Symbiosis driven variation and supply of natural products

Drug discovery and development from marine invertebrates has been fraught with two key problems, namely, the variability of occurrence and limited supply. Bacteria in symbiosis with marine invertebrates have been shown to produce most bioactive natural products isolated from these organisms, and thus are central to addressing questions of occurrence and issues of supply. Specifically, the factors that influence symbiosis influence the distribution and supply of natural products. This dissertation sought to address these two problems through studies in symbiosis and supply of symbiotic natural products. First, the global patterns of chemical symbiosis in marine ascidians, a group of highly prolific producers of natural products, were examined. Symbiosis in ascidians is shown to be host-specific (meaning that similar species of invertebrates contain similar bacterial symbionts); further, microbiomes are shown to be equally diverse regardless of location. Secondary metabolism was also found to be host-specific, but is more sensitive to biogeographical factors as evidenced by the increase in the potency of the secondary metabolites in tropical regions. To address the supply of rare natural products, heterologous expression was used to produce useful quantities of a group of symbiotic natural products, cyanobactins. Using metabolic engineering, a platform was developed to supply cyanobactins in high-titer, and its usefulness showcased in the discovery of ! novel activities of these natural products. Another facet of the supply problem is the substantial difficulty involved in synthesizing derivatives of natural products, which generally requires total chemical synthesis. On this aspect of the supply problem, the capacity of the cyanobactin pathway to generate unprecendented structural diversity by the incorporation of non-proteinogenic amino acids into this multistep, substrate-tolerant biosynthetic pathway was demonstrated.

SYMBIOSIS DRIVEN VARIATION AND SUPPLY OF NATURAL PRODUCTS by Ma. Diarey B. Tianero A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Medicinal Chemistry The University of Utah December 2015 ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! Copyright © Ma. Diarey B. Tianero 2015 All Rights Reserved The University of Utah Graduate School STATEMENT OF DISSERTATION APPROVAL The dissertation of Ma. Diarey B. Tianero has been approved by the following supervisory committee members: Eric Schmidt , Chair 06/18/15 Chris Ireland , Member 06/15/15 Baldomero Olivera , Member 06/15/15 Darrell Davis , Member 06/15/15 Grzegorz Bulaj , Member 06/15/15 and by the Department of Darrell Davis Medicinal Chemistry and by David B. Kieda, Dean of The Graduate School. Date Approved Date Approved Date Approved Date Approved Date Approved , Chair of ABSTRACT Drug discovery and development from marine invertebrates has been fraught with two key problems, namely, the variability of occurrence and limited supply. Bacteria in symbiosis with marine invertebrates have been shown to produce most bioactive natural products isolated from these organisms, and thus are central to addressing questions of occurrence and issues of supply. Specifically, the factors that influence symbiosis influence the distribution and supply of natural products. This dissertation sought to address these two problems through studies in symbiosis and supply of symbiotic natural products. First, the global patterns of chemical symbiosis in marine ascidians, a group of highly prolific producers of natural products, were examined. Symbiosis in ascidians is shown to be host-specific (meaning that similar species of invertebrates contain similar bacterial symbionts); further, microbiomes are shown to be equally diverse regardless of location. Secondary metabolism was also found to be host-specific, but is more sensitive to biogeographical factors as evidenced by the increase in the potency of the secondary metabolites in tropical regions. To address the supply of rare natural products, heterologous expression was used to produce useful quantities of a group of symbiotic natural products, cyanobactins. Using metabolic engineering, a platform was developed to supply cyanobactins in high-titer, and its usefulness showcased in the discovery of ! novel activities of these natural products. Another facet of the supply problem is the substantial difficulty involved in synthesizing derivatives of natural products, which generally requires total chemical synthesis. On this aspect of the supply problem, the capacity of the cyanobactin pathway to generate unprecendented structural diversity by the incorporation of non-proteinogenic amino acids into this multistep, substrate-tolerant biosynthetic pathway was demonstrated. iv ! TABLE OF CONTENTS ABSTRACT .......................................................................................................................iii LIST OF TABLES ............................................................................................................ vii ACKNOWLEDGEMENTS .............................................................................................viii Chapter 1. INTRODUCTION ........................................................................................................... 1 Natural Products variability and supplies ................................................................ 2 References ............................................................................................................. 15 2. SPECIES SPECIFICITY OF SYMBIOSIS AND SECONDARY METABOLISM IN ASCIDIANS............................................................................................................. 22 Abstract.................................................................................................................. 23 Introduction ........................................................................................................... 23 Materials and methods ........................................................................................... 24 Results ................................................................................................................... 26 Discussion.............................................................................................................. 31 References ............................................................................................................. 34 Supporting information ......................................................................................... 37 3. METABOLIC ENGINEERING TO OPTIMIZE HETEROLOGOUS RIPP PRODUCTION IN E. COLI ......................................................................................... 68 Abstract.................................................................................................................. 69 Introduction ........................................................................................................... 70 Results ................................................................................................................... 72 Discussion.............................................................................................................. 87 Materials and methods ........................................................................................... 91 References ............................................................................................................. 97 4. RIBOSOMAL ROUTE TO SMALL MOLECULE DIVERSITY ............................. 101 Abstract................................................................................................................ 102 Introduction ......................................................................................................... 102 Results ................................................................................................................. 103 Discussion............................................................................................................ 107 References ........................................................................................................... 108 Supporting information ....................................................................................... 110 5. CONCLUSIONS ......................................................................................................... 149 Conclusions ......................................................................................................... 150 References ........................................................................................................... 153 vi LIST OF TABLES 2.1 Sample collection details and 6 U51$ BLAST hits…………………………… .40 2.2 Microbiome statistics………………………………………………………………. .41 2.3 List of primers……………………………………………………………………… .41 2.4 6HTXHQFH FRXQWV DQG Dlpha GLYHUVLW\ YDOXHs RI DVFLGLDQ PLFURELRPHV………………42 2.5 +ost species PDSV for networN analysLs RI DVFLGLDQ PLFURELRPHV…………………..43 2.6 Table of shared microbes in locations and KRVW VSHFLHs………………………… .44 2.7 %/$67 KLWV RI 6 U51$ VHTXHQFHV RI eukaryotes associated with ascidians……..46 2.8 Summary of secondary metabolites and associated microbes LQ DVFLGLDQV………… 47 2.9 6WDWLVWLFDO FRPSDULVRQV RI ascidian PLFURELRPHV DFURVs VSHFLHV……………………48 3.1 Additives used in the initial optimization of cyanobactin production in E. coli.........73 3.2 Summary of calcium imaging experiments with patellin 2………………………….8 4.1 Design strategy for mutants………………………………………………………...104 4.2 Expression yields in E. coli…………………………………………………………105 4.3 &DOFXODWHG DQG REVHUYHG PDVVHV IRU FRPSRXQGV V\QWKHVL]HG LQ ( FROL………….. .119 ACKNOWLEDGEMENTS This dissertation would not be possible without the help and support of the following people and funding institutions. First, I would like to thank my advisor, Eric Schmidt, for his guidance, training, and support throughout my graduate research career. I would also like to thank my committee members, Chris Ireland, Baldomero Olivera, Darrell Davis, and Grzegorz Bulaj, for their guidance and interest in my work. I would like to acknowledge NIHGMS for funding all the work in this thesis. I would also like to thank all my collaborators and coauthors. In the ascidian microbiomes and metabolomes project, I would like to thank Tim Bugni and Tom Wyche for performing the LC/MS and PCA analyses of the ascidian extracts, Angela Presson for the assistance in statistical analyses, Mike Koch for performing the bioactivity assays, and Jason Kwan for help in microbiome and bioinformatics analyses. In the heterologous expression of cyanobactins, I would like to thank Elizabeth Pierce for performing the prenyltransferase knock out experiments and assembly of the pat pathway, Duane Ruffner of Symbion Discovery for the tru pathway vector that I used in all optimization and heterologous production experiments, John McIntosh for the initial experiments on the combination of tru and mev pathways, and Shrinivasan Raghuraman for performing all the calcium imaging experiments. In the engineering of cyanobactin pathway, I would like to thank Mohamed Donia for conceiving the project and cloning the derivatives of cyanobactins, as well as Peter Schultz and Travis Young for providing the vectors for unnatural amino acid incorporation. Lastly, I would like to thank my family and friends, especially my husband, John McIntosh, for all the support, patience, and confidence in the making and finishing of this dissertation. ! ix CHAPTER 1 INTRODUCTION ! "! Natural products variability and supply Natural products (NPs) are an excellent source of bioactive lead compounds for drug discovery.1 Among NP producers, marine invertebrates are known for harboring diverse, biologically potent, and structurally complex metabolites that possess drug-like qualities.2 For example, of 36 marine natural products that were in the preclinical/clinical pipeline from 2004-2013, 30 were originally sourced from marine invertebrates.3 However, drug discovery from marine invertebrates has been fraught with two problems, namely, the variability in isolation and limited supply. The isolation of a natural product from an organism, especially from a sessile invertebrate such as a sponge or an ascidian, does not guarantee that one will find the same natural products in subsequent collections.4 Central to these challenges is the role of symbiotic bacteria in the production of marine invertebrate-derived natural products. 5-7Specifically, symbiotic bacteria are central to chemical symbiosis, NP variability, and, as a result, to the supply of NPs. This dissertation addresses these two problems, the distribution and supply of symbiotic natural products. Symbiosis driven variability in natural products The following examples illustrate the influence of symbiosis on the distribution and variability of natural products from three invertebrate phyla: bryozoans, sponges, and ascidians. The identification of the symbiont producers is described in each case, as well as the factors that influence invertebrate-microbe symbiosis resulting in the sporadic occurrence of natural products. These studies emphasize the unappreciated role of cryptic host speciation, symbiont genetics, and geography in the production of compounds. ! "! Lastly, these individual cases, among others, prompted us to examine the global distribution of ascidian microbial symbionts, which is the topic of the first part of my research. Candidatus Endobugula sertula and the bryostatins One of the most well-rounded studies in this field is the investigation of the chemical symbiosis between the bryozoan Bugula neritina and its symbiont Candidatus Endobugula sertula. The bryozoan B. neritina is a cosmopolitan species that is widely distributed in the temperate and tropical ocean regions. 8,9 It is the source of the bryostatins, powerful protein kinase C (PKC) modulators that have been extensively investigated for their anticancer, antiviral, and neuroactive properties. 9 Limited and variable isolation from the bryozoans prompted a series of studies on the biosynthetic origin and ecology of bryostatins, with the goal of finding alternative routes of supplying these high-demand molecules. The structure of bryostatins suggests a microbial origin, specifically via polyketide synthase (PKS) machinery followed by multiple tailoring steps. Indeed, an uncultured gamma-proteobacterium associated with larvae of the bryostatin-producing B. neritina was identified and named Candidatus Endobugula sertula. 10 11 The same bacterium was found to harbor a set of biosynthetic genes consistent with biosynthetic hypotheses as to the origin of of bryostatins. 12,13 Subsequent studies strongly supported the hypothesis that Ca. E. sertula produces bryostatins. 14 Final proof will await attempts to express the bryostatin gene cluster in a heterologous host, which have not been successful to date. Meanwhile, efforts continue to further our understanding of the ! "! genetics of the Ca. E. sertula symbiont. Early studies on the isolation of bryostatins indicated a variable distribution in B. neritina species. 15 Identifying the genetic basis of bryostatins was critical to further progress in the understanding of the distribution of these molecules. Haygood and Davidson (1999) first showed that the B. neritina populations along the coast of California are composed of sibling species. 16 Type D (found in deep waters initially) and type S (found in shallow waters) differed by 8% in their mitochondrial CO1 sequences. Both harbored Ca. E. sertula symbionts that differed in their 16S rRNA gene sequences, and the bryostatin chemistry differed in each type of B. neritina. Subsequently, cryptic species of B. neritina were also identified from the western Atlantic range, 8 and a series of studies have highlighted the influence of geography in the distribution of B. neritina and bryostatin production. In these studies, Cape Haterras, a known biogeographical divide, was used as the reference point in the sampling of B. neritina. Another cryptic species of B. neritina, type N, generally inhabits the northern Atlantic ranges and does not harbor Ca. E. sertula symbionts. 8 This species lacks bryostatins, consistent with a producer role for Ca. E. sertula. Subspecies Type S of B. neritina, similar to the California species, is concentrated in southern areas relative to Cape Haterras and contains both Ca. E. sertula symbionts as well as bryostatins. The presence or absence of symbionts was therefore thought to explain completely the separation between species. Recently, however, more detailed collections have shown that the boundary of N and S species is less distinct than previously thought. 17,18 Although there is a heavy distribution of N type north of Cape Haterras, some N type species were also found south of the Cape and vice versa. Despite this apparent mixing of host species, the Ca. E. sertula symbionts remained geographically restricted to the south, ! "! regardless of the species of B. neritina. 18 This implies that bryostatin production might be strongly influenced by geography. Higher selection pressure in the southern warmer waters may lead to acquisition of Ca. E. sertula symbionts that can provide chemical defense. These studies demonstrate the interplay between different factors controlling the production of bryostatins. Cryptic speciation in B. neritina partly explains the presence and absence of symbionts and compounds, while the host-specificity of symbionts to these cryptic species of B. neritina appears to explain the variation between different bryostatin derivatives. Finally, the local environment may also play a role, presumably via natural selection, in determining the presence or absence of symbiotic natural product producers. Further widespread studies of the bryostatins, B. neritina, and Ca. E. sertula will be needed to attain a complete picture of bryostatin occurence and variation. Prochloron didemni and the cyanobactins Ascidians from the family Didemnidae are among the most chemically prolific groups of marine invertebrates. 2,7 Among other natural products, they produce a family of small cyclic peptides, known as cyanobactins. 19 Genome analysis of the obligate cyanobacterial symbiont of the ascidian Lissoclinum patella, Prochloron didemni, identified the biosynthetic cluster for cyanobactins, which was found to involve a ribosomal mode of biosynthesis, in which a precursor peptide is posttranslationally modified to give bioactive structures. 20 This mode of biosynthesis is now recognized as widespread, and a large class of natural products has been assigned to the growing family of ribosomally synthesized and posttranslationally modified peptides (RiPPs). 21 The cyanobactins are to date the most biotechnologically developed group among ! "! natural products from uncultured symbiotic bacteria. One important milestone in the development of cyanobactins as a platform for drug discovery was the demonstration, coincident with the discovery of their biosynthetic origins, that they could be heterologously expressed in E. coli. 20 Also important was the discovery of the tremendous flexibility of the cyanobactin pathway. 22 This latter discovery was largely enabled by an in-depth understanding of their distribution and variation in Nature. 22,23 Diverse families of cyanobactins had been isolated from different species of didemnid ascidians since the 1980s. 19,24-27 Structural overlaps between these families led to the hypothesis that much of cyanobactin chemical diversity can be traced to hypervariability in precursor peptides. 22 Indeed, analyses of cyanobactin pathways from didemnid ascidians across the tropical Pacific showed that the sequences are virtually identical except for regions that directly encode the cyanobactin backbone.22 In some cases, recombination of pathways leads to altered patters of posttranslational modifications, but even in such cases, precursor peptides from different pathways can be mixed and matched in vitro and in vivo. 23 Producing artificial libraries of cyanobactins would therefore only require minimal manipulation of precursor peptides, without any engineering of modifying enzymes. This in-depth knowledge of the conserved genetic architecture of cyanobactin pathways also accelerated the discovery of cyanobactin derivatives by purely genetic approaches. 28 Cyanobactin families are also variably distributed in didemnid ascidians. For example, trunkamides, patellins, patellamides, and lissoclinamides are often found in different combinations in L. patella ascidians. 22,29,30 To examine this question, three genomes of Prochloron didemni were sequenced and compared.29 This analysis revealed ! "! that the Prochloron didemni strains were largely identical except in regions that encoded secondary metabolites. Subsequently, 24 Prochloron didemni strains from different L. patella samples across the tropical Pacific were analyzed at specific loci as well as for secondary metabolite composition. No apparent distribution patterns could be discerned however. It was hypothesized that the sporadic variation in the pathways and metabolites is ecologically relevant for Prochloron-L. patella symbiosis. Specifically, the presence of different pathways in different ascidians allows access to the suite of cyanobactins as needed. A finer look at the L. patella phylogeny later revealed a previously unrecognized distribution. 31 Specifically, L. patella samples collected from the vast tropical Pacific were identified to derive from cryptic speciation events, varying by 5-10% in their mtCO1 sequences. From this analysis, the phylogeny of L. patella could be used to predict the distribution of cyanobactins, while the Prochloron symbiont was 99% identical in all cases. Thus, in an echo of the bryostatin work described above, host cryptic speciation was found to be a key driving force of the distribution of cyanobactins. A broader sampling of L. patella samples will be needed to show if this ‘host control' phenomenon is widespread and what other factors determine this symbiosis, for example host-factors that allow for recognition of specific Prochloron strains, or perhaps vertical transmission of symbionts. An improved understanding of the distribution of metabolites in L. patella across time and space might help improve our understanding of the ecological role of cyanobactins. ! "! Candidatus Entotheonella sp. and a suite of peptides and polyketides The marine sponges from the family Theonellidae are another group of marine invertebrates that have been studied for their structurally complex and bioactive NPs. 32 Most of these molecules are complex polyketides including the swinholides and discodermolides, as well as peptides such as keramamides, polytheonamides, theopalau amides, or hybrid PKS-NRPS-derived molecules such as the calyculins, onnamides, and theopederins. 32-35 The microbial origin of these metabolites has been proposed early on. Indeed, microscopy 36,37 and later biosynthetic studies confirmed this hypothesis. 38,39 Perhaps the earliest attempt to implicate symbiotic bacteria in the biosynthesis of sponge-derived metabolites was from the work of Bewley et al. in the mid-1990s. 36 Separation of microbial cells by differential centrifugation and subsequent chemical analysis showed that different microbial fractions contained different metabolites. For example, a fraction enriched in unicellular bacteria contained swinholides, while the fraction containing filamentous bacteria contained cyclic peptides of the theopalauamide family. In 2000, Schmidt et al. described the filamentous bacteria as a novel !proteobacteria, named Candidatus Entotheonella palauensis. 40 In this key study, the relationship of chemistry to the symbiont genotype was first proposed. More than a decade later, using a combination of single cell genomics and metagenomics, Wilson et al. confirmed that the Ca. Entotheonella species previously identified by Schmidt et al. are in fact the producers of most Theonella swinhoei-derived natural products. 41 Different strains of Ca. Entotheonella, produce different types of NPs present in the same T. swinhoei sample. Ca. Entotheonella species have also been described from other Theonellid sponges, where they are responsible for making large PKS-NRPS ! "! compounds.42 It is becoming apparent that Ca. Entotheonella strains are diverse and widespread among sponges. In particular, a PCR survey showed that Ca. Entotheonella strains are present in different sponges outside the Theonillidae and also collected from different locations. 41 No apparent patterns of distribution have been identified so far. It is clear, however, that symbiosis with different strains lead to the isolation of different compounds in sponges. Studies on the distribution, host-specificity, and chemistry of Ca. Entoheonella will be critical to advance our understanding of their distribution and to facilitate exploiting their potential in drug discovery. A global view of chemical symbiosis The specific cases above emphasize the roles of symbiotic bacteria in the variability and distribution of natural products. However, they only represent a small subset of marine natural products. The first part of my thesis addresses this problem by examining global patterns in the distribution of chemical symbiosis that could guide natural products discovery. We focused on ascidians, examining the specificity, diversity, and stability of the microbiomes and secondary metabolism across hosts, geography, and time. We hypothesized that in the tropical regions where there is a higher degree of predation and competition, the selective pressure to survive leads to acquisition of more symbionts that could aid with the host' fitness (e.g., chemical defense) and therefore leads to higher microbial diversity. The results are discussed in Chapter 2. ! "#! Supply of symbiotic natural products Natural products that are produced within the context of a symbiotic relationship are often potent because they are evolutionarily refined to interact with relevant biological targets. Consequently, they are often produced in limited amounts in the animals. The bryostatins described above are an excellent example of such compounds. The amount of bryostatins that can be isolated ranges from 10-5 to 10-7 wt% of the animal colony (18 grams from 14 tons of B. neritina in one example). 9 The same is true for L. patella metabolites, the patellazoles, which are products of the endosymbiont, Ca. endolissoclinum patella. 43 The patellazoles are highly cytotoxic, with an IC50 of 332 pM in cell assays. They are isolated at 0.08% dry weight of the ascidian and are sparsely found, severely hampering their development. Another barrier in the supply of NPs from symbiotic bacteria is the inability to cultivate the producing symbionts. Symbiotic bacteria in established relationships with their host often lose functions that are required for independent lifestyles outside the hosts. In extreme cases, the genome of symbiotic bacteria is degraded over time, retaining only the genes that are relevant to the symbiosis such as the secondary metabolic genes, while depending on the host for the rest of its metabolic needs. 44 Chemical synthesis is often the next step in supplying the natural products, especially those that are highly sought for their bioactivities. However, there are obvious drawbacks to total chemical synthesis. Natural products are structurally complex, often containing multiple chiral centers, and often require lengthy synthetic routes, thus posing a significant barrier to achieving a practical supply. 45 The total synthesis of bryostatins, for example, was completed in 79 steps in 1990s, and it was not until recently (2011) that ! ""! several research groups achieved the total synthesis with 36 to 43 steps. 46,47 Heterologous expression of biosynthetic pathways in genetically tractable hosts such as E. coli is one of the most promising alternatives to symbiotic NP supply. 48 49 Although not without challenges, heterologous expression offers many advantages over total chemical synthesis and other complementary approaches such as aquaculture. 50 Once optimized, natural products can be supplied from scaled fermentations without the need for multiple reactions and in between purification steps required of total synthesis. In the second part of my dissertation, we used heterologous expression to address the supply problem in the symbiotic NPs, the cyanobactins. Metabolic engineering and fermentation strategies for NP supply Tools and strategies for the metabolic engineering of natural products expressions in heterologous hosts are becoming increasingly available. 51-53 Some rational techniques involve increasing precursor supply, overexpression of bottleneck enzymes in the pathways, improving regulation of gene expression, and channeling intermediates, precursors into desired pathways. 53 These, in tandem with empirical approaches such as optimization of fermentation conditions, media, additives, cofactors, and other primary metabolites, have been used for the improved supply of natural products in vivo. 54,55 Increasing the precursor supply has been employed in several pathways, including polyketides, terpenoids, and shikimate pathways. As an example, in the production of actinorhodin, the supply of malonyl-CoA from acetyl-CoA was increased through the overexpression of acetyl-CoA carboxylase resulting in a 6-fold increase. 56 Similarly, the overexpression of the first enzyme in the shikimate pathway, DAHP, resulted in an ! "#! increase in the expression of Balhimycin, which required the aromatic amino acids produced through the shikimate pathway. 57 The terpenoid pathways are among the extensively studied tools in metabolic engineering. Terpenoid pathways use both precursors DMAPP and IPP as building blocks for longer terpene compounds. 58 In production of NPs in E. coli, these precursors are usually supplied by improving the efficiency of the native MEP pathway or via the heterologous expression of a mevalonate pathway derived from higher organisms. 59-61 Empirical methods have also been used in the optimization of fermentation conditions for production of NPs. One example is in the optimization of 6deoxyerythronolide B production in E. coli, where the effect of media, and subsequently its individual components were studied. In this example, tryptone, and subsequently, tryptone components, glutamine, and threonine amino acids were found to increase compound production by a further 16-fold. 55 The cyanobactin pathway is one of the few marine NP pathways that have been successfully expressed in a heterologous host. 5,20,28 Since the first expression of cyanobactins in E. coli, there have been continuous efforts to improve and optimize the expression system in order to establish a robust platform that can supply practical amounts of cyanobactins. 28,62However, despite the stability in the detection of compounds from the extracts, the best yield that we were able to obtain previously was 300 µg/L. To improve the production of cyanobactins in E. coli, we used a combination of simple media and fermentation optimization, as well as co-expression of a heterologous mevalonate pathway to supply isoprene precursors. In Chapter 3, we show an overall ! "#! improvement in yield of about 100-fold, which is the highest titer for the expression of cyanobactins so far. With this ability to supply cyanobactins, we started to explore their bioactivities, which resulted in the discovery of a new activity for patellin 2 63 in dorsal root ganglia neurons. Engineering for expansion of cyanobactin diversity One key advantage that synthetic compound possess over natural products is that it is vastly easier to generate derivatives of synthetic compound in order to optimize the bioactivity of a given compound. In contrast, it is much more difficult to discover or engineer natural product derivatives, and natural product chemists generally lack the ability to make deterministic changes in compounds, but are limited to what few derivatives are naturally occurring, or those that can be generated by disabling late-stage biosynthetic enzymes. Within this context, the cyanobactin pathway is particularly notable, in that it can be readily engineered to allow the production of derivative compounds due to the high substrate tolerance of the modifying enzymes. In the first proof of this capacity, the precursor peptide of the patellamide pathway, patE2, encoding the natural product ulithiacyclamide was modified to encode a peptide mimic of the rattlesnake venom peptide, eptifibatide. 22 The artificial cyanobactin was referred to as eptidemnamide, with the replacement of the original disulfide with an amide bond. Successful production of eptidemnamide demonstrated that cyanobactin enzymes are substrate-tolerant beyond just the natural cyanobactin sequences. Eptidemnamide contained charged amino acid residues, arginine, and aspartic acid, and others that were not found in natural ! "#! cyanobactins, specifically, glycine, tryptophan, and glutamine. 22 Successive experiments in the engineering of cyanobactins further supported this finding, including swapping of precursor peptides in the trunkamide pathway, 23 and producing a new cyanobactin called minimide in the tru pathway that was discovered by genome mining. 28 The capacity of the cyanobactin pathway for generating chemical diversity was subjected to a quantitative test with the construction of random mutant libraries based on the sequence of trunkamide. Double and quadruple mutant subsets of a projected library of 3.2 million derivatives were expressed and chemically analyzed to examine the amino acid selectivity in the different positions of the compound. From these subsets, 58% of the unique mutant genes were produced in vivo. Libraries of 108 compounds were made, and by extrapolation from these experiments, a minimum of 106 compounds can be produced from a single cyanobactin pathway. 62 Although these experiments demonstrated that the cyanobactin pathway has a high substrate tolerance for proteinogenic amino acids, this does impose a limitation on the chemical diversity on cyanobactins. Thus, we posed the question of whether it might be possible to incorporate functionalities beyond the 20 proteinogenic amino acids. Incorporation of non-proteinogenic amino acids are often hallmarks of non-ribosomal peptides, among which are some of the most bioactive natural products known. 64 NRPS pathways are widespread, but are extraordinarily difficult to engineer. 65,66Combining multistep, substrate-tolerant, RiPP pathways with the structural complexity of NRPS therefore would be an unprecedented and important advance in the development of natural products as drugs. In Chapter 4, we describe the successful incorporation of unnatural amino acids (UAAs) into cyanobactins. Prior to this report, the incorporation of ! "#! UAAs has only been done in proteins, 67 as well as in lantipeptides in vitro. 68 References 1 Cragg, G. M. & Newman, D. J. Natural products: a continuing source of novel drug leads. Biochim Biophys Acta 1830, 3670-3695, doi:10.1016/j.bbagen.2013.02.008 (2013). 2 Blunt, J. W., Copp, B. R., Keyzers, R. A., Munro, M. H. & Prinsep, M. R. Marine natural products. Nat Prod Rep 30, 237-323, doi:10.1039/c2np20112g (2013). 3 Martins, A., Vieira, H., Gaspar, H. & Santos, S. Marketed marine natural products in the pharmaceutical and cosmeceutical industries: tips for success. Mar Drugs 12, 1066-1101, doi:10.3390/md12021066 (2014). 4 Molinski, T. F., Dalisay, D. S., Lievens, S. L. & Saludes, J. P. Drug development from marine natural products. Nat Rev Drug Discov 8, 69-85, doi:10.1038/nrd2487 (2009). 5 Piel, J. Metabolites from symbiotic bacteria. Nat Prod Rep 26, 338-362, doi:10.1039/b703499g (2009). 6 Florez, L. V., Biedermann, P. H., Engl, T. & Kaltenpoth, M. Defensive symbioses of animals with prokaryotic and eukaryotic microorganisms. Nat Prod Rep, doi:10.1039/c5np00010f (2015). 7 Schmidt, E. W. The secret to a successful relationship: lasting chemistry between ascidians and their symbiotic bacteria. Invertebrate Biology : A Quarterly Journal of the American Microscopical Society and the Division of Invertebrate Zoology/ASZ 134, 88-102, doi:10.1111/ivb.12071 (2015). 8 McGovern, T. M. & Hellberg, M. E. Cryptic species, cryptic endosymbionts, and geographical variation in chemical defences in the bryozoan Bugula neritina. Mol Ecol 12, 1207-1215 (2003). 9 Trindade-Silva, A. E., Lim-Fong, G. E., Sharp, K. H. & Haygood, M. G. Bryostatins: biological context and biotechnological prospects. Curr Opin Biotech 21, 834-842, doi:10.1016/j.copbio.2010.09.018 (2010). 10 Woollacott, R. M. Association of bacteria with bryozoan larvae. Mar Biol 65, 155-158, doi:10.1007/BF00397080 (1981). 11 Haygood, M. G. & Davidson, S. K. Small-subunit rRNA genes and in situ hybridization with oligonucleotides specific for the bacterial symbionts in the ! "#! larvae of the bryozoan Bugula neritina and proposal of "Candidatus endobugula sertula". Appl Environ Microbiol 63, 4612-4616 (1997). 12 Davidson, S. K., Allen, S. W., Lim, G. E., Anderson, C. M. & Haygood, M. G. Evidence for the biosynthesis of bryostatins by the bacterial symbiont "Candidatus Endobugula sertula" of the bryozoan Bugula neritina. Appl Environ Microbiol 67, 4531-4537 (2001). 13 Sudek, S. et al. Identification of the putative bryostatin polyketide synthase gene cluster from "Candidatus Endobugula sertula", the uncultivated microbial symbiont of the marine bryozoan Bugula neritina. J Nat Prod 70, 67-74, doi:10.1021/np060361d (2007). 14 Lopanik, N. B. et al. In vivo and in vitro trans-acylation by BryP, the putative bryostatin pathway acyltransferase derived from an uncultured marine symbiont. Chem Biol 15, 1175-1186, doi:10.1016/j.chembiol.2008.09.013 (2008). 15 Pettit, G. R. in Fortschritte der Chemie organischer Naturstoffe / Progress in the Chemistry of Organic Natural Products Vol. 57 Fortschritte der Chemie organischer Naturstoffe / Progress in the Chemistry of Organic Natural Products (eds W. Herz, G. W. Kirby, W. Steglich, & Ch Tamm) Ch. 3, 153-195 (Springer Vienna, 1991). 16 Davidson, S. K. & Haygood, M. G. Identification of sibling species of the bryozoan Bugula neritina that produce different anticancer bryostatins and harbor distinct strains of the bacterial symbiont "Candidatus Endobugula sertula". Biol Bull 196, 273-280 (1999). 17 Fehlauer-Ale, K. H. et al. Cryptic species in the cosmopolitan Bugula neritina complex (Bryozoa, Cheilostomata). Zool Scr 43, 193-205, doi:10.1111/zsc.12042 (2014). 18 Linneman, J., Paulus, D., Lim-Fong, G. & Lopanik, N. B. Latitudinal variation of a defensive symbiosis in the Bugula neritina (Bryozoa) sibling species complex. PLoS One 9, e108783, doi:10.1371/journal.pone.0108783 (2014). 19 Donia, M. S. & Schmidt, E. W. in Comprehensive Natural Products II (eds Mander Lew, & Liu Hung-Wen) Vol. 2, 539-558 (Elsevier, 2010). 20 Schmidt, E. W. et al. Patellamide A and C biosynthesis by a microcin-like pathway in Prochloron didemni, the cyanobacterial symbiont of Lissoclinum patella. Proc Natl Acad Sci U S A 102, 7315-7320, doi:10.1073/pnas.0501424102 (2005). ! "#! 21 Arnison, P. G. et al. Ribosomally synthesized and post-translationally modified peptide natural products: overview and recommendations for a universal nomenclature. Nat Prod Rep 30, 108-160, doi:10.1039/c2np20085f (2013). 22 Donia, M. S. et al. Natural combinatorial peptide libraries in cyanobacterial symbionts of marine ascidians. Nat Chem Biol 2, 729-735, doi:10.1038/nchembio829 (2006). 23 Donia, M. S., Ravel, J. & Schmidt, E. W. A global assembly line for cyanobactins. Nat Chem Biol 4, 341-343, doi:10.1038/nchembio.84 (2008). 24 Ireland, C., Scheuer,P.J. Ulicyclamide and ulithiacyclamide, two new small peptides from a marine tunicate. J Am Chem Soc 102, 5688-5691 (1980). 25 Ireland, C. M., Durso, A.R., Newman, R.A, Hacker, M.P. Antineoplastic cyclic peptides from the marine tunicate Lissoclinum patella. J Org Chem 47, 18071811 (1982). 26 Carroll, A. R. et al. Patellins 1-6 and trunkamide A: novel cyclic hexa-, heptaand octa-peptides from colonial ascidians, Lissoclinum sp. Aust J Chem 49, 659667 (1996). 27 Degnan, B. M. et al. New cyclic peptides with cytotoxic activity from the ascidian Lissoclinum patella. J Med Chem 32, 1349-1354 (1989). 28 Donia, M. S., Ruffner, D. E., Cao, S. & Schmidt, E. W. Accessing the hidden majority of marine natural products through metagenomics. Chembiochem 12, 1230-1236, doi:10.1002/cbic.201000780 (2011). 29 Donia, M. S., Fricke, W. F., Ravel, J. & Schmidt, E. W. Variation in tropical reef symbiont metagenomes defined by secondary metabolism. PLoS One 6, e17897, doi:10.1371/journal.pone.0017897 (2011). 30 Schmidt, E. W., Donia, M. S., McIntosh, J. A., Fricke, W. F. & Ravel, J. Origin and variation of tunicate secondary metabolites. J Nat Prod 75, 295-304, doi:10.1021/np200665k (2012). 31 Kwan, J. C. et al. Host control of symbiont natural product chemistry in cryptic populations of the tunicate Lissoclinum patella. PLoS One 9, e95850, doi:10.1371/journal.pone.0095850 (2014). 32 Winder, P. L., Pomponi, S. A. & Wright, A. E. Natural products from the Lithistida: a review of the literature since 2000. Mar Drugs 9, 2643-2682, doi:10.3390/md9122643 (2011). ! "#! 33 Fusetani, N. & Matsunaga, S. Bioactive sponge peptides. Chem Rev 93, 17931806, doi:10.1021/cr00021a007 (1993). 34 Bewley, C. A. & Faulkner, D. J. Lithistid sponges: star performers or hosts to the stars. Angew Chem Int Edit 37, 2162-2178, doi:10.1002/(SICI)15213773(19980904)37:16<2162::AID-ANIE2162>3.0.CO;2-2 (1998). 35 Schmidt, E. W., Bewley, C. A. & Faulkner, D. J. Theopalauamide, a bicyclic glycopeptide from filamentous bacterial symbionts of the lithistid sponge Theonella swinhoei from Palau and Mozambique. J Org Chem 63, 1254-1258, doi:10.1021/jo9718455 (1998). 36 Bewley, C. A., Holland, N. D. & Faulkner, D. J. Two classes of metabolites fromTheonella swinhoei are localized in distinct populations of bacterial symbionts. Experientia 52, 716-722, doi:10.1007/bf01925581 (1996). 37 Schirmer, A. et al. Metagenomic analysis reveals diverse polyketide synthase gene clusters in microorganisms associated with the marine sponge Discodermia dissoluta. Appl Environ Microbiol 71, 4840-4849, doi:10.1128/AEM.71.8.48404849.2005 (2005). 38 Piel, J. et al. Antitumor polyketide biosynthesis by an uncultivated bacterial symbiont of the marine sponge Theonella swinhoei. Proc Natl Acad Sci U S A 101, 16222-16227, doi:10.1073/pnas.0405976101 (2004). 39 Freeman, M. F. et al. Metagenome mining reveals polytheonamides as posttranslationally modified ribosomal peptides. Science 338, 387-390, doi:10.1126/science.1226121 (2012). 40 Schmidt, E. W., Obraztsova, A. Y., Davidson, S. K., Faulkner, D. J. & Haygood, M. G. Identification of the antifungal peptide-containing symbiont of the marine sponge Theonella swinhoei as a novel !-proteobacterium, Candidatus Entotheonella palauensis. Mar Biol 136, 969-977, doi:10.1007/s002270000273 (2000). 41 Wilson, M. C. et al. An environmental bacterial taxon with a large and distinct metabolic repertoire. Nature 506, 58-62, doi:10.1038/nature12959 (2014). 42 Wakimoto, T. et al. Calyculin biogenesis from a pyrophosphate protoxin produced by a sponge symbiont. Nat Chem Biol 10, 648-655, doi:10.1038/nchembio.1573 (2014). 43 Kwan, J. C. et al. Genome streamlining and chemical defense in a coral reef symbiosis. Proc Natl Acad Sci U S A 109, 20655-20660, doi:10.1073/pnas.1213820109 (2012). ! "#! 44 McCutcheon, J. P. & Moran, N. A. Extreme genome reduction in symbiotic bacteria. Nat Rev Microbiol 10, 13-26, doi:10.1038/nrmicro2670 (2012). 45 Morrison, K. C. & Hergenrother, P. J. Natural products as starting points for the synthesis of complex and diverse compounds. Nat Prod Rep 31, 6-14, doi:10.1039/c3np70063a (2014). 46 Wender, P. A. & Schrier, A. J. Total synthesis of Bryostatin 9. J Am Chem Soc 133, 9228-9231, doi:10.1021/ja203034k (2011). 47 Lu, Y., Woo, S. K. & Krische, M. J. Total Synthesis of Bryostatin 7 via C-C Bond Forming Hydrogenation: Merged Redox-Construction Events for Synthetic Efficiency. J Am Chem Soc 133, 13876-13879, doi:10.1021/ja205673e (2011). 48 Ongley, S. E., Bian, X., Neilan, B. A. & Muller, R. Recent advances in the heterologous expression of microbial natural product biosynthetic pathways. Nat Prod Rep 30, 1121-1138, doi:10.1039/c3np70034h (2013). 49 Luo, Y., Cobb, R. E. & Zhao, H. Recent advances in natural product discovery. Curr Opin Biotech 30, 230-237, doi:10.1016/j.copbio.2014.09.002 (2014). 50 Leal, M. C. et al. Marine microorganism-invertebrate assemblages: perspectives to solve the "supply problem" in the initial steps of drug discovery. Mar Drugs 12, 3929-3952, doi:10.3390/md12073929 (2014). 51 Zhang, H., Boghigian, B. A., Armando, J. & Pfeifer, B. A. Methods and options for the heterologous production of complex natural products. Nat Prod Rep 28, 125-151, doi:10.1039/c0np00037j (2011). 52 Challis, G. L. Engineering Escherichia coli to produce nonribosomal peptide antibiotics. Nat Chem Biol 2, 398-400, doi:10.1038/nchembio0806-398 (2006). 53 Pickens, L. B., Tang, Y. & Chooi, Y. H. Metabolic engineering for the production of natural products. Ann Rev Chem Biomol Eng 2, 211-236, doi:10.1146/annurevchembioeng-061010-114209 (2011). 54 Sun, L., Zeng, J., Zhang, S., Gladwin, T. & Zhan, J. Effects of exogenous nutrients on polyketide biosynthesis in Escherichia coli. App Microbiol Biot 99, 693-701, doi:10.1007/s00253-014-6212-7 (2015). 55 Pistorino, M. & Pfeifer, B. A. Efficient experimental design and micro-scale medium enhancement of 6-deoxyerythronolide B production through Escherichia coli. Biotechnol Progr 25, 1364-1371, doi:10.1002/btpr.250 (2009). 56 Ryu, Y. G., Butler, M. J., Chater, K. F. & Lee, K. J. Engineering of primary carbohydrate metabolism for increased production of actinorhodin in ! "#! Streptomyces coelicolor. Appl Environ Microbiol 72, 7132-7139, doi:10.1128/AEM.01308-06 (2006). 57 Thykaer, J. et al. Increased glycopeptide production after overexpression of shikimate pathway genes being part of the balhimycin biosynthetic gene cluster. Metab Eng 12, 455-461, doi:10.1016/j.ymben.2010.05.001 (2010). 58 Kuzuyama, T. Mevalonate and nonmevalonate pathways for the biosynthesis of isoprene units. Biosci Biotech Bioch 66, 1619-1627 (2002). 59 Farmer, W. R. & Liao, J. C. Precursor balancing for metabolic engineering of lycopene production in Escherichia coli. Biotechnol Prog 17, 57-61, doi:10.1021/bp000137t (2001). 60 Pitera, D. J., Paddon, C. J., Newman, J. D. & Keasling, J. D. Balancing a heterologous mevalonate pathway for improved isoprenoid production in Escherichia coli. Metab Eng 9, 193-207, doi:10.1016/j.ymben.2006.11.002 (2007). 61 Martin, V. J., Pitera, D. J., Withers, S. T., Newman, J. D. & Keasling, J. D. Engineering a mevalonate pathway in Escherichia coli for production of terpenoids. Nat Biotech 21, 796-802, doi:10.1038/nbt833 (2003). 62 Ruffner, D. E., Schmidt, E. W. & Heemstra, J. R. Assessing the Combinatorial Potential of the RiPP Cyanobactin tru Pathway. ACS Synth Biol 4, 482-492, doi:10.1021/sb500267d (2015). 63 Zabriskie, T. M., Foster, M. P., Stout, T. J., Clardy, J. & Ireland, C. M. Studies on the solution- and solid-state structure of patellin 2. J Am Chem Soc 112, 80808084, doi:10.1021/ja00178a035 (1990). 64 Walsh, C. T., O'Brien, R. V. & Khosla, C. Nonproteinogenic amino acid building blocks for nonribosomal peptide and hybrid polyketide scaffolds. Angew Chem 52, 7098-7124, doi:10.1002/anie.201208344 (2013). 65 Williams, G. J. Engineering polyketide synthases and nonribosomal peptide synthetases. Curr Opin Struc Biol 23, 603-612, doi:10.1016/j.sbi.2013.06.012 (2013). 66 Mootz, H. D., Schwarzer, D. & Marahiel, M. A. Ways of assembling complex natural products on modular nonribosomal peptide synthetases. Chembiochem 3, 490-504, doi:10.1002/1439-7633(20020603)3:6<490::AID-CBIC490>3.0.CO;2N (2002). 67 Liu, C. C. & Schultz, P. G. Adding new chemistries to the genetic code. Ann Rev Biochem 79, 413-444, doi:10.1146/annurev.biochem.052308.105824 (2010). ! 68 "#! Levengood, M. R., Knerr, P. J., Oman, T. J. & van der Donk, W. A. In vitro mutasynthesis of lantibiotic analogues containing nonproteinogenic amino acids. J Am Chem Soc 131, 12024-12025, doi:10.1021/ja903239s (2009). CHAPTER 2 SPECIES SPECIFICITY OF SYMBIOSIS AND SECONDARY METABOLISM IN ASCIDIANS Manuscript reproduced with permission from: Ma Diarey B Tianero, Jason C Kwan, Thomas P Wyche, Angela P Presson, Michael Koch, Louis R Barrows, Tim S Bugni and Eric W Schmidt. Species specificity of symbiosis and secondary metabolism in ascidians. ISME Journal 9, 615-628. Note: My contribution to this paper was in design of the project, collection, processing, analysis of microbiome data, interpretation of results, and writing of the manuscript. ! "#! ! "#! ! "#! ! "#! ! "#! ! "#! ! "#! ! "#! ! "#! Ascidian microbiomes and metabolomes MDB Tianero et al 9 Bioactivity profile of ascidian chemical extracts We assayed the chemical extracts of the ascidians against laboratory strains of microorganisms to test whether there are distinct differences in the potency and toxicity of metabolites from the tropical regions compared with the temperate regions. We tested for activities against different microbes, including Staphylococcus aureus, Escherichia coli and Mycobacterium tuberculosis. We also tested for cytotoxicity against a human CEM-TART cell line and for antifungal activity against Candida albicans. There is a clear geographical distribution of biological activity and potency in these samples, which reflects the results of the secondary metabolite analysis performed as described below (Supplementary Figure S8). In these assays, the most potent extracts were those of tropical ascidians while the least potent are those of temperate ascidians. Importantly, the toxicity of these samples was not correlated to diversity of the underlying microbiome but was independent of this variable. This indicates that strains with more diverse microbiomes do not produce more toxic compounds and serves to further reinforce the role of single, talented bacterial symbionts in the production of toxins (see Discussion). Chemistry of ascidians PCA and hierarchical clustering of the LC/MS profiles showed that ascidians generally grouped by geographical location rather than by species (Figure 5). Three major clusters are formed comprising California, Florida and a combination of Papua New Guinea and Vanuatu in a single cluster. We hypothesized that this pattern was due to lipids in the extracts, which likely vary, because different temperatures would require different components to maintain membrane fluidity (Parrish, 2013). Alternatively, it might be due to other effects, such as available food sources, although this seemed less likely due to the large geographical sampling range in the tropical Pacific. Using the PCA biplot of ascidians and compounds (Supplementary Figure S5A), we performed an initial identification of the extract components that led to the geographical clustering of the ascidians. The chromatographic peaks were assigned using Lipid Maps (Fahy et al., 2007), and the resulting peaks assigned to lipids were then applied to PCA analysis. By using this putative lipid subset, we observed the same clustering by geographical location in the PCA (Supplementary Figure S5B). Conversely, we also subtracted the Lipid Mapsassigned peaks from the whole mzxml files and subjected the remaining data to PCA. Upon removing putative lipids from the analysis, the clustering according to geographical location was no longer observed. Instead, clustering was better correlated with animals species (see below). A weakness of all MS-based approaches is that it is difficult to be sure that an ion has been accurately assigned. We therefore used further methods to determine whether ions preliminarily assigned as lipids did indeed truly represent lipids. We selected representative extracts from each PCA cluster and analyzed them using a validated metabolomics method, in which fragmentation of lipids (MS/MS profile) was compared with a large database of lipid fragment patterns. Through this analysis, we confirmed that the major components responsible for variance in the PCA were lipids. These included phospho-glycerolipids, glycerolipids and some straight chain fatty acids (Figure 5, Supplementary Figure S6). We further analyzed the ascidian chemistry between samples within the same geographical location. We focused on Papua New Guinea samples where we have groups containing replicates of the same species and have characterized the natural products chemistry (Donia et al., 2011b). By PCA, we found that the samples cluster by host species. For example, L. patella samples (one from Fiji and two from Papua New Guinea) formed a separate cluster that was strongly driven by secondary metabolites. These metabolites include members of the cyanobactin family: patellins 1, 2, and 5 (Carroll et al., 1996); and the patellazoles (Corley et al., 1988; Zabriskie et al., 1988) (Figure 6, Supplementary Figure S7). Similarly, L. badium samples grouped together and in correlation with varamines (Molinski and Ireland, 1989) and some other unknown metabolites. When we included the Cystodytes sp. samples from Vanuatu in this analysis, these also strongly clustered together as expected, except for a single sample that was phylogenetically different than the others (Supplementary Figure S1). The grouping of Cystodytes samples was also caused by secondary metabolites, mostly pyridoacridine alkaloids (Marshall and Barrows, 2004), and the variations within the species were due to the abundance of different compound analogs. Specifically, VN_08002 and VN_08041 samples contained major amounts of cystodytins while VN_08019 mainly contained shermilamines in the extract (Cooray et al., 1988; Kobayashi et al., 1991). In cases where they could be identified, secondary metabolites were also noticeably major components of extracts of other ascidians from Florida, such as the eudistomins (Kobayashi et al., 1984) from Eudistoma sp. TB1. The extract of L. bistratum also contained major metabolites, including the bistramides and the bistratamides group of compounds (Degnan et al., 1989; Biard et al., 1994; Perez and Faulkner, 2003). However, in most samples, especially those from California, the majority of the components could not be assigned from databases, such as Antimarin and the Dictionary of Natural Products. Discussion We set out to determine whether there are any correlations between ascidian bacterial communities and the presence of abundant, bioactive The ISME Journal ! "#! ! ""! ! "#! ! "#! ! "#! ! "#! !"#$%#&'&"#$%(%$%)*'+('&*,-%+&%&'./0'&#$+/0.1*',#).-+2%&,'%/'.&$%0%./&' Ma. Diarey B. Tianero, Jason C. Kwan, Thomas P. Wyche, Angela P. Presson, Michael Koch, Louis R. Barrows, Tim S. Bugni, Eric W. Schmidt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scidian tissue (~0.5 cm3) was ground in liquid nitrogen in sterile mortar and pestle. The powder obtained was resuspended in TE (5 mL) with lysozyme (2 mg ml-1) and incubated at 30 ºC for 1 hour. EDTA (0.5 M, 1.2 mL) was added to the solution and gently mixed, after which proteinase K (0.2 mg mL-1, Qiagen) was added and the solution was incubated at 30 ºC for 5 minutes. For cell lysis, SDS (10%, 0.65 mL) was added and the solution was incubated at 37 ºC for 3-4 hours. After visibly complete lysis (clear solution), NaCl (5M, 1.2 mL) was added, mixed, and CTAB/NaCl (1 mL) solution was added before incubation at 65 ºC for 1 hour. Standard phenol/chloroform extraction was then performed (Sambrook and Russell 2006). " " !"#$%&'()+( " Processing and analysis of ascidian chemical extracts. Each ascidian was extracted for 1 hour with methanol (5 mL), and the methanol was removed and dried under vacuum. Each extract was dissolved in methanol:water (1:9; 1.1 mL) and loaded into a Gilson GX-271 liquid handling system. Samples (0.9 mL) were subjected to automated SPE (ThermoFisher Scientific HyperSep C18, 50 mg absorbent mass, 1 mL reservoir volume), washed with H2O (1 mL), and eluted with 95% methanol (1 mL) directly into an LC/MS-certified vial. Following SPE, each extract was quantified using an evaporative light-scattering detector (ELSD) according to previously published methods.(Adnani et al 2012) Each extract was diluted with 95% methanol to a concentration of 30 ng/mL. Methods S3 Representative samples from different locations were analyzed for putative lipid components on an Agilent MS 6520 Q-TOF mass spectrometer fitted to an Agilent 1290 UPLC. The extract components were separated on Acquity UPLC CSH C18 column (1.7 !M, 2.1 X100 mm) using the ! "#! following gradient at 0.3 mL min-1 (mobile phase A:B): t = 0, 85:15; t = 0-4 min, gradient to 70:30; t = 4-5, gradient to 48:52; t = 5-22, gradient to 18:82; t = 22-23, gradient to 1:99; t = 23-30, 1:99. Mobile phase A consisted of acetonitrile:water (60:40 v/v) in 10 mM ammonium formate and 0.1% formic acid, and mobile phase B consisted of isopropyl alcohol:water (90:10 v/v) in 10 mM ammonium formate and 0.1% formic acid. The column temperature was set to 65 oC. Acquisition was performed in positive ESI mode acquiring from m/z 100-1700. The mass spectrometer was operated using the following parameters: dry gas temperature, 350 oC; dry gas flow, 11.1 L/min; nebulizer pressure, 24 psig; Vcap, 5000V; fragmentor, 250 V. Additional parameters (precursors per cycle, 2; absolute threshold, 200) were used for data-dependent MS/MS. Methods S4 Antimicrobial broth dilution assays. The assay was performed as previously described(Koch et al 2010) with the exception that four wells per compound were used for each dilution for MIC determination. DMSO served as a negative control. Bacterial strains used, along with the positive control antibiotic for each strain, were M. tuberculosis H37Ra ATCC 25177 (rifampicin); E. coli ATCC 25922 (gentamicin); B. subtilis ATCC 6633 (gentamicin); and S. aureus ATCC 25923 (gentamicin) and C. albicans ATTC 90028 (itraconazole). Cytotoxicity assay. Cytotoxicity testing was performed using CEM-TART lymphoblastoid cells(Chen et al 1992) cultured in RPMI-1640 with 20% fetal bovine serum and antibiotic/antimycotic supplement at 37 °C with 5% CO2 in moisture saturated atmosphere in 96well culture clusters as previously described (Lin et al 2013). ! ! ! ! ! "#! !"#$%&'(!"!#$%&'(!)*''()+,*-!.(+$,'/!$-.!01#!2345!6(-(!785#9!:,+/;!<!=-.,)$+(/!)*.(/!>/(.!,-! &2(',%,-$2?!+$@*-*%,)!,.(-+,A,)$+,*-!*A!/$%&'(/;! ! Code Collection Date Collection Site, coordinates CA_Ascidia_sp_4-19 20-Feb-11 CA_Botryllus_sp_5-19 12-Oct-11 CA_Didemnum_sp_417 20-Feb-11 CA_Didemnum_sp_4-2 19-Feb-11 CA_Didemnum_sp_4-7 19-Feb-11 CA_Didemnum_sp_5-1 10-Oct-11 CA_Didemnum_sp_511 CA_Didemnum_sp_518 11-Oct-11 12-Oct-11 CA_Pyura_sp_4-21 19-Feb-11 CA_Styela_sp_5-16 FJ_Lissoclinum_patella _06037 FL_Didemnum_sp_TB3 *(D.psammathode) FL_Ecteinascidia_turbi nata_5 FL_Ecteinascidia_turbi nata_TB2 FL_Eudistoma_sp. _004 *(E. hepaticum) FL_Eudistoma_sp_TB1 *(E. olivaceum) FL_Trididemnum_sp_0 01BB *(T. orbicatum) FL_Trididemnum_sp_T B4 *(T. palmae) PNG_Didemnum_sp_1 1038 PNG_Didemnum_sp_1 1055 PNG_Didemnum_sp_1 1089 PNG_Lissoclinum_pate lla_11033 PNG_Lissoclinum_pate lla_11040 PNG_Lissoclinum_vare au_11047 PNG_Lissoclinum_vare au_11062 PNG_Lissoclinum_vare au_11068 11-Oct-11 15-Aug-11 10-Aug-11 15-Aug-11 28-Oct-10 1-Sep-11 11-Oct-10 15-Aug-11 9-Nov-11 10-Nov-11 12-Nov-11 9-Nov-11 9-Nov-11 10-Nov-11 11-Nov-11 11-Nov-11 BLAST hits, GenBank accession number Catalina Island, 33o 26.95 ' N, Ascidia ceratodes, L12378.2 118o 29.94' W Bottrylus planus, Catalina Harbor DQ346653.1 Catalina Island, 33o 26.95' N, Didemnum molle, 118o 29.94' W AB211071.1 Catalina Island, 33o 26.92' N, Didemnum molle, 118o 28.64' W AB211071.1 Catalina Island, 33o 26.92' N, Didemnum sp. DidSA/57, 118o 28.64' W AB211072.1 Catalina Island, 33o 26.69' N, Didemnum sp. DidSA/57, 118o 28.92' W AB211072.1 Catalina Island, 33o 26.87' N, Didemnum molle, 118o 28.61' W AB211071.1 Catalina Island, 33o 26.21' N, Didemnum molle, 118o 27.48' W AB211071.1 Catalina Island, 33o 26.67' N, Pyura haustor, AY903926.1 118o 29.417' W Catalina Harbor Styela plicata, L12444.2 Lissoclinum patella, Fiji, 17o55' S, 177o16' E AB211085.1 Florida Keys, 24 o 41.672' N, Didemnum sp. Whangamata81o 26,797' W EM-2003, AJ579861.1 Florida Keys, 24 o 39.591' N, Ecteinascidia turbinata, 81 o 25.217' W FM244848.1 Florida Keys, 24 o 39.453' N, Ecteinascidia turbinata, 81 o 25.247' W FM244848.1 Florida Keys, 24 o 41.078 N, Eudistoma gilboviride, 81 o 26.957 W AB211069.1 Florida Keys, 24 o 41.685' N, Eudistoma gilboviride, 81 o 26.840' W AB211069.1 Florida Keys, 24 o 37.487' N, Trididemnum paracyclops, 81 o 27.443' W AB211077.1 Florida Keys, 24 o 37.487' N, Trididemnum paracyclops, 81 o 27.443' W AB211077.1 Papua New Guinea, 10° 16' S, Didemnum molle, 145° 38' E AB211071.1 Papua New Guinea, 10° 2' S, Didemnum sp. DidSA/57, 145° 33' E AB211072.1 Papua New Guinea, 9° 36' S, Didemnum molle, 147° 18' E AB211071.1 Papua New Guinea, 10° 16' S, Lissoclinum patella, 145° 38' E AB211085.1 Papua New Guinea, 10° 16' S, Lissoclinum patella, 145° 38' E AB211085.1 Papua New Guinea, 10° 8' S, Lissoclinum badium, 145° 35' E AB211078.1 Papua New Guinea, 10° 2' S, Lissoclinum badium, 145° 32' E AB211078.1 Papua New Guinea, 10° 2' S, Lissoclinum badium, 145° 32' E AB211078.1 % 18S identity primers 99 F1/R1 99 F3/R3 94 AscF2/ AscR5 92 92 92 94 90 F3/R3 AscF2/ AscR5 AscF2/ AscR5 AscF2/ AscR5 AscF2/ AscR5 99 F3/R3 99 F3/R3 99 F1/R1 89 AscF2/ AscR5 99 F3/R3 99 F3/R3 89 95 92 92 94 95 94 AscF2/ AscR5 AscF2/ AscR5 F1/R1 AscF2/ AscR5 AscF2/ AscR5 AscF2/ AscR5 AscF2/ AscR5 99 F1/R1 99 F1/R1 86 F3/R3 86 F3/R3 86 F3/R3 ! "#! PNG_Lissoclinum_vare 12-Nov-11 au_11083 PNG_Lissoclinum_bistr 05-Oct-07 atum *(L. voeltzkowi) VN_Cystodytes_sp_080 28-Oct-08 02 VN_Cystodytes_sp_080 1-Nov-08 19 VN_Cystodytes_sp_080 4-Nov-08 41 Papua New Guinea, 9° 35' S, Lissoclinum badium, 147° 17' E AB211078.1 Papua New Guinea, 4° 8' S, Lissoclinum bistratum, 151° 34' E AB211084.1 Vanuatu, 15° 32.496' S, Cystodytes sp., FM244842.1 167°12.895' E Vanuatu, 15° 36.675' S, 167° Cystodytes sp., FM244842.1 01.258' E Vanuatu, 15° 25.416' S, 167° Cystodytes sp., FM244842.1 12.476' E 87 99 94 91 94 F3/R3 F1/R1 AscF2/ AscR5 AscF2/ AscR5 AscF2/ AscR5 ! ! ! ! !"#$%&'("!#$%&'($')*!+,-,$+,$%+! ! Number of tunicate samples 32 Number of filtered reads 217073 Number of 97 % OTUs 3982 Counts statistics (min/max/average) 572/31660/6784 OTUs statistics (min/max/average) 10/804/207 ! ! ! !"#$%&')*&.$+,!'/!0&$)*&+!1+*2!$3!,4$+!+,125& ! Primer AscF1 AscR1 AscF3 AscR3 AscF2new AscR5new AscR3new 18S1 18S2 27F 1492R 939F ! ! ! Sequence (5'-3') CTGGTTGATCCTGCCAG CACCTACGGRWACCTTG GATCCTGCCAGTAGTBATAT TGATCCTTCTGCAGGTTCA CAAGGAAGGCAGCAGGCGCGCAA AT GCGGTGTGTACAAAGGGCAGGGA AAGGAATTGACGGAAGGGCACCA CCAGGA CCTGGTTGATCCTGCCAG TAATGATCCATCTGCAGG AGAGTTTGATCMTGGCTCAG CGGTTACCTTGTTACGACTT TTGACGGGGGCCCGCACAAG Source Yokobori, 2006 (Yokobori et al 2006) Yokobori, 2006 (Yokobori et al 2006) Yokobori, 2006 (Yokobori et al 2006) Yokobori, 2006 (Yokobori et al 2006) This study This study This study Tsagkogeorga, 2009 (Tsagkogeorga 2009) Tsagkogeorga, 2009 (Tsagkogeorga 2009) Lane, 1991 (Lane 1991) Turner, 1999 (Turner et al 1999) RTL in house primer !! "#! ! ! !"#$%&'()&"#$%#&'#!'(%&)*!+&,!+-./+!,01#2*0)3!1+-%#*!(4!+*'0,0+&!50'2(60(5#*!5#+*%2#,!%*0&7! ,044#2#&)!,01#2*0)3!0&,0'#*8! ! Sample CA_Ascidia_sp_4-19 CA_Botryllus_sp_5-19 CA_Didemnum_sp_4-17 CA_Didemnum_sp_4-2 CA_Didemnum_sp_4-7 CA_Didemnum_sp_5-1 CA_Didemnum_sp_5-11 CA_Didemnum_sp_5-18 CA_Pyura_sp_4-21 CA_Styela_sp_5-16 FJ_Lissoclinum_patella_06037 FL_Didemnum_sp._tb3 FL_Ecteinascidia_turbinata_5 FL_Ecteinascidia_turbinata_tb2 FL_Eudistoma_sp_004 FL_Eudistoma_sp_tb1 FL_Trididemnum_sp_001BB FL_Trididemnum_sp_tb4 PNG_Didemnum_sp_11038 PNG_Didemnum_sp_11055 PNG_Didemnum_sp_11089 PNG_Lissoclinum_badium_11062 PNG_Lissoclinum_badium_11068 PNG_Lissoclinum_badium_11083 PNG_Lissoclinum_badium_11047 PNG_Lissoclinum_patella_11033 PNG_Lissoclinum_patella_11040 PNG_Lissoclinum_sp_7062 PNG_Lissoclinum_bistratum VN_Cystodytes_sp_08002 VN_Cystodytes_sp_08019 VN_Cystodytes_sp_08041 ! ! ! ! & Counts 2213 2406 3992 2959 2681 2478 3312 5068 2841 2037 1594 7454 10015 31660 13886 12822 1007 572 14155 3372 5585 3916 8993 20324 1899 12851 13778 5528 11015 2578 1610 2472 OTU 248 220 136 106 160 170 140 149 58 163 209 429 234 804 41 667 74 10 166 290 236 92 104 278 44 347 186 114 475 106 97 110 Shannon 6.19 5.68 3.27 2.87 4.09 4.86 3.59 3.15 2.54 4.97 4.35 5.62 3.60 4.56 0.78 5.27 3.72 0.61 5.09 6.79 5.38 4.20 3.24 2.40 3.25 3.08 2.55 4.39 3.68 4.51 4.49 4.19 Simpson 0.96 0.94 0.77 0.68 0.82 0.92 0.83 0.67 0.73 0.92 0.84 0.92 0.80 0.85 0.17 0.86 0.84 0.15 0.93 0.98 0.95 0.90 0.80 0.56 0.81 0.65 0.64 0.88 0.64 0.90 0.92 0.89 Chao1 255.40 229.67 156.53 117.12 168.22 182.00 167.04 152.89 60.50 169.88 393.53 446.73 268.55 836.25 50.00 699.72 78.57 10.25 178.55 300.44 241.70 98.18 108.62 284.36 48.50 368.72 195.60 121.00 779.59 109.00 100.62 114.50 ! "#! !"#$%&'(!"#$%&"%'()*(%"+,'%"-$."/(&0$.1%",/,23%*%"$-",%)*4*,/"+*).$5*$+(%" " " !"#$%&'"#& !"#$%&'())%*+,-#-.#/012 !"#3(4'5#-.#6071 !"#8&(,)5#-.#/019 :;#<=&,*>5-=*+*5#&4'?*>5&5#/ :;#<=&,*>5-=*+*5#&4'?*>5&5#&?7 !"#"-=*+*5#-.#6012 :;#<4+*-&%@5#-.#AA6 :;#<4+*-&%@5#-.#&?1 BC#!(-&%+(&,-#-.#ADA12 BC#!(-&%+(&,-#-.#ADAA7 BC#!(-&%+(&,-#-.#ADA61 3CE#;*--%=)*>4@#?5+*4@#11ADF 3CE#;*--%=)*>4@#?5+*4@#11A6G 3CE#;*--%=)*>4@#?5+*4@#11A97 3CE#;*--%=)*>4@#?5+*4@#11A9D :H#;*--%=)*>4@#.5&,))5#A9AFG 3CE#;*--%=)*>4@#.5&,))5#11AFF 3CE#;*--%=)*>4@#.5&,))5#11A6A 3CE#;*--%=)*>4@#I5',54#GA97 3CE#;*--%=)*>4@#?*-&'5&4@ :;#J'*+*+,@>4@#-.#AA1$$ :;#J'*+*+,@>4@#-.#&?6 :;#K*+,@>4@#-.#&?F 3CE#K*+,@>4@#-.#11AD2 !"#K*+,@>4@#-.#601G !"#K*+,@>4@#-.#607 !"#K*+,@>4@#-.#/011 !"#K*+,@>4@#-.#/01D 3CE#K*+,@>4@#-.#11AFD 3CE#K*+,@>4@#-.#11A// !"#K*+,@>4@#-.#/01 !"#K*+,@>4@#-.#60G #"$( ,-./0%%12 301/" !.0&%" 45.&67"25686"9.1/:67"." 45.&67"25686"9.1/:67"." ;25686" 41862.-#"92$9( 41862.-#"92$9) <02.-80.&292$ <02.-80.&292$ <02.-80.&292$ >622-5%671#9:"861# >622-5%671#9:"861# >622-5%671#9:"861# >622-5%671#9:"861# >622-5%671#9$".&%%" >622-5%671#9$".&%%" >622-5%671#9$".&%%" >622-5%671#92$ >622-5%671#9:62./".1# ?/6868G#92$9( ?/6868G#92$9) @68G#92$ @68G#92$ @68G#92$ @68G#92$ @68G#92$ @68G#92$ @68G#92$ @68G#92$ @68G#92$ @68G#92$ #"$) ,-./0%%12 301/" !.0&%" 45.&67"25686"9.1/:67"." 45.&67"25686"9.1/:67"." ;25686" 41862.-#"92$9( 41862.-#"92$9) <02.-80.&292$9=) <02.-80.&292$9=( <02.-80.&292$9=( >622-5%671#9:"861#9=( >622-5%671#9:"861#9=) >622-5%671#9:"861#9=) >622-5%671#9:"861#9=) >622-5%671#9$".&%%" >622-5%671#9$".&%%" >622-5%671#9$".&%%" >622-5%671#92$ >622-5%671#9:62./".1# ?/6868G#92$9( ?/6868G#92$9) @68G#92$9=( @68G#92$9=( @68G#92$9=( @68G#92$9=( @68G#92$9=( @68G#92$9=) @68G#92$9=) @68G#92$9=) @68G#92$9=) @68G#92$9=) #"$* ,-./0%%12 301/" !.0&%" 45.&67"25686"9.1/:67"." 45.&67"25686"9.1/:67"." ;25686" 41862.-#"92$9( 41862.-#"92$9) #"$+ ,-./0%%12 301/" !.0&%" 45.&67"25686"9.1/:67"." 45.&67"25686"9.1/:67"." ;25686" 41862.-#"92$9( 41862.-#"92$9) >622-5%671#9$".&%%" >622-5%671#9$".&%%" >622-5%671#9$".&%%" >622-5%671#92$ >622-5%671#9:62./".1# ?/6868G#92$9( ?/6868G#92$9) @68G#92$ @68G#92$ @68G#92$ @68G#92$ @68G#92$ @68G#92$ @68G#92$ @68G#92$ @68G#92$ @68G#92$ >622-5%671#9$".&%%" >622-5%671#9$".&%%" >622-5%671#9$".&%%" >622-5%671#92$ >622-5%671#9:62./".1# ?/6868G#92$9( ?/6868G#92$9) @68G#92$9=( @68G#92$9=( @68G#92$9=( @68G#92$9=( @68G#92$9=( @68G#92$9=) @68G#92$9=) @68G#92$9=) @68G#92$9=) @68G#92$9=) ABC-2.D 3I"%1& +EFF JEJ() )E(G JEJFF HEH( JEJJF )E+) JEJH AB>-5".6-7D 3I"%1& KE)) JEJJK KE)) JEJJK HE)G JEJ( HE)G JEJ( " " " " " " " " " " " " " " " " " & " ! ""! !"#$%&'(!""#$%&'"()"*+$,'-"./0,(%'*"12(3"456"/7"&(0$2/(7*"189:6"$7-"+(*2*"*3'0/'*"1;9<6"$7-" 2+'/,"3',0'72"$%=7-$70'!" " 8!":$&/)(,7/$"$*0/-/$7*" !"#$%&' 7,*)3)=>?@ 7,*)3)@@F 7,*)3)=GHI 7,*)3)==LH 7,*)3)G? 7,*)3)@HLG 7,*)3)GGK@ 7,*)3)GHLI 7,*)3)==I> ()*+,*+-+%./*,01, A08<,4/0%B;4)<,)608<,4/0%BC0990:4)<,)608<,4/0%B(D4)90</0.,+%B%B%B A08<,4/0%B;4)<,)608<,4/0%BC0990:4)<,)608<,4/0%B2.<,4)9)*070.,+%B2.<,4)9)*0708,0,%BC.08/,8).0%B A08<,4/0%BA08<,4)/7,<,+%BJ.03)608<,4//0%BJ.03)608<,4/0.,+%BJ.03)608<,4/08,0, A08<,4/0%B(M0*)608<,4/0%B(D.)4):.0+<%BN<409,*):/.,+%B%B%B A08<,4/0%BA08<,4)/7,<,+%BJ.03)608<,4//0%BJ.03)608<,4/0.,+%BJ.03)608<,4/08,0,%B%B A08<,4/0%B;4)<,)608<,4/0%B2.:D0:4)<,)608<,4/0%BOD/P)6/0.,+ A08<,4/0%B28/7)608<,4/0%BN30>?GK%BN30>?GK%B%B%B A08<,4/0%B;4)<,)608<,4/0 A08<,4/0%B;4)<,)608<,4/0%BC0990:4)<,)608<,4/0%B2.<,4)9)*070.,+%B"QF>%B%B 23,401,%5% 06-*70*8,% ;4,+,*<%/*% /*%+09:.,+ 5%+09:.,+ >E@F5 ?> >E@>5 F> >E@K5 F> >E>H5 F> >E>F5 K> >E>K5 K> >EG@5 K> >E>F5 K> >E>I5 K> " " >!"<&(,/-$"$*0/-/$7*" !"#$%&' 7,*)3)>?@ 7,*)3)>EE 7,*)3)@@?@ 7,*)3)@HHI 7,*)3)@KGL 7,*)3)@GHM 7,*)3)@LH@ 7,*)3)IIL? 7,*)3)IHI> ()*+,*+-+%./*,01, A08=,4/0%B<4)=,)608=,4/0%BC0990:4)=,)608=,4/0 A08=,4/0%B<4)=,)608=,4/0%BC0990:4)=,)608=,4/0%BF/64/)*0;,+%BF/64/)*08,0,%BF/64/) A08=,4/0%B<4)=,)608=,4/0%BC0990:4)=,)608=,4/0 A08=,4/0%B<4)=,)608=,4/0%B',;=0:4)=,)608=,4/0%B',+-;J)608=,40;,+%B',+-;J)6-;608,0,%B% A08=,4/0%B<4)=,)608=,4/0%BC0990:4)=,)608=,4/0 A08=,4/0 A08=,4/0%B<4)=,)608=,4/0%BC0990:4)=,)608=,4/0%B"8,0*)+:/4/;;0;,+%B./=)4/8);08,0,%B./=)4/8);0% A08=,4/0%B<4)=,)608=,4/0%BC0990:4)=,)608=,4/0%BN0*=O)9)*070;,+%B%B% A08=,4/0%B<4)=,)608=,4/0%BC0990:4)=,)608=,4/0 23,401,%5% 06-*70*8,% <4,+,*=%/*% /*%+09:;,+ 5%+09:;,+ ?D@>5 E> ?D@@5 E> ?D?G5 E> ?D?G5 E> >DH@5 E> ?DHG5 E> ?DHG5 E> ?D?I5 E> ?DMH5 E> " 23,401,%5% 06-*70*8,% A4,+,*?%/*% /*%+09:;,+ 5%+09:;,+ CDEC5 LK KD<G5 LK FDHK5 EE EDEE5 EE CDCL5 EE MDMK5 EE CDFC5 KG CDCK5 KG CDCM5 KG " " :!"?$3=$"@'A"B=/7'$"$*0/-/$7*" " !"#$%&' 7,*)3)<== 7,*)3)FEGH 7,*)3)KG=M 7,*)3)K=HK 7,*)3)KH=M 7,*)3)EMLK 7,*)3)KFH 7,*)3)K<E 7,*)3)EGK ()*+,*+-+%./*,01, >08?,4/0%@A4)?,)608?,4/0%@B0990:4)?,)608?,4/0%@2;?,4)9)*070;,+%@2;?,4)9)*0708,0,%@2;?,4)9)*0+%@ >08?,4/0%@(I0*)608?,4/0%@"+8/;;0?)4/):JI8/7,0,%@(J4))8)880;,+%@A4)8J;)408,0,%@A4)8J;)4)*%@ >08?,4/0%@A4)?,)608?,4/0%@2;:J0:4)?,)608?,4/0 >08?,4/0%@A4)?,)608?,4/0%@',;?0:4)?,)608?,4/0%@',+-;N)608?,40;,+%@O/?4)+:/*08,0,%@O/?4)+:/*0%@ >08?,4/0%@A4)?,)608?,4/0%@B0990:4)?,)608?,4/0%@"8,0*)+:/4/;;0;,+%@"8,0*)+:/4/;;08,0,%@P04/*)9)*0+%@ >08?,4/0%@A4)?,)608?,4/0%@',;?0:4)?,)608?,4/0%@Q30C<GE%@%@%@ >08?,4/0%@>08?,4)/7,?,+%@Q:J/*1)608?,4//0%@Q:J/*1)608?,4/0;,+%@RSJ/7*08,0,%@%@ >08?,4/0%@A4)?,)608?,4/0%@2;:J0:4)?,)608?,4/0%@T)47//9)*070;,+%@T)47//9)*0708,0,%@%@ >08?,4/0 " " ;!"!"##$%&"'()*+,-.&&,"1C,(=3"4D"</C=,'"E>6" !"#$%&' 7,*)3)>?@A 7,*)3)J?@K 7,*)3)JFHF 7,*)3)>M?A 7,*)3)IIM 7,*)3)IMJ 7,*)3)>KAA 7,*)3)JM> 7,*)3)MM@K " ()*+,*+-+%./*,01, B08=,4/0%C(D0*)608=,4/0%C"+8/;;0=)4/):ED8/7,0,%C(E4))8)880;,+%C<4)8E;)408,0,%C<4)8E;)4)*%C B08=,4/0%C<4)=,)608=,4/0%C2;:E0:4)=,)608=,4/0 B08=,4/0%C<4)=,)608=,4/0%CL0990:4)=,)608=,4/0 B08=,4/0%C(D0*)608=,4/0%C"+8/;;0=)4/):ED8/7,0,%C(E4))8)880;,+%C<4)8E;)408,0,%C<4)8E;)4)*%C B08=,4/0%C<4)=,)608=,4/0%C2;:E0:4)=,)608=,4/0 B08=,4/0%C<4)=,)608=,4/0%C2;:E0:4)=,)608=,4/0%CNE)7)+:/4/;;0;,+ B08=,4/0%C(D0*)608=,4/0%C(E;)4):;0+=%CO=409,*):/;,+%C%C%C B08=,4/0%C(D0*)608=,4/0%C"+8/;;0=)4/):ED8/7,0,%C(E4))8)880;,+%C<4)8E;)408,0,%C<4)8E;)4)*%C B08=,4/0%C(D0*)608=,4/0%C"+8/;;0=)4/):ED8/7,0,%C(E4))8)880;,+%C<4)8E;)408,0,%C<4)8E;)4)*%C " 23,401,%5% 06-*70*8,%/*% <4,+,*=%/*% +09:;,+ 5%+09:;,+ F@GHI5 >HH @GI@5 >HH @G>A5 >HH >HGA@5 >HH HG>F5 >HH JGAA5 >HH HG>J5 >HH HG>H5 >HH HGMF5 >HH " ! "#! ! "#!!"##$%&"'()*+,-"()!$%&'()!*+!,-%(&.!/01! !"#$%&' 7,*)3)=>?@ 7,*)3)=>HI 7,*)3)KFHI 7,*)3)@HI@ 7,*)3)@I> 7,*)3)K?II 7,*)3)K?=H 7,*)3)=@E> 7,*)3)@E@> ()*+,*+-+%./*,01, A08<,4/0%B;4)<,)608<,4/0%B',.<0:4)<,)608<,4/0%BC30DEF=%B%B%B A08<,4/0%B;4)<,)608<,4/0%B2.:J0:4)<,)608<,4/0 A08<,4/0%B(L0*)608<,4/0%B"+8/..0<)4/):JL8/7,0,%B(J4))8)880.,+%B;4)8J.)408,0,%B;4)8J.)4)*%B A08<,4/0%B;4)<,)608<,4/0%B',.<0:4)<,)608<,4/0%B',+-.M)608<,40.,+%BN/<4)+:/*08,0,%BN/<4)+:/*0%B A08<,4/0%B;4)<,)608<,4/0 A08<,4/0%B;4)<,)608<,4/0%B2.:J0:4)<,)608<,4/0%BOJ/P)6/0.,+%B;JL..)608<,4/08,0,%BQ,+)4J/P)6/-9%B A08<,4/0%B;4)<,)608<,4/0%BR0990:4)<,)608<,4/0 A08<,4/0 A08<,4/0%B;4)<,)608<,4/0%B2.:J0:4)<,)608<,4/0%BS)47//9)*070.,+%BS)47//9)*0708,0,%B%B ! ! ! ,#!."-/)'()*2)0!$%&'()!/+!,-%(&.!/01! !"#$%&' 7,*)3)=>?@ 7,*)3)=F?D 7,*)3)D@HI 7,*)3)>= 7,*)3)=HI> 7,*)3)DHI= 7,*)3)J=F@ 7,*)3)HIF 7,*)3)JNFI ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ()*+,*+-+%./*,01, A08<,4/0%B;4)<,)608<,4/0%B2.:C0:4)<,)608<,4/0 A08<,4/0%B;4)<,)608<,4/0%BG0990:4)<,)608<,4/0%B;+,-7)9)*070.,+%B;+,-7)9)*0708,0,%B;+,-7)9)*0+ A08<,4/0%B;4)<,)608<,4/0%B',.<0:4)<,)608<,4/0%BA7,..)3/64/)*0.,+%BA7,..)3/64/)*08,0,%BA7,..)3/64/)%B A08<,4/0%B;4)<,)608<,4/0%BG0990:4)<,)608<,4/0%B;+,-7)9)*070.,+%B;+,-7)9)*0708,0, A08<,4/0%B;4)<,)608<,4/0 A08<,4/0%B;4)<,)608<,4/0%B2.:C0:4)<,)608<,4/0%BKC/L)6/0.,+ A08<,4/0%B;4)<,)608<,4/0%BG0990:4)<,)608<,4/0%B"8,0*)+:/4/..0.,+%BM*7)L)/8/9)*08,0,%B%B A08<,4/0%B;4)<,)608<,4/0%BG0990:4)<,)608<,4/0 A08<,4/0%B;4)<,)608<,4/0%BG0990:4)<,)608<,4/0 23,401,%5% 06-*70*8,%/*% ;4,+,*<%/*%5% +09:.,+ +09:.,+ =>GHI5 IF ?GDE5 IF @GF@5 IF >G>>5 IF KGEK5 IF @G>E5 IF DGIE5 IF DGI@5 IF DGFD5 IF ! 23,401,%5% 06-*70*8,%/*% ;4,+,*<%/*%5% +09:.,+ +09:.,+ D?E>=5 DFF >E=F5 DFF DEDD5 DFF FEDI5 DFF FEDJ5 DFF FEFH5 DFF FEDN5 ?N FEF>5 ?N FED=5 ?N ! ! "#! !"#$%&'(!""#$%&'"()*+",-"./&"012%"+3453673+",-"35890:,*3+"9++,7)9*3;"<)*("9+7);)96+"" Ascidian sample Eudistoma sp. 004 Didemnum sp. 4-7 Didemnum sp. 4-17 Didemnum sp. 5-1 Didemnum sp. 5-11 Didemnum sp. 5-18 Cystodytes sp. 08002 Cystodytes sp. 08019 Cystodytes sp. 08041 Didemnum sp. 11038 Didemnum sp. 11055 Didemnum sp.11089 Eudistoma sp. TB1 Didemnum sp. TB3 Trididemnum sp. TB4 Symbiotic eukaryote Aiptasia mutabilis Prostheceraeus vittatus Aiptasia pulchella Grateloupia luxurians Clytia gracilis Clytia sp. AGC-200 Baeria nivea Selaginopsis cornigera Uncultured eukaryote clone SS1_E_02_11 Igernella notabilis Rhacostoma atlantica Melicertum octocostatum Lankesteria ascidiae Uncultured eukaryote clone SS1_E_02_1 Corallochytrium limacisporum Notodelphys prasina Cyclopina gracilis Symbiodinium sp. Notodelphys prasina Lankesteria ascidiae Uncultured eukaryote clone CC02A105.033 Cyclopina gracilis Doridicola agilis Cyclopina gracilis Anticomidae sp. JCC29 Melicertum octocostatum Dictyota sp. YMGTB002 Uncultured alveolate Calomicrolaimus parahonestus Sycettusa aff. hastifera OV-2012 Spongia tubulifera Grantessa sp. OV-2012 voucher GW979 Posidonia australis Spyridia filamentosa Attheyella crassa Biemna fistulosa Thelepus crispus Euplotes charon voucher JJM08102201 Genbank accession number FJ489438.1 AJ312272.1 AY297437.1 U33132.1 DQ068053.1 AF358074.1 AF182191.1 Z92899.1 EU050966.1 EU702420.1 EU305501.1 AY920757.1 JX187607.1 EU050966.1 L42528.1 JF781536.1 JF781537.1 AY165766.1 JF781536.1 JX187607.1 KF031747.1 JF781537.1 JF781541.1 F781537.1 HM564572.1 AY920757.1 AB087109.1 FN263032.1 AY854218.1 JQ272322.1 KC902150.1 JQ272312.1 GQ497582.1 EU718707.1 EU380307.1 FR819688.1 JN936473.1 JF694043.1 ! "#! !"#$%&'(!""#$%%&'(")*"+,-)./&'("%,0&1)230,+"&./"&++)-3&0,/"%3-')1,+"3."&+-3/3&.+"*')%"043+" /&0&"+,0"&./"',2&0,/"+5,-3,+!" " Sample Code Host Taxonomy Location CA_Ascidia_sp_4-19 California CA_Didemnum_sp_4-17 Ascidia ceratodes Bottrylus planus Didemnum sp. CA_Didemnum_sp_4-2 Didemnum sp. California CA_Didemnum_sp_4-7 Didemnum sp. California CA_Didemnum_sp_5-1 Didemnum sp. California CA_Didemnum_sp_5-11 Didemnum sp. California CA_Didemnum_sp_5-18 Didemnum molle Pyura haustor California CA_Styela_sp_5-16 Styela plicata California FL_Didemnum_sp_tb3 Didemnum sp. Florida FL_Ecteinascidia_turbinata_5 Ecteinascidia turbinata Florida ecteinascidins FL_Ecteinascidia_turbinata_t Ecteinascidia b2 turbinata Florida ecteinascidins CA_Botryllus_sp_5-19 CA_Pyura_sp_4-21 Major associated Major symbiont Previous microbiome analysis secondary characterized of related species metabolites California California California plicatamide FL_Eudistoma_sp_004 Eudistoma sp. Florida eudistomins FL_Eudistoma_sp_tb1 Eudistoma sp. Florida eudistomins FL_Trididemnum_sp_001BB Trididemnum Florida sp. FL_Trididemnum_sp_tb4 Trididemnum Florida sp. PNG_Didemnum_sp_11038 Didemnum sp. Papua New Guinea PNG_Didemnum_sp_11055 Didemnum sp. Papua New Guinea PNG_Didemnum_sp_11089 Didemnum sp. Papua New Guinea PNG_Lissoclinum_patella_11 Lissoclinum Papua New 033 patella Guinea cyanobactins, patellazoles PNG_Lissoclinum_patella_11 Lissoclinum 040 patella FJ_Lissoclinum_patella_0603 Lissoclinum 7 patella cyanobactins, patellazoles cyanobactins, patellazoles Papua New Guinea Fiji PNG_Lissoclinum_sp_7062 Lissoclinum sp. Papua New Guinea PNG_Lissoclinum_badium_1 Lissoclinum Papua New 1047 badium Guinea PNG_Lissoclinum_badium_1 Lissoclinum Papua New pyridoacridine alkaloids pyridoacridine Erwin et al, 2013 Ca. Rath 2011, Perez 2007, Moss Endolisoclinum 2003 frumentensis Ca. Rath 2011, Perez 2007, Moss Endolisoclinum 2003 frumentensis Prochloron spp., Schmidt 2005, Donia 2008, Ca. Kwan 2012, Behrendt 2012 Endolissoclinum faulkneri Prochloron spp., Schmidt 2005, Donia 2008, Kwan 2012, Behrendt 2012 Prochloron spp., Schmidt 2005, Donia 2008, Ca. Kwan 2012, Behrendt 2012 Endolissoclinum faulkneri Nitrospina sp. Erwin 2013 Nitrospina sp. Erwin 2013 ! "#! 1062 badium Guinea PNG_Lissoclinum_badium_1 Lissoclinum Papua New 1068 badium Guinea PNG_Lissoclinum_badium_1 Lissoclinum Papua New 1083 badium Guinea PNG_Lissoclinum_bistratum Lissoclinum Papua New bistratum Guinea VN_Cystodytes_sp_08002 Cystodytes sp. Vanuatu VN_Cystodytes_sp_08019 Cystodytes sp. Vanuatu VN_Cystodytes_sp_08041 Cystodytes sp. Vanuatu alkaloids pyridoacridine Nitrospina sp. alkaloids pyridoacridine Nitrospina sp. alkaloids cyanobactins, Prochloron spp., bistramides pyridoacridine alkaloids pyridoacridine alkaloids pyridoacridine alkaloids Erwin 2013 Erwin 2013 !! ! !"#$%&'("!!#$%$&'$&(%)!(*+,%-&'*.'!*/!%'(&0&%.!+&(-*1&*+2'!%(-*''!',2(&2'3!%.0!4&$5&.!$52!'%+2! ',2(&2'!%(-*''!)*(%$&*.'!%.0!*62-!$&+2!7%$!$52!'%+2!)*(%$&*.8"!952-2!4%'!%!'$%$&'$&(%)):! '&;.&/&(%.$!0&//2-2.(2!&.!!%'(&0&%.!+&(-*1&*+2'!%(-*''!',2(&2'3!%.0!4&$5&.!$52!'%+2!',2(&2'! %(-*''!)*(%$&*.'3!1<$!.*!0&//2-2.(2!&.!+&(-*1&*+2'!4&$5&.!%!',2(&2'!(*))2($20!%$!$4*!0&//2-2.$! $&+2!,*&.$'"!=.%):'2'!42-2!,2-/*-+20!*.):!*.!;-*<,'!4&$5!%!+&.&+<+!*/!$4*!'%+,)2'!<'&.;! $52!=>?#@A!+2$5*0!&.!$52!62;%.!B!,%(C%;27=>?#@A!B3!DE6%)<2F8!4&$5!FGG3GGG!,2-+<$%$&*.'"! =!'2.'&$&6&$:!%.%):'&'!4%'!,2-/*-+20!<'&.;!%!'2(*.0!,2-+<$%$&*.!=>?H=!+2$5*0!I%0*.&'I!%)'*! %6%&)%1)2!&.!62;%.!7BJ3!DE6%)<2J8"!!! ! ! ! !"#$%&'()*+,$'.-!"#$%&'! !""#$%&'%#()*#+',-)%## ..# ! !"#$%&'%( :! )"**+,-"&'%./0#"'%( ;! )"**+,-"&'%.102$--0( 9! 5# ! !""#!"#$%&'%# ! 5"$,"%-0&<-),=>&":! ?"%=@+*>="A! 8%",9:;-,%#!"#$%&'%# 2# ! B"%%-C+%%&CD=+>:! !$*=>E-C+%%&CD=+>:! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! /01!23-4/0123# - 567"%,&8/0//3# 4 9567"%,&9/0432# /0//3# /0642# /0/..# /0373# /0/.3# </0/=1# /0455# /0.1.# /01/.# - - - ! "#! References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`(&HBD*=BA@&/H(&3%-$X&)L(&a,-">@,%#&3(&4+,#X$=&'(&KBC$&[(&H1*+B#&N(& NB1_,=C&SI^(&N,->1*&a&5677M9:&!#&1D"$@,"&8E4&=/'!&D+C-B2,#C&BA&@1#%*$@,>&O$>,"&B#&J%T@1=,& $#"&>,*B#"$=C&>@=1*@1=,&JB",->:&@1>"AB(+"@.(+&%:& & 31=#,=&4(&^=C,=&L)(&)%$B&b^Z(&^$-J,=&IN&58MMM9:&Q#<,>@%2$@%#2&N,,D&^+C-B2,#,@%*& /,-$@%B#>+%D>&$JB#2&.C$#BO$*@,=%$&$#"&^-$>@%">&OC&4J$--&41O1#%@&=/'!&4,?1,#*,&!#$-C>%>8:&!" A3<$'8(%"1.*'(7.(+&)&#&V6PFVVE:& & [BXBOB=%&4(&L1=$O$C$>+%&!(&',%-$#&0!(&)$=1C$J$&3(&H%=B>,&S&5677G9:&)1-@%D-,&B=%2%#>&BA&@+,& $>*%"%$#F^=B*+-B=B#&>CJO%B>%>W&JB-,*1-$=&D+C-B2,#C&BA&D+B@B>CJO%B@%*&$#"&#B#F>CJO%B@%*& *B-B#%$-&$>*%"%$#>&%#A,==,"&A=BJ&8E4&=N'!&>,?1,#*,>:&1(+"&48+(C202%"AB(+&)(#&EF8M:& & & ! "#! ! ! ! ! ! Herdmania litoralis Botryllus CA 5.19 Pyura spinifera Megalocercus Styela sp CA 5.16 Herdmania litoralis Pyura haustor Botryllus CA 5.19 Pyura CA 4.21 Pyura sp spinifera Ecteinascidia turbinata Styela sp CA 5.16 Ecteinascidia Pyura haustor turbinata FL 005 Ecteinascidia turbinata FL TB2 Pyura sp CA 4.21 Phallusia fumigata Ecteinascidia turbinata Ascidia ceratodes CA 4.19 Ecteinascidia turbinata FL 005 Ecteinascidia turbinata FL TB2 Ascidia ceratodes Phallusia fumigata Aplidium pliciferum Ascidia ceratodes CA 4.19 Eudistoma sp FL 004 Ascidia ceratodes Eudistoma sp FL TB1 AplidiumEudistoma pliciferum gilboviride Eudistoma sp FL 004 Cystodytes sp VN 08019 Eudistoma sp FLCystodytes TB1 sp TG!2008 Eudistoma gilboviride Cystodytes sp VN 08041 Cystodytes Cystodytessp spVN VN08019 08002 Cystodytes sp TG!2008 Lissoclinum badium PNG 11083 Cystodytes sp VN 08041 Lissoclinum badium Cystodytes sp VN 08002Lissoclinum badium PNG 11062 Lissoclinum badium PNG 11083 Lissoclinum badium PNG 11047 Lissoclinum Lissoclinumbadium badium PNG 11068 LissoclinumLissoclinum bistratum badium PNG 11062 Lissoclinum Lissoclinum patella badium PNG 11047 badium PNG 11068 LissoclinumLissoclinum patella FJ 06037 bistratum PNG Lissoclinum voeltzkawi Lissoclinum patella PNG 11040 Lissoclinum patella FJ 06037 PNG 11033 Lissoclinum voeltzkawi PNG Didemnum sp DidSB Lissoclinum patella PNG 11040 Trididemnum sp FL 01BB Lissoclinum patella PNG 11033 Trididemnum paracyclops Didemnum DidSB Trididemnum sp FLspTB4 Trididemnum Didemnum sp FL TB3sp FL 01BB Trididemnum paracyclops Didemnum sp PNG 11089 Trididemnum Didemnum sp sp FL CATB4 4.17 Didemnum sp FL Didemnum sp CA 4.2TB3 Didemnum sp PNG 11089 CA 5.11 Didemnum sp CAsp4.17 Didemnum CA 5.18 Didemnum sp CA 4.2 molle CA 5.11 Didemnum sp PNG 11038 Didemnum sp CA 5.18 Didemnum sp PNG 11055 Didemnum Didemnum molle sp DidSA Didemnum 11038 Didemnumsp spPNG CA 5.1 Didemnum 11055 Didemnumsp spPNG CA 4!7 Didemnum sp DidSA Didemnum sp CA 5.1 Didemnum sp CA 4!7 01$#23#'()*+,%)- Megalocercus 0.01 !"#$%&#'()*+,%)- 0.01 .,$/'#'()*+,%)- Figure S1. The phylogenetic tree of ascidians using 18S rRNA gene sequences. Sequences obtained from this study are labeled with location (CA, FL, PNG, VN) and sample code. Other sequences were obtained from NCBI with the following accession numbers: Megalocercus huxleyi FM244868.1, Herdmania litoralis AB564300.1, Pyura haustor AY903926.1, Pyura spinifera JF961826.1, Ecteinascidia turbinata FM244848.1, Phallusia fumigata FM244844.1, Ascidia ceratodes L12378.2, Aplidium pliciferum AB211067.1, Eudistoma gilboviride AB211069.1, Cystodytes sp_TG-2008 FM244842.1, Lissoclinum badium AB211078.1, Lissoclinum bistratum AB211084.1, Lissoclinum patella AB211085.1, Didemnum sp DidSB AB211073.1, Trididemnum paracyclops AB211077.1, Didemnum molle AB211071.1, Didemnum sp DidSA/57 AB211072.1 5000 B A o__EC94 )"#&*+,-.%&%# !"#$%&'((%# denovo3155 denovo295 0+123(1+1.#4$"" /%"##!"$-&./)(%," o__Oscillatoriales denovo2642 o__Rhodospirillales denovo3959 denovo1567 denovo1357 o__Entomoplasmatales o__Chroococcales o__Oceanospirillales o__Rhizobiales o__Legionellales denovo1589 denovo932 /%!"#!"$%&'()*()+,+" denovo986 Rank4 Rank4 denovo2079 VN_Cystodytes_sp_08041 VN_Cystodytes_sp_08019 VN_Cystodytes_sp_08002 PNG_Lissoclinum_bistratum PNG_Lissoclinum_sp_07062 PNG_Lissoclinum_badium_11083 PNG_Lissoclinum_badium_11068 PNG_Lissoclinum_badium_11062 PNG_Lissoclinum_badium_11047 PNG_Lissoclinum_patella_11040 PNG_Lissoclinum_patella_11033 PNG_Didemnum_sp_11089 PNG_Didemnum_sp_11055 PNG_Didemnum_sp_11038 FL_Trididemnum_sp_TB4 FL_Trididemnum_sp_01BB FL_Eudistoma_sp_TB1 FL_Eudistoma_sp_004 FL_Ecteinascidia_turbinata_TB2 FL_Ecteinascidia_turbinata_005 FL_Didemnum_sp_TB3 FJ_Lissoclinum_patella_06037 CA_Styela_sp_5-16 CA_Pyura_sp_4-21 CA_Didemnum_sp_5-18 CA_Didemnum_sp_5-11 CA_Didemnum_sp_5-1 CA_Didemnum_sp_4-7 CA_Didemnum_sp_4-2 CA_Didemnum_sp_4-17 CA_Botryllus_sp_5-19 CA_Ascidia_sp_4-19 VN_Cystodytes_sp_08041 VN_Cystodytes_sp_08019 VN_Cystodytes_sp_08002 PNG_Lissoclinum_bistratum PNG_Lissoclinum_sp_07062 PNG_Lissoclinum_badium_11083 PNG_Lissoclinum_badium_11068 PNG_Lissoclinum_badium_11062 PNG_Lissoclinum_badium_11047 PNG_Lissoclinum_patella_11040 PNG_Lissoclinum_patella_11033 PNG_Didemnum_sp_11089 PNG_Didemnum_sp_11055 PNG_Didemnum_sp_11038 FL_Trididemnum_sp_TB4 FL_Trididemnum_sp_01BB FL_Eudistoma_sp_TB1 FL_Eudistoma_sp_004 FL_Ecteinascidia_turbinata_TB2 FL_Ecteinascidia_turbinata_005 FL_Didemnum_sp_TB3 FJ_Lissoclinum_patella_06037 CA_Styela_sp_5-16 CA_Pyura_sp_4-21 CA_Didemnum_sp_5-18 CA_Didemnum_sp_5-11 CA_Didemnum_sp_5-1 CA_Didemnum_sp_4-7 CA_Didemnum_sp_4-2 CA_Didemnum_sp_4-17 CA_Botryllus_sp_5-19 CA_Ascidia_sp_4-19 denovo2061 C 625 denovo2327 denovo3550 denovo1518 denovo3710 625 25 denovo1931 25 denovo2404 denovo1725 10000 abundance 10000 Abundance Abundance denovo2569 Unclassified Alteromonadales Chromatiales Enterobacteriales HOC36 HTCC2188 Legionellales Methylococcales Oceanospirillales Pasteurellales Pseudomonadales Thiotrichales Vibrionales Xanthomonadales 15000 Unclassified [Entotheonellales] Bdellovibrionales Delsulfobacterales FAC87 GMD14H09 Myxococcales NB1-j Spirobacillales Sva0853 Syntrophobacterales abundance "#! ! Figure S2. Abundance and order level composition of Deltaproteobacteria (A) and Gammaproteobacteria (B) showing ascidian species specificity of the most abundant groups. (C) Phylogeny of the top ten most abundant OTUs in L. patella and E. turbinata samples, arrows indicate known natural products producers. Bar plots and phylogenetic trees were generated using Phyloseq R package. ! 5000 Eturbinata_top10_microbes Lpatella_top10_microbes denovo832 ! ! "#! ! Figure S3. Effects of seasonal variation on the composition of ascidian microbiomes. The graph shows that the major phyla (filtered to top 20 percent) are maintained in ascidians collected from fall and spring of 2011 (x-axis), while the abundance of different species of bacteria varies. 10000 Alteromonadales 7500 Bdellovibrionales Abundance Chroococcales Entomoplasmatales Oceanospirillales 5000 Pseudomonadales Rhizobiales Rhodophyta Rhodospirillales 2500 Sva0725 Synechococcales ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! 0).)*#"1()2+ /*%()%&'"()*#'+ ,-'.%&'"()*#'+ !"#$%&'"()*#'+ 0).)*#"1()2+ /*%()%&'"()*#'+ ,-'.%&'"()*#'+ !"#$%&'"()*#'+ 0 ! "#! ! Figure S4. UniFrac-based 3D PCoA plots of all samples showing the clustering of samples according to (A) type of organisms (black: ascidians, red: mollusks) and (B) location (black: CA, green: FL, blue: PNG, light blue: VN, red: FJ). ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! A B ! "#! Figure S5. PCA biplots of (A) ascidian organic extracts showing correlation of samples from the same location (red arrows) with specific extract components (black print indicating retention time and m/z (B) resulting PCA biplot of assigned lipids after assignment of the LCMS data through Lipid Maps database showing the same pattern as A. A B ! ""! Figure S6. Example LC-MS/MS spectra of lipids that were correlated with extracts from different locations (PCA biplots are shown on the left panels). Compounds or compound families were identified by comparison of MS/MS fragmentation patterns and accurate masses to compounds in METLIN database (right panels). ! "#! ! "#! Figure S7A. PCA of samples from Papua New Guinea showing a clustering driven by secondary metabolites. O O O N H NH S N N O O HN NH HN O N O O O !"#$%&#!'$() OH OMe OH HO OH O O O N OOC S O )*#+,--#.%-,() N O N H S HN O )*"/01%#!/010$,)#-2#-%01() ! "#! Figure S7B. Spectrum and selected ion chromatogram of varamine A from Lissoclinum badium samples. -731$<$ -./$ !"#$!"#$%&'()$%%&'($ !"#$!"#$%&'()$%%&')$ !"#$!"#$%&'()#%%&)*$ !"#$!"#$%&'()#%%&+,$ 01213453$4-1$6-738$ ! "#! Figure S7C. Spectrum and selected ion chromatogram of varamine B from Lissoclinum badium samples. :!,(&$<$ !"#$ ./0$!"#$%&'()$11234$ ./0$!"#$%&'()$11235$ ./0$!"#$%&'()#11256$ ./0$!"#$%&'()#11278$ %&'&()*($)!&$+!,(-$ ! "#! Figure S7D. Spectrum and selected ion chromatogram of diplamine from Lissoclinum badium samples. \ &"'() &*+) 234)!"#$%&'())55678) 234)!"#$%&'())55679) 234)!"#$%&'()#5569:) 234)!"#$%&'()#556;<) ,(-('./').&()0&"'1) ! "#! Figure S7E. Spectrum and selected ion chromatogram of lissoclin B from Lissoclinum badium samples. '()( 567( *+,(!"#$%&'()(--./0( *+,(!"#$%&'()(--./1( *+,(!"#$%&'()#--.12( *+,(!"#$%&'()#--.34( 89:9';$'(;59(<5"'=( ! "#! Figure S7F. Spectrum and selected ion chromatogram of patellazole A from Lissoclinum patella samples. 7-38/$9$ +,-$ !"#$!"#$%&'((%#%&%'($ !"#$!"#$%&'((%#))%''$ !"#$!"#$%&'((%#))%*%$ ./0/1231$2+/$4+516$ ! "#! Figure S7G. Spectrum and selected ion chromatogram of patellin 3 from Lissoclinum patella samples. 0$'$ 123$ !"#$!"#$%&'((%#%&%'($ !"#$!"#$%&'((%#))%''$ !"#$!"#$%&'((%#))%*%$ 4-,-0560$51-$71/08$ ! "#! Figure S7H. Spectrum and selected ion chromatogram of cystodytin A from Cystodytes sp. samples. 23#4# 567# !"#!"#$%&"$'##$%&#'('')# !"#!"#$%&"$'##$%&#'('*+# !"#!"#$%&"$'##$%&#'(',*# 89/93203#259#:5;3<# ! "#! Figure S7I. Spectrum and selected ion chromatogram of cystodytin A from Cystodytes sp samples ()*+$,*+-(./. *01. 2(3(-4%-.4*(.5*+-6. ! ""! Figure S8. Bioactivity profile of ascidian extracts showing the distribution of bioactivity across different locations. Activities are shown as percent growth inhibition relative to standard compounds (see methods section). ! "#! Figure S9. Group diversity analysis using the Shannon and Simpson indices. Calculations were made using the mcpHill function in the simboot R package. This function reports multiplicity adjusted p-values from distribution free tests (see methods for details). ! !"#$%&"'()'*+,-.("/(0$+12-03 123'45, <=3':, !@A3'48, 4"0%()'*+,-.("/(0$+12-03 !"#$%&"$'#(#)*('B, +,&'-./-(#)*3'>, +,&'-./-(#)6C;3'B, 01$',.2#1,&,2($/34,.2$23';, 5,##%16,./-(42&,/-3'?, 5,##%16,./-()2$'6623'B, ! !"#$%&'()#**+*, !"#$%&'(-./0+*, 567789 5677:; 5678>? 56>54? 567795 56945? 465555 56:>:9 56559; 465555 56>7:5 569>:9 56777? 56>98> 5655:; 5679;? 569?9; 56:59> CHAPTER 3 METABOLIC ENGINEERING TO OPTIMIZE HETEROLOGOUS RIPP PRODUCTION IN E. COLI Ma. Diarey B. Tianero,1 Elizabeth Pierce,1 John A. McIntosh,1 Shrinivasan Raghuraman,2 Baldomero M. Olivera,2 Eric W. Schmidt1 1 Department of Medicinal Chemistry, L.S. Skaggs Pharmacy Institute, University of Utah, Salt Lake City, UT, USA, 2 Department of Biology, University of Utah, Salt Lake City, UT 84112, USA Note: My contribution to this study was in design of the project, heterologous expression experiments and analysis, interpretation of results, and writing of the manuscript. ! ! ! ! ! ! "#! Abstract The limited supply of natural products remains the single most important bottleneck in natural products drug discovery and development. Heterologous expression of biosynthetic pathways is often perceived as a ready solution to supply natural products from uncultivable microbes, but obtaining useful quantities of metabolites from heterologous expressions is far from straightforward, especially for complex, multistep pathways. Cyanobactins are among the few metabolites from ‘unculturable' bacteria that have been successfully expressed in heterologous hosts. Cyanobactins belong to the growing class of ribosomally synthesized and posttranslationally modified peptides (RiPPs). One notable feature of RiPPs is that in some cases, combinatorial libraries of active compounds can be generated via the straightforward mutagenesis of precursor peptides. Cyanobactins are exemplary in this respect, as a recent study demonstrated that >106 compounds could be produced in this fashion. However, supply issues persist even in heterologous expression experiments, with typical titers ranging from 10-100 µg/L, with the highest reported yield 300 µg/L. Here we present a greatly improved platform for the production of cyanobactins in E. coli, resulting in a 100-fold increase in titer. To showcase this method, we apply heterologously produced cyanobactins to a broad phenotypic assay on mice dorsal root ganglia (DRG) neurons, and discover a new activity for the known natural product, patellin 2. ! "#! Introduction During the past decade, natural products discovery has shifted from purely chemical isolation approaches to a hybrid of chemical and bioinformatic methods.1-3 As a result, the number of biosynthetic pathways for both known and unknown compounds has vastly increased. However, the ability to rapidly discover novel compounds has thrown an old problem back into focus, namely that of how we can supply sufficient quantities of natural products for practical drug discovery and development. This problem is particularly severe for metabolites discovered in rare animals, or for those metabolites predicted from genome sequences. 4 We have provided an important and instructive example of how this problem can be addressed in our work on the discovery, characterization, and heterologous expression of cyanobactin pathways.5-11 Cyanobactins are a broadly-distributed group of small (usually 6-10 amino acid) posttranslationally modified peptides with diverse bioactivities including anticancer,12,13 antimalarial ,14 and antiviral 15 activities. In RiPP biosynthesis, the amino acid sequence of the resulting natural product is encoded on a precursor peptide that, after ribosomal synthesis, is modified by enzymes to add chemical diversity, or to render peptide structures more drug-like. 16 The modifying enzymes generally exhibit relaxed substrate specificity, such that mutations in the precursor peptides are easily tolerated, leading to new compound derivatives.17,10,18 The biosynthetic steps in cyanobactin maturation have been characterized extensively. The promiscuity of the modifying enzymes in the cyanobactin pathways allows for the generation of libraries of derivatives, both natural6 19 and engineered. 6,10,20 Assessment of the capacity of a cyanobactin pathway to generate compound libraries in vivo projects ! "#! that a simultaneous five-site saturation mutagenesis library of one precursor sequence could produce 2 x 105 cyanobactin derivatives.20 The most studied cyanobactin pathways to date are derived from unculturable cyanobacterial symbionts of marine tunicates of the genus Prochloron.21,22 Given this obstacle to traditional fermentative approaches of natural product production and isolation, the alternative routes of chemical synthesis, heterologous expression, or in vitro synthesis are presently the only means of supplying cyanobactins. Considering that the medicinal appeal of cyanobactin pathways is tied to the possibility of creating cyanobactin libraries, heterologous expression is clearly preferable to chemical synthesis or in vitro reconstitution, as it allows the screening of huge libraries in easily transformable E. coli strains. 20 Accordingly, the cyanobactin pathways have been successfully manipulated in E. coli, and the potential of this heterologous system to produce novel derivatives has been demonstrated repeatedly. 23, 17,20 However, the yields are insufficient for practical purposes (drug screening, pharmacokinetics, animal testing etc.), as the best reported fermentation yield of cyanobactins is 300 µg/L. In this work, we describe a greatly improved platform for the production of cyanobactins in E. coli. We achieve this through a combination of both exogenous and endogenous supply of additives and precursors. Specifically, we find that supplementing cultures with L-cysteine and co-expressing a previously reported heterologous mevalonate pathway24 to increase isoprene availability lead to significant increases in compound production; these approaches combine synergistically to net a 100-fold increase in cyanobactin titer. We further demonstrate that mevalonate pathway coexpression improves the yield of both prenylated and unprenylated cyanobactins. We ! "#! showcase the usefulness of this expression system with the heterologous production and resulting discovery of the activity of the natural cyanobactin patellin 2 on dorsal root ganglia (DRG) neurons. Results Optimization of cyanobactin expression conditions In the course of our optimization of the production of cyanobactins in E. coli, we have identified many different parameters that affect compound production. These include the volume of the liquid cultures, fermentation time, temperature, as well as different additives including amino acids, redox reagents, and metals. Previous publications have explored various strategies (i.e., different cell lines, promoters, media) to optimize cyanobactin expression. 23, 20 We compared expressions from 6 mL, 10 mL, 100 mL, and 250 mL expressions. We also performed 1L fermentations in flasks and in a 1 L fermentor. We determined that 6 mL cultures in 24-well plates gave us the maximum amount of compounds per volume of expression. We also compared the number days of expression from 2 to 7 days and determined that the maximum amount of compounds was obtained from 5 day expressions. This experimental set-up also allowed us to study the expressions under many different conditions in replicate experiments reducing the preparation and extraction time as well as materials cost. We next examined how cyanobactin yield was affected by growth media composition as well as additives including metals, redox agents, amino acids and different cofactors (Table 3.1). We first found that a mixture of methionine and cysteine increased compound production. Further analysis with the individual amino acids showed ! "#! Table 3.1. Additives used in the initial optimization of production in E. coli C0D"'+--+E4&1&+%F+0664GH'&+#&'6+F%*+454G0"+%IG34J0G%5+%F+$)05%D0$G5+I*%6#$G%5+45+!"#$%&'### !""#$%&' ()*(&*+,!$)*' !"#$%&'(!")$'*%"+ "0$1%&'+ !")$'*%"+ 0345%+0$46&++ <'8==;++ ?#8==;++ @58==;++ >5?"A+ >'1(?)&++ <>B++ ,-,./(,-./+ ,-2,/+ ,-.,/+ 347+89:;+ 93>+ 93>+ 93>+ 93>+ 2,3>8347;+ 9,,5>+ that cysteine was responsible for the observed effect, and leads to a ~10-fold increase in compounds by LC/MS (Figure 3.1). We therefore added cysteine (5 mM) in all succeeding experiments. The mechanism of increase in compound production due to cysteine will be described elsewhere. Limited supply of isoprene precursors in E. coli We previously observed that production of cyanobactins in E. coli yielded a mixture of cyanobactins with either zero, one, or two isoprene units added to Ser and Thr residues.10 20 This is also apparent in Figure 3.1. We hypothesized that this was due to the limited availability of the necessary isoprene precursor, DMAPP in E. coli, and that increasing the intracellular DMAPP concentration would increase our yield of the desired, doubly prenylated cyanobactins. To increase supply of DMAPP, we coexpressed the second half of the mevalonate pathway (mev) from Saccharomyces cerevisiae that was previously engineered in E. coli for the production of amorphadiene, a precursor to the antimalarial drug artemisinin.24 When using only the latter half of the mevalonate pathway, mevalonate must be supplied exogenously, as E. coli does not !"#$%&'(!"#$%&'()'"*$"%+($,-(%.$"/&,01%.,$"1/,$"(00*.,$",2"%+)3'*$'"45678!"9':' );,<)"3;'"%+($,-(%.$"0'&*=(.=')"4%,6/,1$0"$16-'&8" ! "#! C" 1$/&'$+D(3'0"/(3'DD*$"@"4?8" )&*+'*%&*',-./)$0123"04&%0*5'24*01*%16' 6,$,/&'$+D(3'0"/(3'DD*$"@"4@8" 5" /(3'DD*$"@"4A8" 1$/&'$+D(3'0"/(3'DD*$"A"4B8" B" 6,$,/&'$+D(3'0"/(3'DD*$"A"458" /(3'DD*$"A"4C8" A" @" ?" >" 3&1",$D+" !"#"/(3;<(+",$D+" !"#""E"%+)3'*$'" 3&1"/(3;<(+"E"5"67"%+)3'*$'" Figure 3.1. Increase in cyanobactin production upon addition of cysteine (5mM). Labels indicate the cyanobactin derivatives (compound number). produce this metabolite. Our new expression system therefore involved the coexpression of tru with the mevalonate pathway (mbi) pathway. We supplemented these coexpressions with both mevalonate and cysteine (Figure 3.2). As anticipated, we observed an increased proportion of prenylated products in tru/mbi coexpressions (with added cysteine) when mevalonate was added (Figure 3.3). To further confirm that this was due to the addition of mevalonate, we fed 13C-labeled mevalonate to tru/mbi coexpressions. Mass analysis showed that the labeled 13C-labeled mevalonate was converted to DMAPP and was incorporated into the final compounds as evidenced by a +2 Da shift observable by LC-MS (Figure 3.4). This was also apparent in the masses of the fragment ions, showing the loss of one and two 13C labeled isoprene units (inset). This demonstrated that the mevalonate pathway can be functionally coexpressed with the ! "#! !"#$%&'(!"#$%&'()*$("$+",-)($&),*(")(."%/0)1$()2/"3)245)-6"+$7"24/"37$.8,*$("$+"+811-"37/(-1)2/." ,-)($&),*(6" HO HO N H O O H N N O !"#+3+ !"#+4+ N H O NH NH O N H OH H N O !"#+56+ !"#+57+ O O H N N O SH O O N H OH N S !"#+8+ !"#$%&'()*+,&!-.&/+ S HN OPP O HN HO OH HO 9:867+ N O O O O O N NH O NH O 9:8;+ <=46+ ()(+ '*0&12$&!*+,&!-.&/+ Figure 3.2. Expression system design using the combination of tru and mevalonate pathways for the production of fully prenylated cyanobactins !"#$%&'()'*+',-.%&/0&'"-'1%&-23/45-'3&6&30'"-'78&'1%&0&-.&'59':&6/35-/7&'1/78;/2'/-<' ! :&6/35-/7&'=>?:@+)'A+'B7%$.7$%&'59'.2/-5C/.4-0'1%5<$.&<'"-'!"#$%&')''' *' "#! S A' HN N R1O HN >??D' OR2 NH N 1 R1 R2= H NH O O O O O 2 R1 or R2= 3 R1 R2 = E?D' O R1O O N H NH S N N O ?D' 7%$'1/78;/2'F'E':@' 7%$':&6'1/78;/20'F' ()*+,-.'F' .207&"-&' E:@'.207&"-&'F'>?':@' ()*''F'.207&"-&'' :&6/35-/7&'' .207&"-&'F' O HN NH O HN N OR2 O O :&6/35-/7&'' 4 R1 R2= H 5 R1 or R2= 6 R1 R2 = Figure 3.3 Limited amount of isoprene precursors in E. coli. A) Increase in prenylation levels in the presence of mevalonate pathway in mbi plasmid and mevalonate (10 mM). B) Structure of cyanobactins produced in E. coli. tru pathway in vivo and can be used to increase the fraction of fully prenylated cyanobactins. Effect of mevalonate pathway coexpression on cyanobactin titer Although mbi-coexpression had the desired effect of increasing the fraction of fully prenylated compounds, we were somewhat concerned that metabolic stress due to coexpression of the mevalonate pathway might decrease overall cyanobactin titer. Thus, we were delighted to find that mbi-coexpression also resulted in a robust increase in the ! ""! !"#$%&'($&,-./01/023/-&/4&('5&67899&:-;/&.<2-/=2.3-&10>-<?&@0/A1BC&:-B>;&BD/EB&?/BB&/4&2&?2=>?>F&10>-<?& @0/A1&40/G&7H&402@G>-;23/-$&& !"#$#%%%& !"'$#%()& !"*$#)+)& !""$#%"%&&&&&& !"%$#)'"& m/z Figure 3.4. Incorporation of 13C-DMAPP into cyanobactin prenyl groups; inset shows loss of a labeled prenyl group from MS fragmentation. The peak at 943.5718 m/z shows the unlabeled patellin 3 and the 945.5777 m/z shows the incorporation of two 13CDMAPP in patellin 3. total amount of compounds. In the presence of mevalonate (20 mM) and cysteine (5 mM) (in tru/mbi coexpressions), the total amount of cyanobactins was increased consistently by about 100-fold compared to tru only expressions. To test whether the increased titers were due to the products of the mevalonate pathway (i.e., increased DMAPP levels), and not from background effects due to coexpression with the mbi plasmid, we added increasing amounts of mevalonate into the tru/mbi expressions from 1 to 40 mM. A clear mevalonate-dependent increase in the total amount of cyanobactins was observed (Figure 3.5F-I). In expressions with 40 mM mevalonate, the total amount of compounds was estimated at 27 mg/L, nearly 100-fold higher than in previous reports. Notably, this strong increase in titer was only observed in ! "#! &!" ,-)7;;89#=#>&?# patellin 3 3<9<,*790;-)7@#,-)7;;89#=#>2?# monoprenyl patellin 3 +9,*790;-)7@#,-)7;;89#=#>%?# unprenyl patellin 3 ,-)7;;89#$#>=?# patellin 2 3<9<,*790;-)7@#,-)7;;89#$#>$?# monoprenyl patellin 2 +9,*790;-)7@#,-)7;;89#$#>(?# unprenyl patellin 2 ($!"# !"#$%% %!" (!!"# &'()%(*'(%+,!&,-./#01% %#" !'%#" $#" !" tru-!F1 cys tru-!F1 cys-mev $!" '!"# #" &!"# !" A B C D E F G H I expression conditions %!"#Figure 3.5. Increased cyanobactin production in the presence of mevalonate pathway and mevalonate. Y-axis shows that amount of total compounds produced in each expression condition (columns) in mg/L. Column A) tru B) tru + 5mM cys, C) tru/mbi, D) tru/mbi + 5mM cys, E) tru/mbi +10mM mevalonate, F) tru/mbi +5 mM cys + 5 mM mevalonate, $!"# G) tru/mbi + 5 mM cys + 10 mM mevalonate, H) tru/mbi + 5 mM cysteine + 20 mM mevalonate, I) tru/mbi + 5 mM cysteine + 40 mM mevalonate. Inset: comparison of tru and truDF1 knockout expression showing the relative amounts of unprenylated patellin 3 !"#the presence of both cysteine and mevalonate in tru/mbi expressions. Cysteine by itself )*+#37:#,-)./-06#1#234#506)7897#1#(!#34# )*+#,-)./-0#1#2#34#506)7897# consistently increased compound production, but to a much lesser extent (Figure 3.5B). 37:-;<9-)7## Similarly, mevalonate pathway coexpression in the absence of cysteine gave only small increases in cyanobactin titer (Figure 3.5E). The observation of an epistatic relationship between cys-supplementation and mbi-coexpression (i.e., the fact that mbi-coexpression only improved titer in the presence ! "#! of added cysteine) suggested to us that these two interventions might exert their effects on different stages of biosynthesis, with Cys-supplementation affecting earlier steps and mbi-coexpression affecting later steps of biosynthesis. Such a proposal would be consistent with the accepted order of biosynthesis, in which prenyltransfer is believed to be the final step in biosynthesis. 9 Complicating this straightforward picture is the observation that all compounds, including unprenylated derivatives, are present at increased levels when the mevalonate pathway and mevalonate were added, indicating that the increases in cyanobactin titer cannot be explained solely by the increased availability of isoprene donors such as DMAPP. To further probe the role of the mevalonate pathway, we knocked out the functional pathway prenyltransferase, truF1. Consistent with the above results, tru !F1, showed mevalonate-dependent increases in cyanobactin expression, giving rise to only the unprenylated compounds (Figure 3.5, inset). Certain families of cyanobactins are not modified by isoprene. For example, in the patellamide pathway, pat, threonine and serine residues are heterocylized to yield the corresponding oxazoline derivatives, and as a result are not subject to prenylation; moreover, the prenyltransferase homolog in the pat pathway, PatF, is apparently inactive as a prenyltransferase in vitro. Given the activating effect of the mevalonate pathway with the tru !F1 pathway, we speculated that mbi-coexpression would similarly increase production of patellamides (as observed for unprenylated patellins). Surprisingly, we found that this was indeed the case: patellamide production was increased in proportion to the concentration of added mevalonate, despite the complete lack of prenylation in pat ! "#! pathway products (Figure 3.6). In the presence of cysteine, pat production is minimal and often undetectable and is increased to robustly detectable levels upon mbi-coexpression. The above results suggest that the beneficial effect of mevalonate pathway O expression on cyanobactin production may not beO entirelyNexertedSthrough the pathway H N N itself. Rather, its effect may be at least partly indirect, acting through other metabolic O NH 5/67&"%8:";<=>53?5"8<"2>@9A=B@<"@C"23456637895?"8<"4D5"2>5?5<=5"@C"75E36@<345"234DF3G"3<9"5H45><3 HN O pathways that may benefit from extra isoprenoid precursors. Channeling the precursors N 75E36@<345:"IJ3H8?"?D@F?"4D5">3B@"@C"=@72@A<9"253K"3>53"4@"4D5"8<45><36"?43<93>9:".@6A7<",L"!"#M" H N such as GPP and FPP, into these pathways would conceivably require longer precursors S '7O"=G?458<5M".L"!"#$%&'"N"'7O"=G?458<5M"0L"!"#$%&'"N"'"7O"=G?458<5"N"'"7O"75E36@<345M"1L"!"#$% N O 7O"=G?458<5"N"#!"7O"75E36@<345:"" 25 which would require both DMAPP and IPP for elongation. O #!" O O O +" N H N *" S N N !"#$%#&"#%'()*+,% N O HN O H N S S NH HN )" N H N O NH O N N O (" '" O S O )" *" H N N O O &" O %" O N $" 23456637895"." N H S N O NH #" 23456637895"," HN O N !" S H N N ," -" O O ." /" "-!&"../01%2013/401.% 0" 1" Figure 3.6. Increase in production of patellamides in the presence of mevalonate pathway and external mevalonate. Y-axis shows the ratio of compound peak area to the internal standard. Column A) pat, B) pat + 5mM cysteine, C) pat/mbi + 5mM cysteine, D) pat/mbi + 5mM mevalonate, E) pat/mbi + 5 mM cysteine + 5 mM mevalonate, F) pat/mbi + 5 mM cysteine + 10 mM mevalonate. ! "#! To test whether different isoprenoid species would exhibit defined effects in compound production, we co-expressed tru with three different versions of the mevalonate pathway: mevb, mbi, mbis, which convert mevalonate to IPP, DMAPP, and FPP, respectively. mevb lacks the isopentyl diphosphate isomerase, idi, for the conversion of IPP to DMAPP. We verified that the different expression strains contained these pathways by restriction digest of the overnight seed cultures. All three sets of mevalonate pathway enzymes, in the presence of cysteine and mevalonate (20 mM), increased compound production relative to controls (i.e, tru pathway only and empty plasmid pBBR) (Figure 3.7). Tru/mevb expressions yielded higher proportions of monoprenylated patellin 2, conceivably a result of the lack of idi to supply DMAPP for prenylation. The increase in tru/mbis might indicate that longer isoprenoid precursors could in part be responsible for the unexpected effects of the mevalonate pathway on cyanobactin expressions. Effects of the mevalonate pathway coexpression on growth To determine how the additives cysteine and mevalonate affect growth, we measured the growth (OD600) of E. coli throughout the 5-day expression. Tru only and tru/pBBR expressions in the presence of cysteine, mevalonate, or both displayed similar growth patterns with a sharp log phase starting at 10 to 15 hours (Figure 3.8A) post inoculation. Without any additives, tru/mevb, tru/mbi, and tru/mbis expressions entered log phase at a similar time as controls (A and B, Figure 3.8) but reached a lower OD600 in stationary phase (Figure 3.8C-E), perhaps reflecting the burden of expressing additional enzymes. Cysteine in general did not affect growth measurably, nor did mevalonate. However, in the presence of mevalonate pathway enzymes and both additives (Figure ! "#! Figure 3.7. Increase in production of patellin 2 in the presence of mevalonate pathway variants. Y-axis shows the amount of compounds in mg/L . Expression conditions : A) tru B) tru/pBBR C) tru/ mevb D) tru/mbi E) tru/mbis. Columns 1-5, no additives; columns 6-10, with 5 mM cysteine; columns 11-15, with 20 mM mevalonate; columns 16-20, with 5 mM cysteine and 20 mM mevalonate. 3.8C-E), growth was clearly altered, characterized by a long lag phase after inoculation. Activity of patellin 2 on DRG neurons With a method to supply meaningful quantities of cyanobactins in hand, we sought next to study their bioactivity. While many cyanobactins have been reported to be bioactive, the range of reported activities is limited to relatively standard cytotoxicity and ! "#! ! "#! antibiotic assays.26 Their ecological function within ascidians or free-living cyanobacteria remains unclear, 26 though a variety of hypotheses have been put forward, ranging from settlement inhibitors of competing organisms to antipredation defense.27 To expand the range of potential pharmacological applications of cyanobactins, we chose to screen natural cyanobactins in a high-content phenotypic screen on mice dorsal root ganglion (DRG) neurons, which, given the breadth of the assay, has been described as a ‘constellation pharmacology' approach.28,29 In particular, the assay measures calcium influx into individual DRG neurons in a heterogenous population using a fluorescent probe, and can distinguish a broad variety of responses to compounds, including ionchannel agonists, antagonists, and a diverse array of other receptor ligands.28 The goal was first to identify natural cyanobactin leads that exhibited neuroactivity on a subset of the heterogenous DRG population, and from there to use our heterologous system to supply compounds for follow-up studies into molecular mechanisms of action, and structure-activity relationships. In our initial screen, we tested compounds 2, 3, 6, patellamide C and a mixture (7 and 8), ulithiacyclamide12 (9), lissoclinamide 330 (10), lissoclinamide 431 (11), and trunkamide 32 (12) at 10 µM concentrations (Figure 3.9). We observed clear activity for compounds 3 and 9. Specifically, application of 3 caused changes in the [Ca2+]i shown through enhanced response to KCl (25 mM) in some cells as well as direct perturbation of the baseline in others (Figure 3.10A). Compound 9 similarly enhanced the KClinduced [Ca2+]i upon application. Interestingly, compound 2 was inactive in this assay despite differing from 3 by only a single prenyl group. We chose to pursue 3, as its production was already optimized in our heterologous expression system. ! "#! !"#$%&'(!"#$%&'$&%()"*+"',-.*/-'0.)")'%((.(1"2."$3("456"-))-,!"7*89*&.1)":(%("-))-,(1"-$";<"µ=" '*.'(.$%-0*.)!">?@?-.1"A"-%(")3*:."2."B2C&%(";7!" O O N H O O S S O HN O N N N NH NH O S S N H N O N N HN S S N O H N N O O O ("DE2$32-','E-821(" O O O S N H O )*"F2))*'E2.-821("@" N O HN O O N O N O HN N NH N H N S N H N S O ))"F2))*'E2.-821("G" O HN H N O O O )+"H%&.I-821("J" Figure 3.9. Structures of cyanobactins screened in the DRG assay. Compounds were assayed at 10 µM concentrations. 2,3,and 6 are shown in Figure 3.3B. To rule out the possibility of contamination by another extract component, we tested 3 purified from both the coral reef animal Lissoclinum patella and from recombinant E. coli expression. We observed that both were active and showed the same [Ca2+]i phenotypic response in the same cells (Figure 3.10C). Further dose-response experiments separated the cells into subpopulations according to their sensitivity to 3: in group A were cells that were dose-responsive showing an indirect and increasingly enhanced [Ca2+]i signal from 3 µM 3, in B were cells that were only affected at a high dose (30 µM) with indirect amplification of [Ca2+]i signal, and in C were cells that were directly affected by a firing response at 30 µM (Figure 3.10D). The cells were also challenged with different pharmacological agents: menthol (cold, TRPM8), ATP !! ! ! "#! ! +$! (nociception, P2X), and capsaicin (heat, TRPV1). Table 3.2 shows the percent of cells, as classified by their response to 3 above, that responded to the application of the pharmacological agents. Cells from subpopulations B and C were highly correlated with TRPV1, indicating that 3 could affect nociceptors. The different responses in the subpopulations further suggests that 3 may act on multiple cell types that differ in the expression levels of its molecular target.! Discussion Natural products are indispensible sources of drugs given their complex scaffolds and often potent activities against novel biological targets.33 Most natural products, however, are left undeveloped because we lack a means to supply them. The engineering of natural product biosynthetic pathways into heterologous hosts has the potential to address these challenges by improving the supply of natural products, and also allowing the production of derivatives through simple genetic engineering.34 However, many Table 3.2. Summary of calcium imaging experiments with 3. Values are presented as percentage of cells that exhibited a type of response (A-C) that are sensitive to pharmacological agents (columns 2-4). Response types to 3 are classified as: A, dose responsive from 3 mM to 30 mM; B, indirect through amplification of KCl response; C, direct effects on the baseline [Ca2+]i. ! ""! heterologous expression experiments are conducted primarily for the purposes of verifying bioinformatics predictions of compounds, and thus are unlikely to result in consistent and robust supply of compounds. Thus, the continued development of methods to improve yields in heterologous expression is essential to ease this bottleneck in the search for medicinal natural products. Here we have described a method for the improvement of the heterologous production of cyanobactins, resulting in a stable, 100-fold boost in titer. This increase was brought about through a combination of media optimization and improving the supply of endogenous precursors through the coexpression of an independent biosynthetic pathway. In particular, we have demonstrated that compound production is synergistically increased by the supplementation of the amino acid cysteine and by isoprene precursors supplied by the mevalonate pathway. The effect of mevalonate pathway expression on cyanobactin yield is particularly interesting as we initially intended to increase DMAPP levels to support complete prenylation of cyanobactins. Although we observed a modest shift in the ratio of prenylated/unprenylated compounds upon mevalonate pathway coexpression, a greater effect was observed in the increase of the total amounts of compounds, including those that were unprenylated. We found that this effect depends on the concentration of the added mevalonate. The increase in the amount of all the derivatives indicates that the increase in production is not directly related to compound prenylation. Indeed, knocking out the tru pathway prenyltransferase, truF1, showed that the levels of unprenylated compounds were still increased by mevalonate in a dose-dependent manner. Further, we have shown ! "#! a mevalonate-dependent increase in the production of the patellamides, which do not contain prenyl groups. Thus, it is likely that the effect of mev-coexpression is exerted through multiple channels, some of which act indirectly on the pathway enzymes. For example, excess isoprenoids could be channeled into pathways that in turn affect growth and fitness of the cells under conditions of compound production that are otherwise stressful. Indeed, through coexpressions with tru/mbis, we showed that channeling the isoprene intermediates into longer units also maintained compound production. Given the growing importance of tools such as the mev pathway in metabolic engineering efforts, our results highlight the often unpredictable and nonadditive effects that can result. To this end, we are currently investigating the metabolome of E. coli under these different conditions with the goal of identifying the pathways and metabolites that change in response to mev coexpression. One other intriguing feature of our results is that the beneficial effects of mev-coexpression are dependent on the presence of cysteine. From our quantifications of many different experimental sets, we determined these effects are not merely additive. Cysteine by itself consistently increases compound production, while mevalonate exerts a significant effect only in the presence of added cysteine. Altogether, the unanticipated effects of the mevalonate pathway on cyanobactin production opens new possibilities for heterologous production of RiPPs in E. coli, as even unprenylated cyanobactins were increased upon mev coexpression. To our knowledge, this is the first demonstration of such an ‘off-target' effect of mev pathway coexpression, in which the titer of non-terpenoid natural products is observed to increase. Serendipity aside, this study also emphasizes the combination of both empirical (the ! "#! discovery of cysteine as an additive) and rational (supply of isoprene for full prenylation) approaches in the optimization of cyanobactin production. Having established a high-yielding expression system of cyanobactins, we sought to readdress one of the fundamental goals of natural product discovery and engineering, that of discovering novel bioactivities. Thus, we applied several cyanobactins to a broad phenotypic assay, and uncovered the activity of patellin 2 in provoking action potentials in DRG neurons. Patellin 2 was first isolated and characterized from L. patella ascidian in 199035; however, no biological activity has been reported on this compound until now. Although the neuronal subtype specificity and molecular target of patellin 2 is the subject of ongoing work, based on comparison of both expressed and natural compounds, we were able to establish that patellin 2 is active affecting different types of nociceptors. This represents the first DRG activity reported for a cyanobactin. These results emphasize the importance of broad pharmacological screens since the native targets for the vast majority of natural products are unknown, and further underline the potential of small, highly modified cyclic peptides as neuroactive agents. In summary, we have established a method for the highest titer of cyanobactins so far. We have discovered, serendipitously, the positive effects of mevalonate pathway expressions to cyanobactin production. This method is the first step in removing the bottleneck for the development of cyanobactins as pharmacological and therapeutic agents, and for further exploration of their ecological and functional roles. ! "#! Materials and methods Plasmids and strains Plasmid pTru-SD contained the tru operon in Topo vector. tru contained the precursor peptide gene TruE2, which encodes for patellin 2 (3) and patellin 3 (6) in the core peptides. pMEVB, pMBI, and pMBIS, containing the second half of the mevalonate pathway were obtained from Addgene (plasmids 17819, 17816, and 17817, respectively). pMEVB contains ERG12, ERG8, MVD1; pMBI has the pMEVB genes but with idi appended after MVD1; pMBIS is pMBI with additional ispa. 24 Empty plasmid pBBR with a modified multiple cloning site was constructed from pMBI. pMBI was digested with KpnI and SacI. The vector backbone piece was ligated with a small piece of DNA constructed from overlapping primers to create an expanded multiple cloning site (forward: 5' AGTGTACAGGGCCCCCCCTCGAGGGTATCGATAAGC TTGATATCGAATTCCTGCAGTAGGAGGAATTAACCCATATGTC, reverse: GATGAGCT CCACCGCGGTGGCGGCCGCTCTAGAACTAGTGGATCCCCCGG GTACCATGGACATATGGGTTAATTCCTC). pPat (pat) contains the patellamide pathway genes that were codon-optimized for E. coli. pat is in Topo vector. For the truF1 knockout experiments, modified tru expression plasmid was made from fragments of the previously described vector,17 containing truA-D and truG and gBlocks containing truE, F1, and F2 (IDT, Coralville, IO). The gene sequences in the gBlocks were codonoptimized for expression in E. coli. This plasmid was assembled by recombination in Saccharomyces cerevisiae BJ4741 using previously described methods. 17 truDF1 was made by Gibson assembly 36 of PCR products containing most of the modified tru plasmid in two pieces. The first piece consisted of half of the plasmid backbone and the ! "#! first part of the tru pathway (truA-E ). The second piece contained the end of the tru pathway (truF2 and truG) and the other half of the plasmid backbone. All vectors were partially sequenced to confirm that the pieces had been ligated correctly. All pathways were under the control of lac promoter and expressed constitutively in DH10B cells. Chemicals and other materials L-cysteine hydrochloride was obtained from Amresco (Solon, OH). Ampicillin and tetracycline were obtained from Sigma-Aldrich (St. Louis, MO). All solvents used for silica open column chromatography, HPLC, and HPLC/MS analyses were obtained from Fisher Scientific (Pittsburg, PA). Lissoclinum patella samples that were used for isolation of standards 3 and 6 were collected in Papua New Guinea with proper collection permits. Fermentation conditions and compound expressions Plasmid pTru-SD1 (tru) was transformed alone or co-transformed with pBBR, pMEVB (mevb), pMBI (mbi), or pMBIS (mbis) (Martin 2003). The resulting clones were then inoculated into liquid 2xYT broth (6 mL) in 24-well plates with the addition of appropriate amount of antibiotics (ampicillin: 50 mg/mL, tetracycline: 5 mg/mL), and grown overnight. Growing seed cultures (0.2% v/v) were then inoculated into fresh media containing antibiotics described above. Cysteine (5 mM) was added upon inoculation of expression cultures. Mevalonolactone (Sigma-Aldrich, St. Louis, MO) was hydrolyzed to mevalonate according published methods (Martin 2003) and was also added to the expression cultures at inoculation (5, 10, 20, 40 mM final concentrations). The cultures ! "#! were allowed to shake at 150 rpm, 30 °C for 5 days after which the cells were harvested by centrifugation at 4000 rpm, washed with saline solution (100 mM NaCl), and extracted with acetone (4 mL). The acetone extracts were then dried to yield the organic extracts for each 6 mL culture. In labeling experiments, 13C mevalonate (1 mM) (SigmaAldrich, St. Louis, MO) was added at the beginning of the expressions as described above. To express patellamides, pat was co-transformed pMBI and processed as described above. Isolation of patellin 2 from E. coli expressions Overnight seed cultures of E. coli harboring tru/mbi were grown under appropriate antibiotic selection and inoculated into 1L 2xYT broth with antibiotics, cysteine (5 mM), and mevalonate (10 mM). The broth was distributed into 24 well plates containing 6 mL per well and grown for 5 days at 30 °C while shaking at 150 rpm. The cultures were harvested as described above, and the cells were extracted repeatedly with acetone. The combined acetone fractions were dried and fractionated by silica open column chromatography using previously described gradient and methods .10 Fractions were analyzed by HPLC-MS for the presence of cyanobactins, and the compound containing fractions were subjected to reverse phase HPLC fractionation on an Eclipse XDB, 9.4 x 250 mm, 5m C18 column (Agilent Technologies, Santa Clara, CA) using the following solvent gradient: 5% B to 100% B (0-30 minutes), 100% B (30 to 40 minutes), and 100% to 5% B (40 to 45 mins); Solvent B consisted of acetonitrile and solvent A consisted of water. A flow rate of 2.5 mL min-1 was used. Compound 3 was purified on a Luna, 4.6 x 250 mm, 5 !m C18 Column (Phenomenex, Torrance, CA) using the solvent ! "#! gradient: 80% B to 94% B (0-15 minutes), 94% B (15-17 minutes), and 94% to 80% B (17-20 minutes), at a flow rate of 1ml min-1; Solvent B consisted of acetonitrile and solvent A consisted of water. Quantification by proton NMR 1 H NMR was used to quantify purified patellin 2. This information was used as the basis for the HPLC/MS calibration curve as well as for bioactivity assays. An external standard curve was made using the different amounts of 1,4-dinitrobenzene (SigmaAldrich, St. Louis, MO) using described methods. Briefly, 2-fold dilutions (1.2mM to 79.3 mM) of 1,4-DNB in DMSO-d6 (Cambridge Isotope Laboratories Inc., Tewksbury, MA) were prepared in uniform 3 mM Kontes precision NMR tubes (Sigma- Aldrich, St. Louis, MO). 1H NMR for each DNB sample was acquired on a Varian Inova 500 instrument (Agilent Technologies, Santa Clara, CA) with 32 scans using an 18 s relaxation delay (d1) as calculated from T1 using signal inversion. The integrals of the four DNB protons were obtained and used as the Y-values for the calibration curve (Figure 3.11). Patellin 2 was prepared using the same DMSO-d6. The 1H spectrum of patellin 2 was obtained using 32 scans and 4 s d1, which was determined as described above. Signals from a methyl protons of the valine residue were integrated and the values were compared with the external standards. Quantification of cyanobactins from different expressions using HPLC/MS Extracts were dissolved in methanol (500 mL) and analyzed by HPLC-ESI-MS using a Waters Micromass ZQ mass spectrometer. Samples were injected at uniform 3.45"#%67'"$3"45-"6789:;7<=>"6?;@A"?B9>C"$D2EF9>9G;=:A>HA>A'" ! "#! 1" $'0" $'," $'2" !"#$%% $'1" $" ('0" ('," ('2" ('1" !"#"$%&'()*"+"('(()," -."#"('///0$" (" (" ('((1" ('((2" ('((," ('((0" ('($" &&'(%)*+,-./.0"'1#/2#/#% ('($1" ('($2" Figure 3.11. 1H NMR calibration curve using 1,4-Dinitrobenzene. volumes (40 mL). A synthetic derivative of a heptamer cyclic peptide (7, Figure 3.12B) was used as an internal standard in all HPLC/MS analyses. The extracts were analyzed on a Zorbax Eclipse Plus 4.6 x 150 mm, 5 !m, C18 column (Agilent Technologies, Inc., Santa Clara, CA) using the following solvent gradient: 10% B (0-2 minutes), 10% B to 100% B (2-20 minutes), 100% B (21 to 30 minutes), and 100% to 10% B (30 to 35 mins). Solvent B consisted of acetonitrile with 0.05% (v/v) formic acid; solvent A consisted of water with 0.05% (v/v) formic acid. The peaks of interest were selected from the chromatograms and the resulting peak areas for each compound in the extract were obtained. The ratio of the peak area of the compound of interest to the peak area of the internal standard was then taken in each sample and the amount of each compound was calculated using the calibration curve that was generated using different concentrations of !"#$%&'()!"#$"%&'()"*+,-./+012"*3/45"16"7+85,,-2"9!":;+<-="/57/5=528="8>5"/+01"16"8>5"75+?"+/5+"16"@-A5/528" +B1328="CD;+<-=$"16"7+85,,-2"9"81"8>5"-285/2+,"=8+2@+/@"C7+25,"E$!"E$"=8/3*83/5"16"FG"-285/2+,"=8+2@+/@!" ! "#! E" I5+?"+/5+"7+85,,-2"9'-285/2+,"=8+2@+/@"CFG$" #" O O N H O OH HN O N O HN NH O NH HN H N O O O &1B7132@"*"J" FG" K285/2+,"=8+2@+/@" µH"7+85,,-2"9" Figure 3.12. A) LC/MS calibration curve of patellin 2. Y-axis represents the ratio of the peak area of different amounts (X-axis) of patellin 2 to the internal standard (panel B). B) structure of E8 internal standard. purified patellin 2 (Figure 3.12a). All expressions were done in triplicate. Around 70 independent expression experiments (average of 8 different conditions in each set) were done over the course of optimization, confirming the reproducibility of the results. Growth experiments Samples were grown in 6 mL volumes, in quadruplicates sets, in 24-well plates as described above. Aliqouts (100 mL) were drawn and transferred to a 96-well plate containing 2xYT (100 mL) and the OD was measured on a microplate reader. OD600 was measured every 3-4 hours for 120 hours. The remaining cultures were extracted for compound production and analyzed for relative abundance of cyanobactins between each condition. ! "#! Calcium imaging on mouse dorsal root ganglion (DRG) cells The DRG assay has been described previously.37 28 Briefly, the cells in their growth media were loaded with Fura-2-AM for 1 h at 37° C and incubated further for 30 minutes at room temperature. The media was then replaced with a CSF for calcium imaging at room temperature. Changes in cytosolic calcium concentration were monitored over time through the ratio of fluorescence intensities at 510 nm obtained from intermittent (typically once every 2 s) excitation by 340-nm and 380-nm light (labeled as " 340/380 nm " in the y-axis of calcium-imaging figures. The increase or decrease in calcium levels is represented by an upward or downward deflection of the trace. High concentrations of KCl (25 mM) were used to determine indirect responses elicited by 3. 28 References 1 Luo, Y., Cobb, R. E. & Zhao, H. Recent advances in natural product discovery. Curr Opin Biotechnol 30, 230-237, doi:10.1016/j.copbio.2014.09.002 (2014). 2 Deane, C. D. & Mitchell, D. A. Lessons learned from the transformation of natural product discovery to a genome-driven endeavor. J Ind Microbiol Biotechnol 41, 315-331, doi:10.1007/s10295-013-1361-8 (2014). 3 Helfrich, E. J., Reiter, S. & Piel, J. Recent advances in genome-based polyketide discovery. Curr Opin Biotechnol 29, 107-115, doi:10.1016/j.copbio.2014.03.004 (2014). 4 Wilson, M. C. & Piel, J. Metagenomic approaches for exploiting uncultivated bacteria as a resource for novel biosynthetic enzymology. Chem Biol 20, 636-647, doi:10.1016/j.chembiol.2013.04.011 (2013). Schmidt, E. W. et al. Patellamide A and C biosynthesis by a microcin-like pathway in Prochloron didemni, the cyanobacterial symbiont of Lissoclinum patella. Proc Natl Acad Sci U S A 102, 7315-7320, doi:10.1073/pnas.0501424102 (2005). 5 ! "#! 6 Donia, M. S. et al. Natural combinatorial peptide libraries in cyanobacterial symbionts of marine ascidians. Nat Chem Biol 2, 729-735, doi:10.1038/nchembio829 (2006). 7 McIntosh, J. A. et al. Circular logic: nonribosomal peptide-like macrocyclization with a ribosomal peptide catalyst. J Am Chem Soc 132, 15499-15501, doi:10.1021/ja1067806 (2010). 8 McIntosh, J. A. & Schmidt, E. W. Marine molecular machines: heterocyclization in cyanobactin biosynthesis. Chembiochem 11, 1413-1421, doi:10.1002/cbic.201000196 (2010). 9 McIntosh, J. A., Donia, M. S., Nair, S. K. & Schmidt, E. W. Enzymatic basis of ribosomal peptide prenylation in cyanobacteria. J Am Chem Soc 133, 1369813705, doi:10.1021/ja205458h (2011). 10 Tianero, M. D., Donia, M. S., Young, T. S., Schultz, P. G. & Schmidt, E. W. Ribosomal route to small-molecule diversity. J Am Chem Soc 134, 418-425, doi:10.1021/ja208278k (2012). 11 Sardar, D., Pierce, E., McIntosh, J. A. & Schmidt, E. W. Recognition sequences and substrate evolution in cyanobactin biosynthesis. ACS Synth Biol 4, 167-176, doi:10.1021/sb500019b (2015). Ireland, C., Scheuer,P.J. Ulicyclamide and ulithiacyclamide, two new small peptides from a marine tunicate. J. Am. Chem. Soc. 102, 5688-5691 (1980). 12 13 Ireland, C. M., Durso, A.R., Newman, R.A, Hacker, M.P. Antineoplastic cyclic peptides from the marine tunicate Lissoclinum patella. J. Org. Chem 47, 18071811 (1982). 14 Linington, R. G. et al. Venturamides A and B: antimalarial constituents of the panamanian marine Cyanobacterium Oscillatoria sp. J Nat Prod 70, 397-401, doi:10.1021/np0605790 (2007). 15 Mitchell, S. S., Faulkner, D. J., Rubins, K. & Bushman, F. D. Dolastatin 3 and two novel cyclic peptides from a palauan collection of Lyngbya majuscula. J Nat Prod 63, 279-282 (2000). 16 Arnison, P. G. et al. Ribosomally synthesized and post-translationally modified peptide natural products: overview and recommendations for a universal nomenclature. Nat Prod Rep 30, 108-160, doi:10.1039/c2np20085f (2013). Donia, M. S., Ravel, J. & Schmidt, E. W. A global assembly line for cyanobactins. Nat Chem Biol 4, 341-343, doi:10.1038/nchembio.84 (2008). 17 ! ""! 18 Zhang, Q., Yang, X., Wang, H. & van der Donk, W. A. High divergence of the precursor peptides in combinatorial lanthipeptide biosynthesis. ACS Chem Biol 9, 2686-2694, doi:10.1021/cb500622c (2014). 19 McIntosh, J. A., Lin, Z., Tianero, M. D. B. & Schmidt, E. W. Aestuaramides, a natural library of cyanobactin cyclic peptides resulting from isoprene-derived claisen rearrangements. ACS Chem Biol 8, 877-883, doi:10.1021/cb300614c (2013). 20 Ruffner, D. E., Schmidt, E. W. & Heemstra, J. R. Assessing the combinatorial potential of the RiPP cyanobactin tru pathway. ACS Synth Biol 4, 482-492, doi:10.1021/sb500267d (2015). 21 Donia, M. S. et al. Complex microbiome underlying secondary and primary metabolism in the tunicate-Prochloron symbiosis. Proc Natl Acad Sci U S A 108, E1423-1432, doi:10.1073/pnas.1111712108 (2011). 22 Donia, M. S., Fricke, W. F., Ravel, J. & Schmidt, E. W. Variation in tropical reef symbiont metagenomes defined by secondary metabolism. PLoS One 6, e17897, doi:10.1371/journal.pone.0017897 (2011). 23 Donia, M. S., Ruffner, D. E., Cao, S. & Schmidt, E. W. Accessing the hidden majority of marine natural products through metagenomics. Chembiochem 12, 1230-1236, doi:10.1002/cbic.201000780 (2011). 24 Martin, V. J., Pitera, D. J., Withers, S. T., Newman, J. D. & Keasling, J. D. Engineering a mevalonate pathway in Escherichia coli for production of terpenoids. Nat Biotech 21, 796-802, doi:10.1038/nbt833 (2003). 25 Rodriguez-Concepcion, M., Boronat, A. in Isoprenoid synthesis in plants and microorganisms Vol. 1 (ed T. Bach, Rohmer, M.) 1-16 (Springer-Verlag, 2013). 26 Donia, M. S. & Schmidt, E. W. in Comprehensive Natural Products II (eds Chief Editors in, xA, Mander Lew, & Liu Hung-Wen) 539-558 (Elsevier, 2010). 27 Sera, Y., Adachi, K., Fujii, K. & Shizuri, Y. Isolation of Haliclonamides: new peptides as antifouling substances from a marine sponge species, Haliclona. Mar Biotechnol (NY) 4, 441-446, doi:10.1007/s10126-001-0082-6 (2002). 28 Teichert, R. W. et al. Functional profiling of neurons through cellular neuropharmacology. Proc Natl Acad Sci U S A 109, 1388-1395, doi:10.1073/pnas.1118833109 (2012). 29 Teichert, R. W., Schmidt, E. W. & Olivera, B. M. Constellation pharmacology: a new paradigm for drug discovery. Annu Rev Pharmacol Toxicol 55, 573-589, doi:10.1146/annurev-pharmtox-010814-124551 (2015). ! "##! 30 Rashid, M. A., Gustafson, K. R., Cardellina, J. H., 2nd & Boyd, M. R. Patellamide F, A new cytotoxic cyclic peptide from the colonial ascidian Lissoclinum patella. J Nat Prod 58, 594-597 (1995). 31 Degnan, B. M. et al. New cyclic peptides with cytotoxic activity from the ascidian Lissoclinum patella. J Med Chem 32, 1349-1354 (1989). 32 Carroll, A. R. et al. Patellins 1-6 and trunkamide A: novel cyclic hexa-, heptaand octa-peptides from colonial ascidians, Lissoclinum sp. Aust J Chem 49, 659667 (1996). 33 Cragg, G. M. & Newman, D. J. Natural products: a continuing source of novel drug leads. Biochim Biophys Acta 1830, 3670-3695, doi:10.1016/j.bbagen.2013.02.008 (2013). 34 Ongley, S. E., Bian, X., Neilan, B. A. & Muller, R. Recent advances in the heterologous expression of microbial natural product biosynthetic pathways. Nat Prod Rep 30, 1121-1138, doi:10.1039/c3np70034h (2013). 35 Zabriskie, T. M., Foster, M. P., Stout, T. J., Clardy, J. & Ireland, C. M. Studies on the solution- and solid-state structure of patellin 2. J Am Chem Soc 112, 80808084, doi:10.1021/ja00178a035 (1990). 36 Gibson, D. G. Enzymatic assembly of overlapping DNA fragments. Methods in enzymology 498, 349-361, doi:10.1016/B978-0-12-385120-8.00015-2 (2011). 37 Raghuraman, S. et al. Defining modulatory inputs into CNS neuronal subclasses by functional pharmacological profiling. Proc Natl Acad Sci U S A 111, 64496454, doi:10.1073/pnas.1404421111 (2014). CHAPTER 4 RIBOSOMAL ROUTE TO SMALL MOLECULE DIVERSITY Manuscript reproduced with permission from: Ma. Diarey B. Tianero, Mohamed S. Donia, Travis S. Young, Peter G. Schultz, Eric W. Schmidt. Ribosomal Route to Small-Molecule Diversity. J. Am. Chem. Soc. 2012, 134, 418!425 © 2011 American Chemical Society Note: My contribution to this paper was in the heterologous expression of compounds, analysis of data, interpretation of results, and writing the manuscript. ! "#$! Article pubs.acs.org/JACS Ribosomal Route to Small-Molecule Diversity Ma. Diarey B. Tianero,† Mohamed S. Donia,†,§ Travis S. Young,‡,⊥ Peter G. Schultz,‡ and Eric W. Schmidt*,† † Department of Medicinal Chemistry, University of Utah, Salt Lake City, Utah 84112, United States The Scripps Research Institute, La Jolla, California 92037, United States ‡ S Supporting Information * ABSTRACT: The cyanobactin ribosomal peptide (RP) natural product pathway was manipulated to incorporate multiple tandem mutations and non-proteinogenic amino acids, using eight heterologous components simultaneously expressed in Escherichia coli. These studies reveal the potential of RPs for the rational synthesis of complex, new small molecules over multiple-step biosynthetic pathways using simple genetic engineering. ■ INTRODUCTION A major goal of synthetic biology is to engineer the synthesis of organic compounds in vivo.1 Ribosomal peptide (RP) natural products provide a relatively simple starting point for such studies. In RP pathways, short precursor peptides are ribosomally translated and subsequently modified by enzymes into complex natural products.2 The products of RP pathways are often elaborately tailored, so that in extreme cases (such as pyrroloquinoline quinone),3 the compounds are no longer easily recognizable as arising from amino acids. A further advantage is that RP biosynthetic enzymes generally exhibit relaxed substrate specificity.2,4 Because of this, often sequenceidentical enzymes are capable of accepting a vast array of peptide substrates.5,6 Here, we seek to develop methods for making unnatural RPs and to better understand substrate selectivity of RP pathways. In so doing, we hope to enable synthetic biological approaches to the directed creation of diverse structures using ribosomally translated starting materials. In addition to these design and synthesis goals, RPs are often potently bioactive against diverse targets, and thus understanding and manipulating individual RP pathways is a major goal in its own right. In Nature, RPs occupy many different chemical classes, with many different posttranslational modifications.2 These classes have been examined by observing natural variation, as well as by manipulating biosynthetic genes and by performing biochemical and substrate analyses. Observation of Nature indicates that at least several pathways handle extremely diverse substitutions. For example, >60 cyanobactin genes have been isolated from symbiotic bacteria living in marine animals, where the enzymes are essentially identical yet the products are extremely sequence-diverse.5 RP biosynthetic genes have already been modified in several ways, most of which have focused on point © 2011 American Chemical Society mutations of pathways that require one or two enzymatic steps for maturation.7−10 Single or multiple mutations have been introduced into lantibiotics, and more recently, multiple tandem substitutions have been introduced into lasso peptides.12 These mutations are usually aimed either to improve the known biological activity, or occasionally even to design a new biological activity.13 A few mutational studies have focused on complex pathways with multiple biosynthetic steps.5,14,15 For example, the integrin-binding RGD motif has been introduced into cyclic peptides using cyanobactin pathway enzymes5 and lasso peptide enzymes.13 However, overall the field is still distant from achieving the goal of wholesale introduction of new motifs at will, especially in the more complex, multistep pathways. A further goal in this field has been to incorporate motifs other than the canonical 20 amino acids into the backbone of RPs. Since the posttranslational enzymes exhibit extremely broad substrate tolerance, such changes would greatly extend the capabilities of RP pathways in synthetic biology. In many cases, synthetic substrates have been fed to RP enzymes in vitro. A potentially more powerful approach from the synthetic biology point of view is to simply incorporate nonproteinogenic amino acids in vivo. Indeed, non-proteinogenic amino acids have been added to RPs in vivo, although only in single-step processes involving lanthionine synthase or inteins.16,17 Incorporation of such "unnatural" amino acids is considered desirable because it provides unprecedented control over chemical structure in vivo. Here, we sought to explore the mutation and addition of non-proteinogenic amino acids with a much more complex Received: September 1, 2011 Published: November 22, 2011 418 dx.doi.org/10.1021/ja208278k | J. Am. Chem.Soc. 2012, 134, 418−425 ! "#$! Journal of the American Chemical Society Article pathway to cyanobactins, which are highly modified, macrocyclic RPs.18,19 Cyanobactins usually have four types of posttranslational modifications, with up to eight individual modified amino acids, encompassing up to 18 separate biochemical transformations (Figure 1). Heterocylization and In the tru pathway, originally cloned from uncultivated symbiotic bacteria in a marine animal, a ∼70 amino acid precursor peptide (TruE) is synthesized that encodes two cyanobactins (Figure 1).6 Heterocyclase TruD acts on Cys residues in specific positions, synthesizing thiazoline.23 Subsequently, TruA protease removes the leader sequence and affords two free N-termini for macrocyclization.22 TruG protease cleaves C-terminal recognition sequences in tandem with macrocyclization to provide small, cyclic peptides.20,22,23 Finally, TruF1/TruF2 decorate the cyclic peptides with isoprene on specific Ser and Thr residues.24 In this manner, we have previously successfully produced coral reef-derived marine natural products patellin 2 (1), trunkamide (2) and patellin 3 (3) by heterologous expression in E. coli. Optimal production of wild type compounds 1−3 requires five days, emphasizing the complexity of the process.25 Some ascidian cyanobactins, and especially 2, are considered as potential anticancer leads.26,27 Cyanobactins in the patellamide class are potent inhibitors of drug efflux.28,29 Therefore, synthesis of derivatives in the manner described herein is expected to lead to bioactive molecules that are amenable to drug development. More importantly, as shown by the integration of RGD motifs into cyanobactins, we are interested in exploring the mutability of these pathways because they can be used to install desired or random features into natural product-like structures, using simple genetic engineering. ■ RESULTS Cloning and Expression Strategy. The pUC-based vector ptru-SD1 (Symbion Discovery, Inc.) encodes tru biosynthetic enzymes and a copy of the precursor peptide TruE1, encoding patellins 2 (1) and 3 (3), under control of the lac promoter (Figure 2). To encode new cyanobactin derivatives, we constructed a second vector (ptruE), which is compatible with ptru-SD1 and can be used for coexpression experiments. ptruE contains only the truE gene, under control of the lac promoter. The first cassette of ptruE encodes 3, while the second cassette can be varied to synthesize novel compounds (Figures 1 and 2, Tables 1 and 2). In this system, compounds 1 and 3 are always synthesized in E. coli and serve as internal controls to show that the tru pathway is functional. In addition, because both plasmids are under control of the constitutive lac promoter, no induction is necessary, and optimum production requires five days of fermentation at 30 °C. The general system was previously optimized, and we showed that addition of inducers or repressors serve to decrease yields.25 Thus, after five days without induction, internal controls 1 and 3 are produced, and possibly new derivatives encoded on ptruE. In this study, the sequence in the second cassette of ptruE was manipulated using cloning, while the first cassette was kept constant to produce positive control 1. The fidelity of this process was determined by gene sequencing, where the ptruE derivatives were Sanger sequenced. Thus, each E. coli expression clone encoded all necessary tru enzymes, internal standards leading to production of 1 and 3, and a gene for one of the new compounds 4−22. In expression experiments, production of 1 and 3 would indicate that enzymes were functional and active and would also provide internal calibration for yield determination, while new compounds would be expressed only if the sequences were substrates for all tru enzymes. Figure 1. Biosynthesis of tru pathway derivatives in Escherichia coli. Numbered components indicate eight individual elements including enzymes, tRNA, or precursor peptide (TruE variants) that are required for synthesis of mature natural products. The orthologous tRNA and amino acyl tRNA synthetase components required to introduce nonproteinogenic amino acids are shown in blue. Enzymes from the tru pathway are indicated in orange. The precursor peptide TruE is shown with a helical, 35-amino acid leader sequence indicated schematically.11 X indicates any amino acid. oxidation, N- and C-terminal proteolysis, macrocyclization, and prenylation lead to final natural products.5,6,20−24 In the pat and tru cyanobactin pathways, core sequences encoding six- to eight-amino acid cyclic peptides are contained in defined cassettes in a precursor peptide (Figures 2 and 3).5,6 The 18− 24 nucleotide hypervariable regions are flanked by invariant sequences that direct enzymatic modification.5 Thus, the synthesis of new compounds requires only simple genetic engineering.6 419 dx.doi.org/10.1021/ja208278k | J. Am. Chem.Soc. 2012, 134, 418−425 ! "#$! Journal of the American Chemical Society Article Figure 2. Expression strategy. Enzymes and two control molecules are synthesized constitutively from the vector ptru-SD1 (top). An internal control sequence and a variable region, that can lead to peptide libraries, is synthesized from ptruE (bottom). Optionally, non-proteinogenic amino acids can be included using a third vector, pEVOL (right). Table 1. Design Strategy for Mutants Figure 3. Expression of mutant cyanobactins. E. coli cultures expressing 20 (left) and 15 (right) were harvested and analyzed by FT-MS. Extracted ion chromatograms showed that each culture expressed control compounds 3 (A) and 1 (B). In cells containing plasmids encoding 20, compound 20 could be detected (C), but not 15 (D), nor any other cyanobactin 4−22. Similarly, in cells encoding 15, only compound 15 (D) but not 20 (C), nor any other cyanobactin, could be detected. Thus, each experiment contained two internal positive controls and 19 external negative controls. a .Wild-type compounds 1−3 were mutated, with changes to wild type shown in bold. b.Compounds were designed based on known natural products that had not been previously expressed, or based on sequences identified by genome mining (see Discussion for full description). c.Macrocylic ring size, in number of amino acids. scientific questions about the cyanobactins pathways (see Discussion). Thus, the results of these experiments would have broad application in synthetic biology and specific application to understanding this interesting group of natural products, which are very broadly distributed. Chemical Analysis and Isolation. After five days of fermentation, the pelleted E. coli cells were extracted with methanol. The organic extracts were partially purified and then analyzed using HPLC-ESI-MS. In all cases, heterologous expression of control compounds 1 and 3 was confirmed, with the compounds eluting with the same profile as standards of authentic 1 and 3, which we obtained from a marine animal We constructed mutants based upon known natural products, compounds 1-3, which we previously identified in uncultivated symbiotic bacteria living in marine animals and expressed successfully in E. coli.6 In addition, 1−3 are hexa-, hepta-, and octameric, respectively, so that mutations explored a range of different substrate and product sizes. We sought primarily to make a series of mutants that would broadly explore sequence selectivity of the tru pathway. Out of the constellation of possible mutations that could answer this question, we picked representative derivatives that were also interesting to us because they helped to answer other pressing 420 dx.doi.org/10.1021/ja208278k | J. Am. Chem.Soc. 2012, 134, 418−425 ! "#$! Journal of the American Chemical Society Article Table 2. Expression Yields in E. coli a Yields shown are for new compounds, compound 1, compound 3, and the total of all compounds found in a single expression experiment. Yields are reported as fully prenylated/−1 prenyl; i.e., for compound 4, 8.8 μg L−1 of diprenylated and 16 μg L−1 of monoprenylated compounds were produced. bRel yield (%) indicates the total amount of new compounds, divided by the total amount of compound 1, found in an individual sample (times 100 to give percentage). This gives the most accurate assessment of relative yield of new compounds between experiments. cOnly the nonprenylated 18 was identified, indicating that 18 is not a substrate for TruF1/F2. as previously described (Figure 3).6 We previously showed that isoprene is readily lost from Ser and Thr cyanobactin derivatives under standard MS conditions.25 Thus, loss of isoprene reliably indicates the formation of mature cyanobactins, and this loss is not observed in any natural E. coli compound. In addition, we observed ions representing a total of 16 out of 22 recombinant cyanobactins. A table was constructed in which the recombinant sequence determined by DNA sequencing was used to predict a unique mass for the new cyanobactin and for the loss of one or more isoprene groups from each predicted new compound. The prediction ions were only observed in expression experiments involving the sequence in question, and not in other experiments, so that we essentially had 19 negative control experiments for each compound produced (Figure 3). In addition, if prenylated, the ions readily fragmented to lose the predicted numbers of isoprene groups, in contrast to all other E. coli metabolites. Finally, in most cases we observed incomplete prenylation, so that mono-, di- and sometimes triprenylated derivatives were formed in E. coli. In total, we identified 21 new compounds, produced in E. coli, that matched the masses predicted from DNA sequencing. To further confirm the expression of the predicted compounds, they were subjected to analysis by high-resolution LC-FT-ICR-MS/MS, using previously established methods.24,25,30 The ions did indeed reflect the predicted compounds, to <2 ppm value, with loss of isoprene observed in MS/MS. When no isoprene was present on a compound, the fragmentation pattern reflected the sequence of the peptide. Thus, because of the numerous internal and external controls, the sequencing data, and the well-validated MS data, we had high confidence about the identity of expressed products. Finally, to further demonstrate that the recombinant compounds were successfully produced, we selected a set of representative compounds, 1, 10, and 14, for NMR analysis. These compounds were purified to homogeneity from the E. coli cell pellets, and their 1H NMR spectra were obtained (Figures S4−S7 and Table S1 in Supporting Information [SI]). In all cases, the NMR spectra matched those predicted for the new compounds, showing that they were the predicted cyanobactins. Expression of Mutant Cyanobactins. Hexameric patellin 2 (1) was manipulated to create compounds 4−8, which were triple or quadruple mutants of the wild type sequence (Tables 1 and 2). The triple mutants (4−6) were successfully synthesized in E. coli, while the quadruple mutant 7 was not detected. We previously reported successful expression of 4,31 while all other compounds 5−22 are reported here for the first time. In contrast to hexapeptide derivatives, for which diverse sequences are known, there are few known heptameric tru derivatives. Therefore, we created single-point mutants 13−15 of heptameric trunkamide (2). Only 13 and 14 were successfully processed in vivo. Pentuple mutants 16 and 17 were also not formed. Octapeptide selectivity was examined by synthesizing double mutants 19 and 20, pentuple mutant 21, and hextuple mutant 22. Surprisingly in comparison with our experience with heptapeptides, 19, 20, and 22 were synthesized. If more than one prenylation event was possible, we detected both singly and multiply prenylated derivatives. (For derivatives 9, 11, and 12 we detected only monoprenylated compounds, while for 18, no prenylation was detected, indicating that the molecule was not a substrate for TruF1/F2.) This was true even for the wild type compounds in E. coli, showing that complete prenylation is a limiting step in this host. We also observed this pattern in systems that produce a lower cyanobactin yield, indicating that it may be an intrinsic pathway property and not linked to the amount of dimethylallylpyr421 dx.doi.org/10.1021/ja208278k | J. Am. Chem.Soc. 2012, 134, 418−425 ! "#$! Journal of the American Chemical Society Article arabinose). Optimal incorporation of p-chloro-phenylalanine was achieved in the absence of arabinose, while increasing arabinose concentrations and thus the flux of aaRS incrementally decreased product yield (Figure 4). With optimized conditions in hand, pEVOL-pCNF was used in the preparative synthesis of 9−12, in which Leu of patellin 2 (1) ophosphate available in E. coli. In known natural products, TruD heterocyclase only modifies the C-terminal Cys residue. However, we previously showed that TruD modifies internal Cys residues in unnatural substrates in vitro.16 The primary sequence of compound 22 was derived from the pat biosynthetic pathway,23 in which all Cys and Thr/Ser residues are heterocyclic. However, in 22, two Cys residues were heterocyclic, but the two Thr residues were prenylated. This reveals that the pat and tru pathways may be hybridized to create diverse new derivatives. Incorporation of Non-Proteinogenic Amino Acids. We sought to incorporate non-proteinogenic amino acids using an orthogonal tRNA/aminoacyl-tRNA synthetase (tRNA/aaRS) pair that incorporates a specific unnatural amino acid in response to nonsense or frameshift mutation.24 First, we showed that non-proteinogenic amino acids could be incorporated into the TruE precursor peptide. The plasmid pEVOL-pAcF which can specifically incorporate The nonproteinogenic amino acid p-acetyl-phenylalanine (p-AcF) was coexpressed with plasmids encoding truE derivatives, which were C-terminally His-tagged and in which each position in patellin 2 was individually mutated to the amber codon TAG (Figure S3, SI). p-AcF was added to the fermentations. SDSPAGE of Ni purified lysates revealed that p-AcF could be incorporated in all six positions in the context of a precursor peptide (Figure S3, SI). These experiments showed that there was no problem in placing non-proteinogenic amino acids into the TruE precursor peptide backbone, in the absence of modifying enzymes. However, we still needed to determine whether the modifying enzymes could accept the highly unnatural amino acids and carry them through 4−5 biosynthetic steps to the mature unnatural natural products. We therefore attempted to coexpress this same system with ptru-SD1 and other ptru vectors. However, in extensive initial experiments with non-proteinogenic amino acids, including pAcF and others, could not be incorporated into cyanobactins using standard conditions. Several enzymes must be temporally coordinated in order to produce the mature cyanobactins over the course of our five-day fermentation. In extensive previous work with this pathway, changing the coordinated regulation of this pathway in any way severely impacted product yield.25 For example, even providing each gene in the eight-gene tru pathway with its own heterologous promoter was unsuccessful. It is clear that the complex regulation of this multistep pathway remains poorly understood. Previously, it was noted that incorporation of nonproteinogenic amino acids into single proteins over relatively long (14 h) induction times could be optimized with two copies of aaRS, one of which was constitutive and one of which was inducible with arabinose.32 In fact, ∼1 day is the longest expression period to which this methodology has been previously applied, to the best of our knowledge. The fiveday fermentation time for heterologous expression of tru is much longer than this, and in addition, the complex pathway regulation of tru is poorly understood. Therefore, we reasoned that adjustment of aaRS induction might improve cyanobactin production. Specifically, we thought that constitutive expression of aaRS might be better matched to the constitutive expression of tru. The plasmid pEVOL-pCNF (which encodes a polyspecific aaRS that incorporates a variety of Phe derivatives but not Phe or Tyr to any significant extent)25 was used. We varied expression of the arabinose-inducible copy of aaRS in pEVOL-pCNF using different concentrations of inducer (L- Figure 4. Optimization of non-proteinogenic amino acid incorporation. Compound 10 was synthesized in E. coli and analyzed by LCESI-MS. Optimum production was achieved in the absence of inducer (arabinose), whereas higher concentrations of inducer completely repressed synthesis. Two peaks are present because multiple stereoisomers are present (the α-proton adjacent to thiazoline is labile, and Pro undergoes cis−trans isomerization in this family).26. was replaced with substituted phenylalanine derivatives, and 18, in which Phe of trunkamide (2) was replaced with pbromophenylalanine (Figure 2 and Table 1). Out of the many possible amino acids that might be incorporated into engineered natural products using pEVOL-pCNF, we chose a limited subset for this work not including p-AcF. Yield Measurement. Yields were determined for all heterologous expression experiments. First, purified 1 was quantified by NMR integration using several different concentrations of the standard, 1,4-dinitrobenzene,33 to give an isolated yield of 260 μg from a 10-L fermentation. This method is much more accurate than other methods for microgram quantities of material, since it measures the relative molar amounts of compound directly using a calibrant. With a standard concentration of 1 available, we used known concentrations of 1 as internal and external standards for HPLC-ESI-MS experiments. These experiments used crude fractions containing recombinant peptides, so that they much more closely resembled the native production yield, and not merely the isolated yield. For example, the area under the curve was integrated to show that 1.7 mg of 1 was produced in the 10 L E. coli expression described above, but the isolated yield was only ∼10% after several purification steps. An additional 1.3 mg of the monoprenylated variant of 1 was also produced in this 10 L fermentation, leading to an overall yield of 3 mg of 1 variants. 422 dx.doi.org/10.1021/ja208278k | J. Am. Chem.Soc. 2012, 134, 418−425 ! "#$! Journal of the American Chemical Society Article Figure 5. Compounds synthesized in this study. Yellow indicates wild-type compounds, while pink bubbles indicate mutations that deviate from wild type. Hexa- (left), hepta- (mid), and octa- (right) peptide derivatives were synthesized using these methods. (*) Indicates compounds for which only monoprenylated (9, 11, and 12) or nonprenylated (18) derivatives were identified. For all other compounds, both singly, doubly, and sometimes triply prenylated products were identified. Because 1 was produced in all expression experiments reported herein, the compound could also be used to estimate the production level of other recombinant products. Yields of most compounds were in the range of 1−170 μg L−1. (see Table 1). Compound 2 was also present as an internal control. A potential weakness of this method is that the relative ionization efficiencies of the compounds are likely to differ. However, the ratios were also compared to those achieved using FT-ICR-MS, showing a consistent relative ratio of ions in both methods. Moreover, in previous studies we have not seen a large range of ionization efficiencies for these similar compounds, indicating that the reported yields are good estimates. The relative yields determined by MS were also reflected in the isolated yields of 10 and 14. ■ ification. In these cases, the relative simplicity of the heterologous expression systems made them very tolerant of overexpression. In more complex systems, our data suggest that producing too much of the aaRS (or any other component in a complex pathway) is detrimental to efficient product synthesis. There are now many different methods for implanting nonproteinogenic amino acids into ribosomal peptides, some of which involve changing the ribosomal code and some of which involve synthetic modifications.34 Notably, recently the concepts of multiple orthogonal substitution and residuespecific substitution have been elaborated.35,36 The former enables different non-proteinogenic amino acids to be introduced into single peptides, while the latter enables all residues of a specific type to be replaced by non-proteinogenic amino acids. These methods are compatible with the ideas developed here, but they will undoubtedly require some further study to be applied to highly modified RPs, since so far they have not been applied to such compounds. In addition to exploring incorporation of diverse substitutions, these experiments were designed to provide proof-ofconcept for key technological advances in marine natural products, in five major areas. First, we have recently applied a genomic method, in which we sequence genes from marine animals and their symbiotic bacteria to discover new TruE-like peptides.25,30 Subsequently, we find that the predicted peptides are the major natural products found in whole animals. However, the compounds are exceptionally limited in supply, and at most micrograms are available from the natural sources. As mentioned above, previously we expressed 4 to supply the compound. Here, we wished to test the broad applicability of the method by expressing compounds 4-6, which were first identified by genomics and which were successfully made by E. coli here. Second, there were a series of previously described, rare compounds from Nature, including mollamide (encoded by 16 and 17) and keenamide (encoded by 8).19,37 These compounds were reported to exhibit interesting bioactivity but are in very short supply. However, these compounds were not produced by our tru pathway in E. coli, indicating that they are probably synthesized by a different variation of the tru pathway found in DISCUSSION In this study, we exploited the relaxed substrate selectivity of the tru pathway to generate diverse mutants, despite the need to go through a complex biosynthetic route. Although not all derivatives can be made, this study shows that the tru posttranslational machinery exhibits broad substrate tolerance. We also report the first in vivo incorporation of nonproteinogenic amino acids into multistep RP products. Notably, control of six different enzymes, two precursor peptides, and a tRNA molecule was required for synthesis of the reported derivatives. Incorporated, non-proteinogenic amino acids were successfully carried through a multistep pathway with 12 individual enzymatic transformations. This was accomplished using eight enzyme tru domains encoded in six tru proteins, whereas previously at most one enzyme with two domains has been used.11 We expect that the methods here will be widely useful in the synthesis of diverse RP derivatives encoding nonproteinogenic amino acids. A simple but important advance was the observation that low-level, constitutive expression provides excellent incorporation of non-proteinogenic amino acids in natural product derivatives. Prior to this report, the described technology had been used routinely for incorporation of amino acids into proteins, and a few reports exist for incorporation into short peptides which undergo one further posttranslational mod423 dx.doi.org/10.1021/ja208278k | J. Am. Chem.Soc. 2012, 134, 418−425 ! "#$! Journal of the American Chemical Society Article California, San Francisco, San Francisco, California 94158, United States. ⊥ Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, Boston, Massachusetts 02115, United States. Nature. Third, we wished to express known marine animal compounds, patellins 4 and 5 (19 and 20),19,38 for which we had not previously found genes; these were successfully made. Fourth, trunkamide is a potential anticancer agent of interest because of its unusual profile in the National Cancer Institute's 60-cell line panel. We tested its potential for derivatization by making a series of point mutants, 13−15. Interestingly, addition of a Ser led to a third prenylation event in 14, and 13 was successfully produced. However, a simple Ser-Val substitution in 15 led to loss of product. Fifth, for compounds 21 and 22, we wished to cross the tru pathway with the pat pathway that leads to patellamides. Patellamides are not prenylated, and instead Ser and Thr residues are cyclized to yield oxazoline residues.5,39,40 Moreover, often 2 thiazole residues exist in the pat products, instead of 1 thiazoline as found in all known tru pathway relatives. The genetic and biochemical basis of these differences have been thoroughly established. We have been very interested in determining the "portability" of enzymes; that is, can individual RP enzymes be moved from one pathway to another? As a first step to establish this fact, we cloned patellamide precursor peptides that would normally yield the compounds ulithiacyclamide and patellamide C into the tru pathway background, to give derivatives 21 and 22. 21 could not be produced, but 22 was successfully synthesized and contained 2 thiazoline residues, the first time this modification has been reported in tru derivatives. Both Thr residues were prenylated, rather than heterocyclic as they would be in the natural product. This report represents a significant expansion of the toolkit available for directed, posttranslational modification of peptides in living cells. Additionally, it provides the first technical guidelines for performing complex manipulations with multistep RP pathways. Ultimately, we hope to improve control over RP processing for directed synthetic biology of fine chemicals and pharmaceuticals. This technology has numerous possible applications, including supply of marine natural products and improvement of anticancer properties of cyanobactins. More importantly, such technology allows simple genetic engineering tools to be applied to creation of wholly new drug motifs that combine the advantages of peptide technology and synthetic chemistry. For example, intein-circularized and phage display libraries encode an enormous sequence diversity which can be used to attack diverse biological targets,40 but the resulting compounds are generally not drugs because of poor pharmacological properties. The addition of many posttranslational modifications promises to afford the same sequence diversity, but with products that are more "drug-like". Here, we use the multistep cyanobactin pathway to explore the chemistry that necessarily underlies these downstream applications. ■ Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes E.W.S. is a co-owner of Symbion Discovery, Inc. ■ ACKNOWLEDGMENTS This work was funded by NIH GM071425 and GM071425-S1 to E.W.S. and NIH GM097206 to P.G.S. We thank K. Parsawar and C. Nelson (University of Utah Mass Spectrometry Core Facility) for acquiring high-resolution mass spectra, J. Skalicky (University of Utah) for help with NMR, and D. E. Ruffner (Symbion Discovery, Inc.) for providing ptru-SD1. ■ ASSOCIATED CONTENT S Supporting Information * Methods, supporting data, and tables. This material is available free of charge via the Internet at http://pubs.acs.org. ■ REFERENCES (1) Mitchell, W. Curr. Opin. Chem. Biol. 2011, 15, 505. (2) McIntosh, J. A.; Donia, M. S.; Schmidt, E. W. Nat. Prod. Rep. 2009, 26, 537. (3) Velterop, J. S.; Sellink, E.; Meulenberg, J. J.; David, S.; Bulder, I.; Postma, P. W. J. Bacteriol. 1995, 177, 5088. (4) Oman, T. J.; van der Donk, W. A. Nat. Chem. Biol. 2010, 6, 9. (5) Donia, M. S.; Hathaway, B. J.; Sudek, S.; Haygood, M. G.; Rosovitz, M. J.; Ravel, J.; Schmidt, E. W. Nat. Chem. Biol. 2006, 2, 729. (6) Donia, M. S.; Ravel, J.; Schmidt, E. W. Nat. Chem. Biol. 2008, 4, 341. (7) Widdick, D. A.; Dodd, H. M.; Barraille, P.; White, J.; Stein, T. H.; Chater, K. F.; Gasson, M. J.; Bibb, M. J. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 4316. (8) Kelleher, N. L.; Hendrickson, C. L.; Walsh, C. T. Biochemistry 1999, 38, 15623. (9) Willey, J. M.; van der Donk, W. A. Annu. Rev. Microbiol. 2007, 61, 477. (10) Melby, J. O.; Nard, N. J.; Mitchell, D. A. Curr. Opin. Chem. Biol. 2011, 15, 369. (11) Houssen, W. E.; Wright, S. H.; Kalverda, A. P.; Thompson, G. S.; Kelly, S. M.; Jaspars, M. ChemBioChem 2010, 11, 1867. (12) Pan, S. J.; Link, A. J. J. Am. Chem. Soc. 2011, 133, 5016. (13) Knappe, T. A.; Manzenrieder, F.; Mas-Moruno, C.; Linne, U.; Sasse, F.; Kessler, H.; Xie, X.; Marahiel, M. A. Angew. Chem., Int. Ed. 2011, 50, 8714. (14) Bowers, A. A.; Acker, M. G.; Koglin, A.; Walsh, C. T. J. Am. Chem. Soc. 2010, 132, 7519. (15) Li, C.; Zhang, F.; Kelly, W. L. Mol. Biosyst. 2011, 7, 82. (16) Shi, Y.; Yang, X.; Garg, N.; van der Donk, W. A. J. Am. Chem. Soc. 2011, 133, 2338. (17) Young, T. S.; Young, D. D.; Ahmad, I.; Louis, J. M.; Benkovic, S. J.; Schultz, P. G. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 11052. (18) Sivonen, K.; Leikoski, N.; Fewer, D. P.; Jokela, J. Appl. Microbiol. Biotechnol. 2010, 86, 1213. (19) Donia, M. S.; Schmidt, E. W. In Comprehensive Natural Products II; Mander, L, Liu, H.-W., Eds.; Elsevier: Oxford, 2010; Vol 2, p 539. (20) Lee, J.; McIntosh, J.; Hathaway, B. J.; Schmidt, E. W. J. Am. Chem. Soc. 2009, 131, 2122. (21) McIntosh, J. A.; Donia, M. S.; Schmidt, E. W. J. Am. Chem. Soc. 2010, 132, 4089. (22) McIntosh, J. A.; Robertson, C. R.; Agarwal, V.; Nair, S. K.; Bulaj, G. W.; Schmidt, E. W. J. Am. Chem. Soc. 2010, 132, 15499. (23) McIntosh, J. A.; Schmidt, E. W. ChemBioChem 2010, 11, 1413. (24) McIntosh, J. A.; Donia, M. S.; Nair, S. K.; Schmidt, E. W. J. Am. Chem. Soc. 2011, 133, 13698. AUTHOR INFORMATION Corresponding Author [e-mail] Present Addresses § Department of Bioengineering and Therapeutic Sciences and California Institute for Quantitative Biosciences, University of 424 dx.doi.org/10.1021/ja208278k | J. Am. Chem.Soc. 2012, 134, 418−425 ! "#$! Journal of the American Chemical Society Article (25) Donia, M. S.; Ruffner, D. E.; Cao, S.; Schmidt, E. W. ChemBioChem 2011, 12, 1230. (26) Carroll, A. R.; Coll, J. C.; Bourne, D. J.; MacLeod, J. K.; Zabriskie, T. M. Aust. J. Chem. 1996, 49, 659. (27) Wipf, P.; Uto, Y. J. Org. Chem. 2000, 65, 1037. (28) Aller, S. G.; Yu, J.; Ward, A.; Weng, Y.; Chittaboina, S.; Zhuo, R.; Harrell, P. M.; Trinh, Y. T.; Zhang, Q.; Urbatsch, I. L.; Chang, G. Science 2009, 323, 1718. (29) Williams, A. B.; Jacobs, R. S. Cancer Lett. 1993, 71, 97. (30) Donia, M. S.; Fricke, W. F.; Ravel, J.; Schmidt, E. W. PLoS One 2011, 6, e17897. (31) Donia, M. S.; Schmidt, E. W. Chem. Biol. 2011, 18, 508. (32) Young, T. S.; Ahmad, I.; Yin, J. A.; Schultz, P. G. J. Mol. Biol. 2010, 395, 361. (33) Rundlof, T.; Mathiasson, M.; Bekiroglu, S.; Hakkarainen, B.; Bowden, T.; Arvidsson, T. J. Pharm. Biomed. Anal. 2010, 52, 645. (34) Budisa, N. Angew. Chem., Int. Ed. 2004, 43, 6426. (35) Neumann, H.; Wang, K.; Davis, L.; Garcia-Alai, M.; Chin, J. W. Nature 2010, 464, 441. (36) Johnson, J. A.; Lu, Y. Y.; Van Deventer, J. A.; Tirrell, D. A. Curr. Opin. Chem. Biol. 2010, 14, 774. (37) Wesson, K. J.; Hamann, M. T. J. Nat. Prod. 1996, 59, 629. (38) Fu, X.; Do, T.; Schmitz, F. J.; Andrusevich, V.; Engel, M. H. J. Nat. Prod. 1998, 61, 1547. (39) Schmidt, E. W.; Nelson, J. T.; Rasko, D. A.; Sudek, S.; Eisen, J. A.; Haygood, M. G.; Ravel, J. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 7315. (40) Tavassoli, A.; Benkovic, S. J. Nat. Protoc. 2007, 2, 1126. (41) Zabriskie, T. M.; Foster, M. P.; Stout, T. J.; Clardy, J.; Ireland, C. M. J. Am. Chem. Soc. 1990, 112, 8080. (42) Donia, M. S.; Wang, B.; Dunbar, D. C.; Desai, P. V.; Patny, A.; Avery, M.; Hamann, M. T. J. Nat. Prod. 2008, 71, 941. (43) Carroll, A. R.; Bowden, B. F.; Coll, J. C.; Hockless, D. C. R.; Skelton, B. W.; White, A. H. Aust. J. Chem. 1994, 47, 61−69. (44) Ireland, C.; Scheuer, P. J. J. Am. Chem. Soc. 1980, 102, 5688. (45) Ireland, C. M.; Durso, A. R.; Newman, R. A.; Hacker, M. P. J. Org. Chem. 1982, 47, 1807. 425 dx.doi.org/10.1021/ja208278k | J. Am. Chem.Soc. 2012, 134, 418−425 ! ""#! Supporting Information for: Ribosomal route to small molecule diversity Ma. Diarey B. Tianero, Mohamed S. Donia, Travis S. Young, Peter G. Schultz, and Eric W. Schmidt Table of Contents 1. Materials and Methods……S2 2. Figure S1. 1H NMR spectra of 1,4-dinitrobenzene and patellin 2 (1)…….S9 3. Figure S2. Standard curve for LC/MS quantification……..S10 4. Table S1. MS analysis of compounds expressed in this study……S11 5. Figure S3. Western blot analysis of TruE and tRNA/aaRS co-expression……S12 6. Figure S4. FT-MS/MS spectra for all compounds expressed in this study……S13 7. Figure S5. 1H NMR and COSY spectra of compound 11……S35 8. Figure S6. 1H NMR and COSY spectra of compound 15……S37 9. Figure S7. 1H NMR spectrum of compound 1……S39 10. References…….S40 ! "#! ! """! Materials and Methods Plasmids. ptru-SD1 was obtained from Symbion Discovery, Inc., in which the tru operon is controlled by lac. This vector performs identically to the previously reported TOPOtru, except that yield is improved. ptruE is derived from the previously reported pRSFtrue2 vector in which the TruE peptide is present in the pRSF vector backbone and TruE expression is controlled by lac.1 Restriction sites were introduced so that the second cassette of TruE2 could be replaced with desired variants. Top strand and bottom strand oligonucleotides encoding the desired cyanobactin variant were synthesized at the University of Utah Peptide Core Facility. The two strands were annealed ligated to ptruE digested with the corresponding restriction enzymes. T4 ligase (NEB) was used for cloning and E. coli TOP10 (Invitrogen) was used for the propagation of the plasmids. All constructs were verified by DNA sequencing. Using this conventional cloning method, ptruE-X1-ptruEX18 and ptruE-amber1 and ptruE-amber2 were constructed and verified by sequencing. These vectors are identical to previously reported pRSF-true2,1 except that the second cassette encodes the new cyanobactin variant. Expression of cyanobactin variants. ptru-SD1 and ptruE wild type or mutant vectors were co-transformed into Top10 E. coli cells (Invitrogen) following the manufacturer's protocol and plated on LB agar plates supplemented with ampicillin (50 µg ml-1) and kanamycin (25 µg ml-1). Because expression from individual colonies is variable,1 for each mutant five to ten colonies were picked and grown overnight in 2xYT broth (5 ml). These seed cultures were pooled, and pooled broth (1 ml) was inoculated into 2xYT broth (100 ml) in a 250 ml flask. All 2xYT media contained ampicillin (50 µg ml-1), kanamycin ! "#! ! ""#! (25 µg ml-1), and a proprietary Symbion Mix that slightly improves yield but does not affect lac expression. The cells were cultivated for five days at 30 o C and 200 rpm, after which they were collected by centrifugation at 3320 x g for 20 minutes. The cell pellet was washed with NaCl (100 mM) and extracted by sonication in methanol (75 ml) for 30 minutes. The suspension was filtered, Diaion HP20 (5% v/v) was added to the filtrate, and the mixture was dried under vacuum to remove most of the methanol. The resin was transferred into vac elut cartridges (Varian) and washed with water, followed by a step gradient of methanol (25%, 50%, 100%) in water. The 100% methanol fraction was dried under vacuum and re-suspended in methanol (500 µl). An aliquot (50 µl) of this extract was passed through a small plug of end-capped C18 resin and analyzed by HPLC-ESI-MS using a Waters Micromass ZQ mass spectrometer and a Phenomenex C18 column (10 cm x 4.6 mm, 5µm). A linear gradient was employed at 0.2 ml min-1, from 10-100% acetonitrile in water (0.1% formic acid) over 20 minutes, followed by 15 minutes at 100% acetonitrile. Co-expression of truE with amber suppressor tRNA/aaRS. Seven pRSF-TruE constructs containing all six possible amber stop codon mutants in the TVPTLC sequence and the wild type (WT) sequence were independently co-transformed with pEVOL-pAcF (incorporates p-acetylphenylalanine in response to the amber stop codon) in BL21(DE3) cells. Single colonies were picked for expression in LB media containing kanamycin (25 µg ml-1), and chloramphenicol (25 µg ml-1) and grown overnight. Cultures (250 ml) were grown to OD600=0.8 and induced with IPTG (1 mM) and arabinose (0.02%) and supplemented with pAcF (1 mM). One culture of the pRSF-WT variant was not induced ! "#! ! ""#! with IPTG (indicated as uninduced in Figure S3). After 12 hours, cultures were pelleted, lysed in Bugbuster cell lysis reagent (5 ml, Novagen), and purified using Ni-NTA resin (0.25 ml, Qiagen) following protocols in the Qiagen protein purification handbook for native purification. Eluents were concentrated 5-fold before loading onto a 4-20% SDSPAGE gel (Novagen) with DTT (20 mM). The gel was stained with a solution of Coomassie Blue R-250 (0.25%, Sigma) and washed before imaging. The presence of full-length TruE protein in constructs containing an amber mutation indicated successful incorporation of the pAcF unnatural amino acid. In vivo synthesis of cyanobactins with non-protein amino acids. ptru-SD1 (Symbion Discovery, Inc.), ptruE-amber, and pEVOL-pCNF containing the amber suppresor tRNA/aaRS were co-transformed in Top10 E. coli cells following the manufacturer's protocol and plated on LB agar plates supplemented with ampicillin (50 µg ml-1), kanamycin (25 µg ml-1), and chloramphenicol (25 µg ml-1). The resulting clones were treated as described in Expression of cyanobactin variants, except that chloramphenicol (12.5 µg ml-1) and non-proteinogenic amino acids (Chemimpex; 2 mM) were also added to the culture broth. To optimize induction, experiments were run as described above, except that growth was monitored until OD600=0.8, at which time L-arabinose was added at varying concentrations (0%, 2x10-4%, 2x10-3%, 2x10-2%, 2x10-1%). Expression using pEVOL was optimized using p-chloro-L-phenylanine (p-ClF) (2mM), which was added in batches. At inoculation, p-ClF was added to 0.7 mM and growth was monitored until ! "#! ! ""#! OD600=0.8 at which time L-arabinose and p-ClF (1.3 mM) were added. Later, this batch addition was found not to make a difference, so that exogenous amino acids (2 mM) were subsequently added upon inoculation. Characterization of fermentation products. Fractions containing new and known cyanobactin derivatives were partially purified using HP20 and C18 resins, as described above, and then analyzed by HPLC FT-ICR-ESI-MS at the University of Utah Mass Spectrometry Core Facility using a LTQ-FT (ThermoElectron), with MS/MS using collision induced dissociation (Table S2 and Figure S4). Preparative fermentation of TVPT-pClF-C (11) and TSIASFC (15). Seed cultures were used to inoculate 2xYT broth (10 l) containing additives as above in a 14 l fermentor (Bioflo 110, New Brunswick Scientific). Fermentation was performed over five days at 30 oC and 200 rpm, after which the cells were collected by centrifugation at 4424 x g and washed with NaCl (100 mM). The pellet was extracted with acetone (1.5 l). Diaion HP20 (5% v/v) was added, and the slurry was dried under vacuum. The resin was then washed extensively with water, followed by 40% acetone. Finally, an extract enriched with desired compounds was obtained by elution with acetone (500 ml). The vacuum-dried acetone extract was fractionated by flash chromatography on silica gel (50 g, 230-400 mesh) pre-equilibrated in 9:1 hexane: ethyl acetate. A step gradient was employed consisting of the following solvents: 1) 9:1 hexane: ethyl acetate (100 ml); 2) 1:1 hexane: ethyl acetate (100 ml); 3) ethyl acetate (150 ml); 4) 10:1 dichloromethane: methanol (200 ml); 5) 5:1 dichloromethane: methanol (200 ml). Fractions 3 and 4, containing cyanobactin derivatives, were further purified by HPLC on a Hitachi LaChrom Elite ! "#! ! ""#! system using two rounds of chromatography. The first employed a gradient of 35-100 % acetonitrile (aq) over 40 minutes (2 ml min-1) on a Supelco HS Discovery (25 cm x 10 mm, 5 µm) C18 HPLC column. The second used 72 % acetonitrile (aq; 1 ml min-1) on a Phenomenex Gemini (25 cm x 4.6 mm, 5µm) C18 HPLC column. NMR data were obtained using a Varian INOVA 600 (1H 600 MHz). For spectral data, see Figures S5 and S6. Yield determination.Yield of purified products could be determined by the usual methods, but we also sought more accurate determination using spectroscopic techniques, detailed below. Quantification by NMR. 1H NMR quantification was performed on a Varian INOVA digital 500 MHz NMR spectrometer at 25 o C. Quantification was performed using an external standard, 1,4-dinitrobenzene (DNB) (Sigma-Aldrich), following a previously validated method.2 The standard 1 and different dilutions of DNB were prepared in DMSO-d6 (Cambridge Isotopes) (120 mL) in identical, vacuum-dried NMR tubes (3mm). All samples were treated identically by placing into the probe and tuning and matching to identical values. All 1H NMR experiments were run using the same conditions, with repeated single 90o pulse sequence, D1(5T1) of 4s, and 512 scans as determined for appropriate signal averaging. Calibration of the p90o pulse length for each samples was done by determination of the null (360o) pulse length (pw=pw/4). T1 relaxations for both the 1 and DNB were obtained by inversion recovery method. ! "#! ! ""#! A two-fold dilution series of DNB (6 samples from 0.4 to 0.0125 mg) was prepared in DMSO-d6 from a stock solution (34 mg/mL). The singlet signal at 8.4 ppm corresponding to four equivalent protons of DNB was integrated and used to generate a standard curve under various conditions (Figure S1). The 1H NMR spectrum of 1 was measured under identical conditions, and well-separated NH (d 7.4X ppm) and methyl (d 0.73 ppm) signals were integrated and compared to the standard curve of DNB (Figure S1). Under several different conditions, reproducible values were obtained using this method, which met the previously described accuracy level.2 Quantification by HPLC-MS. The quantified NMR sample measured above was used directly from the NMR tube to generate highly accurate standards for MS. To generate a standard curve, a serial dilution of known concentrations of 1 (five different concentrations, from 0.26 to 0.032 mg in 40 mL MeOH) was used (Figure S2). This curve was generated both in pure MeOH and by adding the diluted standard to crude of the E. coli expression constructs reported above. Samples were then run under the LCMS conditions described above. The area under the curve was determined by obtaining the spectrum corresponding to 1 over its ~2.5 min elution time, then multiplying this time by the average signal intensity at m/z = 733.5 over that time. Analysis was performed in duplicate and the integrations for each point were averaged. A plot of area vs. concentration was generated, indicating that the signal response to concentration was linear and that 1 provided a robust internal standard over the concentration range being measured. The resulting linear equation was used to calculate the concentrations of the recombinant and standard compounds (Figure S2). ! "#! ! ""#! Figure S1. A. Stacked 1H-NMR spectra and integration values of 1,4-dinitrobenzene at different concentrations. B. Integration of sample 1 (CH3). ! "#! ! ""#! Figure S2. Standard curve with internal standard 1 (0.262 to 0.032 µg) added to IVVPFC crude extracts. ! "#$! ! ""#! Table S1. Calculated and observed masses for compounds synthesized in E. coli. Prenylation state varied, and often derivatives were found that were not completely prenylated. Compounds with 1, 2 or 3 prenyl groups were identified as mixtures in single expression experiments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a) b) c) d) ! all masses observed with +/- 2 ppm error (i) indicates compounds produced with 1 less prenyl group - C10H16 indicates loss of 2 prenyl groups (MS/MS) - C5H8 indicates loss of 1 prenyl group (MS/MS) "##! ! "#$! WT induced WT uninduced TVPTLX TVPXLC TVXTLC TXPTLC XVPTLC TVPTXC Figure S3. Co-expression of precursor peptide constructs with pEVOL-pAcF where X indicates the position of the amber stop codon in the truE primary amino acid sequence. The wild type (WT) sequence contains no amber stop codon (sequence TVPTLC). The arrow indicates the position of full length TruE protein. ! "#$! ! "#"! Figure S4. A1 TLATIC (4) chromatogram A2 TLATIC (4) spectrum A3 TLATIC (4) MS-MS spectrum B1 TLATIC (4) (i) chromatogram B2 TLATIC (4) (i) spectrum B3 TLATIC (4) (i) MS-MS spectrum C1 IVPPFC (5) chromatogram C2 IVPPFC (5) spectrum C3 IVPPFC (5) MS-MS %&'()*+,! D1 TTVTAC (7) chromatogram D2 TTVTAC (7) spectrum D3 TTVTAC (7) MS-MS %&'()*+,! E1 TTVTAC (7) (i) chromatogram E2 TTVTAC (7) (i) spectrum E3 TTVTAC (7) (i) MS-MS %&'()*+,! F1 TVPT-pBrF-C (9) (i) chromatogram F2 TVPT-pBrF-C (9) (i) spectrum F3 TVPT-pBrF-C (9) (i) MS-MS %&'()*+, G1 TVPT-pClF-C (10) chromatogram G2 TVPT-pClF-C (10) spectrum G3 TVPT-pClF-C (10) MS-MS %&'()*+, H1 TVPT-pClF-C (10) (i) chromatogram H2 TVPT-pClF-C (10) (i) spectrum H3 TVPT-pClF-C (10) (i) MS-MS %&'()*+, I1 TVPT-pOMeF-C (11) (i) chromatogram I2 TVPT-pOMeF-C (11) (i) spectrum I3 TVPT-pOMeF-C (11) (i) MS-MS %&'()*+, J1 TVPT-pN3F-C (12) (i) chromatogram J2 TVPT-pN3F-C (12) (i) spectrum J3 TVPT-pN3F-C (12) (i) MS-MS %&'()*+, K1 TSIAPLC (13) chromatogram K2 TSIAPLC (13) spectrum K3 TSIAPLC (13) MS-MS %&'()*+, L1 TSIAPLC (13) (i) chromatogram L2 TSIAPLC (13) (i) spectrum L3 TSIAPLC (13) (i) MS-MS spectrum and assignments M1 TSIASFC (14) chromatogram M2 TSIASFC (14) spectrum M3 TSIASFC (14) MS-MS %&'()*+, N1 TSIASFC (14) (i) chromatogram N2 TSIASFC (14) (i) spectrum N3 TSIASFC (14) (i) MS-MS %&'()*+, ! "#$! ! "##! O1 O2 O3 P1 P2 P3 Q1 Q2 Q3 R1 R2 R3 S1 S2 S3 T1 T2 T3 ! TSIAS-pBrFC (18) (i) chromatogram TSIAS-pBrFC (18) (i) spectrum TSIAS-pBrFC (18) (i) MS-MS %&'()*+, TLPVPTVC (19) chromatogram TLPVPTVC (19) spectrum TLPVPTVC (19) MS-MS %&'()*+, TLPVPTVC (20) (i) chromatogram TLPVPTVC (20) (i) spectrum TLPVPTVC (20) (i) MS-MS %&'()*+, TVPVPSFC (21) (i) chromatogram TVPVPSFC (21) (i) spectrum TVPVPSFC (21) (i) MS-MS %&'()*+, VTACITFC (22) chromatogram VTACITFC (22) spectrum VTACITFC (22) MS-MS %&'()*+, VTACITFC (22) (i) chromatogram VTACITFC (22) (i) spectrum VTACITFC (22) (i) MS-MS %&'()*+, "#$! ! "#$! S4. A1. LC-FT-ICR %&'&()&*!+,-!(./,01),2/10!3,/!TLATIC (4)! A2. LC-FT-ICR spectrum of TLATIC (4) A3. MS-MS spectrum of TLATIC (4) !!"#!$%&'()*+,-&)*.,+/'(&!01&23&45 #6!$60"!!7!8""8"97:; <&8 0=>$"717=#>=" L287=>=?M= A6NO7D0!>=@"""1D0!>=#"""7 %87%&;P7Q7R7LP47%IEE7AS7T7 @#">""1!="">""U77;P7 !!"#!$%&'()*+,-&)*.,+/' (&!01&23&45 =!>@! !"" $" <,E*F)G,73HI.J*.K, 9" D" ?" #" =#>!D =" @" 0" !" =!>#" 0$>9" " 0? 09 @" @0 @= @? &)A,7BA).C @9 =" =0 == !!"#!$%&'()*+,-&)*.,+/'(&!01&23&457V#DD=1#909 <&8 =!>!@1=!>== 3W8 $ L28 9>?=M@ &8 %&;P7Q7R7LP47%IEE7AS7T7@#">""1!="">""U D0!>=@@@# !"" $" S <,E*F)G,73HI.J*.K, 9" HN D" O ?" HN #" =" D00>=@#$D N O O O O O O NH N H @" NH 0" !" " D!= D!# D!? D!D D!9 D!$ D0" D0! D00 A6N D0@ D0= D0# D0? D0D D09 D0$ !!"#!$%&'()*+,-&)*.,+/'(&!01&23&457V#DD=1#909 <&8 =!>0$1=!>@= 3W8 0 L28 @>9=M@ &8 4&;P7Q7K7LP47J7%IEE7AS07D0!>=@X@#>""7T7!9#>""1D@#>""U #9#>@"D?= !"" ?#@>0#=D9 $" <,E*F)G,73HI.J*.K, 9" 4!7!5/&-6'! D" ?" #" =" 4!5/&-6'! @" 0" !" " 0"" ! 0#" @"" @#" ="" =#" A6N "#$! #"" ##" ?"" ?#" D"" ! "#$! S4. B1. LC-FT-ICR %&'&()&*!+,-!(./,01),2/10!3,/!TLATIC (4) (i)! B2. LC-FT-ICR spectrum of TLATIC (4) (i) B3. MS-MS spectrum of TLATIC (4) (i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spectrum of IVPPFC (5) !!"#$"%&'()*+,-&)*.,+/'(&!0'1233%4 C3. MS-MS spectrum of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spectrum of TTVTAC (7) D3. MS-MS spectrum of TTVTAC (7) !"#$$%&'&()*+,-)-,./01*+)*$2*))3)45 &6'&6'%$$72"'8"987:; <)" %=%%7>7?@='@ ?8=&2 LM"7'=%D!? A6NO7@89=&9>@89=&&7("7 ();P7Q7R7LPS7(IEE7AT7U7 9?%=%%>$&%%=%%V77;P7 $$%&'&()*+,-)-,./01*+ )*$2*))3)45 $%% D% 'D=%' </E,F-G/74HI.J,.K/ 2% @% 8% '2=2D ?% &% 'D=%D ?8=$@ 9% '% '$=D@ $% % % ? $% $? '% '? 9% )-A/7BA-.C 9? &% &? ?% ?? $$%&'&()*+,-)-,./01*+)*$2*))3)45 7WD&D8>D?'2 <)" ?8=9@>?8=?@ 43" @ LM" D=&9!& )" ();P7Q7R7LPS7(IEE7AT7U79?%=%%>$&%%=%%V @89=&&9%2 $%% S D% HN </E,F-G/74HI.J,.K/ 2% O @% HN 8% ?% &% N O NH N H @8&=&&89' NH O O O O O O 9% '% $% % @8$ @8' @89 @8& @8? A6N @88 @8@ @82 @8D $$%&'&()*+,-)-,./01*+)*$2*))3)45 7WD&D8>D?'2 <)" ?8=&?>?8=&D 43" ' LM" 9=D?!9 )" S);P7Q7K7LPS7J7(IEE7AT'7@89=&&X9?=%%7U7'%%=%%>@@?=%%V ??D='? $%% D% %!-&'()*+! </E,F-G/74HI.J,.K/ 2% 8'@='$ @% 8D?='% 8% %!,&'()*+! ?% &% 9% '% %&'()*+! 9D9=99 $% % '?% 9%% 9?% &%% &?% ?%% A6N ! "#$! ??% 8%% 8?% @%% @'@=$@ ! "#$! S4. E1. LC-FT-ICR %&'&()&*!+,-!(./,01),2/10!3,/!TTVTAC (7) (i)! E2. LC-FT-ICR spectrum of TTVTAC (7) (i) E3. MS-MS spectrum of TTVTAC (7) (i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pBrF-C (9) (i)! F2. LC-FT-ICR spectrum of TVPT-pBrF-C (9) (i) F3. MS-MS spectrum of TVPT-pBrF-C (9) (i) !"#$$%&'()*+,-.*.-/012+,*+31+*45*67 &8'(8'%$$9'"(%"$:95; <*" '=>=?9@9?A>%: N6"9:>=(!( B8OP9AAA>'(@AAA>'A9)"9 )*;Q9R9S9NQT9)KFF9BU9V9 :(%>%%@$&%%>%%W99;Q9 $$%&'()*+,-.*.-/012+, *+31+*45*67 (%>$? $%% =% <0F-G.H09IJK/L-/M0 E% A% ?% (%>&% (% &% :% '% &=>E% $% (%>A( % :% :( &% &( (% *.B09CB./D (( ?% ?( $$%&'()*+,-.*.-/012+,*+31+*45*67 9X?&?(@?&E$ <*" (%>$$@(%>$= I4" & N6" $>$%!( *" )*;Q9R9S9NQT9)KFF9BU9V9:(%>%%@$&%%>%%W AAA>'?&&' AA=>'?$=E $%% Br =% <0F-G.H09IJK/L-/M0 E% A% S ?% HN (% HO &% HN :% N AAE>'?AA( AE%>'?(($ OH NH N '% NH O O O O O $% AE$>'?E(E % A?? A?E AA% AA' AA& AA? AAE AE% AE' AE& AE? AEE B8O $$%&'()*+,-.*.-/012+,*+31+*45*67 9X?&?(@?&E$ <*" (%>$(@(%>'% I4" ' N6" &>=&!& *" T*;Q9R9M9NQT9L9)KFF9BU'9AA=>'?Y:(>%%9V9'%%>%%@A=%>%%W A$$>$( $%% =% <0F-G.H09IJK/L-/M0 E% A% ?% %&'()*+! (% &% :% '% $% % '%% ! '(% :%% :(% &%% &(% (%% B8O "#$! ((% ?%% ?(% A%% A(% ! "#$! S4.! G1. LC-FT-ICR ,(+(-.(/!01)!-2'134.15'43!61'!TVPT-pClF-C (10)! G2. LC-FT-ICR spectrum of TVPT-pClF-C (10) G3. MS-MS spectrum of TVPT-pClF-C (10) !!"#$%&'()*+,-.'*+/-,0(1+23450,0 #6$%6$"!!7$8$98""7:; <'8 ==>$!7?79@>A= KL87!>%BM9 C6NO7@"!>=B?@"!>A"7&87 &';P7Q717KPR7&I557CS7T7 ="">""?!A"">""U77;P7 !!"#$%&'()*+,-.'*+/-,0 (1+23450,0 B!>A# !"" B!>"# %" <-5+2*F-7GHI/J+/3- @" #" 9" B" A" =" $" !" 9$>B$ " =B A" AB B" '*C-7DC*/E BB 9" 9B !!"#$%&'()*+,-.'*+/-,0(1+23450,07V9@%A?9%@" <'8 B!>"!?B!>=9 GW8 # KL8 $>"%MB '8 &';P7Q717KPR7&I557CS7T7="">""?!A"">""U @"!>=#%== !"" Cl %" <-5+2*F-7GHI/J+/3- @" S #" HN N 9" B" O @"$>=@$9$ O O O O O HN A" @"=>=#9"" =" O NH N $" NH @"A>=#%!B !" " #%$ #%A #%9 #%@ @"" @"$ @"A C6N @"9 @"@ @!" @!$ @!A @!9 @!@ !!"#$%&'()*+,-.'*+/-,0(1+23450,07V9%"#?%=#= <'8 B!>"=?9A>=% GW8 !! KL8 $>!@MA '8 GF-,+X-7S1-32,IC7;P$7@"!>=@7D9%"#?%=#=E 99B>$@ !"" %" <-5+2*F-7GHI/J+/3- @" #==>$" %!#&'()*+! #" 9" B" %&'()*+! A" =" $" !" " $B" ! ="" =B" A"" AB" B"" C6N "#$! BB" 9"" 9B" #"" #B" @"" ! "#$! S4. H1. LC-FT-ICR $%&%'(%)!*+,!'-.+/0(+1.0/!2+.!TVPT-pClF-C (10) (i)! H2. LC-FT-ICR spectrum of TVPT-pClF-C (10) (i) H3. MS-MS spectrum of TVPT-pClF-C (10) (i) !!"#$%&'()*+,-.'*+/-,0(1+23450,0 #6$%6$"!!7$8$98""7:; <'8 =>?9=7@79!?9% KL87B?A#M9 C6NO7#==?=!@#==?=$7&87 &';P7Q717KPR7&I557CS7T7 =""?""@!>""?""U77;P7 !!"#$%&'()*+,-.'*+/-,0 (1+23450,0 >#?9! !"" %" <-5+2*F-7GHI/J+/3- A" #" 9" B" >#?AA >" =" $" >#?$9 !" " =9 =A >" >$ >> >9 >A B" '*C-7DC*/E B$ B> B9 BA 9" !!"#$%&'()*+,-.'*+/-,0(1+23450,07V9$!%@9$B% <'8 >#?>A@>#?99 GW8 B KL8 $?!9M9 '8 &';P7Q717KPR7&I557CS7T7=""?""@!>""?""U #==?=!BA$ !"" Cl %" <-5+2*F-7GHI/J+/3- A" S #" HN 9" HO B" >" #=>?=!%#> HN #=B?=!$## =" N OH NH N $" NH O O O O O #=9?=!9>" !" " #$9 #$A #=" #=$ #=> #=9 C6N #=A #>" #>$ #>> !!"#$%&'()*+,-.'*+/-,0(1+23450,07V9$$"@9$B% <'8 >#?B"@>#?B> GW8 $ KL8 %?A"M> '8 R';P7Q737KPR7J7&I557CS$7#==?=$X=B?""7T7!%"?""@#>B?""U 99B?$# !"" %" <-5+2*F-7GHI/J+/3- A" #" 9" B" 34.%,5&! >" =" $" !" " $"" $B" ="" =B" >"" >B" B"" C6N ! "##! BB" 9"" 9B" #"" ! "#"! S4. I1. LC-FT-ICR ,(+(-.(/!01)!-2'134.15'43!61'!TVPT-pOMeF-C (11) (i)! I2. LC-FT-ICR spectrum of TVPT-pOMeF-C (11) (i) I3. MS-MS spectrum of TVPT-pOMeF-C (11) (i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pN3F-C (12) (i)! J2. LC-FT-ICR spectrum of TVPT-pN3F-C (12) (i) J3. MS-MS spectrum of TVPT-pN3F-C (12) (i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spectrum of TVPVPSFC (20) Q3. MS-MS spectrum of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i)! R2. LC-FT-ICR spectrum of TVPVPSFC (20) (i) R3. MS-MS spectrum of TVPVPSFC (20) (i) !!"#$%&'()*+,-.'*+/-,0()'($"('12123&4 #5$%5$"!!6!$7$#78$629 :'7 8$;8<6=6<%;!# LM76$;?$N# A5OP6@@!;>?=@@!;>#6&76 &'936Q6R6L3S6&IDD6AT6U6 8"";""=!>"";""V66936 !!"#$%&'()*+,-.'*+/-,0( )'($"('12123&4 ><;%> !"" %" :-D+E*F-6GHI/J+/K- @" #" ?" <" >" 8" ><;"" $" !" " 8> 8? 8@ >" >$ >> >? '*A-6BA*/C >@ <" <$ <> !!"#$%&'()*+,-.'*+/-,0()'($"('12123&46W@!<!=@$<" :'7 ><;?@=>?;!" G17 % LM7 $;8%N? '7 &'936Q6R6L3S6&IDD6AT6U68"";""=!>"";""V @@!;>?"8$ !"" O <@ O HO %" S N H NH @" :-D+E*F-6GHI/J+/K- <? N #" N ?" O @@$;>?>"? <" >" O HN NH HN O N O OH O 8" $" @@8;>??8# !" " @#? @## @#@ @#% @@" @@! @@$ @@8 @@> A5O @@< @@? @@# @@@ @@% @%" @%! !!"#$%&'()*+,-.'*+/-,0()'($"('12123&46W@!<!=@$<" :'7 ><;@<=><;%" G17 $ LM7 !;#!N? '7 S'936Q6K6L3S6J6&IDD6AT$6@@!;>?X8<;""6U6$8";""=@%<;""V @!8;>8 !"" %" :-D+E*F-6GHI/J+/K- @" #" ?" <" 45/&-6'! >" 8" $" !" " $<" ! 8"" 8<" >"" ><" <"" <<" A5O "#$! ?"" ?<" #"" #<" @"" @<" ! "#"! S4.! S1. LC-FT-ICR ,'*'-.'/!01(!-2&134.15&43!61&!VTACITFC (22)! S2. LC-FT-ICR spectrum of VTACITFC (22) S3. MS-MS spectrum of VTACITFC (22) !!"#$"%&'()*+,-&)*.,+/'(&$#'0&123&%2 #4$"4$"!!5!$6!76895:; <&6 $7=>95?5#"=#" LM657=7>N7 A4OP5>8>=7#?>8>=7@5%65 %&;Q5R5S5LQ35%IEE5AT5U5 8#"=""?!7""=""V55;Q5 !!"#$"%&'()*+,-&)*.,+/' (&$#'0&123&%2 79=87 !"" >" <,E*F)G,51HI.J*.K, @" D" 9" 77=@$ #" 7" 8!=9D 8" $" 79=9" 79=D> !" 8@=D@ " $9 $@ 8" 8$ 87 89 8@ &)A,5BA).C 7" 7$ 77 79 7@ #" !!"#$"%&'()*+,-&)*.,+/'(&$#'0&123&%25W9@$D?9>7D <&6 7#=@@?79=## 106 !8 LM6 >=@9N8 &6 %&;Q5R5S5LQ35%IEE5AT5U58#"=""?!7""=""V >8>=7@#8! !"" >" <,E*F)G,51HI.J*.K, O O @" O D" NH S N H N 9" #" O NH >7"=7@@$$ O 7" 8" HN N HN H N S O O O $" >7!=7@D>D !" " >8" >8$ >87 >89 >8@ >7" >7$ A4O >77 >79 >7@ >#" >#$ >#7 !!"#$"%&'()*+,-&)*.,+/'(&$#'0&123&%25W##9@?9>#8 <&6 8@=@9?79=9" 106 > LM6 8=D7N8 &6 1G,+*X,5TS,KF+IA5;Q$5>8>=#!5B##9@?9>#8C @"8=8$ !"" >" $!+!%&'()*! <,E*F)G,51HI.J*.K, @" D" 9" @D!=$D #" 7" $%&'()*! 8" $" !" " $#" ! 8"" 8#" 7"" 7#" #"" ##" 9"" A4O "##! 9#" D"" D#" @"" @#" >"" >#" ! "#$! S4. T1. LC-FT-ICR %&'&()&*!+,-!(./,01),2/10!3,/!VTACITFC (22) (i)! T2. LC-FT-ICR spectrum of VTACITFC (22) (i) T3. MS-MS spectrum of VTACITFC (22) (i) !!"#$"%&'()*+,-&)*.,+/'(&$#'0&123&%2 #4$"4$"!!5!$6!76895:; <&6 8!=8$5>577=?! LM65!=A8N# B4OP5?@!=7$>?@!=785%65 %&;Q5R5S5LQ35%IEE5BT5U5 8#"="">!7""=""V55;Q5 !!"#$"%&'()*+,-&)*.,+/' (&$#'0&123&%2 89=?? !"" A" 89="" <,E*F)G,51HI.J*.K, ?" @" 9" #" 7" 8" $" !" " 8$ 88 87 8# 89 8@ 8? 8A &)B,5CB).D 7" 7! 7$ 78 77 !!"#$"%&'()*+,-&)*.,+/'(&$#'0&123&%25W#"9#>#!"! <&6 8#=?@>89="7 106 7 LM6 !=8!N# &6 %&;Q5R5S5LQ35%IEE5BT5U58#"="">!7""=""V ?@!=7$$#! !"" A" ?" <,E*F)G,51HI.J*.K, O HO O S N H NH @" N 9" #" O NH ?@$=7$##$ O 7" 8" HN N $" HN H N S O OH O ?@8=7$99? !" " ?## ?9" ?9# ?@" ?@# ??" B4O ??# ?A" ?A# A"" A"# !!"#$"%&'()*+,-&)*.,+/'(&$#'0&123&%25W#"9#>#!"! <&6 8#=?#>8#=?A 106 $ LM6 $=A?N7 &6 3&;Q5R5K5LQ35J5%IEE5BT$5?@!=7$X8#=""5U5$$#="">??#=""V ?"8=88 !"" A" <,E*F)G,51HI.J*.K, ?" @" 9" #" 7" 45/&-6'! 8" $" !" " $#" ! 8"" 8#" 7"" 7#" #"" ##" B4O "#$! 9"" 9#" @"" @#" ?"" ?#" ! "#$! O HN O NH O O NH O N H H N O N S HN O O O Figure S5. A1. 1H NMR spectrum of TSIASFC (15) in DMSO%!" &'!())!*+,! ! "#$! ! "##! Figure S5. B1. COSY spectrum of TSIASFC (15) in DMSO%!" &'!$((!)*+!#! ! ! "#$! ! "#$! NH HO HN N N O O O O O S HN Cl NH OH !"#$%&'()*'' %&'!!&(!)*+!,-./0123!45!67869-:;<9:=++>!in DMSO9!" ?0!@AA!*(B! ! ! "#$! ! ! ! "#$! ! !"#$%&'!()*!! !%&'!()"*!+,-./012!34!56758,(9:8(;++<!in DMSO8!" =/!>??!@AB!C!! ! ! ! "#$! ! "#$! NH O HN N O O O O O N HN S NH O !!"#$%&'()*'' !%&'!()"*!+,-./012!34!56/-7789!:!;+<!in DMSO=!" 6/!>??!@AB! ! ! "#$! ! "#$! %&'()*+,-!./!"/0!1233*45-!(/!6/0!7,)-!"/0!"89:+;<-!6/!=/!!"#$%&'!"#$!!"##-!()-!&>?$/! %>'@25<)*-!A/!=/-!B2+CC+,:-!./D/-!=,C<45-E/D/!*+,-./!"#$.!!""$-!00-!?&>?/! %?'=+F3-!G/0!H<)-!I/!1./234./!"#$.!!"""-!56-!&$?J/! %#'K,L5+MN+4-!O/!./0!P)M<45-!./!G/0!"<)2<-!O/!E/0!7C,5;Q-!E/0!A54C,*;-!7/!./!1./*$./!"#$./7'8.! #%%"-!(()-!R$R$/! ! ! ! ! ! "#$! CHAPTER 5 CONCLUSIONS ! ! ! ! ! ! ! ! ! "#$! Conclusions We set out to address the two outstanding problems in natural products drug discovery: the variability in occurrence and supply of natural products. Both of these challenges create a significant hindrance to the development of natural products into useful drugs. Although geographical and sample-to-sample variability in occurrence can occur in terrestrial natural products, 1 it can be particularly widespread for those natural products derived from marine organisms. Several reasons may explain variability of occurrence for marine natural products, such the complexity of ocean habitats 2,3 (local predation, competition, flux of microbes from seawater) as well as difficulties assigning species among heteromorphic organisms such as sponges. 4,5 Answering the question of ‘who makes what, when, and why?' is therefore highly pertinent to understanding the distribution of natural products and their potential. There is increasing evidence that symbiotic bacteria are the actual producers of most marine invertebrate-derived NPs. 6-9 We thus proposed that symbiotic bacteria are central to distribution of marine natural products. Further, we proposed that the factors that control symbiosis in turn control the distribution of natural products. Piece by piece, this hypothesis has been addressed in individual studies of chemical symbiosis. We asked if there are global patterns and what they would mean for drug discovery - global in this sense means examining all symbionts and chemicals within different animals and the changes in their distribution across time, location, and hosts. It was therefore critical that we use sequencing platforms capable of identifying thousands of bacteria in each sample. This approach allowed us to compare their general characteristics as a community to the other variables that were being tested. Statistical methods that are independent of ! "#"! metabolite identity were also critical for the overall analysis of chemistry between many samples. We found that in ascidians, the hosts generally select their symbionts and secondary metabolites. In known cases, the symbionts that are natural product producers are the most abundant components of the microbiome. This is an important finding in the era of genomics-based natural product discovery, not just in ascidians but also in other marine invertebrates. Marine invertebrates are microbial hotspots, 10,11 containing hundreds or thousands of different species of bacteria. Here we show that in mining for microbial producers of natural products, efforts should be focused on the most abundant microbes in the composition. Studying the latitudinal influence of geography to the symbionts and chemistry was very important as it revealed that the selection pressure between these regions could influence chemistry and symbiosis in imbalanced ways. While the symbionts generally maintained host specificity, the potency of secondary metabolites increased in tropical samples. Therefore, within the limits of our sampling and analysis, symbiosis is generally host-specific while secondary metabolism can be location specific when there are large biogeographical divides such as water temperature and species diversity. While it is tempting to generalize our results to a broader section of marine invertebrates, their symbionts, and associated metabolites, it is important to reiterate the limits of our data. In particular, our tests on the potency of extracts were limited to a select few strains, which may not capture ecologically relevant activities of the extracts from temperate ascidians. Additionally, since almost none of the chemistry can be identified from the temperate ascidians in the database collections, it is difficult to say ! "#$! precisely what differences in chemistry exist between closely related species in temperate and tropical regions. In this respect, methods that can reveal structural relatedness without dependence on complete identification and isolation of compounds, such as molecular networking, will be useful in future studies. 12 To address the supply of symbiotic natural products, we used the heterologous expression of cyanobactins as a model case. Cyanobactin pathways are not only amenable to library generation; they are also amenable to metabolic engineering, specifically through precursor supply and simple fermentation optimization. Serendipitously, we discovered that increasing the precursor DMAPP improved cyanobactin production, in a manner that is independent of compound prenylation. The scope of this effect is of great interest in that it might be applicable to other heterologous systems such as to other RiPPs, or other natural product pathways in general. The ability to supply cyanobactins will now allow further practical investigations on their chemistry and biology. Indeed this method allowed us to uncover the first activity for patellin 2 using a high content phenotypic assay. This opens the possibility of finding more activities for ‘orphan' cyanobactins and may lead to a clearer picture of their ecological roles. We further demonstrated the capacity of cyanobactin pathways to generate unusual derivatives containing nonproteinogenic amino acids. By incorporating unnatural amino acids to the cyanobactin backbone, we successfully combined the structural complexity of non-ribosomal peptides with a multistep, substrate tolerant ribosomal pathway. ! "#$! References 1 Marienhagen, J. & Bott, M. Metabolic engineering of microorganisms for the synthesis of plant natural products. J Biotech 163, 166-178, doi:10.1016/j.jbiotec.2012.06.001 (2013). 2 Hay, M. E. Marine chemical ecology: what's known and what's next? J Exp Mar Biol Ecol 200, 103-134, doi:http://dx.doi.org/10.1016/S0022-0981(96)02659-7 (1996). 3 Li, J. W. & Vederas, J. C. Drug discovery and natural products: end of an era or an endless frontier? Science 325, 161-165, doi:10.1126/science.1168243 (2009). 4 Cardenas, P., Perez, T. & Boury-Esnault, N. Sponge systematics facing new challenges. Adv Mar Biol 61, 79-209, doi:10.1016/B978-0-12-387787-1.00010-6 (2012). 5 Morrow, C. & Cardenas, P. Proposal for a revised classification of the Demospongiae (Porifera). Front Zool 12, 7, doi:10.1186/s12983-015-0099-8 (2015). 6 Florez, L. V., Biedermann, P. H., Engl, T. & Kaltenpoth, M. Defensive symbioses of animals with prokaryotic and eukaryotic microorganisms. Nat Prod Rep, doi:10.1039/c5np00010f (2015). 7 Piel, J. Metabolites from symbiotic bacteria. Nat Prod Rep 26, 338-362, doi:10.1039/b703499g (2009). 8 Schmidt, E. W. Trading molecules and tracking targets in symbiotic interactions. Nat Chem Biol 4, 466-473, doi:10.1038/nchembio.101 (2008). 9 Schmidt, E. W., Donia, M. S., McIntosh, J. A., Fricke, W. F. & Ravel, J. Origin and variation of tunicate secondary metabolites. J Nat Prod 75, 295-304, doi:10.1021/np200665k (2012). 10 Hentschel, U., Piel, J., Degnan, S. M. & Taylor, M. W. Genomic insights into the marine sponge microbiome. Nat Rev Microbiol 10, 641-654, doi:10.1038/nrmicro2839 (2012). 11 Erwin, P. M., Pineda, M. C., Webster, N., Turon, X. & Lopez-Legentil, S. Down under the tunic: bacterial biodiversity hotspots and widespread ammoniaoxidizing archaea in coral reef ascidians. Isme J, doi:10.1038/ismej.2013.188 (2013). ! 12 "#$! Nguyen, D. D. et al. MS/MS networking guided analysis of molecule and gene cluster families. Proc Natl Acad Sci U S A 110, E2611-2620, doi:10.1073/pnas.1303471110 (2013).