12, and 13). The percent conversion of unsubstituted benzyl pyrophosphate (1) was approximately 20%, while 3-halogenated analogs (6, 7, and 8) showed less than 20% conversion. Notably, CdpNPT efficiently transferred benzylic donors with electron-donating groups at the para position, such as methoxy (2) and methyl (5) groups, with 57% and 78% conversions respectively. However, the conversion with 2-methoxy (3) and 3,4-dimethoxy (11) were approximately 44% and 24% respectively. These findings suggest that substrate scope for the benzyl analogs was influenced by a balance between electron donation/withdrawal and steric bulk.

Interestingly, piperonyl pyrophosphate (18) produced ~80% product yield, the highest among the benzyl donors tested. This is surprising given that the di-methoxy analog (11) showed approximately 24% turnover. The steric differences between the two moieties likely account for this phenomenon. The oxygen atom para to the benzyl tail in both cases provides resonance stabilization to the carbocations. However, the drastic difference in activity likely results from the conformational constraints of the benzodioxole ring reducing the entropic barrier for binding compared to the more flexible and bulkier methoxy groups of 11. These observations highlight the potential steric limitations of the binding pocket and the need for stabilization of the benzylic carbocation in CdpNPT-catalyzed reactions.

Among the heterocyclic analogs, those bearing heteroatoms at the 2-position of the ring, such as 2-furan (14), 2-thiophene (16), and 2-benzofuran (20), were successfully transferred by CdpNPT. In contrast, those with heteroatoms at the 3-position were poorly accepted or not accepted at all. For example, 3-thiophene (17) was poorly accepted (<10%), while 3-furan (15), 3-benzopyrrole (19), and 3-benzothiophene (21) did not serve as substrates. This may be because the lone pair of electrons on adjacent heteroatoms can stabilize the positive charge more effectively compared to when it is one carbon away. Notably, while 2-benzofuran (20) gave a single product with good conversion (~75%), the five-membered heterocycles (14 and 16) produced more than one product, as indicated by multiple peaks on LC chromatograms corresponding to the product mass (See Supporting Information Figure S2). This may be due to suboptimal binding of the donor and acceptor in the active site, resulting in low regiospecificity and hence multiple products.

These patterns indicate a strong dependence of the CdpNPT-alkylation reaction on carbocation stability and available space in the active site. Differences in turnover between the heterocycles with heteroatoms at the 2- and 3-positions of the alkyl-PPs are due to better carbocation stabilization provided by adjacent lone pairs in 2-heterocycles. Thus, even among analogs with minimal variation, activity appeared to result from the complex interaction of three-dimensional structure and electronic stabilization. These are crucial factors to consider for late-stage diversification efforts using CdpNPT and the prenyltransferase class in general.

Structural Characterization of Select Alkyl-Daptomycin Analogs

The analytical scale reactions yielding ≥50% product (Figure 2, alkyl-PPs 2, 5, 14, 16, 18, and 20) were scaled up using uniform standard conditions. Analog 2, from our previous study, served as a reference point for structural and activity studies. The newly generated alkyl-daptomycin analogs were purified via step gradient RP-HPLC, facilitating the separation of different regio-isomers. All purified daptomycin analogs were then fully characterized by LCMS and NMR spectroscopy (See Supporting Information Figures S2–S14, Table S2 and Figures S16–S45). The regio-chemistry of alkyl-group attachment on Trp1 was determined by COSY, TOCSY, HSQC, and HMBC experiments (Supporting Information Figures S16–S45) and verified by HMBC correlation between H1' of the alkyl group and the functionalized indole carbon.

NMR analyses of all alkylated daptomycin products revealed that CdpNPT catalyzes C6-attachment for the para-methylbenzyl (5) and 2-benzofuran (20) analogs, while it catalyzed both C5- and C6-alkylation for the para-methoxybenzyl (2) and benzodioxole (18) analogs (Figure 3). The smaller heterocycles 14 and 16 produced 3 products. Attempts to purify all regio-isomers from the CdpNPT reactions with 14 and 16 resulted in yields sufficient for full NMR characterization of C2 and N1-alkylated daptomycin for 16, and the C2-alkylated daptomycin for 14. The purification yields of remaining products of 14 and 16 were insufficient for NMR characterization. The C5-/C6-regioisomers of 2 and 18 (boxed in Figure 3, Supporting Information Table S2) could not be separated via RP-HPLC, as they eluted at the same retention time and were inseparable. After multiple attempts to separate these combinations of regioisomers failed, it was decided that any further evaluation of their structures or antibacterial activities would be conducted using the mixtures.

Figure 3.

NMR characterized structures of alkyl-daptomycin analogs. Boxed structures represent inseparable mixtures of C5-/C6- regio-isomers (see Supporting Information). Compound 22 was previously characterized analog of daptomycin, used here as reference.

Antibacterial Activity of Alkyl-Daptomycin Analogs

Once structurally characterized, the entire set of daptomycin analogs was tested for antibacterial activity against a panel of fourteen Gram-positive bacterial strains, as well as a Gram-negative control (Table 1). The assay included the previously reported analog 22^[9b] for comparison purposes and daptomycin as a control. As expected, none of the compounds exhibited antimicrobial activity against Gram-negative E. coli.

Table 1.

MIC (in μg mL-¹) of alkyl-daptomycin analogs against a panel of eleven DapS and three DapR strains (highlighted with grey background) obtained using a standard microdilution assay as MIC₉₀ after incubation at 35 °C for 22 h. Where, Sa1=S. aureus (ATCC 25923); Sa2=S. aureus, MRSA (ATCC 700787); Se3=S. epidermidis (ATCC 12228); Ef4=Enterococcus faecalis, VRE (ATCC 700802); Bs5=Bacillus subtilis (ATCC 6051); Sa6= S. aureus, Strain SR1129 (NR-50506); Sa7=S. aureus, Strain SR2852 (NR-50508); Sa8=S. aureus, Strain SR2609 (NR-50507); Sa9=S. aureus, Strain SR3777 (NR-50509); Sa10=S. aureus, VISA, Strain SR220 (NR-50512); and Sa11=S. aureus, Strain SR4035 (NR-50510). The DapR strains are Ef12=Enterococcus faecalis, Strain S613 (HM-334); Sa13=S. aureus, Strain JE2, Transposon Mutant NE573 (NR-47116); and Sa14=S. aureus, Strain JE2, Transposon Mutant NE1656 (NR-48198).

Strain	Dap[a]	22[a]	23	24	25	26	27	28
Sa1	0.6	0.3	0.1	0.1	0.1	1.0	1.0	0.3
Sa2	0.6	0.3	0.5	0.5	0.3	1.0	4.0	1.0
Se3	0.3	0.4	0.3	0.3	0.3	0.5	2.0	0.5
Ef4	1.5	1.4	0.1	0.8	0.4	1.5	2.0	1.0
Bs5	2.0	2.2	2.0	2.0	2.0	2.0	2.0	1.0
Sa6	2.0	0.6	0.3	0.5	0.3	0.5	2.0	0.5
Sa7	2.0	1.1	0.3	0.1	0.1	0.5	2.0	0.5
Sa8	4.0	0.6	0.3	0.5	0.3	0.5	2.0	0.5
Sa9	2.0	2.2	0.3	0.1	0.3	0.5	2.0	0.5
Sa10	4.0	0.6	0.3	0.1	0.1	0.5	2.0	0.5
Sa11	8.1	4.2	1.0	0.5	0.5	2.0	4.0	2.0
Ef12	16.2	1.1	0.3	0.1	0.1	0.5	2.0	0.5
Sa13	16.2	1.1	0.3	0.3	0.1	0.5	1.0	0.5
Sa14	16.2	1.1	0.5	0.5	0.1	1.0	1.0	0.3

^[a]Taken from previous studies.^9b

All daptomycin analogs, irrespective of the regiochemistry and the nature of the attached group, showed similar or enhanced activity against all Gram-positive strains compared to parent daptomycin. Surprisingly, the C2-Trp-substituted daptomycin analogs, 26 and 27, displayed similar potency to the C5-, C6-, or N1-substituted analogs. This contrasts with previous observations where daptomycin analogs with diene moieties attached at the C2 position of the indole showed worse activity.^[9b] This reduced activity in those cases might be due to the diene moieties being prone to conjugate addition reactions, leading to their degradation. While all Trp1-substituted analogs showed good potency, the most potent alkyl-daptomycin analogs against both daptomycin-susceptible (DapS) and resistant strains (DapR) were either single C6-alkyl-substituted daptomycin (23 and 24) or a mixture of C5- and C6-regioisomers (25). These studies highlight the potential of Trp1-aromatic substituted daptomycin analogs to become new drug leads.

Conclusion

In summary, we assessed the possibility of generation of daptomycin analogs using CdpNPT reaction along with a library of 21 benzylic and heterocyclic alkyl-PP analogs, resulting in 13 positive reactions. These analytical-scale reactions revealed the previously unreported substrate scope of CdpNPT with daptomycin. Structural elucidation of six scaled-up daptomycin analogs indicated that CdpNPT could install aromatic groups at N1, C2, C5, and C6 positions of Trp1 moiety of daptomycin. Subsequent antibacterial assays indicated that attaching aromatic moieties to Trp1 enhances the activity of the parent compound against Gram-positive strains, regardless of the regiochemistry of the attached group. In general, C5- and C6-substituted analogs exhibited higher potency against both DapS and DapR strains compared to the parent molecule. This study further establishes CdpNPT as a valuable biocatalytic tool for daptomycin diversification. While the current lack of regioselectivity limits the approach, future studies focusing on protein engineering are likely to improve this aspect of the methodology.

Experimental Section

Synthesis and Characterization of Alkyl-PP Analogs

All alkyl-PP analogs were synthesized and characterized as previously reported.^[10]

CdpNPT Assays

CdpNPT was over-expressed and purified using Ni-NTA chromatography as described previously.^[9b] Analytical-scale, in vitro CdpNPT reactions were conducted in a volume of 20 μL with 1.2 mM alkyl-PP (compounds 1–21, Figure 2A), 1 mM daptomycin and 5 μM purified CdpNPT in a reaction buffer (25 mM Tris pH 8.0, 5 mM CaCl₂, 50 mM KCl). Reactions were incubated at 35 °C for 16 h and were quenched by adding an equal volume of methanol. After quenching, the samples were centrifuged at 10,000 g for 15 minutes to remove precipitated protein. Product formation for each reaction was analyzed by RP-HPLC using Method A (see Supporting Information). The percentage yield was calculated by dividing the integrated area of individual product peaks by the total integrated area of products and remaining substrate. All putative products were subsequently confirmed by LCMS using positive (+) and/or negative (−) modes (see Supporting Information Figures S2–S14). Reactions with >50% product yield were scaled up using standard conditions (7.5 mM alkyl-PP analogue, 5 mM daptomycin, 5 μM CdpNPT, 5 mM CaCl₂, 50 mM KCl, 25 mM Tris-HCl buffer pH 7.5, total volume 5.0 mL, incubated 16 h at 35 °C) and purified by semi-prep RP-HPLC using Method B (see Supporting Information). Putative products were confirmed by LCMS and NMR (see Supporting Information Table S2).

NMR Spectroscopy

NMR spectra (300–350 μL final volume) were collected at 25 °C in 5 mm Shigemi tubes in 99.9% DMSO-d₆ with 0.05% v/v TMS on a 600 MHz Varian VNMRS spectrometer equipped with a z-axis gradient 5 mm HCN cold probe at the National Magnetic Resonance Facility at Madison (NMRFAM). For each compound, a set of 1D-proton, ¹H-¹H-2D-COSY, ¹H-¹H-2D-TOCSY, ¹H-¹³C-2D-HSQC, and ¹H-¹³C-2D-HMBC were recorded for resonance assignments.

Antibacterial Activity Assays

All daptomycin analog stocks were calibrated at absorbance (Ɛ_366nm = 4,000 L mol⁻¹ cm⁻¹), and all bioactivity assays were conducted in triplicate. The bacterial strains for which MICs were determined were obtained from the American Type Culture Collection (ATCC) or the Biodefense and Emerging Infections Research Resources Repository (BEI Resources). Minimum inhibitory concentration (MIC) testing against all strains was performed in Mueller-Hinton Broth (MHB) medium supplemented with 50 mg L⁻¹ of calcium (MHBc) using NCCLS guidelines (see Supporting Information).