{"paper_id":"1ae7d498-d463-45ea-aa15-1dec71e5ecc1","body_text":"1 \n \nDiscovery and Biosynthesis of Nitrilobacillins by Post-\ntranslational Introduction of C-Terminal Nitrile Groups \n \nLide Cha1,2, Chuyang Qian1,2, Chandrashekhar Padhi1, Lingyang Zhu3 and Wilfred A. \nvan der Donk1,2* \n \n1 Department of Chemistry and Howard Hughes Medical Institute,  \nUniversity of Illinois at Urbana−Champaign, Urbana, Illinois 61801, United States. \n2 Carl R. Woese Institute for Genomic Biology, University of Illinois at Urbana-Champaign,  \nUrbana, Illinois, 61801, United States. \n3 School of Chemical Sciences NMR Laboratory, University of Illinois at Urbana-Champaign, \nUrbana, 61801, IL, United States. \n* E-mail: vddonk@illinois.edu \n \nAbstract \nNitrile-containing natural products are produced in all kingdoms of life . Despite the \nwide application of nitrile -containing peptide scaffolds in medicinal chemistry , the \npresence of the nitrile group is unprecedented in ribosomally synthesized and post -\ntranslationally modified peptide s (RiPPs). In this work, we report the identification and \ncharacterization of a RiPP biosynthetic gene cluster (BGC) , where an asparagine \nsynthetase-like (AS-like) protein encoded in the BGC converts the C-terminal carboxylate \nof the precursor peptide to a nitrile. Furthermore, a multinuclear nonheme iron-dependent \noxidative enzyme  (MNIO) and  an α-ketoglutarate-dependent HExxH motif -containing \nenzyme (αKG-HExxH) perform stereoselective β -hydroxylation of aspartate and proline \nresidues, respectively. The final product is a cysteine protease inhibitor and shows that \nNature makes similar warheads as found in synthetic therapeutics such as the active \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n2 \n \ningredient of Paxlovid. These findings extend our understanding of the structural and \nfunctional diversity of RiPPs. \n \nIntroduction \nSince the discovery of allyl cyanide in 1863,1 nitrile-containing natural products have \nbeen reported to be produced by animals, plants and microorganisms.2-5 The electrophilic \nnature of the nitrile carbon contributes to the activity of nitrile-containing metabolites in \nbiology and medicinal chemistry.6-8 In nature, production of the nitrile group is achieved \nmostly via dehydration of the corresponding aldoximes catalyzed by various enzymes \n(Figure S1a ).9,10 Alternatively, nitrile formation proceed s through oxidative or ATP -\ndependent mechanisms (Figure S1a).5,11-15 Nitrile groups have been identified in diverse \nclasses of natural products  including glycosides, alkaloids, and terpenes, but  nitrile-\ncontaining peptide natural products are rare, with auranthine the only example discovered \nthus far (Figure S1b). In contrast, use of the nitrile functional group is common in synthetic \npeptides designed to inhibit proteases, with one notable example being nirmatrelvir, a key \ningredient in the SARS-CoV-2 therapeutic Paxlovid that inhibits the viral main protease.16 \nRecent estimates suggest that a very large fraction of natural products (up to 97%) \nremains to be discovered.17 Peptide secondary metabolites are mostly derived from non-\nribosomal or ribosomal pathways. 18,19 Compared with non -ribosomal pathways , which \ninclude the biosynthesis of auranthine, ribosomally synthesized and post -translationally \nmodified peptides (RiPPs) are differentiated by the presence of precursor peptides that \nare genetically encoded. Estimations of the abundance of various natural product classes \nin defined natural environments show that RiPPs are a major family ,20 and in some \nenvironments such as the human microbiome, RiPPs appears to be the most prevalent \nclass of natural products. 21 Maturation of RiPPs involve s diverse post-translational \nmodifications (PTMs) of the precursor peptide.22 Around 50 classes of RiPPs have been \nreported to date  that are  categorized by class -defining PTMs. The number of distinct \nPTMs continues to expand as a result of the  rapid increase in genome sequences over \nthe past two decades.23 A large number of different enzyme families have been linked to \nRiPP biosynthesis.24 Because the substrate sequences of RiPP modifying enzymes are \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n3 \n \nencoded within the BGC, f unctional characterization of RiPP biosynthetic enzymes for \nwhich activity has not been reported previously is less challenging compared with their \ncounterparts in the biosynthe sis of other natural products . These features have made \nRiPP BGCs excellent repositories to discover novel enzyme chemistries.23 \nRecent studies have expanded the chemical space of the products of non-heme iron \ndependent enzymes in RiPP biosynthesis ,25-32 including RiPP BGCs that encode MNIO \nand αKG -HExxH proteins. In the current work, we focused on a BGC that contains \nmembers of both enzyme classes from Peribacillus simplex VanAntwerpen02. We \ndemonstrate that the MNIO and HExxH enzymes  catalyze β-hydroxylation of aspartate \nand proline residues, respectively. The more unique reaction in the pathway features an \nAS-like enzyme that unexpectedly installs a nitrile group at the C -terminus of the \nprecursor peptide. We therefore termed the products from this BGC nitrilobacillins \nbecause orthologous pathways were only identified in Bacillus-related genera. Production \nof the nitrile-containing peptide seems to be regulated by two different mechanisms. First, \nnitrile installation requires Asp hydroxylation by the MNIO. Second , an apparent pseudo \nenzyme related to the MNIO protein is encoded within the BGC, which can compete with \nthe MNIO protein  in binding to the MNIO partner protein , thereby preventing nitrile \nbiosynthesis. The RiPP products are inhibitors of several cysteine proteases tested. This \nwork therefore expands RiPP biosynthe sis with a  pharmaceutically significant \npharmacophore. \n \nResults \nIdentification of the pes BGC \nDuring our genome mining efforts for novel RiPP BGCs encoding MNIOs, a BGC \n(pes cluster) from Peribacillus simplex VanAntwerpen02 was identified. The BGC \nencodes the αKG-HExxH protein PesO, two precursor peptides PesA1/A2, a n MNIO \nenzyme PesH, a putative MNIO partner protein PesI, a n AS-like enzyme PesC, a \nhypothetical protein PesX, two M16 peptidases PesP1/P2, and a transporter protein PesT \n(Figure 1a). Unlike previously reported MNIO enzymes which require a partner protein \nthat is encoded adjacently within the BGC ,25,27,30,33-36 the putative MNIO partner protein \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n4 \n \ngene pesI is positioned next to pesX. Initial bioinformatic analysis of PesX suggested that \nit is a hypothetical protein from an unknown protein family, but structural analysis with \nAlphaFold 337 as well as a Foldseek38 search revealed that PesX likely possesses a \ntriose-phosphate isomerase (TIM)-barrel fold like MNIOs and shares structural similarity \nto another MNIO protein MbnB (Figure S2). However, a sequence alignment of PesX with \ncharacterized MNIO proteins demonstrated that it lacks the iron-binding residues that are \nconserved in MNIOs (Figure S3).39 Therefore, the role and function of PesX was unclear.  \nA sequence h omology search of the enzymes in the pes cluster using BLASTp, \nfollowed by genome neighborhood analysis using RODEO40 lead to the discovery of 26 \ngene clusters that resemble the pes cluster composition. These orthologous clusters all \ncontain homologs of PesX, PesI, PesH and PesC.  αKG-HExxH proteins are only present \nin ~ 75% of the clusters and when they are absent, they are  replaced by a predicted \narginase (e.g. Figure S4). Alignment of the precursor peptides from these homologous \nclusters reveals several conserved residues including a conserved C-terminal R(V/T)DN \nmotif (Figure 1b). Notably, a proline residue before the R(V/T)DN motif only co-occurs \nwhen an αKG-HExxH protein is encoded in the BGC, hinting that the latter enzyme may \nmodify this Pro residue.  \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n5 \n \n \nFigure 1. a) Composition of the pes BGC from Peribacillus simplex VanAntwerpen02 and \nthe amino acid sequence of precursor peptides PesA1/2. The numbering is based on the \nC-terminal fragment obtained upon GluC-digestion; b) Sequence logo of precursor \npeptides from BGCs homologous to the pes BGC (4 2 total sequences, 21 unique \nsequences (Figure S5), identical sequences were removed). \n \nCharacterization of the product of the pes BGC \nBecause the  native pes BGC encoding strains  were not available to us , we \ninvestigated the functions of the encoded enzymes by heterologous expression in E. coli. \nA His 6-tag was appended to the N -terminus of the precursor peptides, and they were \nexpressed separately or co -expressed with a subset of putative modifying enzymes. \nFollowing immobilized metal affinity chromatography (IMAC)  purification, the purified \npeptides were digested with endoproteinase GluC and characterized by mass \nspectrometry. \nWe co-expressed both precursor peptides PesA1 and PesA2 individually with the \nvarious enzymes encoded in the pes BGC with essentially the same results. We will \ndescribe the data with PesA2 here and refer to the Supporting Information for the \nanalogous data for PesA1. Co-expression of PesA2 with the MNIO protein PesH and its \npotential partner PesI  and analysis by  liquid chromatography-high resolution mass \nspectrometry (LC-HRMS) revealed the emergence of a new species with a +16 Da mass \nshift with respect to the unmodified PesA2 (Figure 2). The site of PesHI modification was \nlocated to Asp21 by high-resolution tandem mass spectrometry (HR-MS/MS) (Figure S6). \nMarfey’s analysis  was used to determine the position and stereochemistry of the \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n6 \n \noxidation.41,42 After hydrolysis of the PesHI-modified peptide, the hydroxylated aspartate \nwas reacted with 1-fluoro-2-4-dinitrophenyl-5-L-alanine amide (L-FDAA) and analyzed by \nliquid chromatography coupled to mass spectrometry detection ( LC-MS). The product \ncoeluted with an L-threo-3-hydroxyaspartate standard that was derivatized in the same \nmanner (Figure 3a  and S7). Collectively, these observations demonstrate that PesHI \nfacilitate stereoselective hydroxylation of Asp21. \nNext, we included the αKG -HExxH protein PesO in the co -expression system  in \naddition to PesHI, for which we will use the designation PesOHI . LC-HRMS analysis of \nthe endoproteinase GluC-digested product peptide PesA2-PesOHI revealed the \ndominant production of a M+32 Da species (Figure 2). This additional +16 Da mass shift \ncompared to the PesHI modification was assigned to the Pro1 8 residue by HR-MS/MS \nanalysis (Figure S6). We again performed Marfey’s analysis  on the PesOHI -modified \npeptide to elucidate the identity of the hydroxyproline residue  (Figure 3b  and S7). \nCoelution with L-3S-hydroxyproline derivatized with the M arfey’s reagent indicated  that \nPesO stereoselectively hydroxylates carbon 3 of Pro18.  \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n7 \n \n \nFigure 2. HR-MS/MS analysis of unmodified precursor peptide PesA2 or PesA2 co-\nexpressed with the indicated modifying enzymes. The MS/MS fragmentation pattern of \neach modified peptide is shown (for tandem MS spectra, see Figure S6). Prior to analysis, \nthe peptides were digested with endoproteinase GluC. Fragment ion annotation w as \nperformed using the interactive peptide spectral annotator 43 with residues indicated in \nlower case p and d entered as hydroxylated residue s (M+16), and in lower case d n as \ntwo residues that were hydroxylated and nitrile containing (+16 and −19 to give a net \nchange of −3 Da).  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n8 \n \n \nFigure 3. Structural determination of PesA2 co-expressed with  PesOHIC. a) LC-MS \nanalysis of hydroxyaspartate authentic standards or hydrolyzed PesA2 -PesHI and \nPesA2-PesOHI peptides derivatized with L-FDAA; for coinjections, see Fig. S7. b) LC-MS \nanalysis of L-FDAA derivatized hydroxyproline from authentic standards or hydrolyzed \nPesA2-PesHI and PesA2-PesOHI peptide s; c) 1H−13C HMBC spectrum of digested \nPesA2-PesHIC highlighting the C-terminal residue, and d) proposed structure of digested \nPesA2-PesHIC peptide 1 and PesA2-PesOHIC peptide 2 based on NM R, MS/MS and \nMarfey’s analysis. Key HMBC correlations from c) are labeled. \n \nIncorporation of the AS-like enzyme PesC into the co-expression system resulted in \na new product with a −19 Da mass shift relative to the PesOHI  modified peptide. HR -\nMS/MS analysis demonstrated that  a −3 Da mass shift had occurred to the C -terminal \nAsp21-Asn22 sequence (Figure 2 and S6). Asparagine synthetases usually catalyze the \namidation of aspartate to yield asparagine, 44 while in natural product biosynthesis, \nmembers of this enzyme family also catalyze lactam formation. 45,46 Therefore, we \nenvisioned that an amidation reaction of a carboxyl group (−1 Da) and a dehydration \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n9 \n \nevent ( −18 Da) would account for the observed overall mass shift  of −19 Da . When \ncoupled to the PesHI catalyzed hydroxylation of Asp21  (+16 Da) , amidation and \ndehydration would explain the net −3 Da shift of the Asp21 -Asn22 motif. Within this \nsequence, the dehydration could originate in a dehydroAsp, or in the  formation of ester, \nimide, or nitrile group s ( Figure S 8). The t andem MS/MS data were inconclusive in \ndistinguishing these possibilities, and we therefore turned to analysis by nuclear magnetic \nresonance (NMR) spectroscopy described in the next section. \nWe also integrated the hypothetical protein PesX into the co-expression system. Co-\nexpression of PesX in any combination with the other enzymes  did not lead to the \nemergence of any new species derived from PesA2. These results imply that PesX was \neither non-functional in the E. coli heterologous host or the protein serves a non-catalytic \nrole in the pes BGC. As described below, a potential function of PesX is suggested from \nin vitro experiments. \nAs noted above, c o-expression experiments of PesA1 with Pes enzymes yielded \nsimilar modification patterns as PesA2 (Figure S9), indicating that the less conserved \nprecursor N-termini as well as the residue at position 20 (Thr/Val) have minimal impact \non the function of the modifying enzymes within the pes BGC. \n \nStructural Elucidation of Modified PesA2 \nThe data described thus far suggest  that PesC catalyze s consecutive \namidation/dehydration reactions on the precursor peptide  that is hydroxylated by PesHI \non Asp21 , yet the precise form of dehydration remain ed unclear. For NMR structural \ndetermination, we first attempted to process the modified PesA2 peptides with the native \nprotease PesP1/P2. While the expression of this heterodimer in E. coli yielded soluble \nprotein (Figure S10), prolonged incubation of the peptides and PesP2/P1 led to a range \nof proteolytic products, with a 25-mer the major product (Figure S11). The poor in vitro \nactivities of PesP2/P1 limited their application in this study, and therefore we mutated \nGln14 to Lys in PesA2 with the aim of generating a short peptide upon digestion with \nendoproteinase LysC (for residue numbering, see Figure 1a). We first explored the impact \nof this mutation on the PTM process, as well as the minimal enzyme requirement for \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n10 \n \ncomplete C -terminal modification. HR-MS/MS analysis of this Q14K  mutant after co -\nexpression with PesHIC verified that the product still contained the −3 Da change in the \nAsp21-Asn22 sequence (Figure S12), indicating that Gln14 is not required for PesHIC \nmodification. Therefore, we conducted large scale preparation of PesHI C-modified \nPesA2-Q14K in E. coli. After IMAC purification and LysC digestion , the  resulting \noctapeptide was purified by high-performance liquid chromatography (HPLC) and used \nfor NMR spectroscopic analysis. \nA combination of 1D and 2D NMR experiments was conducted to  elucidate the \nstructure. Based on interactions observed in 1H-1H TOCSY, 1H-1H NOESY, 1H-13C HSQC \nand 1H-13C HMBC spectra (Figures S13-S15), all proton and carbon signals from each of \nthe eight residues were assigned ( Table S1). Importantly, all side chain protons and \ncarbons of the Asp21-Asn22 sequence displayed chemical shift deviation compared to \nunmodified Asp/Asn residues. For Asp21, the β-carbon shifted downfield to 71.2 ppm and \nhas only one proton attached to it, affirming that PesHI catalyze d β-hydroxylation o f \nAsp21. Thus, the NMR data ruled out dehydration involving the hydroxyl group introduced \nby PesHI, and in turn implicated that the −19 Da change must occur entirely on the C -\nterminal Asn. T he α-proton of Asn22 shifted downfield to 5.13 ppm, indicating a more \nelectron-withdrawing environment. Together with the overall mass change, these findings \nare consistent with co nversion of the C -terminal carboxylate into  a nitrile group. \nConsistent with this hypothesis , the 13C NMR spectrum showed only 10 rather than the \nexpected 11 signals in the range of 150-220 ppm (10 amide/carboxylate carbonyl carbons \nand one guanidine carbon), and instead a characteristic nitrile carbon signal at 118.0 ppm \nwas observed. In the 1H-13C HMBC spectrum, this carbon signal showed correlations with \nthe α and β protons and the amide proton of former Asn22 (Figure 3c). Collectively, these \nobservations from 1D and 2D NMR experiments unambiguously show that in the PesA2-\nPesHIC peptide, the C-terminal carboxylate is converted to a nitrile group. \nWe also isolated and characterized the precursor peptide PesA2-Q14K after \nPesOHIC modification  and LysC digestion by NMR spectroscopy  (Figure S1 6-S18). \nCompared with the PesA2 -PesHIC peptide, the signals characteristic  of a  β-\nhydroxyaspartate as well as the C-terminal nitrile group were retained, and additional \nchemical shift deviations from unmodified residues were mainly observed for protons and \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n11 \n \ncarbons associated with the Pro residue (Table S2). In the 1H-1H COSY and 1H-13C HSQC \nspectra, the observed correlations clearly assigned the hydroxy group to the β position \n(C3), consistent with the conclusions from Marfey’s analysis . This modification resulted \nin the conversion of a CH2 group to a CH group, accompanied by a downfield shift of the \nβ-proton to 4.43 ppm and the corresponding β-carbon to 73.2 ppm. \n \nCharacterization of the Nitrile Synthetase PesC \nWith the C-terminal nitrilation activity of PesC established, we examined the reaction \nin vitro . For typical AS-like enzymes , the N -terminus is buried inside the enzyme .44 \nTherefore, we constructed a PesC-encoding plasmid with a C -terminal His 6-tag for \nexpression in E. coli. In the canonical AS catalytic cycle, L-aspartate is converted to L-\nasparagine via ATP-dependent adenylation of the side chain carboxylate , followed by \namidation using L-glutamine as the ammonia donor.44 Because PesC shows sequence \nhomology to asparagine synthetases, IMAC-purified PesC was incubated with Mg2+, ATP, \nL-glutamine and the PesA2-PesOHI peptide for 6 hours. LC-MS analysis of the GluC -\ndigested product clearly demonstrated the formation of the C-terminal nitrile (Figure 4a). \nWhen ATP was omitted from the reaction, nitrilation activity was abolished. When the \nassay was performed for 1 hour, a product with a mass shift of −1 Da was observed by \nhigh resolution electron spray ionization mass spectrometry (HR-ESI MS) (Figure 4a) \nconsistent with the formation of a n intermediate C-terminal amide. To confirm the origin \nof the nitrogen atom in both products, the assay was also performed with L-glutamine-\n(amide-15N). The resulting C-terminal amide and nitrile containing peptides were both \ndetected by LC-HRMS with a +1 Da mass shift (Figure 4b), indicating that in the PesC \nreaction, the inserted nitrogen is indeed derived from the side chain amide nitrogen of \nglutamine.  \nNext, we explore d the substrate requirements of PesC. The importance of prior \nPesHI/PesO modification for PesC activity was first examined. When unmodified \nprecursor PesA2 , PesA2-PesHI, and PesA2-PesOHI were incubated individually with \nPesC, the production of the corresponding nitrile products was only observed with PesHI- \nand PesOHI-modified PesA2 (Figure 4a, S19). For unmodified PesA2, only trace amounts \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n12 \n \nof the amide species were formed. Thus, the hydroxylation of the adjacent Asp residue \non the substrate is essential for efficient amidation and for dehydration of the intermediate \namide to occur, while hydroxylation by PesO is not required. This conclusion is consistent \nwith the gene composition of orthologous BGCs, because pesHI homologs always co -\noccur with pesC homologs while pesO is only partially conserved (e.g. Figure S4). \nThe leader peptide dependence of PesC was examined  using PesA1 as substrate, \nwhich provided more PesOHI -modified product from the co -expression experiment. We \npurified PesOHI-modified, endoproteinase LysC digested PesA1 -Q14K as well as \nPesOHI-modified, endoproteinase AspN digested PesA1. Reaction of the C-terminal 14-\nmer (AspN product) with PesC under the standard reaction conditions resulted in nitrile \nproduction (Figure S20). The C-terminal 8-mer (LysC product) was converted by PesC to \nthe corresponding amide with a small amount of nitrile formed (Figure S20). These results \nsuggest that PesC can still function in the absence of a leader peptide, but the attenuation \nin PesC efficiency suggests that the enzyme  likely acts prior to proteolysis in the \nbiosynthetic pathway. \n \nFigure 4. a) LC-MS analysis of the in vitro PesC reaction with PesA2 co-expressed with \nPesOHI after GluC digestion (assay conditions: 100 μM PesA2-PesOHI peptide, 100 μM \nPesC, 5 mM ATP, 10 mM L-glutamine, 20 mM  Mg2+); b) HR-ESI MS analysis of the \nresulting amide or nitrile species when L-glutamine or L-glutamine-(amide-15N) was used \nin the assay  (calculated mass for amide product [ M+2H]2+: 1111.5798, nitrile product \n[M+2H]2+: 1102.5745). \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n13 \n \nBased on these observations, a mechanism for the PesC catalyzed nitrilation can be \nproposed (Figure S21). The reaction is initiated by ATP-dependent adenylation of the C-\nterminal carboxylate, followed by the nucleophilic attack of glutamine-derived ammonia \nto form a C-terminal amide intermediate. The amide is then activated by a second \nequivalent of ATP to generate an AMP-imidate, from which AMP is eliminated to yield the \nC-terminal nitrile. This mechanism is similar to that previously proposed for AS-like nitrile \nsynthetases in other biosynthetic contexts such as the NRPS product auranthine and 7-\ncyano-7-deazaguanine (preQ0).13,14,47 \nWith the aim of  potentially uncovering more AS-like enzymes that are capable of  \ninstalling nitrile groups on ribosomal peptides, we performed a bioinformatic analysis on \nPesC. A BLASTp search with the PesC sequence as query only recovered high-identity \nhomologs of PesC located in orthologous clusters of the pes BGC. Other less highly \nconserved hits were not related to RiPP biosynthesis, e.g. enzymes linked to the \nbiosynthesis of N-acetylglutaminylglutamine amide.48 We then constructed a sequence \nsimilarity network (SSN) of the asparagine synthase protein family (PF00733) , and \nexamined their co -occurrence with RiPP biosynthesis elements ( Figure S22).49,50 This \nanalysis identified ~3,800 candidates, of which  ~3,500 were related to lasso peptide \nbiosynthesis. Among the remaining enzymes, TsrC51 and ScdTA52 are known to catalyze \namidation of the C-terminal carboxylate of their substrate peptides. The remaining ~300 \nuncharacterized candidates suggest a potential underexplored functional space of AS-\nlike enzymes in RiPP biosynthesis. These enzymes may catalyze amidation, nitrilation, \nlactam formation, or currently uncharacterized reactions. \n \nIn vitro Characterization of MNIO Proteins \nThe co-expression experiments were unable to assign a  role for the MNIO-like \nprotein PesX. We therefore conducted in vitro characterization. Considering the prevalent \npartner dependence of MNIO proteins in the literature,39 we constructed PesHI and PesXI \nexpression plasmid s with a  His6-tag on the N -terminus of either PesH or PesX. As \nanticipated, PesI was co-purified with both PesH and PesX (Figure 5b). Reaction of \nPesA2 with as -purified PesHI resulted in hydroxylation similar to the co -expression \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n14 \n \nstudies. When the assay was performed with PesH (no PesI) or with PesXI, no activity \nwas observed by ESI-MS (Figure 5a). These findings are consistent with the absence of \nthe iron-binding amino acids that are characteristic for MNIOs in the sequence of PesX  \n(Figure S3), but do not provide any insights regarding the potential function of the protein. \nThe ability of PesX to co-purify with partner protein PesI led us to speculate that PesX \nmight serve to regulate PesH activity, and thereby nitrile formation . Structural models of \nthe PesXI and PesHI heterodimer s using AlphaFold 3, and alignment of these two \nstructures revealed strong similarity (Figure 5c). Similar to reported structures of the TglHI \nheterodimer,53 only the helical N-terminus of PesI interacts with PesH or PesX . The \noverlapping binding interface suggests that PesX can compete with PesH in binding to \nthe partner protein PesI, which is critical for PesH activity as shown above. This proposal \nis supported by the observation that introduction of  PesX in the PesHI activity assay \nresulted in diminished  hydroxylation activit y (Figure 5a) . Moreover, addition of Strep-\ntagged PesX to a solution of  the His6-PesH/PesI heterodimer, and subsequent addition \nto Ni-NTA resin resulted in the complex of PesX and PesI eluting from the column (Figure \n5b). These observations indicate that PesX compete s with PesH in binding with partner \nprotein PesI, thus inhibiting its aspartate hydroxylation activity. Considering Asp2 1 \nhydroxylation is essential for PesC to perform nitrile formation, expression of PesX can \nslow down  the production of the nitrile product. If correct, this regulatory mechanism \nsuggests that the nitrile product may be toxic to t he producing organism and  that its \nproduction must be delicately regulated. \nWe performed bioinformatic analysis on PesX  to gauge its distribution . A BLASTp \nsearch with an E value cutoff of 0.05 retrieved only PesX homologs from orthologous pes-\nlike BGCs, suggesting that the sequence of PesX is highly specific to the nitrile formation \npathway. \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n15 \n \n \nFigure 5. PesX inhibits the activity of PesH. a) ESI-MS analysis of the in vitro PesH/X/I \nreaction with PesA2 peptide  (calculated mass for unmodified product [ M+7H]7+: \n1002.6476, hydroxylated product [ M+7H]7+: 110 4.9326); b) SDS-PAGE analysis of \npurified His 6-PesH/PesI, His 6-PesX/PesI, and Strep-PesX. Also shown are the \nflowthrough (FT) and elution (E) fractions of His6-PesH-PesI and Strep-PesX loaded on \nan IMAC column; c) Overlayed AlphaFold 3 models of the PesHI and PesXI complexes. \nNitrilobacillin Inhibits Cysteine Proteases \nWhile nitrilobacillin represent s the first peptide natural product with a C -terminal nitrile \ngroup, such scaffolds have been extensively explored in medicinal chemistry. As seen in \nthe therapeutic peptides vildagliptin and nirmatrelvir (Figure 6),6,7 nitrile groups are \ncommon warheads in peptide -derived compounds that function by covalently inhibiting \ncysteine/serine proteases.  To assess such activity of pes BGC products, peptide 2 \n(Figure 3d), was evaluated with a panel of proteases. Compound 2 exhibited little to no \ninhibitory activity against trypsin, chymotrypsin or angiotensin-converting enzyme (ACE) \nat 25 μM concentration. Instead, it inhibited the cysteine proteases cathepsin L, cathepsin \nB and papain with IC 50 values of 1.1 μM, 9.0 μM and 9.7 μM, respectively (Figure 7) . \nThese observed inhibitory activities align well with the reported activities of synthe tic \npeptidyl nitriles ,7 and highlight the poten tial of engineering nitrilobacillins as cysteine  \nprotease inhibitors. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n16 \n \n \nFigure 6.  Proposed biosynthetic pathway of nitrilobacillin . Representative examples of \ntherapeutic peptides with C-terminal nitrile groups are shown in the box. \n \nFigure 7.  Protease inhibition assay of compound 2 with human cathepsin L (top), human \ncathepsin B (middle) and papain (bottom). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n17 \n \nDiscussion \nIn genome mining campaigns for RiPP BGCs, a common strategy features a RiPP \nclass-defining enzyme family as query.23 This approach has proven an efficient means to \ndiscover new and divergent compounds in existing RiPP families. Several ubiquitous \nfamilies of RiPP post -translational modification enzymes  are not currently identified as \nclass-defining because they occur in many different RiPP classes. Two such enzyme \nfamilies are the MNIO and αKG-HExxH enzymes.28,39 These proteins catalyze a variety \nof different transformations that at present cannot be predicted. In the current study, we \ninvestigated a representative  BGC featuring members of both protein families that lead \nto the identification of the nitrilobacillins (Figure 6). Since the nitrile functional group is \nnew to RiPPs and most likely functions as the pharmacophore of the mature products, \nwe suggest the name nitrilotides for the wider group of products and the AS-like enzymes \nas the class-defining enzyme. \nIn medicinal chemistry, nitrile groups are common ly used as warheads including in \nin peptide-derived drug candidates. For instance, the nitrile s effect covalent inhibition of \ncysteine/serine proteases in vildagliptin and nirmatrelvir (Figure 6).6,7 Similarly, the C -\nterminal nitrile containing peptide product nitrilobacillin (compound 2) characterized in this \nwork exhibited micromolar inhibition activities against several cysteine proteases. These \nobservations highlight the convergence between synthetic , designed inhibitors  and \nbiosynthetic evolution. Currently, only moderate inhibitory activity was observed with \npeptide 2, likely because the investigated proteases are not its physiological targets. \nTypically, peptidic protease inhibitors have structures at the P1-P3 position that resemble \nthe native substrates. For instance, the pyrrolidone moiety at the P1 position of \nnirmatrelvir mimics the glutamine residue recognized by SARS -CoV-2 main protease. 16  \nThese observations suggest that the  currently unknown  physiological targets of \nnitrilobacillins are proteases that cleave at asparagine residues.  \nαKG-HExxH proteins were recently identified to be iron and α -ketoglutarate \ndependent dioxygenases. 28 Unlike canonical Fe /αKG dependent enzymes, proteins \nwithin this family utilize a conserved HExxH motif to bind iron , yielding a new active site \narchitecture.28,32 The reactivity of PesO described in the current work is consistent with \nthe reported β-hydroxylation activities of previous studies on members of the αKG-HExxH \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n18 \n \nenzyme family,28,32 while revealing proline as a new site of modification since previous \nmembers hydroxylated His, Asp, Phe and Gln residues. Hydroxylations of proline at the \nβ/γ carbons are common modifications  in the contest of single amino acid and protein \nside chain s, and are usually catalyzed by canonical Fe/ αKG enzymes. 54 In RiPP \nbiosynthesis, proline β,γ-dihydroxylation has been reported in the biosynthesis of \nmicrobisporicin and is catalyzed by the cytochrome P450 enzyme MibO.55  \nThe AS-like enzyme PesC catalyzes the nitrilation at the C-terminal carboxylate in \nthe PesA substrates expanding the previously reported small number of ATP-dependent \nnitrilation enzymes, ToyM/QueC47 and ArtA/NitB.13,14 These enzymes introduce nitriles \ninto nucleoside bases and the side chains of amino acids. Other pathways to nitrile \ncontaining natural products involve oxidation of amino groups by hemeproteins9,10 and \nFMOs,12 as well as  oxidative rearrangement processes of -amino acids by non -heme \niron dependent enzymes  (Figure S1).56 Given the structures of their  substrates, from a \nbiocatalyst development perspective, PesC may present the most promising starting point \nfor enzymatic installation of nitriles at the C-terminus of peptides. The reaction catalyzed \nby PesC has some differences compared to the two previous ATP-dependent systems. \nQueC/ToyM utilizes ammonia as the nitrogen source  instead of Gln used by P esC, \nwhereas ArtA/NitB catalyzes only the amide-to-nitrile dehydration step. Crystallographic \nstructures of QueC and ArtA have been solved,13,57 and alignment of the AlphaFold 3 \npredicted structures of PesC with the crystal structures of QueC or ArtA reveals that, even \nthough chemically the transformation catalyzed by PesC and QueC are more similar, the \noverall fold of PesC is better aligned with ArtA (Figure S23). More specifically, the ATP \nbinding domain of PesC exhibit s structural similarity with th at of ArtA. However, the \nsubstrate binding pocket  of PesC is different from the active site of ArtA, and the key \namide-interacting residues (Q146, S197, D199) identified through docking and mutational \nanalysis of ArtA13 are not conserved in PesC. Instead, the active site of PesC adopts a \nmuch more open conformation  in the model , likely to allow  accommodation of large \npeptide substrates (Figure S23). \nEnzymes from the AS protein family catalyze key transformations in the biosynthesis \nof bioactive natural products, such as β-lactamization in the biosynthesis of several β-\nlactam antibiotics.45,58,59 In RiPP biosynthesis, the lasso peptide cyclases belong to th e \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n19 \n \nAS enzyme family .46,60 In addition, several characterized asparagine synthetase -like \nenzymes such as TsrC51 and ScdTA52 amidate the C-termini of their substrate peptides. \nThe nitrilation reactivity of PesC identified herein expand s the diversity of chemical \noutcomes of asparagine synthetase-like enzymes  in RiPP biosynthesis . Our current \ngenome mining efforts identified hits that are likely involved in novel RiPP biosynthesis, \nbut no homologs were predicted with high confidence to catalyze nitrile formation, in part \nbecause it is currently not clear what controls enzymatic one-step amidation versus two-\nstep nitrilation.   \nMultinuclear non-heme iron dependent oxidative enzymes (MNIOs) are an emerging \nclass of metalloenzymes that contain two or three iron ions in their active site.39 Although \nthis protein family comprises more than 14,000 members, only a handful have been \nfunctionally characterized. MNIOs are strongly associated with the tailoring of RiPPs. \nThey catalyze a wide array of novel oxidative modifications to construct unusual scaffolds \nincluding oxazolones and thioamides,33 thiooxazoles,35,36,61 and ortho-tyrosines.30 PesHI \nis shown here to catalyze a comparatively  simpler transformation, the stereoselective -\nhydroxylation of aspartate  to 3 S-hydroxyAsp. This same reaction was also recently \nreported for an MNIO -nitroreductase fusion enzyme PflD where the MNIO domain \ncatalyzes the transformation.32 In another related system, ApyHI catalyzes the oxidation \nof a C-terminal Asp residue to the corresponding  alpha-keto acid with a β-hydroxylated \naspartate as a proposed intermediate.27 Sequence alignment of Pes H, ApyH and PflD \nrevealed less than 30 % identity for the MNIO domain . Notably, the conserved binding \nHxD motif that normally binds the third iron (Fe3) in MNIOs is missing for both PesH and \nPflD (Figure S3). Previous mutational studies on the MNIOs MbnB62 and TglH53 showed \nthat mutating some Fe3 binding residues lead to decreased but not abolished enzymatic \nactivity. Thus, the role of Fe3 in characterized MNIO enzymes remains elusive. \nThe function of the MNIO-like protein PesX is intriguing. No activity was detected in \nvitro or in E. coli, consistent with the absence of the metal binding residues in its sequence \nthat are conserved in all other MNIOs.  Heterodimeric proteins composed of  an inactive \nhomolog with an active enzyme is not uncommon and is observed for instance with  the \nFe/αKG enzyme system CorB-CorD,31 and with several examples of metalloproteases.63-\n66 However, PesX appears to function in a different manner in that it competes with PesH \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n20 \n \nfor PesI, which is required for PesH activity . The identification of separate operon s \ngoverning pesH and pesXI expression, whereas in other MNIO systems the HI proteins \nare often under control of a single operon (Figure S24), suggests that PesH activity is \nregulated in a more complicated manner. PesC catalyzed nitrilation only after \nhydroxylation o f Asp21 by PesHI, suggesting that this hydroxylation might serve as a \ngatekeeping event. The observed competition of PesX fo r binding PesI could therefore \nbe relevant to tightly control nitrile formation within the native host, possibly to avoid \ntoxicity. In this model, PesX and PesI would be expressed resulting in a non -functional \nheterodimer, which then would be converted to a fraction of ac tive PesHI heterodimer \nupon initiation of PesH expression. Experimental test of this suppression strategy in \nspecies harboring pes-like BGCs requires investigations with native producing strains.  \n \nConclusion \nIn this work, we identified and characterized a novel RiPP BGC from Peribacillus simplex. \nThe activity of modifying enzymes w as assigned through a combination of HR -ESI-\nMS/MS, NMR and Marfey’s analysis . MNIO and αKG-HExxH proteins were shown  to \nperform stereoselective β-hydroxylation on aspartate and proline residues, respectively , \nand the AS-like enzyme PesC was demonstrated to catalyze two-step nitrile formation at \nthe C-terminus of ribosomally produced peptides. The resulting peptide exhibits inhibitory \nactivity against several cysteine proteases.  This study therefore  expands the list of \nribosomal peptide PTMs  with nitrilations , and unveils a convergence between rational \nmolecule design and natural evolution. \n \nSupporting Information \nExperimental procedures, Figures S 1-S24 showing spectroscopic data , AlphaFold  3 \nmodeling, and Tables S 1-S3 listing NMR annotations  and protease inhibition assay \nconditions.  \n \nFunding \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n21 \n \n This manuscript is the result of funding in whole or in part by the National Institutes \nof Health (NIH ; grant R37 GM058822 ). It is subject to the NIH Public Access Policy. \nThrough acceptance of this federal funding, NIH has been given a right to make this \nmanuscript publicly available in PubMed Central upon the Official Date of Publication, as \ndefined by NIH . A Bruker UltrafleXtreme mass spectrometer used was purchased  with \nsupport from the Roy J. Carver Charitable Trust (Grant No. 22 -5622). W.A.v.d.D. is an \nInvestigator of the Howard Hughes Medical Institute. \nNotes \nThe authors declare no competing financial interest. \nAcknowledgements \n The authors thank Dr. Dinh T. Nguyen for helpful discussions, and Prof. Angad P . \nMehta for access to an Agilent Synergy H1 plate reader. \n \n \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n22 \n \nReferences \n(1)  Pomeranz, C. Ueber allylcyanid und allylsenföl. Liebigs Ann. Chem. 1907, 351, 354-\n62. \n(2)  Scotti, C.; Barlow, J. W. Natural products containing the nitrile functional group and \ntheir biological activities. Nat. Prod. Commun. 2022, 17, 1934578X221099973. \n(3)  Gleadow, R. M.; Møller, B. L. Cyanogenic glycosides: synthesis, physiology, and \nphenotypic plasticity. Annu. Rev. Plant Biol. 2014, 65, 155-85. \n(4)  Yulvianti, M.; Zidorn, C. Chemical diversity of plant cyanogenic glycosides: an \noverview of reported natural products. Molecules 2021, 26, 719. \n(5)  Liu, M.; Li, S. Nitrile biosynthesis in nature: how and why? Nat. Prod. Rep. 2024, 41, \n649-71. \n(6)  Wang, X.; Wang, Y.; Li, X.; Yu, Z.; Song, C.; Du, Y. Nitrile-containing \npharmaceuticals: target, mechanism of action, and their SAR studies. RSC Med. Chem. \n2021, 12, 1650-71. \n(7)  Bonatto, V.; Lameiro, R. F.; Rocho, F. R.; Lameira, J.; Leitão, A.; Montanari, C. A. \nNitriles: an attractive approach to the development of covalent inhibitors. RSC Med. \nChem. 2023, 14, 201-17. \n(8)  Rakshit, A.; Dhara, H. N.; Sahoo, A. K.; Patel, B. K. The renaissance of organo \nnitriles in organic synthesis. Chem.: Asian J. 2022, 17, e202200792. \n(9)  Kahn, R. A.; Bak, S.; Svendsen, I.; Halkier, B. A.; Moller, B. L. Isolation and \nreconstitution of cytochrome P450ox and in vitro reconstitution of the entire biosynthetic \npathway of the cyanogenic glucoside dhurrin from sorghum. Plant Physiol. 1997, 115, \n1661-70. \n(10)  Oinuma, K.; Hashimoto, Y.; Konishi, K.; Goda, M.; Noguchi, T.; Higashibata, H.; \nKobayashi, M. Novel aldoxime dehydratase involved in carbon-nitrogen triple bond \nsynthesis of Pseudomonas chlororaphis B23. Sequencing, gene expression, \npurification, and characterization. J. Biol. Chem. 2003, 278, 29600-8. \n(11)  Yamaguchi, T.; Asano, Y. Nitrile-synthesizing enzymes and biocatalytic synthesis \nof volatile nitrile compounds: a review. J. Biotechnol. 2024, 384, 20-8. \n(12)  Peng, M.; Wu, Q.; Ma, L.; Teng, Z.-J.; Hou, X.; Zhu, H.; Ju, J. Bacterial \nbiosynthesis of nitrile-containing natural products: basis for recognition of diversified \nsubstrates. ACS Catal. 2024, 14, 17780-93. \n(13)  Zeng, Y.; Zhang, K.; Lu, T.; Yin, X.; Wang, Q.; Xiong, L.; Guo, H.; Li, J.; Lu, X.; Liu, \nL.; Ma, H.; Gao, Z. Structural and mechanistic basis for nitrile synthetase by an \nargininosuccinate synthetase-like enzyme. ACS Catal. 2025, 15, 16254-67. \n(14)  Kishimoto, S.; Tamura, T.; Okamoto, T.; Watanabe, K. Enantioselective \nbiosynthesis of (+)- and (−)-auranthines. J. Am. Chem. Soc. 2025, 147, 10612-7. \n(15)  Cha, L.; Tang, Y.; Guo, Y.; Chang, W.-C. In Comprehensive Natural Products III; \nLiu, H.-W., Begley, T. P., Eds.; Elsevier: Oxford, 2020, p 191-214. \n(16)  Owen, D. R.; Allerton, C. M. N.; Anderson, A. S.; Aschenbrenner, L.; Avery, M.; \nBerritt, S.; Boras, B.; Cardin, R. D.; Carlo, A.; Coffman, K. J.; Dantonio, A.; Di, L.; Eng, \nH.; Ferre, R.; Gajiwala, K. S.; Gibson, S. A.; Greasley, S. E.; Hurst, B. L.; Kadar, E. P.; \nKalgutkar, A. S.; Lee, J. C.; Lee, J.; Liu, W.; Mason, S. W.; Noell, S.; Novak, J. J.; \nObach, R. S.; Ogilvie, K.; Patel, N. C.; Pettersson, M.; Rai, D. K.; Reese, M. R.; \nSammons, M. F.; Sathish, J. G.; Singh, R. S. P.; Steppan, C. M.; Stewart, A. E.; Tuttle, \nJ. B.; Updyke, L.; Verhoest, P. R.; Wei, L.; Yang, Q.; Zhu, Y. An oral SARS-CoV-2 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n23 \n \nM(pro) inhibitor clinical candidate for the treatment of COVID-19. Science 2021, 374, \n1586-93. \n(17)  Gavriilidou, A.; Kautsar, S. A.; Zaburannyi, N.; Krug, D.; Müller, R.; Medema, M. \nH.; Ziemert, N. Compendium of specialized metabolite biosynthetic diversity encoded in \nbacterial genomes. Nat. Microbiol. 2022, 7, 726-35. \n(18)  Dang, T.; Süssmuth, R. D. Bioactive peptide natural products as lead structures for \nmedicinal use. Acc. Chem. Res. 2017, 50, 1566-76. \n(19)  Zimmer, L.; Crüsemann, M. Amino acid tailoring strategies in peptide natural \nproduct biosynthesis. Trends Chem. 2025, 7, 705-18. \n(20)  Padhi, C.; Field, C. M.; Forneris, C. C.; Olszewski, D.; Fraley, A. E.; Sandu, I.; \nScott, T. A.; Farnung, J.; Ruscheweyh, H.-J.; Narayan Panda, A.; Oxenius, A.; Greber, \nU. F.; Bode, J. W.; Sunagawa, S.; Raina, V.; Suar, M.; Piel, J. Metagenomic study of \nlake microbial mats reveals protease-inhibiting antiviral peptides from a core \nmicrobiome member. Proc. Natl. Acad. Sci. USA 2024, 121, e2409026121. \n(21)  Zhang, J.; Zhang, D.; Xu, Y.; Zhang, J.; Liu, R.; Gao, Y.; Shi, Y.; Cai, P.; Zhong, \nZ.; He, B.; Li, X.; Zhou, H.; Chen, M.; Li, Y.-X. Large-scale biosynthetic analysis of \nhuman microbiomes reveals diverse protective ribosomal peptides. Nature Commun. \n2025, 16, 3054. \n(22)  Montalbán-López, M.; Scott, T. A.; Ramesh, S.; Rahman, I. R.; van Heel, A. J.; \nViel, J. H.; Bandarian, V.; Dittmann, E.; Genilloud, O.; Goto, Y.; Grande Burgos, M. J.; \nHill, C.; Kim, S.; Koehnke, J.; Latham, J. A.; Link, A. J.; Martínez, B.; Nair, S. K.; Nicolet, \nY.; Rebuffat, S.; Sahl, H.-G.; Sareen, D.; Schmidt, E. W.; Schmitt, L.; Severinov, K.; \nSüssmuth, R. D.; Truman, A. W.; Wang, H.; Weng, J.-K.; van Wezel, G. P.; Zhang, Q.; \nZhong, J.; Piel, J.; Mitchell, D. A.; Kuipers, O. P.; van der Donk, W. A. New \ndevelopments in RiPP discovery, enzymology and engineering. Nat. Prod. Rep. 2021, \n38, 130-239. \n(23)  Nguyen, D. T.; Mitchell, D. A.; van der Donk, W. A. Genome mining for new \nenzyme chemistry. ACS Catal. 2024, 14, 4536-53. \n(24)  Richter, D.; Piel, J. Novel types of RiPP-modifying enzymes. Curr. Opin. Chem. \nBiol. 2024, 80, 102463. \n(25)  Ting, C. P.; Funk, M. A.; Halaby, S. L.; Zhang, Z.; Gonen, T.; van der Donk, W. A. \nUse of a scaffold peptide in the biosynthesis of amino acid-derived natural products. \nScience 2019, 365, 280-4. \n(26)  Ayikpoe, R. S.; Zhu, L.; Chen, J. Y.; Ting, C. P.; van der Donk, W. A. \nMacrocyclization and backbone rearrangement during RiPP biosynthesis by a SAM-\ndependent domain-of-unknown-function 692. ACS Cent. Sci. 2023, 9, 1008-18. \n(27)  Nguyen, D. T.; Zhu, L.; Gray, D. L.; Woods, T. J.; Padhi, C.; Flatt, K. M.; Mitchell, \nD. A.; van der Donk, W. A. Biosynthesis of macrocyclic peptides with C-terminal β-\namino-α-keto acid groups by three different metalloenzymes. ACS Cent. Sci. 2024, 10, \n1022-32. \n(28)  Morishita, Y.; Ma, S.; De La Mora, E.; Li, H.; Chen, H.; Ji, X.; Usclat, A.; Amara, P.; \nSugiyama, R.; Tooh, Y. W.; Gunawan, G.; Pérard, J.; Nicolet, Y.; Zhang, Q.; Morinaka, \nB. I. Fused radical SAM and αKG-HExxH domain proteins contain a distinct structural \nfold and catalyse cyclophane formation and β-hydroxylation. Nat. Chem. 2024, 16, \n1882-93. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n24 \n \n(29)  Kim, H. W.; Kang, S.; Kim, S.; Lee, H.; Hur, Y.; Song, W. J.; Oh, D. C.; Kim, S. \nDiscovery of a two-step enzyme cascade converting aspartate to aminomalonate in \npeptide natural product biosynthesis. J. Am. Chem. Soc. 2025, 147, 20909-18. \n(30)  Padhi, C.; Zhu, L.; Chen, J. Y.; Huang, C.; Moreira, R.; Challis, G. L.; Cryle, M. J.; \nvan der Donk, W. A. Biosynthesis of biphenomycin-like macrocyclic peptides by \nformation and cross-linking of ortho-tyrosines. J. Am. Chem. Soc. 2025, 147, 23781-96. \n(31)  Xia, Y.; Shi, Y.; Xiao, Z.; Xie, X.; Wang, B.; Yao, H.; Wang, H. α-Ketoglutarate-\ndependent dioxygenase-catalyzed macrocyclization in RiPPs. J. Am. Chem. Soc. 2025, \n147, 26693-703. \n(32)  Ouyang, Y.; Yu, Y.; Zhu, L.; Nguyen, D. T.; van der Donk, W. A. Oxidative peptide \nbackbone cleavage by a HEXXH enzyme during RiPP biosynthesis. J. Am. Chem. Soc. \n2026, 148, 3551-61. \n(33)  Kenney, G. E.; Dassama, L. M. K.; Pandelia, M. E.; Gizzi, A. S.; Martinie, R. J.; \nGao, P.; DeHart, C. J.; Schachner, L. F.; Skinner, O. S.; Ro, S. Y.; Zhu, X.; Sadek, M.; \nThomas, P. M.; Almo, S. C.; Bollinger, J. M., Jr.; Krebs, C.; Kelleher, N. L.; Rosenzweig, \nA. C. The biosynthesis of methanobactin. Science 2018, 359, 1411-6. \n(34)  Yu, Y.; van der Donk, W. A. Biosynthesis of 3-thia-α-amino acids on a carrier \npeptide. Proc. Natl. Acad. Sci. U. S. A. 2022, 119, e2205285119. \n(35)  Leprevost, L.; Jünger, S.; Lippens, G.; Guillaume, C.; Sicoli, G.; Oliveira, L.; \nFalcone, E.; de Santis, E.; Rivera-Millot, A.; Billon, G.; Stellato, F.; Henry, C.; Antoine, \nR.; Zirah, S.; Dubiley, S.; Li, Y.; Jacob-Dubuisson, F. A widespread family of ribosomal \npeptide metallophores involved in bacterial adaptation to metal stress. Proc. Natl. Acad. \nSci. U. S. A. 2024, 121, e2408304121. \n(36)  Tang, Y.; Zhong, W.; Fu, L.; Asante, E.; Kostenko, A.; Vidya, F. N. U.; Mandelare-\nRuiz, P.; Adeogun, T. T.; Anderson, G. P.; Edmonds, B. E.; Fang, O.; Han, M.; \nHollingsworth, A. S.; Ingham, A. R.; Kirby, C. R.; Landrum, A.; Mack, C. R.; Nobari, N. \nS.; Oswald, E. J.; Polevoy, C. L.; Sharifian, Y.; So, T. J.; Stokes, J. R.; Thompson, R. \nS.; Vuthamaraju, R.; Wang, E. C.; Yang, W. H.; Onstine, A. E.; Paul, V. J.; Wu, R.; \nAron, A. T.; Agarwal, V. Phylogenomic identification of a highly conserved copper-\nbinding RiPP biosynthetic gene cluster in marine Microbulbifer bacteria. ACS Chem. \nBiol. 2025, 20, 2462-74. \n(37)  Abramson, J.; Adler, J.; Dunger, J.; Evans, R.; Green, T.; Pritzel, A.; Ronneberger, \nO.; Willmore, L.; Ballard, A. J.; Bambrick, J.; Bodenstein, S. W.; Evans, D. A.; Hung, C.-\nC.; O’Neill, M.; Reiman, D.; Tunyasuvunakool, K.; Wu, Z.; Žemgulytė, A.; Arvaniti, E.; \nBeattie, C.; Bertolli, O.; Bridgland, A.; Cherepanov, A.; Congreve, M.; Cowen-Rivers, A. \nI.; Cowie, A.; Figurnov, M.; Fuchs, F. B.; Gladman, H.; Jain, R.; Khan, Y. A.; Low, C. M. \nR.; Perlin, K.; Potapenko, A.; Savy, P.; Singh, S.; Stecula, A.; Thillaisundaram, A.; Tong, \nC.; Yakneen, S.; Zhong, E. D.; Zielinski, M.; Žídek, A.; Bapst, V.; Kohli, P.; Jaderberg, \nM.; Hassabis, D.; Jumper, J. M. Accurate structure prediction of biomolecular \ninteractions with AlphaFold 3. Nature 2024, 630, 493-500. \n(38)  van Kempen, M.; Kim, S. S.; Tumescheit, C.; Mirdita, M.; Lee, J.; Gilchrist, C. L. \nM.; Söding, J.; Steinegger, M. Fast and accurate protein structure search with Foldseek. \nNat. Biotechnol. 2024, 42, 243-6. \n(39)  Chen, J. Y.; van der Donk, W. A. Multinuclear non-heme iron dependent oxidative \nenzymes (MNIOs) involved in unusual peptide modifications. Curr. Opin. Chem. Biol. \n2024, 80, 102467. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n25 \n \n(40)  Tietz, J. I.; Schwalen, C. J.; Patel, P. S.; Maxson, T.; Blair, P. M.; Tai, H. C.; Zakai, \nU. I.; Mitchell, D. A. A new genome-mining tool redefines the lasso peptide biosynthetic \nlandscape. Nat. Chem. Biol. 2017, 13, 470-8. \n(41)  Marfey, P. Determination of D-amino acids. II. Use of a bifunctional reagent, 1,5-\ndifluoro-2,4-dinitrobenzene. Carlsberg Res. Commun. 1984, 49, 591-6. \n(42)  Ayon, N. J.; Sharma, A. D.; Gutheil, W. G. LC-MS/MS-based separation and \nquantification of Marfey's reagent derivatized proteinogenic amino acid DL-\nstereoisomers. J. Am. Soc. Mass Spectrom. 2019, 30, 448-58. \n(43)  Brademan, D. R.; Riley, N. M.; Kwiecien, N. W.; Coon, J. J. Interactive peptide \nspectral annotator: A versatile web-based tool for proteomic applications. Mol. Cell. \nProteom. 2019, 18, S193-S201. \n(44)  Lomelino, C. L.; Andring, J. T.; McKenna, R.; Kilberg, M. S. Asparagine \nsynthetase: function, structure, and role in disease. J. Biol. Chem. 2017, 292, 19952-8. \n(45)  Bachmann, B. O.; Li, R.; Townsend, C. A. β-Lactam synthetase: A new \nbiosynthetic enzyme. Proc. Natl. Acad. Sci. U. S. A. 1998, 95, 9082-6. \n(46)  Maksimov, M. O.; Pan, S. J.; Link, A. J. Lasso Peptides : Structure, Function, \nBiosynthesis, and Engineering. Nat. Prod. Rep. 2012, 29, 996-1006. \n(47)  Nelp, M. T.; Bandarian, V. A single enzyme transforms a carboxylic acid into a \nnitrile through an amide intermediate. Angew. Chem. Int. Ed. 2015, 54, 10627-9. \n(48)  Sagot, B.; Gaysinski, M.; Mehiri, M.; Guigonis, J.-M.; Le Rudulier, D.; Alloing, G. \nOsmotically induced synthesis of the dipeptide N-acetylglutaminylglutamine amide is \nmediated by a new pathway conserved among bacteria. Proc. Natl. Acad. Sci. U. S. A. \n2010, 107, 12652-7. \n(49)  Oberg, N.; Zallot, R.; Gerlt, J. A. EFI-EST, EFI-GNT, and EFI-CGFP: Enzyme \nFunction Initiative (EFI) web resource for genomic enzymology tools. J. Mol. Biol. 2023, \n435, 168018. \n(50)  Zallot, R.; Oberg, N.; Gerlt, J. A. Discovery of new enzymatic functions and \nmetabolic pathways using genomic enzymology web tools. Curr. Opin. Biotechnol. \n2021, 69, 77-90. \n(51)  Liao, R.; Liu, W. Thiostrepton maturation involving a deesterification-amidation \nway to process the C-terminally methylated peptide backbone. J. Am. Chem. Soc. 2011, \n133, 2852-5. \n(52)  Crnkovic, C. M.; Braesel, J.; Krunic, A.; Eustáquio, A. S.; Orjala, J. Scytodecamide \nfrom the cultured Scytonema sp. UIC 10036 expands the chemical and genetic diversity \nof cyanobactins. Chem. Bio. Chem. 2020, 21, 845-52. \n(53)  Zheng, Y.; Xu, X.; Fu, X.; Zhou, X.; Dou, C.; Yu, Y.; Yan, W.; Yang, J.; Xiao, M.; \nvan der Donk, W. A.; Zhu, X.; Cheng, W. Structures of the holoenzyme TglHI required \nfor 3-thiaglutamate biosynthesis. Structure 2023, 31, 1220-32. \n(54)  Hausinger, R. P. In 2-Oxoglutarate-dependent oxygenases; Schofield, C., \nHausinger, R., Eds.; The Royal Society of Chemistry: 2015, p 1-58. \n(55)  Foulston, L. C.; Bibb, M. J. Microbisporicin gene cluster reveals unusual features \nof lantibiotic biosynthesis in actinomycetes. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, \n13461-6. \n(56)  Adak, S.; Ye, N.; Calderone, L. A.; Duan, M.; Lubeck, W.; Schäfer, R. J. B.; \nLukowski, A. L.; Houk, K. N.; Pandelia, M.-E.; Drennan, C. L.; Moore, B. S. A single \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint \n\n26 \n \ndiiron enzyme catalyses the oxidative rearrangement of tryptophan to indole nitrile. Nat. \nChem. 2024, 16, 1989-98. \n(57)  Cicmil, N.; Huang, R. H. Crystal structure of QueC from Bacillus subtilis: an \nenzyme involved in preQ1 biosynthesis. Proteins 2008, 72, 1084-8. \n(58)  Miller, M. T.; Bachmann, B. O.; Townsend, C. A.; Rosenzweig, A. C. Structure of \nβ-lactam synthetase reveals how to synthesize antibiotics instead of asparagine. Nat. \nStruct. Mol. Biol. 2001, 8, 684-9. \n(59)  Miller, M. T.; Bachmann, B. O.; Townsend, C. A.; Rosenzweig, A. C. The catalytic \ncycle of β-lactam synthetase observed by x-ray crystallographic snapshots. Proc. Natl. \nAcad. Sci. U. S. A. 2002, 99, 14752-7. \n(60)  Barrett, S. E.; Mitchell, D. A. Advances in lasso peptide discovery, biosynthesis, \nand function. Trends Genet. 2024, 40, 950-68. \n(61)  Gadgil, M. G.; Dommaraju, S. R.; Liu, X.; Battiste, A. J.; Bregman, M. H.; Mitchell, \nD. A. Biosynthesis of peptidic thiooxazole metallophores installed by multinuclear \nnonheme iron enzymes. bioRxiv 2025, 10.1101/2025.11.18.689057. \n(62)  Park, Y. J.; Jodts, R. J.; Slater, J. W.; Reyes, R. M.; Winton, V. J.; Montaser, R. A.; \nThomas, P. M.; Dowdle, W. B.; Ruiz, A.; Kelleher, N. L.; Bollinger, J. M., Jr.; Krebs, C.; \nHoffman, B. M.; Rosenzweig, A. C. A mixed-valent Fe(II)Fe(III) species converts \ncysteine to an oxazolone/thioamide pair in methanobactin biosynthesis. Proc. Natl. \nAcad. Sci. U. S. A. 2022, 119, e2123566119. \n(63)  Allali, N.; Afif, H.; Couturier, M.; Van Melderen, L. The highly conserved TldD and \nTldE proteins of Escherichia coli are involved in microcin B17 processing and in CcdA \ndegradation. J. Bacteriol. 2002, 184, 3224-31. \n(64)  Tsibulskaya, D.; Mokina, O.; Kulikovsky, A.; Piskunova, J.; Severinov, K.; \nSerebryakova, M.; Dubiley, S. The product of Yersinia pseudotuberculosis mcc operon \nis a peptide-cytidine antibiotic activated inside producing cells by the TldD/E protease. \nJ. Am. Chem. Soc. 2017, 139, 16178-87. \n(65)  Xue, D.; Older, E. A.; Zhong, Z.; Shang, Z.; Chen, N.; Dittenhauser, N.; Hou, L.; \nCai, P.; Walla, M. D.; Dong, S. H.; Tang, X.; Chen, H.; Nagarkatti, P.; Nagarkatti, M.; Li, \nY. X.; Li, J. Correlational networking guides the discovery of unclustered lanthipeptide \nprotease-encoding genes. Nat. Commun. 2022, 13, 1647. \n(66)  Ramos-Figueroa, J. S.; Zhu, L.; Halliman, M.; van der Donk, W. A. The \nbiosynthetic pathway to the pyrroloiminoquinone marine natural product ammosamide \nC. J. Am. Chem. Soc. 2025, 147, 46315-23. \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.11.711119doi: bioRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}