A widespread family of bacterial gene clusters produces ClpP inhibitors

Intracellular proteolytic complexes play an essential role in modeling the proteome in both bacteria and eukaryotes. ClpP is the protease subunit of one such highly conserved proteolytic complex that, despite its potential, remains unexploited as a drug target. Here we describe a target-directed genome mining strategy to identify ClpP targeting compounds from the bacterial order Actinomycetales. By searching for biosynthetic gene clusters that contain duplicated copies of ClpP as putative antibiotic resistance genes, we identify a family of ClpP-associated clusters that are widespread across phyla, including environmental and pathogenic bacteria. While numerous bacterial pyrrolizidine alkaloids produced by these gene clusters are known, their connection to ClpP has never been made. We show that these previously characterized molecules do not affect ClpP function but are shunt metabolites derived from the genuine product of these gene clusters, a reactive covalent ClpP inhibitor. Focusing on one such cryptic gene cluster from Streptomyces cattleya DSM 46488, we use heterologous expression to purify the relevant ClpP inhibitor, which we name clipibicyclene. We show in vitro and in vivo that clipibicyclene is a potent covalent inhibitor of ClpP and that cluster-associated ClpPs provide resistance. ClpP inhibition results in antibacterial activity against actinobacteria, including Mycobacterium smegmatis, and inhibition of virulence factor production by Staphylococcus aureus. Finally, we solve the crystal structure of clipibicyclene-modied Escherichia coli ClpP. Clipibicyclene’s discovery deconvolutes the actual function of a family of natural products widespread in nature. It provides a novel scaffold for therapeutic ClpP inhibitor development, making these ndings signicant from the perspective of their discovery and their clinical potential. as Strains streaked on agar. Mixed Cac16/ClpP2 uorogenic

Targeting of ClpP is attractive in the discovery of both anticancer and antibacterial therapeutic candidates. For example, several synthetic β-lactones and phenyl esters have been developed that inhibit bacterial and mitochondrial ClpP [5][6][7][8][9] . In bacterial pathogens where it is dispensable (e.g., Staphylococcus aureus), ClpP inhibition can be used as an anti-virulence strategy 6,10 , while if essential (e.g., Mycobacteria tuberculosis), it can act as a traditional antibiotic 5 . In the context of anticancer drugs, human mitochondrial ClpP is dispensable but appears to be more important to the function of cancer cells than normal cells 11 . For example, ClpP inhibition can selectively kill acute myeloid leukemia cells versus normal hematopoietic cells 12 .
While ClpP inhibition is the aim of many drug discovery campaigns, activation of the complex is also attractive. By mimicking AAA+ ATPase binding and inducing widening of the ClpP tetradecamer's axial pore, ClpP activators stimulate nonspeci c intracellular proteolytic activity and cause cell death 13,14 .
Imipiridones such as ONC201 15 activate mitochondrial ClpP and are currently being studied in numerous anticancer clinical trials 16 . ONC201 also has antibacterial activity via the same mechanism 17 . Most ClpP campaigns focus on synthetic compounds, and only a single natural product, acyldepsipeptide antibiotic (ADEP) 18 , is known to speci cally target this important protease complex. ADEP appears rare in nature as the producer organism, Streptomyces hawaiiensis NRRL15010, remains the only organism in GeneBank containing the ADEP biosynthetic gene cluster (BGC) 19 . Given ClpP's promise as a therapeutic target, we wondered if other natural products that target ClpP were yet to be discovered. Such compounds could identify new chemical scaffolds with application in antibacterial and anticancer drug discovery.
Target-directed genome mining for ClpP directed natural products To avoid self-intoxication, BGCs contain resistance genes against their encoded antibiotic. When this resistance gene takes the form of an antibiotic-insensitive copy of the targeted protein, searching for BGCs containing that target gene can speci cally lead to the identi cation of new antibiotics. So-called "target-directed genome mining" was applied in the discovery of thiolactomycin inhibitors of fatty acid synthases FabB /F 20,21 . The ADEP BGC encodes a resistant copy of ClpP, demonstrating that this protection mechanism from self-intoxication is also relevant to ClpP 19 . We, therefore, applied targetdirected genome mining to identify novel natural products targeting ClpP.
To identify ClpP-associated clusters, we used the collection of Actinobacterial ClpP proteins from pFAM as queries to extract 5928 clpP genes from RefSeq ( Figure 1a). We examined genomic regions 50 kb upstream and downstream of each clpP homolog and identi ed BGCs in 1096 of these regions using AntiSMASH 22 . These BGCs were then ltered for those that contained ClpP within cluster boundaries, as identi ed through tblastn of S. coelicolor ClpP1, leaving 145 candidate hits. We next sought to differentiate cluster-associated from housekeeping clpPs coincidentally in proximity to a BGC, an exercise complicated because actinomycetes encode up to six clpP homologs per genome. Using the characteristic genetic context of each clpP homolog to lter out housekeeping copies (see methods), we landed on a nal prioritized list of ten BGCs associated with ClpPs ( Figure 1a, Supplementary Table 1).
While our target-directed genome mining approach is similar to the idea employed by the Antibiotic Resistant Target Seeker (ARTS) 21 , the latter focuses on a single genome/group of genomes, and using a reference 'core' genome, identi es any duplicated essential gene within these genomes' BGCs. Conversely, our approach focuses on a single target, ClpP, and searches all genomes in RefSeq. Furthermore, given the multiple and a variable number of clpP copies per genome, speci c knowledge of genetic context was vital to differentiating housekeeping from cluster-associated paralogs.
Interestingly, six of our ten nal prioritized BGCs contained a bimodular non-ribosomal peptide synthetase (NRPS) predicted to activate proline and serine. We identi ed 306 BGCs in unique species' genomes from RefSeq with this bimodular NRPS (Supplementary Table 2). These BGCs were missed in our initial analysis due to cluster fragmentation, upload date, or lacking ClpP association. As a comparison, BGCs for streptothricin-family compounds, which are estimated to be produced by 1 in 10 antibiotic-producing Streptomyces spp. 23 , are present in RefSeq only 50 times. In addition to its abundance, this bimodular NRPS is widespread across genera of the order Actinomycetales (Streptomyces, Saccharothrix, Nocardia, etc.) as well as across phyla in important human pathogens, including Pseudomonas aeruginosa, Acinetobacter baumannii, and Enterobacter cloacae (Supplementary Table 2). While not all of the identi ed BGCs were associated with ClpPs, many were located close to other serine hydrolases such as S8-family peptidases and β-lactamases. These results suggest that this bimodular NRPS BGC family is unusually common, widespread in environmental and clinical bacteria, and may have a conserved function of inhibiting Ser hydrolases.
To visualize the relationship between ClpP association and BGCs in this family, we selected 30 exemplar BGCs and built a multi-locus phylogeny using the tool CORASON (CORe Analysis of Syntenic Orthologs to prioritize Natural product biosynthetic gene clusters) 24 . This analysis effectively revealed three distinct groups among our collection of BGCs ( Figure 1b). In the rst group, the bimodular NRPS is adjacent to a Baeyer-Villiger avin-containing monooxygenase (FMO) in a pseudomonal strain, with members of this group encoding the known phospholipase inhibitor SB-253514 (also called brabantamide, Supplementary gure 1a) 25,26 . In the second group, the bimodular NRPS is adjacent to a condensation domain containing protein, with no product having been characterized. The third group includes BGCs associated with an FMO and ClpP, with members known for producing a family of bacterial pyrrolizidine alkaloids (Supplementary gure 1a). The biosynthesis of some of these compounds has been characterized, such as for the bohemamines produced by Streptomyces sp. CB02009 27 and azabicyclenes (also called azetidomonamide B) produced by P. aeruginosa PAO1 28,29 . However, to the best of our knowledge, the obligate association of these BGC's with ClpP has not been previously explored. We, therefore, set out to further examine this widespread family of clusters and investigate their association with ClpP.
In surveying the unknown members of this family of ClpP-associated BGCs, we became interested in a cluster from Streptomyces cattleya DSM 46488 that, in addition to the bimodular NRPS, also contains a type I PKS (Figure 1c, Supplementary Table 4). We named the cluster cac for ClpP Associated Cluster. The cac BGC was transcriptionally silent under several culturing conditions tested. Consequently, we used heterologous expression by directly capturing the 44 kb cluster on a shuttle vector (pCGW-cac) and moving it to the superhost Streptomyces coelicolor M1154. To transcriptionally activate the cluster, we rst attempted to overexpress the cluster-situated XRE family transcriptional activator, Cac15. Still, we did not observe production of any new metabolites from this engineered strain, consistent with incomplete transcriptional activation (Figure 1c, Supplementary Figure 1b). Therefore, we refactored cac by inserting non-native promoters in front of four putative transcriptional units (Supplementary Figure 1c; pCGW-cac-LHK). This approach successfully activated the full BGC and supported the production of numerous metabolites (Supplementary Figure 1b, Figure 1d).
We puri ed and solved the structure of major peaks by NMR and/or LC-MS/MS (Supplementary Figures  2-6, Supplementary Tables 5-8, Supplementary Discussion). The primary product identi ed was a novel azabicyclene with a hydroxylated decatriene acyl tail (Figure 1e, compound 5). We also identi ed congeners lacking the hydroxyl group (compound 6), harboring a ketone at the same position (compound 4), and primary amide variants of the respective acyl tails (compounds 1-3). We named these molecules azabicyclenes B (4), C (5), and D (6), following the nomenclature of the compound azabicyclene (which we will hereto refer to as azabicyclene A) produced by P. aeruginosa PAO1 (Supplementary Figure   1a) 28, 29 . The biosynthesis of azabicyclenes B-D is predicted to be similar to bohemamines and azabicyclene A 27,29 (Supplementary Discussion, Supplementary Figure 7).

An elusive product of the cac BGC results in ClpP inhibition
Having identi ed the azabicyclenes as metabolites produced by cac, we next tested the hypothesis that they disrupt the function of ClpP. However, azabicyclene C and D did not robustly inhibit the growth of eukaryotic cells or bacterial species where ClpP is essential (Supplementary Figure 8a). Azabicyclene D has weak growth inhibitory activity (MIC = 64-128 μg/mL) but is equally active against species where ClpP is dispensable (S. aureus, Bacillus subtilis) and clpP knockout strains (B. subtilis ΔclpP), indicating that its activity is independent of ClpP. Azabicyclenes C and D also do not impact ClpP activity in vitro (Supplementary Figure 8b). Similarly, bohemamines puri ed from Streptomyces sp. NBRC110035 did not affect ClpP (Supplementary Figure 8c), consistent with no previously reported biological activity 30 .
Since azabicyclenes and previously identi ed pyrrolizidine alkaloids do not perturb ClpP, we hypothesized that another undetected product of the BGC might. To comprehensively assay metabolites produced by cac, we, therefore, probed ClpP function directly in the heterologous host, S. coelicolor M1154 pCGW-cac-LHK. In S. coelicolor, as in many Streptomyces spp., there are ve copies of clpP in three operonic units: clpP1clpP2, clpP3clpP4, and clpP5. Each operon encodes subunits capable of forming a functional hetero-tetradecameric complex except for ClpP5, whose catalytic triad is irregularly spaced, and proteolytic activity remains unknown 31 . ClpP1P2 complexes are susceptible to ADEP activation while ClpP3P4 is resistant (Figure 2a) 32 . Due to this differential susceptibility, ADEP can be used to probe the inhibition of ClpP1P2 in S. coelicolor; if ClpP1P2 is functional, nonspeci c degradation induced by ADEP will result in cell death, while if ClpP1P2 is inhibited, ADEP will be unable to trigger nonspeci c degradation and ClpP3P4 will support the growth of cells. S. coelicolor M1154 pCGW-cac-LHK was more resistant to ADEP than the empty vector control, S. coelicolor M1154 pCGW, indicating that ClpP1P2 was inhibited in the heterologous producer ( Figure 2b). Furthermore, deletion of the cluster-associated ClpPs homologous to clpP1 and clpP2, cac16, and cac17 respectively, provided further ADEP resistance. This nding supports that Cac16/17 are susceptible to ADEP but resistant to inhibition by cac BGC products ( Figure 2a).
Having shown that harboring the cac BGC results in ClpP1P2 disruption in the heterologous producer, we next sought to develop an assay to easily determine whether this was through the production of a secreted metabolite. To this end, we made use of the transcriptional regulation of ClpP homologs in S. coelicolor. Under normal conditions, ClpP1P2 is the primary housekeeping complex, and in conjunction with ClpX targets substrates for degradation, including ClgR, its own transcriptional activator, and PopR, the activator of clpP3clpP4 (Figure 2c) 31,33,34 . Since PopR is usually degraded, the clpP3clpP4 operon is not expressed and is only activated upon PopR accumulation due to ClpP1P2 disruption 31 . Validating this reasoning, we observed activation of clpP3 by reverse-transcriptase PCR (RT-PCR) in the heterologous producer (Supplementary Figure 9a). Therefore, to easily probe for ClpP1P2 disruption in S. coelicolor not harboring cac, we created the construct pIJGUS-pClpP3, where the clpP3 promoter drives gusA expression, a β-glucuronidase that hydrolyzes the chromogenic substrate X-gluc resulting in the production of a blue pigment 35 . Co-streaking this indicator strain next to S. coelicolor M1154 pCGW-cac-LHK resulted in P clpP3 activation, as shown by formation of the blue pigment ( ClpP1P2. Using this same assay in the context of another ClpP-associated BGC known to produce bohemamines, which have no ClpP activity (Supplementary gure 9d), co-streaking the producer organism S. sp. NBRC110035 revealed P clpP3 activation (Supplementary Figure 9e). Therefore, the production of ClpP inhibitors is a widespread trait of this family of BGCs.
Next, we tested whether our elusive ClpP1P2 inhibitor could kill actinobacterial species where ClpP is essential. However, in Kirby-Bauer assays using agar plugs inoculated with S. coelicolor M1154 pCGWcac-LHK, minimal zones of inhibition were produced against wildtype S. coelicolor (Figure 2e). We hypothesized that these small zones resulted from intrinsic resistance provided by ClpP3P4, as has been observed for other β-lactone ClpP inhibitors 5 . Consistent with this hypothesis, S. coelicolor M1154 pCGWcac-LHK Δcac16-17 is viable even though ClpP1P2 is inhibited in this strain. Moreover, S. coelicolor ΔclpP3P4 exconjugants with pCGW-cac-LHK Δcac16-17 could not be isolated. Indeed, in agar plug assays with S. coelicolor M1154 pCGW-cac-LHK, S. coelicolor ΔclpP3P4 was more susceptible than wildtype to growth inhibition (Figure 2e). In a collection of six other Streptomyces spp., those that naturally lacked clpP3clpP4 were overall more susceptible to growth inhibition than those that encode these homologs (Supplementary Figure 9f), suggesting that this is a general resistance mechanism in the environment. Notably, the growth of Mycobacterium smegmatis was also inhibited by the ClpP inhibitor ( Figure 2e). Covalent modi cation of ClpP guides puri cation of Clipibicyclene Given our strong evidence for the production of a ClpP inhibitor by cac family BGCs, we wondered why we and others had been previously unable to detect this bioactivity or the responsible metabolite. The ClpP1P2 inhibitory activity of spent media from S. coelicolor M1154 pCGW-cac-LHK decreased throughout a ve-day fermentation and was lost upon incubation overnight at room temperature or nbutanol extract generation (Supplementary Figure 9g). In contrast, azabicyclenes accumulate in fermentation broth over time (results not shown), suggesting that they are not responsible for the activity, and are in fact, shunt metabolites. The decreasing activity we observe from spent media and butanolic extract over time indicates that the active metabolite is unstable. This led us to hypothesize that the compound may possess a reactive warhead that inhibits ClpP by covalent modi cation of the active site Ser. The active metabolite's instability compared to inert azabicyclenes explains how it evaded detection and may explain why previous research has focused on stable pyrrolizidine alkaloids like bohemamine rather than the active metabolites produced by these BGCs.
To test for covalent modi cation of ClpP, we rst reconstituted the in vitro activity of recombinant S. cattleya ClpP1 and ClpP2 hetero-tetradecamers (hereafter referred to as ClpP1P2 scatt ). Peptidase activity required both subunits to be present and was stimulated by ADEP (Figure 3a), similar to M. smegmatis ClpP1P2 36 . ClpP proteins are produced as pro-peptides and require N-terminal processing to form active peptidases. ClpP1 and ClpP2 were co-incubated to predict processing sites and then analyzed by intact protein liquid chromatography-mass spectrometry (LC-MS). We observed no evidence of ClpP1 scatt processing while ClpP2 scatt was cleaved only in the presence of ClpP1 scatt , indicating cross-processing (Supplementary Figure 10a). Processing did not require the presence of cofactors such as ADEP, corresponding closely to what is observed in M. smegmatis 36 . We further characterized the role of agonist peptides, substrate selection, and asymmetric interaction with AAA+ ATPases (Supplementary discussion, Supplementary Figure 10b,c). This represents the rst time ClpP from an Actinomycetaceae species has been reconstituted in vitro.
Next, we incubated recombinant ClpP1P2 scatt in spent media from S. coelicolor M1154 pCGW-cac-LHK containing the active metabolite. Based on our time course results (Supplementary gure 9g), spent media was harvested after just 24 hours, diverging from our standard protocol for fermentation involving growth for 5-14 days. After recovery of ClpP from spent media, peptidase activity was assayed. The activity of both ClpP1P2 scatt and Escherichia coli ClpP (ClpP ec ) was abolished (Figure 3b We then analyzed spent media containing the active compound by LC-MS and detected a peak with an m/z of 329.14 Da, which corresponds to the mass shift observed from the modi cation of ClpP. This peak was associated with a parent ion with an m/z = 347.15, suggesting m/z = 329.14 Da is a dehydration product (Figure 3e). Further, this species was removed from spent media after incubation with recombinant enzyme, con rming that it was the active metabolite (Supplementary Figure 11a). We subsequently isolated the active compound, which we named clipibicyclene, by organic extraction of the supernatant after a 24 hr fermentation. The pure compound was relatively stable once removed from the fermentation broth and could covalently modify and inhibit ClpP1P2 scatt and ClpP ec in vitro  Figure 3g). Future studies will focus on the biosynthesis of the bicyclocarbamate warhead and the installation of the hydroxyl group on azabicyclene C's acyl tail (See supplementary discussion). Notably, the discovery that clipibicyclene is a ClpP inhibitor represents the identi cation of the biologically relevant metabolite produced by cac, and more broadly, this BGC family.
Since we observed growth inhibition of Streptomyces and Mycobacterial strains using agar plugs, we tested whether puri ed clipibicyclene could also kill bacteria in liquid culture. However, we were unable to observe growth inhibition at concentrations up to 512 μg/mL by microbroth dilution. Clipibicyclene was also non-toxic to mammalian cells (HEK293) up to 250 μg/mL. This lack of activity may result from the instability of clipibicyclene in media over the >24 hr incubations required for these assays (Supplementary Figure 11c). We also tested clipibicyclene's ability to inhibit S. aureus production of hemolysin, which is attenuated in ClpP mutants 6 . We observed that 64-128 μg/mL of puri ed clipibicyclene reduced hemolysis induced by S. aureus MRSA strains (Supplementary Figure 11d). We hypothesize that modifying clipibicyclene's scaffold may afford more stable derivatives for further clinical development.
Cluster-associated ClpPs are resistant to clipibicyclene inhibition A fundamental postulate of ClpP-directed genome mining is that cluster-associated ClpPs, in this case, Cac16 and Cac17, provide resistance to the produced antibiotic. Consistently, expression of cac16-17 in S. coelicolor provided resistance to growth inhibition by clipibicyclene (Figure 4a). It also protected cells from P clpP3 activation using our S. coelicolor pIJGUS-pClpP3 indicator strain (Figure 2d), demonstrating that it performs an analogous function to ClpP1P2 in the ClgR regulatory network ( Figure  2b).
To investigate clipibicyclene resistance in vitro, we rst reconstituted the activity of recombinant Cac16 and Cac17 and characterized their activity. Both Cac16 and Cac17 were N-terminal processed in trans (Supplementary Figure 10a), and analogous to ClpP1P2 scatt , peptidase activity required both subunits to be present and was stimulated by ADEP ( Figure 3a). Substrate selectivity and asymmetric interaction with ADEP were also assessed (see Supplementary Discussion and Supplementary Figure 10b,c).
Measuring peptidase activity, Cac16/17 complexes were signi cantly more resistant to clipibicyclene inhibition than either ClpP1P2 scatt or ClpP ec (Figure 4b). We also assessed the susceptibility of different ClpP isoforms to covalent modi cation by intact protein LC-MS using 100 μM clipibicyclene (Figure 4c). ClpP1 scatt was modi ed to a greater extent than ClpP2 scatt , and ClpP2 scatt was not modi ed at all when co-incubated with ClpP1 scatt . Cac16 was almost wholly resistant to modi cation by 100 μM clipibicyclene, while Cac17 was partially modi ed. Together with cell-based assays, these results show that Cac16/17 provide resistance to clipibicyclene through reduced susceptibility to covalent inactivation. Clipibicyclene's speci city for different ClpP subunits, including ClpP ec , ClpP1, ClpP2, Cac16, and Cac17, may result from differences in active site shape or reactivity and warrants further investigation. We next wondered whether both Cac16 and Cac17 are required for resistance to clipibicyclene. To address this, we tested growth inhibition of S. coelicolor ΔclpP3P4, expressing various combinations of Cac16, Cac17, and catalytically dead mutants (Cac16 S108 , Cac17 S129A,S130A ). Cac16 alone was necessary and su cient to provide resistance to clipibicyclene and required that the enzyme be catalytically active ( Figure 4a). In contrast, Cac17 was insu cient to provide resistance. We sought to explain this discrepancy by rst investigating the minimal ClpP subunits required to support the growth of Streptomyces. To this end, we placed the clpP1clpP2 operon under control of a tightly repressed cumate inducible promoter 37 at its native locus (P A26-cmtO :clpP2clpP2; Figure 4d). In a ΔclpP3P4 background, this allows controlled expression of these essential genes. Complementation was then assessed in the absence of inducer by de ned ClpP subunits expressed in trans. Since hetero-complexes are required for tetradecamer formation, we tested which subunits were essential for growth by generating inactive variants of each isoform's partner. Both ClpP1 S99A /ClpP2 and ClpP1/ClpP2 S132A supported growth ( Figure   4e), showing that only a single catalytically active ClpP isoform is required S. coelicolor growth. Coexpression of Cac16/17 supported growth, as did Cac16/Cac17 S129A,S130A , but Cac16 S108 /Cac17 did not ( Figure 4e). Therefore, Cac17 lacks the required catalytic activity or biological function for it to support growth. Combined with in vitro data showing partial modi cation of Cac17 by clipibicyclene, these observations explain Cac17's inability to provide resistance.
Finally, we wondered whether functional tetradecamers were formed strictly between ClpP1/ClpP2 and Cac16/Cac17 or if 'mixed-complexes' of ClpP1/Cac17 and Cac16/ClpP2 were also relevant for resistance and growth. Supporting this hypothesis, mixed complexes were active in in vitro peptidase assays ( Figure  4f). While complexes containing Cac16 or Cac17 were less active in vitro than ClpP1P2 scatt , this may be an artifact of sub-optimal reaction conditions. In vivo, using our inducible system for testing ClpP function in a clean background, Cac16/ClpP2 and ClpP1/Cac17 could support growth (Figure 4e), while only the former provided resistance, consistent with our previous results (Figure 4a). Therefore, ClpP subunits can mix and match in the cell, and these mixed complexes are physiologically relevant. This provides the rst evidence of the complexity of ClpP function in an organism containing more than two isoforms.
Crystal structure of ClpP ec in complex with clipibicyclene We next determined an x-ray crystal structure of ClpP ec in complex with clipibicyclene to uncover the molecular basis for the observed irreversible inhibition. ClpP ec was modi ed using pure clipibicyclene until intact MS analysis determined it to be fully modi ed. Using this material, a ClpP ec :clipibicyclene complex was solved to 2.95 Å resolution using the apo form of ClpP ec as a search model 38 (PDB 1TYF, (Supplementary Table 10). The ClpP ec inhibitor complex crystallized with two complete tetradecamers in the asymmetric unit with all catalytic serine residues (S97) modi ed (Supplementary Figure 13a). The individual tetradecamer complexes were formed by two heptameric rings that produced the typical barrellike quaternary structure of ClpP 38 . However, the N-terminal residues (7-17) that form each heptameric ring's axial pores were disordered and could not be fully modeled (Supplementary Figure 13b). The Nterminal region of ClpP is known to exhibit high conformational exibility, and thus it is unlikely that disorder in these residues was caused by clipibicyclene binding 39 . Moreover, each ClpP ec monomer (18-193) adopted the typical α/β fold and did not signi cantly differ from the apo form as judged by superposition of the tetradecamer (RMSD = 0.38 Å over 16,148 atoms; Supplementary Figure 13c).
Positive peaks in the electron density map extended from the catalytic Ser97 residues in all 28 chains, indicating covalent modi cation and full occupancy of the clipibicyclene adduct (Supplementary Figure  14). The structures of the modi ed Ser97 residues were built and restrained under our proposed reaction mechanism and t well within the electron density map (Figure 3g). The adducts were de ned by a carbamoyl linkage between the O of Ser97 and the nitrogen of the azetine ring ( Figure 5a). As expected, the seven-membered ring of clipibicyclene is no longer intact. Instead, it is replaced by an imide moiety connecting the azetine portion of the adduct to the linear aliphatic chain, which showed a high degree of exibility ( Figure 5a). Many of the aliphatic chains possessed elevated B-factors and weak electron density, which precluded them from being built. However, three complete adduct complexes could be placed (chains A, O, and W), revealing different aliphatic tail conformations (Supplementary Figure 14). We use the adduct from chain A as the representative conformer to describe the ClpP ec :clipibicyclene complex.
In the active site context, the clipibicyclene adduct points clockwise relative to the interior circumference of the heptameric ring of ClpP ec (Figure 5b). The azetine ring is positioned between sheets β4 and β9 in the hydrophobic S1 subsite and forms van der Waals contacts with I70, M98, P124, and L125 ( Figure  5c,d). The carbonyl oxygen of the carbamate linkage forms a critical hydrogen bond with the backbone amide of Gly68, which usually serves as an H-bond donor in the oxyanion hole (Figure 5d). The linear portion of the adduct extends toward the S' subsites over the carbonyl carbon across the β4 strand and partially protrudes into the solvent. To our knowledge, this is the rst example of an inhibitor that occupies this region of a ClpP active site (Figure 5c, d). Only one potential H-bond is formed between the side chain of Q134 and the imide moiety, and a van der Waals contact is likely made with P66 and the aliphatic tail. This limited interaction network along the tail probably explains the disorder associated with the linear aliphatic portion of the adduct (Figure 5c,d). Except for a few outliers, these interactions were observed in the majority of chains. Based on this model, irreversible inhibition can be partly explained by both the inherent stability of the carbamoyl moiety and/or steric exclusion of water from the catalytic base (His122), preventing any base-catalyzed hydrolysis of the adduct.

Discussion
By applying target-directed genome mining, here we describe the association of a widespread family of BGCs encoding bimodular NRPSs with the highly conserved and often essential protease ClpP. Despite being a relatively well-known BGC family for their production of pyrrolizidine alkaloids, the function and biologically relevant molecules produced by this group of BGCs has previously evaded discovery. Using a combination of heterologous expression, in vivo assays focused on Streptomyces biology and in vitro enzyme reconstitution to guide inhibitor puri cation, we identi ed clipibicyclene, a representative ClpP inhibitor, as the active metabolite produced by cac. The production of ClpP inhibitors is likely a governing property of this family of BGCs. The presence of peptidases in other bimodular NRPS BGCs suggests that they may produce other Ser hydrolase inhibitors such as the brabantamides 25,26 . Ser hydrolase inhibitors hold potential for various clinical indications, and as a ClpP inhibitor, clipibicyclene may hold promise as a scaffold for antibiotics against the important human pathogen, Mycobacterium tuberculosis, or virulence inhibitors for S. aureus. In addition to drug development, the abundance and widespread nature of this family of BGCs, from Streptomyces to Pseudomonas, is unusual compared to other BGCs families and suggests that the metabolites produced are important to bacterial physiology. In addition to informing eco-evolutionary hypotheses about the ubiquity of these BGCs, our identi cation of clipibicyclene as a ClpP inhibitor is essential to deciphering their function in human pathogens, including P. aeruginosa, K. pneumoniae, and E. cloacae. Further investigation is warranted, as production of metabolites from the ClpP-associated aze BGC in P. aeruginosa PAO1 causes reduced growth and virulence in a Galleria mellonella model 28 .
Numerous antibiotics act on and remodel the transcriptional and translational landscape of the cell, and by doing so, play essential roles during infection, adaptation, and survival [40][41][42] . Post-translational regulation is likely no less critical in coordinating cellular activities, and so it should be expected that specialized metabolites have also evolved to modulate this level of regulation. Fully elucidating the function, biosynthesis, and mechanism of clipibicyclene and related compounds will be the rst step towards understanding post-translational control of the proteome through ClpP.
ADEP was puri ed from S. hawaiiensis NRRL15010 as previously described with minor modi cations 19  and puri ed using a procedure similar to that for azabicyclenes described below.
MICs were assessed using standard microbroth dilution protocols. Streptomyces spp. were assayed in TSB, M. smegmatis in 7H9 media, and all other strains in Mueller-Hinton Broth (BD Biosciences).
In silico detection and analysis of ClpP associated clusters First, we collected all ClpP homologs from Actinobacteria in PFAM (PF00574) to assemble a complete and diverse set of relevant ClpP-like proteins. This collection of ClpP protein sequences were used to query prokaryotic RefSeq proteins using blastp, and the top ve hits for each query were taken. Proteins were then mapped to RefSeq genomes with the identical protein report function. Using a custom python script, 50 kb upstream and downstream of each clpP hit was collected, and BGCs were identi ed using AntiSMASH 22 . Further ltering was performed in Geneious V8.9.1. Clusters containing ClpP homologs were identi ed by tblastn of S. coelicolor ClpP1 (NP_626855.1). Distinctive genetic context for clpP paralogs is as follows: clpP1clpP2 is associated with the AAA+ ATPase clpX, clpP3clpP4 is associated with the transcriptional regulator popR, and clpP5 is associated with a distinctive gene encoding a protein of unknown function. Tblastn was used to identify these paralogs from protein sequences: ClpX (NP_626853.1), PopR (NP_631335.1, WP_067791847.1, WP_028564420.1), ClpP5-associated hypothetical protein (NP_625527.1). Eight BGCs where clpP was located on the contig edge were also discounted.
To count how many Ser-Pro bimodular NRPS or streptothricin-type BGCs were present in RefSeq, key proteins (WP_014143997.1 and WP_037694042.1, respectively) were taken as BlastP queries against the NCBI non-redundant protein sequence database (accessed January 7, 2021), allowing up to 5000 hits and excluding P. aeruginosa due to its overwhelming abundance. Amino acid % identity cut-offs were set manually by determining which cut-off separated hits belonging to the BGC family from unrelated clusters. To this end, hits were rst sorted by genus. Identical protein sequences in RefSeq were taken to obtain genomic context, and ~10 kb on either side of the query protein was analyzed by AntiSMASH to determine if the hit belonged to the BGC family. Hits above the % identity cut-off were taken, and identical sequences were removed to give a nal count of BGCs in unique species. 30 BGCs containing the biomodular NRPS for CORASON analysis were collected using an arbitrary amino acid identity cut-off of 54% vs. Cac9 (WP_014143997.1; Supplementary Table 2). Five Pseudomonas spp. BGCs with amino acid identity <54% were also included for comparison. CORASON was run on contigs containing these 30 BGC using Cac9 (NRPS) as the query gene, cac from S. cattleya as the reference BGC, and a bit score cut-off of 1000.
Cloning cac using TAR pCGW is a capture construct derived from pCAP03 20 , and modi ed to use the 'oriV-ori2-repE-sopABC' single copy origin of replication from pBAC-lacZ as previously described 44 . pCGW was maintained in E. coli EPI300, which controls the expression of trfA required for high copy ampli cation from oriV with an arabinose inducible promoter. When necessary for miniprep and plasmid mapping, high copy number was induced using 1 mM arabinose. The vector backbone contains all the elements required for propagation and selection in E. coli (pUC ori and Kan R ), yeast (ARSH4/CEN6 and TRP1 auxotrophic marker), and Streptomyces (φC31 integration, oriT, and Kan R ).
The boundaries of cac were de ned by comparison with homologous clusters in Streptomyces pini PL19 NRRL B-24728, Streptomyces barkulensis RC1831, and Kitasatospora phosalacinea NRRL B-16228 (Supplementary Figure 15). 50 bp homology arms anking cac (orf-3 to cac18) were concatenated with an MssI site in between, and 18 bp overlaps with the pCGW backbone were added to either end. The gBlock (Integrated DNA Technologies; IDT) was cloned into pCGW linearized with NdeI/XhoI by Gibson assembly (Supplementary Table 12). Before TAR cloning, the capture vector was digested with MssI to release the hooks.
HMW genomic DNA was prepared from S. cattleya DSM 46488 and digested with XhoI and NdeI to cut surrounding, but not within, the BGC. Saccharomyces cerevisiae VL648N sphereoplasting and transformation were carried out as previously described 45 . Twenty-two22 colonies were restreaked on SD-Trp plates and screened by colony PCR for the desired insertion using diagnostic primers located inside cac, giving one positive colony. DNA was extracted from the positive yeast colony by zymolyase treatment, followed by phenol/chloroform extraction and ethanol precipitation. This DNA was transformed into E. coli EPI300 cells by electroporation and plasmids were selected with kanamycin. E. coli EPI300 was induced for high copy number using 1 mM arabinose, and the resulting construct, pCGWcac, was con rmed by restriction mapping.
Refactoring pCGW-cac pCGW-cac was refactored using a combination of yeast recombination and λ-red recombineering in E. coli. All primers, gblocks, strains, and plasmids are listed in Supplementary Tables 11-13. First, Leu2 and His3 auxotrophic selection cassettes were constructed to contain orthogonal promoters and terminators using pSASS series plasmids, a gift from Dr. Mike Tyers (University of Montreal). The leu2 open reading frame was PCR ampli ed from pRS316 and inserted into pSASS5, and his3 was PCR ampli ed from pYAC10 (Mike Tyers) and inserted into pSASS4, using Gibson assembly. Selection cassettes containing the marker along with yeast promoter and terminator sequences were subsequently PCR ampli ed from pSASS5-leu2 and pSASS4-his3. For replacement of the cac14 promoter, XNR_1700p was PCR ampli ed from Streptomyces albus gDNA 46 , and stitched to the leu2 selection cassette with 500 bp homology arms to cac on either side through overlap extension PCR. The resulting refactoring cassette was cotransformed with pCGW-cac into Saccharomyces cerevisiae SASy31/SASy35 by standard lithium acetate/single-stranded carrier DNA/PEG mediated transformation 47 . Recombinants were selected on SDtrp-leu plates, screened by colony PCR, and the resulting construct, pCGW-cac-L was recovered and con rmed in E. coli EPI300, similar to initial TAR cloning. Replacement of cac8 and cac9 promoters were achieved similarly using the His3 selection marker sandwiched by synthetic Streptomyces promoters, A26 and A35, incorporated into primer sequences 48 . This time, S. cerevisiae SASy31 transformants were selected using SD-trp-leu-his plates, and the resulting construct, pCGW -cac-LH, was moved to E. coli EPI300.
After refactoring cac8, cac9, and cac14, RT-PCR revealed that cac5 and the putative operon, including cac4, cac3, cac2, and cac1, was still poorly transcribed (data not shown). Since few selection markers were left for use in S. cerevisiae, we chose to use E. coli λ-Red recombineering. The promoter kasOp* and T7 terminator were designed adjacent to the apramycin resistance gene aac(3)IV anked by PmeI/HpaI restriction sites to remove aac(3)IV after recombination. Thirty-nine bp homology arms were added to each end to direct recombination. The recombineering cassette was synthesized as a gBlock by IDT, and recombineering was carried out using standard protocols 49 . pCGW-cac-LH was transformed into E. coli BW25113/pKD46, selected using kanamycin and ampicillin, and grown at 30°C to maintain pKD46. The strain was grown in LB overnight at 30°C, then sub-cultured in SOB without MgSO 4 with the addition of 10 mM arabinose to induce expression of red genes from pKD46. After reaching OD 600 = 0.6, the cells were recovered by centrifugation, washed twice with ice-cold 10% glycerol, and electroporated with 1 μg of linear refactoring cassette. Successfully recombinants were selected with apramycin and grown at 37°C to promote loss of pKD46. The resulting construct, pCGW-cac-LHK-apra, was extracted from E. coli BW25113, transformed into E. coli EPI300, and veri ed by restriction mapping. Finally, aac(3)IV was removed by digestion with HpaI and PmeI followed by intramolecular blunt end ligation with T4 DNA ligase, generating pCGW-cac-LHK.

RT-PCR measurement of cac and clpP regulon expression
Strains for analysis were inoculated 1:100 from a saturated TSBY seed culture into 60 mL Bennett's media in 250 mL ba ed asks, and pellets were taken at 24 hr. Cells were lysed by bead beating mycelium with 4 mm glass beads in 5 mL TRIzol reagent (Invitrogen), and RNA was extracted using the manufacturer's recommendations. RNA from the resulting aqueous phase was extracted a second time using acid phenol/chloroform, then combined with a half volume of anhydrous ethanol, and nally puri ed using PureLink RNA Mini Kit (Invitrogen). Maxima H Minus First Strand cDNA synthesis kit with dsDNase (Thermo Scienti c) was used for cDNA synthesis, and PowerUp SYBR Green master mix (Applied Biosystems) was used for RT-PCR quanti cation on a BioRad CFX96 real-time system. Primers targeting genes of interest (Supplementary Table 11) were designed, and 80-100% e ciency was veri ed before quanti cation. Analysis was performed on three or four independent fermentations and quanti ed in technical duplicate. Technical duplicates for each biological replicate were averaged, then fold change expression for each replicate was calculated by normalizing to hrdB expression using the ΔCt method. Statistical analysis of clpP3 gene expression was performed using GraphPad Prism V6. Multiple comparisons to pCGW were made using a two-sided Kruskal-Willis analysis with Dunn's test for multiple comparisons (n = 3).

Crude extract LCMS analysis
To prepare crude extracts, strains were fermented in 50 mL media for 7 days, unless otherwise indicated, extracted with 15 mL n-butanol, dried under vacuum, and resuspended in 100 µL DMSO. Extracts were analyzed on an Agilent 1290 UPLC with G6550A Q-TOF using a ZORBAX StableBond C18 column (Agilent, 4.6 x 150 mm, 3.5 µm, 80 Å, 0.4 mL/min, Buffer A water + 0.1% formic acid, Buffer B acetonitrile Puri cation and structural elucidation of azabicyclenes S. coelicolor pCGW-cac-LHK seed culture was grown in TSB with 50 μg/mL kanamycin for 3 days, then spent media was removed, and 50 mL worth of mycelium was inoculated into each of 25x600 mL Bennett's media in 3 L asks. Fermentations were grown for 7 days at 30°C, 250 rpm. 13 L of spent media was harvested and extracted with 390 g HP-20 resin (Diaion). The resin was washed with 10% methanol (MeOH), then eluted with 100% MeOH and concentrated under vacuum. Dry material was extracted with ethyl acetate:MeOH (1:1) and dried onto 5 g silica gel (Sigma) under vacuum. Normal phase vacuum liquid chromatography was performed using the following stepwise gradient: 1. hexanes, 2. ethyl acetate, 3. ethyl acetate:MeOH (3:1), 4. ethyl acetate:MeOH (1:1). Fractions 2-4 were combined, dried, and using DMSO, applied to an 86 g reverse-phase CombiFlash ISCO (RediSep Rf C18, Teledyne) eluted with a linear gradient system (5-45% water/acetonitrile, 0.1% formic acid). Fractions containing azabicyclenes were pooled and further puri ed by semipreparative HPLC using a ZORBAX StableBond C18 column (Agilent, 9.4 x 150 mm, 5 µm, 3 mL/min, Buffer A water + 0.1% formic acid, Buffer B acetonitrile + 0.1% formic acid). Compound 2 was puri ed by the following gradient Puri cation and structural elucidation of clipibicyclene S. coelicolor pCGW-cac-LHK culture was grown in the same way as for azabicyclene puri cation, except for that the fermentation was harvested after 24 hr. After the growth of a 2 L culture, mycelium was removed, and the supernatant was extracted twice by a total volume of 4 L dichloromethane. The organic phases were combined and concentrated gently in vacuo to obtain a yellow oil. The crude extract was resuspended in 200 μL DMSO and puri ed on a Waters XSelect CSH preparative C-18 HPLC column (10 × 100 mm, 5 µm, 130 Å, 3 mL/min, Buffer A water + 0.1% formic acid, Buffer B acetonitrile + 0.1% formic acid)) using the following method

Kirby-Bauer assays
Indicator Streptomyces strains were grown for 2-3 days in TSBY media with antibiotic selection as necessary until saturated. Cultures were diluted to OD 600 = 0.1 and streaked on Bennett's agar using a sterile cotton swab. For assays involving ADEP, 50 μg compound in DMSO was placed on a sterile cellulose disk. For assays involving agar plugs, S. coelicolor pCGW-cac-LHK or S. coelicolor pCGW was inoculated on Bennett's agar from a saturated TSBY starter culture and grown for 24 hr before removal of an agar plug that was placed on a plate inoculated with the indicator strain.

GusA indicator strain construction and assays
A cassette consisting of the T7 terminator placed upstream of P clpP3 :gusA was constructed in the pIJ10257 backbone. Using the primers listed in Supplementary Table 11, parts PCR were ampli ed as follows: clpP3 promoter region lacking its RBS from S. coelicolor gDNA, gusA including its RBS from pGUS 35 , and T7 terminator from pDR3K 50 . PCR products were cloned into pIJ10257 digested with EcoRV using Gibson assembly to generate the construct pIJGUS-pClpP3. The construct was moved into S. coelicolor M1154 in addition to pSET152 (empty vector) or pSET152-cac16-17, cloned between XbaI and EcoRI restriction sites using primers listed in Supplementary Table 11.
For time-course analysis of clipibicyclene production, 250 mL asks containing 50 mL Bennett's media were inoculated with S. coelicolor M1154 pCGW-cac-LHK on sequential days. Cultures that had been growing for 1-5 days were harvested simultaneously, and mycelium was removed by centrifugation. Cultures were split and either stored overnight at -80°C, at room temperature, or extracted with butanol, dried under vacuum, and resuspended at 1/150 the original volume in DMSO. 2.5 μL of each sample was spotted on an agar plate to test for P clpP3 activation as above.
Hemolysis assay S. aureus hemolysis assays were performed with strains USA-300 JE2, NE912 (JE2 clpP::ΝΣ) 51 , and the clinical isolate CMRSA-3, which was isolated from Mount Sinai Hospital, Toronto, ON. Overnight cultures of each strain were subcultured in MHB medium at 37 ºC until OD 600 of 0.6 was reached. The bacteria were then pelleted by centrifugation and resuspended in an equal volume of fresh MHB medium. The cell suspension was dispensed in a 96-well round bottom plate and mixed with 2-fold serial dilutions of inhibitor or vehicle. The nal concentration of DMSO did not exceed 1% (v/v). The bacteria were grown for a further 5 hr and then centrifuged (5000 x g, 10 min) to remove the cells. The supernatant was added to a 5% (w/v) suspension of sheep erythrocytes in phosphate buffered saline and incubated at 37 ºC for 1 hr. To determine the relative extent of hemolysis, any intact erythrocytes were pelleted by centrifugation (5000 x g, 10 min) and the released haemoglobin was measured in the supernatant by monitoring absorbance at a wavelength of 545 nm. The experiments were performed on two independent occasions.
Expression and puri cation of ClpPs Untagged E. coli ClpP was expressed from E. coli BL21(DE3) pET9a-EcClpP, a kind gift from Dr. Walid Houry (University of Toronto), and puri ed as previously described with minor modi cations 52 . S. cattleya ClpP1 (SCATT_17350), ClpP2 (SCATT_17340), Cac16 (SCATT_32700) and Cac17 (SCATT_32710) were codon-optimized for expression in E. coli and synthesized as gBlocks by IDT (Supplementary Table 12). Synthesized DNA was cloned into pET-28 digested with NcoI and NotI using Gibson assembly to be expressed with a C terminal 6xHis tag. Constructs were transformed into E. coli BL21(DE3) 1146D (ΔclpP::cm r ). For protein expression, overnight cultures were inoculated into 1 L LB with appropriate antibiotics in a 3 L ask, grown at 37°C until cultures reached OD 600 = 0.6-0.8, then induced with 0.5 mM IPTG. Cultures were grown for 18 hr at 17°C before harvesting cell pellets by centrifugation.
Cell pellets were resuspended in 20 mL lysis buffer (20 mM Tris (pH 8), 300 mM KCl, 10 mM imidazole, 10% glycerol), treated with 10 mg/mL lysozyme and 5 µg/mL DNase, and lysed on ice by sonication. substrates. Quenching of AMC uorescence was observed above 50 μM azabicyclene/bohemamines, so (Suc-LLVY)2-Rhodamine110 was used for all assays containing these compounds. Where indicated, agonist peptides (Sigma) were tested at 500 μM. Where indicated, clipibicyclene, azabicyclene, or bohemamines were preincubated with enzyme for 10 min before substrate addition. 100 μL reactions were initiated with the addition of substrate, carried out at room temperature, and tracked by uorescence excitation/emission 360 nm/460 nm (AMC) or 485 nm/525 nm (rhodamine 110). Rates were calculated using the slope over the rst 15 min of the reaction.

Thermal shift assays for ADEP binding
Melt curves were performed using the following conditions: 5 μM ClpP subunit (monomer), 5x SYPRO orange dye (5000x stock, ThermoFisher), 100 μM ADEP (or DMSO), in ClpP reaction buffer. 50 μL per well were assayed on a BioRad CFX96 real-time system, ramping from 35°C to 95°C in increments of 0.5°C for 10 seconds, reading SYPRO uorescence after every increment. Melting temperatures were calculated using CFX Maestro software. In general, 1 μL of ~30-40 μM protein was injected for analysis. Spectra were deconvoluted, and gures were generated using UniDec software 53 . To assess cross-processing, subunits were mixed in a 1:1 molar ratio and incubated overnight at room temperature. Processing sites were predicted using ExPASy FindPept.
To label ClpPs in spent media, the fermentation broth was rst ltered through an Amicon Ultra centrifugal lters with 5 K cut-off to remove secreted proteins present in the media and neutralized by adding 50 mM Tris-HCl pH 7.5. Next, 5 μg/mL ClpP, either total (ClpP ec ) or each subunit (ClpP1P2 scatt ), was added and incubated for 1.5 hr at room temperature. ClpP protein was recovered, and buffer exchanged using an Amicon Ultra centrifugal lters with 30 K cut-off before LCMS analysis or peptidase assays. To label ClpP with pure clipibicyclene, 20 μM of each ClpP subunit was rst incubated in ClpP reaction buffer for 3 hr at room temperature to allow processing to occur before the addition of 100 μM clipibicyclene and LCMS analysis.

Peptide mapping LC-MS/MS
To identify the site of ClpP1 modi cation, modi ed enzyme was prepared using a modi ed version of the method used prior to intact protein LC-MS. ClpP1 (50 μg) was incubated with spent media ltrate of S. coelicolor pCGW or S. coelicolor pCGW-cac-LHK cultures (10 mL) for 1.5 h at 25º C and concentrated to 1 mg/mL. Five times the sample volume of cold acetone (-20 ºC) was added, and the mixture was and incubated for 10 min at -20 ºC to completely precipitate the protein. The protein oc was collected by centrifugation (12,000 x g, 2 min), and the pellet was resuspended in cold acetone (1 mL). This was repeated three times to remove residual media components. The supernatant was removed from the nal pellet, and the protein was dissolved in 50 mM ammonium bicarbonate buffer pH 8 (20 μL). To proteolytically digest ClpP1 in-solution for peptide mapping, sequencing-grade trypsin:ClpP1 ratio of 1:25 (w/w) was used, and the reaction was incubated at 37 ºC for 16 h. The resultant peptides were analyzed by LC-MS/MS with an Agilent G6550A Q-TOF in positive ion mode. Peptides were separated with a C18 column (Agilent XDB C18 100 mm x 2.1 mm; 2.7 μm) equilibrated with 5% acetonitrile in 0.1 % formic acid. A linear gradient to 95% acetonitrile in 0.1 formic acid was applied over 8 min at a ow rate of 0.5 mL/min. The mass spectrometer was operated with the following parameters: Gas Temp: 350°C, Gas CRISPR-editing and complementation of S. coelicolor ClpPs CRISPR-editing constructs containing sgRNA targeting sequences were cloned using the pCRISPomyces-2 backbone using primers listed in Supplementary Table 11 via golden-gate assembly 54 . S. coelicolor clpP3clpP4 homology arms were designed to introduce a scarless in-frame deletion and cloned using Gibson assembly. For inducible control of clpP1clpP2, a cassette consisting of the cumate-responsive repressor cymR, the strong synthetic promoter A26, and the cymR operator sequence cmtO was designed and synthesized as a gBlock (IDT ; Supplementary Table 12) 37,48 . This cassette was PCR ampli ed using primers with suitable homology arms for Gibson assembly. Construction of the nal pCRISPomyces-2 construct was performed in two Gibson assembly steps: rst homology arms for clpP1clpP2 were introduced anking a HindIII site, and second, the promoter cassette was introduced into the HindIII site.
CRISPR-editing was performed as previously described 55 . Brie y, pCRISPomyces constructs were conjugated to S. coelicolor M1154; exconjugants were restreaked and screened using PCR for successful editing, pCRISPomyces-2 was cured by growth at 37°C, and exconjugants were veri ed again by PCR.
Conjugation and maintenance of the clpP1clpP2 edited strain were performed on media containing 100 μM cumate.
Complementation plasmids pIJ-ncac16-17, pIJ-nclpP1P2, pIJ-nclpP3P4, and pIJ-clpP1-cac17 were cloned by PCR amplifying inserts from S. coelicolor (clpP1clpP2 and clpP3clpP4) or S. cattleya (cac16-17) gDNA using primers listed in Supplementary Table 11 before Gibson assembly into pIJ10257 digested with KpnI and HindIII. For pIJ-cac16-clpP2, the plasmid backbone including cac16 was PCR ampli ed from pIJ-ncac16-17, and clpP2 was PCR ampli ed from gDNA for Gibson assembly. All remaining constructs, including pIJ-ncac16, pIJ-ncac17, and catalytically inactive variants (pIJ-ncac16d17, pIJ-cac16-17d, pIJ-clpP1dP2, pIJ-clpP1P2d) were cloned using site-directed mutagenesis by PCR amplifying the full constructs followed by DpnI digestion and ligation. We observed that native promoters provided superior expression than the commonly used constitutive Streptomyces promoter ermEp*, so they were retained in these constructs. Since Cac17 contains two adjacent serine residues in its active site, both were converted to alanine to ensure the enzyme was catalytically inactive.
Crystallization and structure determination of ClpP ec :clipibicyclene Preparation of ClpP ec :clipibicyclene for crystallization involved incubating puri ed enzyme (2 mg/ml) with 0.5 mM clipibicyclene (5-fold molar excess ) in ClpP reaction buffer for 1 hour at room temperature.
Intact protein ESI-MS veri ed complete covalent modi cation. Before crystallization, the covalent complex was buffer exchanged into 10 mM HEPES pH 7.5, 100 mM NaCl using a BioGel P6-DG column, and was concentrated to 12 mg/ml. Crystals of ClpP ec :clipibicyclene were grown at a 2:1 precipitant to protein ratio using 0.1 M sodium acetate pH 5.6, 30% (v/v) MPD, and were ash cooled in a N 2 cryo stream.
Data collection was conducted at the Canadian Light Source CMCF-BM (08IB1), Saskatoon, SK, Canada. The x-ray data was processed using autoPROC 56 , XDS 57 , and CCP4 58 . The structure of ClpP ec :clipibicyclene determined by molecular replacement with Phenix 59 using apo ClpP ec (PDB ID: ITYF) as the search model. Model building and re nement were carried out using Coot 60 and Phenix 59 with TLS groups determined automatically using the TLSMD webserver 61 . Ramachandran statistics were calculated in Phenix using MolProbity, which gave 97.4 % total favored assignments and 0.21% outliers. To ensure that the stereochemistry of the clipibicyclene adduct remained well-restrained during re nement, complete amino acid restraints were generated for a modi ed serine residue using the GradeWebServer

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.