Discovery and Derivatization of Tridecaptin Antibiotics with Altered Host Specificity and Enhanced Bioactivity

The prevalence of multidrug-resistant (MDR) pathogens combined with a decline in antibiotic discovery presents a major challenge for health care. To refill the discovery pipeline, we need to find new ways to uncover new chemical entities. Here, we report the global genome mining-guided discovery of new lipopeptide antibiotics tridecaptin A5 and tridecaptin D, which exhibit unusual bioactivities within their class. The change in the antibacterial spectrum of Oct-TriA5 was explained solely by a Phe to Trp substitution as compared to Oct-TriA1, while Oct-TriD contained 6 substitutions. Metabolomic analysis of producer Paenibacillus sp. JJ-21 validated the predicted amino acid sequence of tridecaptin A5. Screening of tridecaptin analogues substituted at position 9 identified Oct-His9 as a potent congener with exceptional efficacy against Pseudomonas aeruginosa and reduced hemolytic and cytotoxic properties. Our work highlights the promise of tridecaptin analogues to combat MDR pathogens.


Resin swelling
The resin was swollen in 10 ml of DMF for 300 s prior to the first coupling.

Automated coupling protocol
Step

Analytical RP-HPLC
Shimadzu Prominence-i LC-2030 system with a Dr. Maisch ReproSil Gold 120 C18 column (4.6 × 250 mm, 5 µm) at 30 °C and equipped with a UV detector monitoring 214 nm and 254 nm.The following solvent system, at a flow rate of 1 mL/min, was used: solvent A, 0.

Lipid II antagonization assay
The minimum inhibitory concentrations (MICs) of Oct-TriA 1 , Oct-TriA 5 and vancomycin against Staphylococcus aureus USA300 (MRSA) were determined according to Clinical and Standards Laboratory Institute (CLSI) guidelines.Gram-positive lipid II, containing lysine at position 3 of the pentapeptide, was prepared by total chemical synthesis. 1,2Lipid II in 1 : 1 chloroform : methanol (v/v) was added to a polypropylene 96-well plate (5-fold molar excess compared to test antibiotics) and the organic solvent was allowed to evaporate.Oct-TriA 5 and vancomycin in MHB (50 µL, 16xMIC) were added to the wells with the 5-fold molar excess of pure Gram-positive lipid II in triplicates and to the control wells without lipid II.Single colony of S. aureus USA300 (MRSA) from a fresh blood agar plate was suspended in TSB and grown to an OD 600 of 0.5.The bacterial culture then was diluted in MHB with 0.002% polysorbate-80 to reach 10 6 CFU/mL and 50 µL were mixed with test compounds to achieve the final concentration of 8xMIC for all tested compounds.The plate was incubated at 37°C for 18 h with constant shaking (600 rpm) and subsequently inspected for visible bacterial growth.To visualize the viability of the indicator strain, resazurin sodium salt solution (Merck, Dorset, UK) was added to the cultures to achieve the final concentration of 0.0015% (w/v) and incubated at RT for 1 h.
60°C then 300 s at 70°CAfter coupling of the final residue on the synthesizer, the resin was washed with DCM, filtered and treated with 3 mL of TFA : TIPS : H 2 O (95 : 2.5 : 2.5, v/v) for 90 min.The reaction mixture was filtered through cotton, the filtrate was precipitated in MTBE : petroleum ether (1 : 1, v/v) and centrifuged (4500 rpm, 5 min).The pellet was then resuspended in MTBE : petroleum ether (1 : 1, v/v) and centrifuged again (4500 rpm, 5 min).Finally the pellet containing the crude lipopeptide was dissolved in tBuOH : H 2 O (1 : 1, v/v) and lyophilized overnight.The crude mixtures were subsequently purified by RP-HPLC.Fractions were assessed by HPLC and LC-MS and product containing fractions were pooled, frozen and lyophilized to yield the pure lipopeptides in >95% purity (determined by HPLC).

Figure S1 .
Figure S1.MS/MS spectrum of tridecaptin A 5 produced by of Paenibacillus sp.JJ-21 (precursor ion [M + 3H] 3 ⁺ m/z 539.9712).The assignment of the sequence of amino acid residues at the top of the spectrum is based on the mass differences between the consecutive y and b ions.

Figure S2 .
Figure S2.HPLC trace showing the reinjection of purified Oct-TriA 1 (1).The peptide eluted as a single peak at 15.524 min using the HPLC method A.

Figure S3 .Figure S4 .
Figure S3.HPLC trace showing the reinjection of purified Oct-TriA 5(2).The peptide eluted as a single peak at 15.993 min using the HPLC method A.

Figure
Figure S5. A. Gram-positive lipid II binding assays with S. aureus USA300 (MRSA) to show lipid II binding of Oct-TriA 5 .Vancomycin was used as a positive control for lipid II binding.Resazurin was used to visualize the viability of the indicator strain (pink colour indicates growth and blue means inhibition of growth).The addition of lipid II significantly reduced the efficacy of Oct-TriA5 and vancomycin, as the growth of S. aureus USA300 (MRSA) was not inhibited at the concentration of 8x MIC.This indicates that Oct-TriA5 binds to Lipid II of Gram-positive bacteria.B. MIC assays for Oct-TriA 1 , Oct-TriA 5 and vancomycin against S. aureus USA300.The wells with the MIC of the test compounds are highlighted with red edge.

Figure S6 .
Figure S6.HPLC trace showing the reinjection of purified Oct-Gly9 (4).The peptide eluted as a single peak at 14.814 min using the HPLC method A.

Figure S7 .Figure S8 .
Figure S7.HPLC trace showing the reinjection of purified Oct-Ala9 (5).The peptide eluted as a single peak at 14.863 min using the HPLC method A.

Figure S9 .Figure S10 .
Figure S9.HPLC trace showing the reinjection of purified Oct-Ile9 (7).The peptide eluted as a single peak at 26.567 min using the HPLC method C.

Figure S11 .Figure S12 .
Figure S11.HPLC trace showing the reinjection of purified Oct-Tyr9 (9).The peptide eluted as a single peak at 15.201 min using the HPLC method A.

Figure S13 .
Figure S13.HPLC trace showing the reinjection of purified Oct-His9 (11).The peptide eluted as a single peak at 24.303 min using the HPLC method C.

Figure S14 .
Figure S14.HPLC trace showing the reinjection of purified Oct-Dab9 (12).The peptide eluted as a single peak at 14.366 min using the HPLC method A.

Table S1 A
-domain specificity analysis of NRPS BGCs from the genome of Paenibacillus sp.JJ-21.

Table S2 A
-domain specificity comparison of reference tridecaptin A 1 BGC and tridecaptin cluster from the genome of Paenibacillus sp.JJ-21.

Table S3
Predicted amino-acid specificities of the adenylation domains present in representative BGCs of tridecaptin gene cluster families (GCFs) in a BiG-SCAPE sequence similarity network.

Table S4
HPLC and HRMS analysis of peptides.