Graduate Research Summary – Cyanobacterial Biosynthetic Pathways Deepti Sharma Overarching Summary Recently, I have successfully defended a Ph. D. thesis titled “Harnessing Genomes and Building Molecules for Investigating Biosynthetic Mechanisms in Model Group V Cyanobacteria”. Herein, I present an introduction and highlights of key discoveries I developed during my thesis studies. My thesis unraveled the biosynthetic pathways encoded by Fischerella genus and is currently either published or in the process of preparation for publications.

Introduction Traditionally, microbes and plant-derived small

molecules have contributed to a significant ratio of therapeutics currently in use. Between the years 1990-2005, the field of natural product-based drug discovery was sidelined (in preference to synthetic libraries) due to slow progress attributed to the tediousness of extraction and purification and/or synthesis of these complex molecules. The past ten years, however, has witnessed a renaissance for natural product chemistry, spurred from to the marriage between traditional tools (synthesis, extraction, structure elucidation) and powerful modern technologies (shotgun sequencing, bioinformatics, in vitro expression of genes and high-level accuracy of mass spectrometry). Given this general excitement, I chose to study terrestrial cyanobacteria belonging to the Group V classification for their potential biosynthetic prowess, for my thesis project. The premise of my research is succinctly presented in Figure 1.

Results – Genome Sequencing and Assembly

At the start of my thesis, practically no genomes from Group V cyanobacteria were publically available, and therefore I invested a significant portion of efforts to generate draft genomes of: Fischerella sp. ATCC43239 and Fischerella ambigua UTEX1903. Whole genome sequencing using Illumina GAIIx platform was performed on paired end libraries obtained using Nextera DNA sample preparation kits. The project involved two more Group V cyanobacterial strains: Westiella intricata UH strain HT-29-1 and Hapalosiphon welwitschii UH strain IC-52-3 and their genome drafts were generated by our collaborator’s group (UWS). Biosynthetic Characterization through Gene Cluster Identification

Aforementioned cyanobacterial strains along with all publically available Group V cyanobacterial genomes were mined for their genetic potential to produce secondary metabolites. A putative gene cluster for the cyclic depsipeptide hapalosin, known to reverse P-glycoprotein multiple drug resistance, was identified within three Group V cyanobacterial genomes, including the producing cyanobacterium H. welwitschii UH strain IC-52-3 as presented in Figure 2.

A number of orphan NRPS/PKS gene clusters and ribosomally synthesised gene clusters (including cyanobactin, microviridin and bacteriocin gene clusters) were identified. Furthermore, gene clusters encoding the biosynthesis of microcystin, mycosporine-like amino acids, scytonemin, hydrocarbons (alkanes) and terpenes were also identified and compared. Identification of microcystin and alkane biosynthetic gene clusters was also complemented by direct observation of metabolites through metabolic extractions and mass spectrometry.

Figure 1. Schematic representation of pathway discovery from filamentous terrestrial cyanobacteria. My thesis focused on unraveling novel biosynthetic machinery encoded by terrestrial Group V cyanobacteria.

Figure 2. Hapalosin biosynthetic gene cluster from H. welwitschii IC-52-3, W. intricata HT-29-1 and F. sp. PCC 9431

Graduate Research Summary – Cyanobacterial Biosynthetic Pathways Deepti Sharma Biosynthetic Characterization of Hapalindole Family of Terpenoid Indole Alkaloids The hapalindole family of natural products is a group of hybrid isoprenoid-indole alkaloids (with over 80 variations identified to date) produced solely by members of the subsection V cyanobacterial strains. This family broadly includes the hapalindoles, welwitindolinones, fisherindoles and ambiguines, all of which contain an isonitrile- or isothiocyanate-containing indole alkaloid skeleton, with a cyclized isoprene unit. The hapalindoles are diversified into the welwitindolinones, fischerindoles and ambiguines through the employment of tailoring oxygenase, methyltransferase and prenyltransferase enzymes. A feeding study identified putative precursors of the hapalindole pathway in 1988, but the biosynthetic genes and enzymes remained to be unambiguously characterized. The genome mining of the four aforementioned Group V cyanobacterial strains led to the identification of a candidate hapalindole-type gene cluster (hpi) in each strain as presented in Figure 3. This study presents a detailed comparative bioinformatic analysis of the hpi gene clusters. WelI1 and WelI3 are characterized for the biosynthesis of a pivotal indole-isonitrile intermediate. Reconstitution studies and metabolite extractions from two hapalindole-producing Fischerella strains indicate the presence of cis and trans indole-isonitriles as biosynthetic intermediates. This study illuminated the initial stage of the pathway and also reports findings which are essential to the overall elucidation of the unusual mechanism of the biosynthesis of hapalindole family of compounds. The study shows promising leads to combinatorially using these genes and enzymes for hapalindole library construction.

Regioselective Cope Rearrangement and Prenyl Transfers on Indole Scaffold Mimicking Fungal and Bacterial Dimethylallyltryptophan Synthases Tryptophan-containing alkaloids and peptidyl natural products are prenylated by a monophyletic sub-class of aromatic prenyltransferases in cyanobacteria, fungi and a few plants. HpiP1/AmbP1/WelP1 from hapalindole type natural products gene clusters and LtxC from the lyngbyatoxin gene cluster are few examples from cyanobacteria. These enzymes lack an obvious prenyl diphosphate binding site and exhibit catalytic activities in the absence of divalent cations. Both of these enzymes catalyze regiospecific addition of a 10-carbon geranyl group to L-tryptophan scaffold; however not much is known about the catalytic mechanism. Another class of these enzymes is DMATS (dimethylallyl tryptophan synthase) and these enzymes catalyze the transfer of a prenyl group from dimethylallyl diphosphate onto the indole ring of tryptophan (Figure 4). At a fundamental level, the mechanism of C4-prenylation of L- tryptophan recently has surfaced to engage a debate between a “direct” electrophilic alkylation mechanism (for wt DMATS and FgaPT2) versus an indole C3-C4 “Cope” rearrangement followed by rearomatization (for mutant FgaPT2).

Synthetic mimicry of these selective biocatalysts has been rare primarily due to the challenge in taming regio and stereoselectivities of such transformations In this project, regioselectively tunable conditions for a Cope rearrangement between C3 and C4 positions and biomimetic conditions effecting [3,3]-sigmatropic shift are studied. These conditions are expected to find use in natural product total synthesis considering the urgent need for stereoselective methods for complex alkaloid construction.

Figure 3. 2 Illustration of the hapalindole (hpi), ambiguine (amb) and welwitindolinone (wel) biosynthetic gene clusters

Figure 4. Biomimicry of enzymatic reaction-catalyzed by C4-Dimethylallyl Tryptophan Synthase (4-DMATS) through regioselective Cope rearrangement and prenyl transfers on indole scaffold.

Graduate Research Summary – Cyanobacterial Biosynthetic Pathways Deepti Sharma Future Studies

The accumulation of large amount of sequencing data in public databases has revealed that many microbial genomes contain diverse gene clusters which possibly encode for natural product, however, they are not associated with the biosynthesis of known natural products. These are referred to as Orphan biosynthetic pathways and number of these pathways outnumbers the currently known metabolites for a given organism. Multiple orphan hybrid NRPS/PKS gene clusters have been identified from aforementioned Group V cyanobacterial strains. In silico analysis enabled the predictions of the domains within each NRPS and PKS module along with the substrate preference of each acyltransferase (AT) and adenylation (A) domain. The identification of signature amino acid sequences suggests catalytically active domains. These substrate specificity predictions combined with the co-linearity logic of NRPS and PKS systems have led to preliminary results with the ability to derive structural elements encoded by these systems; however, these predictions require further characterizations which can be achieved by following complementary techniques under the umbrella of genome mining. One of the approaches will involve MS imaging, where metabolite profile of the cyanobacterial cells will be captured in vivo directly from the petri dish. This will effectively eliminate the need for any sample preparation steps that may complicate molecular characterization accuracy. Further, cloning and expression of the intact biosynthetic gene cluster in an amenable heterologous host such as S. coelicolor is expected to be useful in generating reasonable quantity (5-50 mg) of the respective product for characterization by NMR.

Publications

This graduate work comprises data for four independent research publications In addition; my graduate work also involves publishing, upon invitation, a contemporary review on current advances in natural product biosynthesis, based on discussions in the introduction of my thesis.

Melinda L. Micallef*, Deepti Sharma*, Brittney M. Bunn, Lena Gerwick, Rajesh Viswanathan, Michelle C. Moffitt. "Comparative analysis of hapalindole, ambiguine and welwitindolinone gene clusters and reconstitution of indole-isonitrile biosynthesis from cyanobacteria." BMC Microbiology, 14, 213, 2014. * Equal contribution

Karthikeyan Thandavamurthy, Deepti Sharma, Suheel, K. Porwal, Dale Ray, Rajesh Viswanathan*. “Regioselective Cope Rearrangement and Prenyl Transfer on Indole Scaffold Mimicking Fungal and Bacterial Dimethylallyltryptophan Synthase.” J. Org. Chem, 21, 10049, 2014. Deepti Sharma*, Melinda L. Micallef*, Ormond Brathwaite, Paul M. D’Agostino, Michelle C. Moffitt, Rajesh Viswanathan “Identification Metabolite Profiling of Mycosporine, Microcystin and Alkane Biosynthetic Gene Clusters Group V cyanobacteria." manuscript in preparation. Melinda L. Micallef, Paul M. D’Agostino, Deepti Sharma, Rajesh Viswanathan, Michelle C. Moffitt. “Genome mining for natural product biosynthetic gene clusters in the subsection V cyanobacteria." manuscript submitted.