General experimental procedures
All chemicals were purchased from Sigma-Aldrich, Acros Organics or Iris BIOTECH. Isotope-labelled chemicals were purchased from Cambridge Isotope Laboratories. Genomic DNA of selected Xenorhabdus and Photorhabdus strains was isolated using the Qiagen Gentra Puregene Yeast/Bact Kit. DNA polymerases (Taq, Phusion and Q5) and restriction enzymes were purchased from New England Biolabs or Thermo Fisher Scientific. DNA primers were purchased from Eurofins MWG Operon. PCR amplifications were carried out on thermocyclers (SensoQuest). Polymerases were used according to the manufacturers’ instructions. DNA purification was performed from 1% Tris-acetate-EDTA (TAE) agarose gel using an Invisorb Spin DNA Extraction Kit (STRATEC Biomedical AG). Plasmids in E. coli were isolated by alkaline lysis. HPLC-UV-MS analysis was conducted on an UltiMate 3000 system (Thermo Fisher) coupled to an AmaZonX mass spectrometer (Bruker) with an ACQUITY UPLC BEH C18 column (130 Å, 2.1 mm × 100 mm, 1.7-μm particle size, Waters) at a flow rate of 0.6 ml min−1 (5–95% acetonitrile/water with 0.1% formic acid, vol/vol, 16 min, UV detection wavelength 190–800 nm). HPLC-UV-HRMS analysis was conducted on an UltiMate 3000 system (Thermo Fisher) coupled to an Impact II qTof mass spectrometer (Bruker) with an ACQUITY UPLC BEH C18 column (130 Å, 2.1 mm × 100 mm, 1.7-μm particle size, Waters) at a flow rate of 0.4 ml min−1 (5–95% acetonitrile/water with 0.1% formic acid, vol/vol, 16 min, UV detection wavelength 190–800 nm). Flash purification was performed on a Biotage SP1 flash purification system (Biotage) by a C18 main column (Interchim, PF50C18HP-F0080, 120 g) with a self-packed pre-column (Interchim, PF-DLE-F0012, Puriflash dry-load empty F0012 Flash column) coupled with a UV detector. HPLC purification was performed on preparative and semipreparative Agilent 1260 systems coupled to a diode array detector (DAD) and a single quadrupole detector with a C18 ZORBAX Eclipse XDB column (9.4 mm × 250 mm, 5 μm, 3 ml min−1; 21.2 mm × 250 mm, 5 μm, 20 ml min−1; 50 mm × 250 mm, 10 μm, 40 ml min−1). Freeze drying was performed using a BUCHI Lyovapor L-300 Continuous system. NMR experiments were carried out on a Bruker AVANCE 500-, 600- or 700-MHz spectrometer equipped with a 5-mm cryoprobe. 2R,3S-IOC (1) and GameXPeptide A (16) were synthesized by WuXi App Tec following the literature (ref. 73 for 2R,3S–1 and ref. 74 for 16).
Genome sequencing, assembly and annotation
Isolated DNA was sequenced on the Illumina NextSeq 500 platform. DNA libraries were constructed using the Nextera XT DNA preparation kit (Illumina) and whole-genome sequencing was performed using 2× 150-bp paired-end chemistry. A sequencing depth of >50× was targeted for each sample. Adapters and low-quality ends were trimmed with Trimmomatic 0.39 (ref. 75) and the parameters [2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:10 MINLEN:12] using a database of adapter sequences as provided by Illumina. All genomes were assembled using SPAdes v. 3.10.1 (ref. 76) executed with the following parameters: –cov-cutoff auto –careful in paired-end mode plus mate pairs (in cases where accompanying mate-pair libraries were available). Genome annotation was performed using Prokka v. 1.12 (ref. 77) with the following parameters: –usegenus–genus GENUS–addgenes–evalue 0.0001–rfam–kingdom Bacteria–gcode 11–gram –mincontiglen 200. Geneious Prime 2021 was used in genome visualization and analysis.
antiSMASH annotations and BiG-FAM preliminary classification
The antiSMASH 5.0 (ref. 23) web server was employed to mine all the genome sequences for the presence of putative natural-product BGCs. The annotations were conducted using default settings with the extended parameters of ClusterBlast, Cluster Pfam analysis and Pfam-based GO term annotation. The annotated BGCs were summarized for each strain (Supplementary Fig. 1) and visualized in the anvi’o 6.1 (refs. 28,78) layers (Fig. 1 and Supplementary Figs. 3 and 5). We then submitted the antiSMASH job IDs to the biosynthetic gene cluster families database (BiG-FAM 1.0.0)25 for preliminary GCF explorations and classifications of annotated BGCs (Supplementary Table 3), followed by BiG-SCAPE 1.0.0 (ref. 42) refinement with a cutoff of 0.65 (Source Data Fig. 3). The GCFs were double-checked manually via the interactive network (Fig. 3), and corrections were made if necessary. A putative thiopeptide BGC (Xszus_1.region006, Xsze_2.region003, Xsto_4.region001, Xpb_30.3_21.region001, Xmir_10.region001, Xmau_6.region001, Xkoz_3.region001, Xjap_NZ_FOVO01000011.region001, Xish_1.region003, Xhom_ANU1.region005, Xhom_2.region003, Xets_11.region001, XenKK7.region002, XenDL20_c00108_NODE_12.region001, Xekj_19.region001, Xehl_28.region001, Xe30TX1_c0031_NODE_38.region001, Xdo_HBLC131_1.region001, Xdo_FRM16.1.region005, Xbov_NC_013892.1.region004, Ptem_HBLC135_17.region001, Ppb6_4.region001, Plum_TT01_1.region008, Pthr_PT1.1_23.region001, Plau_IT4.1_12.region001, Plum_IL9_35_scf0001.region001, Pbod_HU2.3_20.region001, Plau_HB1.3_105.region001, Plum_EN01_24_scf0009.region001, Pbod_DE6.1_24.region001, Plau_DE2.2_108.region001, Phpb_1.region001, Pbod_LJ_007.region001, Pbod_CN4_25_scf0020.region001, Paeg_BKT4.5_19.region001, P_tem_1.region017 and so on) that exists throughout 45 XP genomes was excluded in the analysis, because it turned out that its annotation by antiSMASH 5.0 is a false positive and early reports suggest that this cluster is responsible for ribosomal methylthiolation79,80. Two BGCs, Xdo_HBLC131_4.region001 encoding the biosynthesis of glidobactins in X. doucetiae HBLC131 and Ptem_HBLC135_2.region002 encoding the biosynthesis of ririwpeptides in P. temperata HBLC135, were artificially integrated into their respective genome by CRAGE39 previously, and thus the two BGCs were also excluded in our analysis.
Pangenome analysis
Biosynthetic gene cluster boundary definition
The cluster boundary was defined by antiSMASH with the start nucleotide of the first biosynthetic gene (5′ end) and the stop nucleotide of the last biosynthetic gene (3′ end), and was manually corrected if necessary. Non-structural genes (such as transporters, regulators, transposases and so on) on the outer periphery of an operon were excluded. We compiled a table with contigs of all BGCs encoded by a given genome, BGC start and stop nucleotide positions, BGC classifications by antiSMASH and BiG-SCAPE (see the BiG-SCAPE analysis section), and possible biosynthetic pathways that the BGCs encode (Source Data Fig. 1). These tables would be integrated into the contigs databases of the pangenome for filtering the biosynthetic genes and monitoring distributions of biosynthetic gene homology groups.
Interface generation
All genomes were obtained from the National Center for Biotechnology Information (NCBI). Supplementary Table 1 reports their accession numbers. The pangenome analysis herein mainly followed the anvi’o 6.1 pangenomic workflow28,78. After simplifying the header lines of 45 FASTA files for genomes using ‘anvi-script-reformat-fasta’, we converted FASTA files into anvi’o contigs databases by the ‘anvi-gen-contigs-database’ and then decorated the contigs database with hits from HMM models by ‘anvi-run-hmms’. The program ‘anvi-run-ncbi-cogs’ was run to annotate genes in the contigs databases with functions from the NCBI’s Clusters of Orthologous Groups (COGs). Tables of gene caller IDs with start and stop nucleotide positions were exported by ‘anvi-export-table’. By linking the gene caller IDs with BGCs via the start and stop nucleotide positions, genes that fell within a given BGC boundary were considered to be natural product biosynthetic genes (Source Data Fig. 1). Thereafter, the biosynthetic genes were furnished with a classification and a possible compound name, both of which were derived from the BGC that the biosynthetic genes made up. The obtained tables were imported back to contigs databases by ‘anvi-import-functions’. External genome storage was created by ‘anvi-gen-genomes-storage’ to store DNA and amino-acid sequences, as well as functional annotations of each gene. With the genome storage in hand, we used the program ‘anvi-pan-genome’ with the genomes storage database, the flag ‘–use-ncbi-blast’ and the parameter ‘–mcl-inflation 8’. The results were displayed in an interface by ‘anvi-display-pan’. The organization of the pangenome interface as shown in the dendrogram in the centre was represented by ‘presence/absence’ patterns. The core gene bin was characterized by searching the gene homology group (gene homology group represents amino-acid sequences from one or more genomes aligned by muscle81) using filters with ‘Min number of genomes gene homology group occurs, value = 45’. The singleton bin was identified by ‘Max number of genomes gene homology group occurs, value = 1’. The rest of the gene clusters that were neither sorted into the core gene bin nor the singleton bin were appended to the accessory bin. The single-copy-core-gene (scg) bin was found by ‘Min number of genomes gene homology group occurs, value = 45’ and ‘Max number of genes from each genome, value = 1’. The scg bin was refined by ‘Max functional homogeneity index 0.9’ and ‘Min geometric homogeneity index 1’. The resulting protein sequences were exported by ‘anvi-get-sequences-for-gene-clusters’ and aligned using ClustalW 1.2.2, which is incorporated in Geneious Prime 2021. Phylogenetic trees were generated using the Geneious tree builder utilizing the Jukes–Cantor distance model and the unweighted pair group method with arithmetic mean (UPGMA), and subsequently imported back to anvi’o by ‘anvi-import-misc-data’ and visualized by the interface. The statistical data of BGCs obtained from antiSMASH 5.0 (ref. 23) and BiG-SCAPE42 were imported to the layers of the interface by ‘anvi-import-misc-data’ for visualization.
Biosynthetic gene and biosynthetic gene cluster filtering
The bin summary (scg, core, accessory and singleton) with BGC classifications was exported by ‘anvi-summarize’ to monitor the distributions of the biosynthetic gene homology group in the pangenomes (Source Data Fig. 1 and Supplementary Data). In the Excel sheets, ‘core’ and ‘scg’ filters were selected from the ‘bin_name’ column, and the ‘(Blank)’ filter from the ‘BGC_classification’ column was unselected. The table was then sorted by ‘genome_name’ and ‘gene_callers_id’ columns in ascending order. This then displayed consecutive core biosynthetic genes that could possibly make up a BGC. The same procedure was used to filter BGCs in the accessory or singleton region.
BiG-SCAPE analysis
BGCs in all genome sequences obtained from antiSMASH 5.0 (ref. 23) analyses were compared to reference BGCs from MIBiG repository 2.0 (refs. 41,82) using BiG-SCAPE 1.0.0 (ref. 42) with the PFAM database 32.0 (ref. 83). The analysis was conducted using default settings with the mode ‘auto’, mixing all classes and retaining singletons. Networks were computed for raw distance cutoffs of 0.30–0.95 in increments of 0.05. Results were visualized as a network using Cytoscape 3.7.2 (ref. 84) for a cutoff of 0.65 (Fig. 3 and Source Data Fig. 3). Statistical data for the BGCs were analysed and evaluated using Origin 2020b and Excel from Microsoft Office 365.
Strain and culture conditions
Wild-type strains and the mutants thereof and E. coli (Supplementary Table 14) were cultivated on lysogeny broth (LB) agar plates at 30 °C overnight, and subsequently inoculated into liquid LB culture at 30 °C with shaking at 200 r.p.m. For compound production, the overnight LB culture was transferred into 5 ml of LB, XPP19 or Sf-900 II SFM medium (1:100, vol/vol) with 2% (vol/vol) Amberlite XAD-16 resins, 0.1% l-arabinose as the inducer for mutants with a PBAD promoter, and selective antibiotics such as ampicillin (Am, 100 µg ml−1), kanamycin (Km, 50 µg ml−1) or chloramphenicol (Cm, 34 µg ml−1) at 30 °C, with shaking at 200 r.p.m.
Culture extraction and HPLC-UV-MS analysis
The XAD-16 resins were collected after 72 h and extracted with 5 ml of methanol or ethyl acetate. The solvent was dried under rotary evaporators, and the dried extract was resuspended in 500 μl of methanol or acetonitrile/water (1:1 vol/vol for photoxenobactins), of which 5 μl was injected and analysed by HPLC-UV-MS or HPLC-UV-HRMS. Unless otherwise specified, HPLC-UV-MS and HPLC-UV-HRMS chromatograms in the figures are shown on the same scale. Bruker Compass DataAnalysis 4.3 was used for data collection and analysis of chromatography and MS. MetabolicDetec 2.1 was utilized to differentiate MS profiles between induced and non-induced promoter insertion mutants for identifying possible metabolites produced by targeted BGCs.
Construction of PBAD promoter insertion mutants
A 500–800-bp section upstream of the target gene (lpcS, pxbF, rdb1A and xvbA) was amplified with a corresponding primer pair as listed in Supplementary Table 15. The resulting fragments were cloned using Hot Fusion85 into a pCEP_kan or pCEP_cm backbone that was amplified by pCEP_Fw and pCEP_Rv. After transformation of a constructed plasmid into E. coli S17-1 λ pir, clones were verified by PCR with primers pCEP-Ve-Fw and pDS132-Ve-Rv. A wild-type strain (X. bovienii SS-2004, X. szentirmaii DSM 16338, X. budapestensis DSM 16342 or X. vietnamensis DSM 22392) or a deletion mutant (X. szentirmaii ∆hfq, X. budapestensis ∆rdb1P or X. budapestensis ∆rdb1P ∆hfq) was used as a recipient strain. The recipient strain was mated with E. coli S17-1 λ pir (donor) carrying a constructed plasmid (Supplementary Table 16). Both strains were grown in the LB medium to an optical density at 600 nm (OD600) of 0.6 to 0.7, and the cells were washed once with fresh LB medium. Subsequently, the donor and recipient strains were mixed on an LB agar plate in ratios of 1:3 and 3:1, and incubated at 37 °C for 3 h followed by incubation at 30 °C for 21 h. After that, the bacterial cell layer was collected with an inoculating loop and resuspended in 2 ml of fresh LB medium. A 200-μl sample of the resuspended culture was spread out on an LB agar plate with Am/Km or Am/Cm and incubated at 30 °C for two days. Individual insertion clones were cultivated and analysed by HPLC-UV-HRMS, and the genotype of all mutants was verified by plasmid- and genome-specific primers.
Construction of deletion mutants
A ~1,000-bp upstream and a ~1,000-bp downstream fragment of hfq in X. budapestensis DSM 16342 were amplified using the primer pairs listed in Supplementary Table 15. The amplified fragments were fused using the complementary overhangs introduced by primers and cloned into the pEB17 vector that was linearized with PstI and BglII by Hot Fusion85. Transformation of E. coli S17-1 λ pir with the resulting plasmid (Supplementary Table 16) and conjugation with X. budapestensis DSM 16342, as well as the generation of double crossover mutants via counterselection on LB plates containing 6% sucrose, were carried out as previously described86. The deletion mutant was verified via PCR using the primer pairs listed in Supplementary Table 15, which yielded a ~2,000-bp fragment for mutants genetically equal to the WT strain and a ~1,000-bp fragment for the desired deletion mutant. The same procedure was used to generate Δrdb1P mutants, during which E. coli S17-1 λ pir carrying pEB17 rdb1P was mated with the X. budapestensis DSM 16342 wild-type and X. budapestensis ∆hfq mutant.
Labelling experiments for structural elucidation of photoxenobactins C and D by MS
The cultivation of strains for labelling experiments was carried out as described above. For photoxenobactin C (6) labelling experiments, the overnight culture was transferred into LB medium additionally fed with 4-fluorosalicylate-SNAC, l-methionine-(methyl-d3), l-[U-13C,15N]cysteine and l-[U-34S]cysteine at a final concentration of 1 mM. In terms of inverse feeding experiments, cell pellets of the 100-μl overnight culture were washed once with ISOGRO 13C or 13C,15N medium (100 μl) and resuspended in the corresponding isotope labelling medium (100 μl). The feeding culture in the isotope labelling medium (5 ml) was inoculated with a washed overnight culture (50 μl) and additional l-cysteine was added at a final concentration of 1 mM.
For photoxenobactin D (7) labelling experiments, the cell pellets of the 100-μl overnight culture were washed once with ISOGRO 13C or 15N medium (100 μl) and then resuspended in the corresponding isotope labelling medium (100 μl). A 5-ml isotope labelling medium was inoculated with a washed overnight culture (50 μl).
Isolation and purification
For photoxenobactin isolation, 10 ml of LB medium was inoculated with a colony of the X. szentirmaii PBAD pxbF ∆hfq mutant from an LB agar plate and cultivated overnight. A 10-ml culture was taken to inoculate 2 × 100 ml of LB medium (OD600 ≈ 0.1). The 2 × 100-ml cultures were incubated overnight and the whole culture volume (200 ml) was used to inoculate a 20-l LB fermenter (Braun) supplemented with 2% XAD-16 and 0.2% arabinose (antifoam was added when required). Fermenter settings were as follows: 30 °C without pH control, three six-blade impellers 150 r.p.m. After 24 h, 10 l of the culture was collected from the fermenter, and the XAD resins were separated from the cells by filtration. (1) The XAD resins were extracted with 2 × 1 l of ethyl acetate with 1% formic acid, and the combined organic phase was dried under reduced pressure. (2) The culture without XAD was centrifuged and the supernatant was extracted with 3 × 5 l of ethyl acetate with 1% formic acid, and the combined organic layers were dried under reduced pressure. (3) The cell pellet was extracted with 2 × 1 l of ethyl acetate with 1% formic acid, and the organic supernatant was dried under reduced pressure. After 48 h, the remaining 10 l of bacterial culture were extracted as described in steps (1) to (3). The combined extracts from 20 l of culture were fractionated by a flash purification system with a C18 column with a gradient elution of acetonitrile/water 20–100% at 20 ml min−1 (every 10% gradient step was performed with five column volumes, except the 60–70% step, which was performed with ten column volumes). Fractions containing photoxenobactins were combined and dried under reduced pressure. Final purification was achieved via preparative and semipreparative HPLCs with a gradient of 30% acetonitrile/water (0–30 min) and 30–100% acetonitrile/water (30–40 min). The fractions were combined in brown flasks and were immediately freeze-dried to afford photoxenobactin A (4, 0.8 mg), photoxenobactin B (5, 0.6 mg), photoxenobactin C (6, 1.2 mg) and photoxenobactin E (8, 2.2 mg).
For the isolation and purification of lipocitides A and B, 2% of XAD-16 resins from a 6-l LB culture of the X. bovienii PBAD lpcS mutant induced by l-arabinose were collected after 72 h of incubation at 30 °C with shaking at 120 r.p.m., and were washed with water and extracted with methanol (3 × 1 l) to yield a crude extract (5.3 g after evaporation). The extract was dissolved in methanol and was subjected to preparative HPLC with a C18 column using an acetonitrile/water gradient (0.1% formic acid) for 0–32 min, 55–80%, 40 ml min−1 to afford lipocitides A (17, 4.8 mg) and B (18, 9.0 mg).
Two percent of XAD-16 resins from a 12-l LB culture of the X. budapestensis PBAD rdb1A ∆rdb1P ∆hfq mutant induced by l-arabinose were collected after 72 h of incubation at 30 °C with shaking at 120 r.p.m. and washed with water and extracted with methanol (3 × 2 l) to yield a crude extract (15.3 g after evaporation). The extract was subject to a Sephadex LH-20 column eluted with methanol. The fraction (2.8 g) containing pre-rhabdobranins was subjected to preparative HPLC with a C18 column using an acetonitrile/water gradient (0.1% formic acid) for 0–20 min, 15–35%, 40 ml min−1 to afford a fraction (206 mg) mainly containing pre-rhabdobranin D, which was further purified by semipreparative HPLC with a C18 column using an acetonitrile/water gradient (0.1% formic acid) for 0–24 min, 5–53%, 3 ml min−1 to afford pre-rhabdobranin D (27, 59.1 mg).
Benzobactin A (28) and its methyl ester (29), which were detected in X. vietnamensis PBAD xvbA, were also produced by Pseudomonas chlororaphis subsp. piscium DSM 21509 (unpublished). Owing to the high production level in Pseudomonas chlororaphis subsp. piscium DSM 21509, 28 and 29 were isolated from the Pseudomonas strain. Four percent of XAD-16 resins from a 12-l XPP culture of Pseudomonas chlororaphis subsp. piscium DSM 21509 PBAD pbzA mutant induced by l-arabinose were collected after 72 h of incubation at 30 °C with shaking at 120 r.p.m., and washed with water and extracted with methanol (3 × 2 l) to yield a crude extract (95.4 g after evaporation). The extract was dissolved in methanol and subjected to preparative HPLC with a C18 column using an acetonitrile/water gradient (0.1% formic acid) for 0–18 min, 5–59%, 20 ml min−1 to afford ten fractions. Fractions 2 (95.6 mg) and 3 (50.7 mg) were further purified by semipreparative HPLC with a C18 column using an acetonitrile/water gradient (0.1% formic acid) for 0–35 min, 5–95%, 3 ml min−1 to afford benzobactin A (28, 3.2 mg) and its methyl ester (29, 0.9 mg), respectively.
NMR spectroscopy
Measurements were carried out using 1H and 13C NMR, 1H-13C heteronuclear single quantum coherence (HSQC), 1H-13C heteronuclear multiple bond correlation (HMBC), 1H-1H correlation spectroscopy (COSY), 1H-13C heteronuclear multiple quantum correlation/1H-1H correlation spectroscopy (HMQC-COSY) and 1H-13C heteronuclear single quantum coherence/1H-1H total correlation spectroscopy (HSQC-TOCSY). Chemical shifts (δ) were reported in parts per million (ppm) and referenced to the solvent signals. Data are reported as follows: chemical shift, multiplicity (br = broad, s = singlet, d = doublet, t = triplet, dd = doublet of doublet, m = multiplet and ov = overlapped) and coupling constants (in hertz). Bruker TopSpin 4.0 was used for NMR data collection and spectral interpretation.
General synthetic procedures
The Fmoc protecting group was removed with 2 ml of 40% piperidine/dimethylformamide (DMF; 5 min) followed by 2 ml of 20% piperidine/DMF (10 min). Washings between coupling and deprotection steps were performed with DMF (five syringe volumes) and dichloromethane (DCM) (five syringe volumes). Resin loadings were determined by Fmoc cleavage from a weighted resin sample87. The combined filtrates containing Fmoc cleavage products were quantified spectrophotometrically at 301 nm using a UV–vis spectrophotometer with Hellma absorption cuvettes with a path length of 1 cm. Loadings were calculated (in mmol resin) using Lambert–Beer’s law with ɛ = 7,800 M−1 cm−1: loading (mmol) = ({frac{{{rm{Abs}},{({rm{sample}})}}}{{varepsilon l}}} times V), where ɛ is the molar extinction coefficient, V is the sample volume in liter and l is the optical path length in cm. Final cleavage was achieved by shaking the resin in 2 ml of a mixture of TFA/TIPS/H2O (95:2.5:2.5) for 1 h. The filtrate was then collected and the resin washed three times (2 ml each) with DCM, and the combined filtrates were dried under reduced pressure.
Syntheses of lipocitide A
Fmoc-protected Rink Amide resin (192 mg, 0.52 mmol g−1, 0.1 mmol) was placed in a polypropylene 6-ml syringe vessel fitted with polyethylene porous filter disks and swollen in 3 ml of DMF for 10 min. Subsequently, the Fmoc-protected resin was deprotected and then washed as described in the general synthetic procedures. Fmoc-d-Cit-OH (198.0 mg, 0.5 mmol, 5 equiv.), 1-hydroxy-7-azabenzotriazole (HOAT, 0.83 ml, 0.5 mmol, 5 equiv.), hexafluorophosphate azabenzotriazole tetramethyl uronium (HATU, 190.5 mg, 0.5 mmol, 5 equiv.) and N,N-diisopropylethylamine (DIPEA, 170 μl, 1.0 mmol, 10 equiv.) were dissolved in 1.5 ml of dry DMF. After 5 min, the clear solution was added to the resin and shaken at room temperature overnight. The resin was washed and loading was calculated (79.2%) as described in the general synthetic procedures. Acylation of Fmoc-l-Ala-OH (74.1 mg, 0.24 mmol, 3 equiv.), Fmoc-d-Leu-OH (84.8 mg, 0.24 mmol, 3 equiv.) and myristic acid (54.8 mg, 0.24 mmol, 3 equiv.) were carried out using the abovementioned procedure. Final cleavage was performed as described in the general synthetic procedures, and the crude product (70.8 mg) was purified by HPLC to obtain lipocitide A (17, Supplementary Fig. 100; 24.3 mg, 54.0%) as a white solid.
Syntheses of lipocitide B
2-CTC resin (63 mg, 1.6 mmol g−1, 0.1 mmol) was placed in a polypropylene 6-ml syringe vessel fitted with polyethylene porous filter disks. The resin was incubated with Fmoc-d-Cit-OH (119.0 mg, 0.3 mmol, 3 equiv.) and DIPEA (153 μl, 0.9 mmol, 9 equiv.) in 1.5 ml of dry DCM at room temperature overnight. The resin was washed and loading was calculated (56.7%) as described in the general synthetic procedures. Acylations of Fmoc-l-Ala-OH (52.9 mg, 0.17 mmol, 3 equiv.), Fmoc-d-Leu-OH (60.1 mg, 0.17 mmol, 3 equiv.) and myristic acid (38.9 mg, 0.24 mmol, 3 equiv.) were performed with additional HOAT (0.47 ml, 0.28 mmol, 5 equiv.), HATU (108 mg, 0.28 mmol, 5 equiv.) and DIPEA (96 μl, 0.56 mmol, 10 equiv.). Final cleavage was carried out as described in the general synthetic procedures, and the crude (54.2 mg) was purified by HPLC to obtain lipocitide B (18, Supplementary Fig. 101; 18.6 mg, 57.6%) as a white solid.
Synthesis of S-(2-acetamidoethyl)4-fluoro-2-hydroxybenzothioate (4-fluorosalicylate SNAC)
To a solution of 4-fluorosalicylic acid (156 mg, 1.0 mmol, 1.0 equiv.) and hydroxybenzotriazole (HOBt, 162 mg, 1.2 mmol, 1.2 equiv.) in 45 ml of THF, N,N′-dicyclohexylcarbodiimide (DCC, 248 mg, 1.2 mmol, 1.2 equiv.) was added, followed by N-acetylcysteamine (112 µl, 1.0 mmol, 1.0 equiv.). After 1 h at room temperature, K2CO3 (138 mg, 1.0 mmol, 1.2 equiv.) was added and the reaction was stirred for an additional 2 h. The reaction mixture was then filtered and concentrated by rotary evaporation. The solid residue was dissolved in ethyl acetate and washed with sat. NaHCO3 (50 ml) and water (50 ml). The organic layer was dried over MgSO4, concentrated, and purified by flash chromatography (1–10% MeOH in CHCl3) to give 26 mg (10%) S-(2-acetamidoethyl)4-fluoro-2-hydroxybenzothioate (Supplementary Fig. 102).
IC50 value determination with the purified yeast 20S proteasome core particle
Yeast 20S proteasome core particle (yCP) from Saccharomyces cerevisiae was purified according to previously described methods88,89. The concentration of purified yCP was determined spectrophotometrically at 280 nm. yCP (final concentration: 0.05 mg ml−1 in 100 mM Tris-HCl, pH 7.5) was mixed with dimethyl sulfoxide (DMSO) as a control or serial dilutions of IOC (1) in DMSO, thereby not surpassing a final concentration of 10% (vol/vol) DMSO. After an incubation time of 45 min at room temperature, fluorogenic substrates Boc-Leu-Arg-Arg-AMC (AMC, 7-amino-4-methylcoumarin), Z-Leu-Leu-Glu-AMC and Suc-Leu-Leu-Val-Tyr-AMC (final concentration of 200 µM) were added to measure the residual activity of caspase-like (C-L, β1 subunit), trypsin-like (T-L, β2 subunit) and chymotrypsin-like (ChT-L, β5 subunit), respectively. The assay mixture was incubated for another 60 min at room temperature, then diluted 1:10 in 20 mM Tris-HCl, pH 7.5. The AMC molecules released by hydrolysis were measured in triplicate with a Varian Cary Eclipse fluorescence spectrophotometer (Agilent Technologies) at λexc = 360 nm and λem = 460 nm. Relative fluorescence units were normalized to the DMSO-treated control. The calculated residual activities were plotted against the logarithm of the applied inhibitor concentration and fitted with GraphPad Prism 9.0.2. IC50 values were deduced from the fitted data. These depend on enzyme concentration and are comparable within the same experimental settings.
Crystallization and structure determination of the yCP in complex with IOC (1)
Crystals of the yCP were grown in hanging drops at 20 °C, as previously described88,89. The protein concentration used for crystallization was 40 mg ml−1 in Tris/HCl (20 mM, pH 7.5) and EDTA (1 mM). The drops contained 1 μl of protein and 1 μl of the reservoir solution (30 mM magnesium acetate, 100 mM 2-(N-morpholino)ethanesulfonic acid (pH 6.7) and 10% (wt/vol) 2-methyl-2,4-pentanediol). Crystals appeared after two days and were incubated with 1 at a final concentration of 10 mM for at least 24 h. Droplets were then complemented with a cryoprotecting buffer (30% (wt/vol) 2-methyl-2,4-pentanediol, 15 mM magnesium acetate, 100 mM 2-(N-morpholino)ethanesulfonic acid, pH 6.9) and vitrified in liquid nitrogen. The dataset from the yCP:IOC complex was collected using synchrotron radiation (λ = 1.0 Å) at the X06SA-beamline (Swiss Light Source). X-ray intensities and data reduction were evaluated using the XDS program package version 5 February 2021 (Supplementary Table 17)90. Conventional crystallographic rigid body, positional and temperature factor refinements were carried out with REFMAC5 5.0.32 (ref. 91) and the CCP4 Program Suite 7.1.016 (ref. 92) using coordinates of the yCP structure as the starting model (PDB 5CZ4)50. Model building was performed by the programs SYBYL-X and COOT 0.8.7 (ref. 93). The final coordinates yielded excellent residual factors, as well as geometric bond and angle values. Coordinates were confirmed to fulfil the Ramachandran plot and have been deposited in the RCSB (PDB 7O2L).
Haemocyte-spreading assays
Spodoptera exigua larvae were collected from Welsh onion (Allium fistulsum L.) fields in Andong, Korea. Insects were reared in the laboratory under the following conditions: 25 ± 2 °C constant temperature, 16:8 h (light/dark) photoperiod and 60 ± 5% relative humidity. Larvae were reared on an artificial diet94 and 10% sucrose solutions were fed to adult insects. Fifth instar larvae were used in all experiments. For analysing haemocyte behaviours in vivo, fifth instar larvae of S. exigua were co-injected with 1 µl of heat-killed (95 °C for 10 min) E. coli TOP10 (2.4 × 104 cells per larva) with the test compound (0–1,000 ng per larva) by using a Hamilton microsyringe (Reno). At 1 h post-injection, 10 µl of haemolymph from each larva was collected on the glass slide and incubated for 5 min inside a dark wet chamber at room temperature. The medium was replaced with 3.7% of formaldehyde dissolved in phosphate buffered saline (PBS) and incubated for 10 min. After washing three times with PBS, cells were permeabilized with 0.2% Triton X-100 in PBS for 2 min at room temperature. After incubation, the slides were washed with PBS three times. Blocking was performed using 5% skimmed milk (Invitrogen) dissolved in PBS, followed by incubation for 10 min. After washing once with PBS, the cells were incubated with fluorescein isothiocyanate (FITC)-tagged phalloidin in PBS for 1 h at room temperature. After washing three times, the cells were incubated with 4′,6-diamidino-2-phenylindole (DAPI, 1 mg ml−1, Thermo Scientific) in PBS for nucleus staining. Finally, after washing twice in PBS, cells were observed under a fluorescence microscope (DM2500, Leica) at ×400 magnification. Haemocyte spreading was determined by the extension of F-actin out of the original cell boundary. For the in vitro assay, ~100 μl of haemolymph was collected into 400 μl of anticoagulation buffer (ACB; 186 mM NaCl, 17 mM Na2EDTA, 41 mM citric acid, pH 4.5). After adding ACB, the medium was incubated for 30 min on ice. After centrifugation at 300g for 5 min, 400 μl of supernatant was discarded. The rest of the suspension was gently mixed with 200 μl of TC100 insect tissue culture medium (Welgene). From this suspension, 10 µl of haemolymph was collected on the glass slide. The slides were co-injected with 1 µl of E. coli TOP10 (2.4 × 104 cells per larva) with the test compound (0–1,000 ng per larva), followed by the procedure described above. Means were compared by a least squared difference (LSD) test of one-way analysis of variance (ANOVA) using POC GLM of the SAS program (SAS Institute, 1989) and discriminated at type I error = 0.05.
Nodulation assays
E. coli TOP10 was heat-killed by incubating at 95 °C for 10 min. Fifth-instar larvae of S. exigua were injected with 1 µl of bacteria (2.4 × 104 cells per larva) using a Hamilton microsyringe along with 1 µl of different concentrations (10, 50, 100, 500 and 1,000 ppm) of inhibitors. Control larvae were injected with bacteria and DMSO. At 8 h after bacterial injection, nodules were counted by dissecting larvae under a stereomicroscope (Stemi SV 11, Zeiss) at ×50 magnification.
Phenoloxidase activity assays
The PO activity from plasma was estimated as previously described95. Briefly, DOPA (l-3,4-dihydroxyphenylalanine) was used as a substrate for determining PO activity from treated larvae plasma. For PO activation, each fifth-instar larva of S. exigua was challenged with 2.4 × 104 cells of heat-killed E. coli TOP10. Different inhibitors were co-injected (1 µg per larva) along with E. coli TOP10. After 8 h of bacterial challenge, haemolymph was collected from treated larvae in a 1.5-ml tube containing a few granules of phenylthiocarbamide (Sigma-Aldrich) to prevent melanization. Haemocytes were separated from plasma by centrifuging at 4 °C for 5 min at 300g. A reaction volume of 200 µl consisted of 180 µl of 10 mM DOPA in PBS (pH 7.4) and 20 µl of plasma. Absorbance was measured using a VICTOR multi-label plate reader (PerkinElmer) at 490 nm. PO activity was expressed as ΔABS per min per µl of plasma. Each treatment was replicated three times with independent samples.
Measurement of nitric oxide
The NO was indirectly quantified by measuring its oxidized form, nitrate (NO3−), using the Griess reagent of a Nitrate/Nitrite Colorimetric Assay Kit (Cayman Chemical). Fifth-instar larvae were injected with 1 µl of heat-killed E. coli TOP10 (2.4 × 104 cells per larva) using a Hamilton microsyringe along with 1 µl of the test compound. Haemolymph was collected from each sample 1 h post infection. A 150-μl volume of haemolymph from three L5 larvae was collected and homogenized in 350 μl of 100 mM PBS pH 7.4 with a homogenizer (Ultra-Turrax T8, Ika Laboratory). After centrifugation at 14,000g for 20 min at 4 °C, the supernatant was used to measure the nitrate amounts, and the total protein was measured in each sample by a Bradford assay. The samples were analysed in a 200-μl final reaction volume. Briefly, 80 μl of samples were added to the wells, then 10 μl of enzyme cofactor mixture and 10 μl of nitrate reductase mixture were added. After incubation at room temperature for 1 h, 50 μl of Griess reagent R1 and immediately 50 μl of Griess reagent R2 were added to each well. The plate was left at room temperature for 10 min for colour development. For a standard curve to quantify the nitrate concentrations of the samples, nitrates with final concentrations of 0, 5, 10, 15, 20, 25, 30 and 35 μM in a 200-μl reaction volume were used. The absorbance was recorded at 540 nm on a VICTOR multi-label plate reader. Our measurements used three larvae per sample, and we repeated the treatment with three biological samples.
Galleria injection assays
Precultures of X. szentirmaii DSM wild-type strain and the mutants thereof were grown in LB medium and inoculated into fresh cultures at an OD600 of 0.1. Cells were grown to exponential phase (OD600 ≈ 1) and then diluted to an OD600 of 0.00025. A 5-µl volume of the diluted bacterial culture was injected into the last left pro-leg of the larvae (LB medium as a negative control). G. mellonella larvae were kept at 4 °C for 10 min before injection. After infection, the larvae were incubated at 25 °C. Dead Galleria larvae were frozen at −20 °C, then at −80 °C, and freeze-dried for one day. Freeze-dried larvae were ground. Every injection experiment was aliquoted into two portions, one of which was extracted with 25 ml of acetone/ethyl acetate (vol/vol, 1:1) while the other one was extracted with acetone/methanol. Extracts were dried and resuspended in 3 ml of acetonitrile/water (1:1, vol/vol) with a tenfold dilution for HPLC-MS-UV analysis. To compare the survival percentage of G. mellonella larvae infected with the WT strain and mutants and to determine median lethal time (LT50) values, Kaplan–Meier curves were generated by GraphPad PRISM 8.4.3.
Cytotoxicity assays
HepG2 cells (hepatoblastoma cell line; ACC 180, DSMZ) were cultured under conditions recommended by the depositor, and cells were propagated in Dulbecco’s modified Eagle medium supplemented with 10% fetal bovine serum. To determine the cytotoxicity of test compounds, cells were seeded at 6 × 103 cells per well of 96-well plates in 120 μl of complete medium. After 2 h of equilibration, compounds were added in serial dilution in 60 µl of complete medium. Compounds as well as the solvent control and doxorubicin as an in-assay positive control (IC50 of 0.06 ± 0.01 µg ml−1) were tested as duplicates in two independent experiments. After 5 days of incubation, 20 μl of 5 mg ml−1 MTT (thiazolyl blue tetrazolium bromide) in PBS was added per well, and the cells were further incubated for 2 h at 37 °C. The medium was then discarded and cells were washed with 100 μl of PBS before adding 100 μl of 2-propanol/10 N HCl (250:1) to dissolve the formazan granules. The absorbance at 570 nm was measured using a microplate reader (Tecan Infinite M200Pro with Tecan iControl 2.0), and cell viability was expressed as a percentage relative to the respective solvent control. IC50 values were determined by sigmoidal curve fitting using GraphPad PRISM 8.4.3.
Statistical analysis
In Fig. 5d,e,g,j,k, means were compared using an LSD test of one-way ANOVA using POC GLM of the SAS program (SAS Institute, 1989) for continuous variables and discriminated at type I error = 0.05. The results were plotted using Sigma Plot 12.0.
Reporting Summary
Further information on research design is available in the Nature Research Reporting Summary linked to this Article.
Source: Ecology - nature.com
