Blueprints of Blue From evolutionary origins to biosynthetic pathway engineering: exploring the sustainable production of the fungal cyan pigment xylindein

Blueprints of Blue Yanfang Guo 2025 Blueprints of BLUE From evolutionary origins to biosynthetic pathway engineering: exploring the sustainable production of the fungal cyan pigment xylindein Yanfang Guo Propositions 01. Beyond its potential as a natural dye, xylindein represents a promising material for sustainable electronics, particularly for organic solar cell applications. 02. The slowing pace of new bioactive natural product discovery highlights the greater value of elucidating biosynthetic pathways of known compounds and developing their sustainable production. 03. Genome mining and comparative genomics are indispensable tools for uncovering the evolutionary origin and diversification of fungal secondary metabolism. Chapter 2 04. There is no perfect host for heterologous expression because different biosynthetic pathways have different requirements. 05. Combining genes from different biosynthetic pathways provides a strategy to generate novel metabolites and enhance the properties of known compounds. Chapter 2 & Chapter 3 06. Gene duplication and differential gene recruitment, as major evolutionary mechanisms, drive the diversification of fungal fatty acid synthases and their associated secondary metabolite biosynthetic pathways. Chapter 4 07. Completing a project is often harder than initiating it, and this is true for many other aspects of life. 08. A PhD is full of challenges, each like a stepping stone, never an obstacle and always a chance to rise and grow. Propositions belonging to the thesis, entitled: Blueprints of Blue From evolutionary origins to biosynthetic pathway engineering: exploring the sustainable production of the fungal cyan pigment xylindein Yanfang Guo Groningen, 1 December 2025 Blueprints of Blue From evolutionary origins to biosynthetic pathway engineering: exploring the sustainable production of the fungal cyan pigment xylindein Yanfang Guo 2025 The research presented in this thesis was conducted in the Fungal Natural Products group at Westerdijk Fungal Biodiversity Institute, and in the Department of Molecular Microbiology, part of the Groningen Biomolecular Sciences and Biotechnology Institute (GBB), University of Groningen, The Netherlands. The PhD was financially supported by the China Scholarship Council fellowship (CSC No. 202107720104). Author: Yanfang Guo Email: [email protected] WeChat: 15137184208 Layout: Zhenguo Wang Cover: Zhenguo Wang, Yanfang Guo Printed by: ProefschriftMaken, Utrecht - https://www.proefschriftmaken.nl/ Copyright © 2025 by Yanfang Guo All rights reserved. No part of this thesis may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording or any information storage or retrieval system, without prior permission of the author. Blueprints of Blue From evolutionary origins to biosynthetic pathway engineering: exploring the sustainable production of the fungal cyan pigment xylindein PhD thesis to obtain the degree of PhD of the University of Groningen on the authority of the Rector Magnificus Prof. J.M.A. Scherpen and in accordance with the decision by the College of Deans. This thesis will be defended in public on Monday 1 December 2025 at 11:00 hours by Yanfang Guo born on 9 December 1993 in Zhengzhou, China Thesis Committee Promoter: Prof. Dr. A.J.M. (Arnold) Driessen University of Groningen Co-promoter: Dr. J. (Jérôme) Collemare Westerdijk Fungal Biodiversity Institute Opponents: Prof. Dr. S. (Soizic) Prado, Museum National d'Histoire Naturelle, Paris Prof. Dr. Ir. M.W. (Marco) Fraaije, University of Groningen Prof. Dr. A.F.J. (Arthur) Ram, Leiden University Dr. K. (Kristina) Haslinger, University of Groningen Dr. K.J. (Kira) Tiedge, University of Groningen Contents Chapter 1 General introduction 7 Chapter 2 Identification of a biosynthetic gene cluster for the production of the blue-green pigment xylindein by the fungus Chlorociboria aeruginascens 33 Chapter 3 Characterisation of tailoring enzymes in xylindein biosynthesis reveals a functionally conserved pyranone reductase and a putative neofunctionalised carbonic anhydrase 75 Chapter 4 Evolution of fungal fatty acid synthases from primary to secondary metabolism 115 Chapter 5 General discussion and summary Samenvatting 中文总结 159 Appendix Acknowledgements About the author List of publications 181 6 7 Chapter 1 General Introduction Chapter 1 14 In fungi, genes encoding both core and tailoring enzymes involved in a given biosynthetic pathway are typically co-regulated and co-localised in the genome, forming BGCs (Figure 3).8,27 Most BGCs contain a single core gene along with a variable number of tailoring genes; however, some BGCs harbour multiple core genes that enable the production of hybrid molecules. For example, the biosynthesis of macrolide lactones involves a conserved pair of nrPKS and rPKS enzymes.28 Similarly, the production of meroterpenoids requires the coordinated activity of both PKSs and TCs.29 In addition to biosynthetic enzymes, BGCs may also encode pathway-specific transcription factors that regulate gene expression30, protein decoys that offer self-protection31, and transporter proteins involved in SM export or cellular detoxification.32 Fungi have a remarkable ability to produce a wide variety of pigments, thanks to the diverse combinations of BGCs in their genomes. With the rapid development of next-generation sequencing technologies and lower sequencing costs, more and more fungal genomes are being decoded. This growing collection of genomic data is opening exciting new opportunities to explore fungal pigments. To date, over 36,000 SMs have been recorded in the open-access Natural Product Atlas.33 At the same time, bioinformatics tools like antiSMASH, the MiBIG database, and CBlaster are making it easier to identify and understand these biosynthetic pathways by comparing them with already known BGCs.34,35 Heterologous expression of fungal BGCs and pathway engineering One concern regarding pigment production through fungal fermentation is the potential co-production of mycotoxins. This issue can be mitigated through bioengineering strategies such as heterologous expression, where target genes or entire BGCs are transferred from the native strain into a genetically tractable, nontoxigenic host.36 By expressing BGCs in well-characterised hosts, the safety and efficiency of compound production can be significantly improved. For example, in the total biosynthesis of pleuromutilin, a potent inhibitor of the prokaryotic ribosome, heterologous expression of the seven-gene BGC in Aspergillus oryzae achieved a yield of 84 mg L-1, compared to just 4 mg L-1 from the original producer, the basidiomycete Clitopilus passeckerianus.37,38 Commonly used heterologous expression hosts include Escherichia coli, yeast, and filamentous fungi. These organisms are favoured due to their rapid growth, genetic amenability, and compatibility with laboratory and industrial workflows.39 In particular, A. oryzae, a food-grade filamentous fungus used in traditional General Introduction 15 fermentation, is an excellent host for expressing fungal SM pathways.40 It is generally recognised as safe (GRAS) and does not naturally produce mycotoxins, simplifying downstream purification. Another industrial workhorse is Penicillium rubens, widely used for β-lactam antibiotic production. A strain derivative has been developed in which four BGCs, including the β-lactam BGC, have been deleted to generate a clean metabolic background for heterologous expression.41 42 Similar efforts in hosts like Trichoderma reesei demonstrated that natural products such as 3-methylorcinaldehyde and pretenellin A can be produced directly from waste materials like fruit peels and barley straw, offering a sustainable way to convert biomass into valuable compounds.15 Yeast systems have also gained attraction in natural product research. For example, a synthetic biology platform known as heterologous expression synthetic biology platform (Hex) has enabled the expression of 41 fungal BGCs in Saccharomyces cerevisiae, with over half yielding detectable products, demonstrating the utility of yeast in fungal metabolite discovery.43 Fungi are estimated to comprise 2.2 to 3.8 million species in nature, and each fungal genome can harbour 10 to 80 BGCs.11,22 However, many of these BGCs remain silent under standard laboratory conditions.22 Several strategies have been developed to activate these cryptic pathways and uncover novel secondary metabolites, including epigenetic manipulation, co-cultivation with other microbes, manipulation of transcriptional regulators, and heterologous expression.22 Among them, heterologous expression stands out as a powerful approach for both exploring and efficiently producing fungal natural products. For example, activation of a silent azaphilone BGC in Aspergillus niger ATCC 1015 led to the discovery of six azaphilone compounds, azanigerones A–F.44 The biosynthetic steps involve coordinated actions of a highly reducing PKS and a non-reducing PKS, while a flavin-dependent monooxygenase was shown to catalyse hydroxylation and pyran ring formation, establishing the core azaphilone scaffold.44 Heterologous expression enables the total biosynthesis of complex metabolites in a single fermentation process, bypassing the limitations of native producers. Classic examples include the transfer of the full penicillin BGC from P. rubens into A. niger, and the expression of the citrinin BGC from Monascus purpureus in A. oryzae, demonstrating production in safe, genetically tractable hosts.45,46 This approach also supports pathway elucidation. By expressing genes stepwise in heterologous systems, the functions of individual biosynthetic enzymes can be determined. For instance, the pleuromutilin BGC from Clitopilus pseudopinsitus was dissected in A. oryzae, Chapter 1 16 revealing the roles of seven genes, including terpene synthases, cytochrome P450s, a dehydrogenase, and an acyltransferase.38 Similarly, the biosynthesis of viriditoxin by Paecilomyces variotii was fully elucidated through the expression of the biosynthetic genes in A. nidulans.47,48 Finally, heterologous expression opens the door to pathway engineering, allowing for the creation of novel chemical diversity. This can be achieved by truncating biosynthetic pathways, which results in the accumulation of intermediates or shunt metabolites. Alternatively, mixing and matching genes from related pathways can yield new congeners with modified structures. Modifications to core biosynthetic enzymes such as PKSs can alter the carbon backbone of the product, while engineering tailoring enzymes offers opportunities to fine-tune functional groups and stereochemistry.16 Altogether, these strategies highlight the versatility of heterologous hosts not only as expression platforms but also as powerful tools for synthetic biology and natural product innovation. The pyranone chemical family Fungal pigments represent a chemically diverse group of natural products, encompassing various structural families such as polyketides, terpenoids, and nitrogen-containing compounds. Within these groups, several classes are wellknown for producing vibrant colours, including azaphilones with their orange-red hues, naphthoquinones that exhibit deep reds and purples, and anthraquinones known for bright yellows and oranges. Among these colourful metabolites, compounds containing a pyranone ring stand out for their wide colour range, biosynthetic versatility, and potential industrial relevance. A diversity of fungal pyranone pigments Pyranones are a distinct class of compounds featuring a pyranone ring, a sixmembered ring structure that incorporates one oxygen atom and one conjugated double bond. These compounds are commonly found in fungi. The pyranone family encompasses a range of fungal natural pigments known for their striking colours and notable biological activities (Figure 4). For instance, aurofusarin, produced by Fusarium graminearum, exhibits pHdependent colour changes, appearing yellow in acidic conditions and reddish in alkaline environments (Figure 4).49 This compound also demonstrates bioactivities such as antibiotic, cytotoxic and genotoxic effects, and the ability to induce oxidative stress.50,51 Another notable example is xylindein, a cyan pigment secreted by the wood-staining fungus Chlorociboria aeruginascens and related species, historically General Introduction 17 Figure 4. Representative fungal pyranone compounds, their associated colours, and reported bioactivities. used in the art of wood spalting due to its vibrant blue-green colouration (Figure 4).10,52,53 Similarly, viriditoxin, a well-known mycotoxin, exhibits antibacterial and other bioactive properties (Figure 4).47 Furthermore, laccaridiones A, B, and C, purple pyranones produced by Laccaria amethystina and Montagnula species, are notable for their potent biological activities, including protease inhibition, antifungal effects, and anticancer potential (Figure 4).54,55 These examples illustrate the diversity of pyranone pigments in terms of both colour and biological function, highlighting their potential applications in natural pigment production and pharmaceutical development. In addition to the monomeric pyranones like rubrofusarin56,57 and obionin A58, many pyranone dimers have been isolated and characterised (Figure 4). Some of these dimers feature 6,6’-linkage, including compounds such as asteromine57, talaroderxines59,60, pigmentosins59,61, and aschenaphtopyrone A62, while others possess 8,8’-linkage, such as vioxanthin61,63, mycopyranone64, aschenaphtopyrone B62, and lichenocholin A65 (Figure 4). A particularly rare 5,8’-linkage has been reported in lulworthinone, which was isolated from a marine-derived fungus Lulworthia medusa66,67 (Figure 4). Chapter 1 18 Biosynthetic pathways of fungal pyranone pigments Despite the isolation and characterisation of numerous pyranones, only a few biosynthetic pathways have been investigated. The elucidation of the biosynthetic routes for compounds like rubrofusarin, aurofusarin, viriditoxin, and citrinin has provided valuable insights into the enzymatic processes involved (Figure 5). In rubrofusarin biosynthesis, Pks12 generates the initial polyketide intermediate YWA1, which undergoes cyclisation to form the backbone. Subsequent tailoring steps include dehydration by AurZ and methylation by AurJ to yield rubrofusarin (Figure 5).68,70 The pathway for aurofusarin begins with the same steps and includes additional modifications by other tailoring enzymes. In particular, dimerisation occurs via an oxidative coupling catalysed by the laccase enzyme Gip1, leading to the characteristic dimeric structure of aurofusarin.68,70 In the case of viriditoxin, the nrPKS VdtA first synthesises the core structure, which is then modified by a series of methylation, oxidation and reduction mediated by VdtC, VdtE and VdtF, respectively. Additional coupling synthesised by the laccase VdtB and auxiliary protein VdtD results in the bioactive viriditoxin (Figure 5).47,48 For citrinin, the biosynthetic pathway involves the synthesis of an unreduced trimethylated pentaketide by the nrPKS CitS.69 The nrPKS is assisted by a hydrolysis step catalysed by CitA to release the first intermediate. Then, the non-heme iron oxidase CitB oxidises the methyl group of the intermediate to an alcohol. Subsequent steps are catalysed by CitC which oxidises the alcohol to an aldehyde, and CitD which converts the aldehyde to a carboxylic acid. The final reduction by CitE yields citrinin (Figure 5).69,71 The different release mechanisms from the nrPKS result in different pyrone products as aurofusarin is a gamma-pyrone, viriditoxin is an alpha-pyrone, while the pyranone ring in citrinin is due to CitE and not to the PKS. These pathways highlight the complexity and diversity of pyranone production in fungi, with key biosynthetic enzymes such as nrPKSs and laccases playing crucial roles in the formation of these pigments. The understanding of these pathways is essential for exploring the potential of pyranones in biotechnological applications, including the production of novel bioactive compounds and the engineering of natural pigments with additional functional properties. General Introduction 19 Figure 5. Biosynthetic gene clusters and biosynthetic routes of rubrofusarin, aurofusarin, viriditoxin and citrinin. (Adapted from Frandsen et al., 2011; Hu et al., 2019; He and Cox, 2016)48,68,69 A Unique fungal pigment with potential optoelectronic application: xylindein Among fungal pyranones, xylindein, a blue-green pigment produced by the fungus C. aeruginascens (Figure 6A), has attracted significant interest due to its remarkable stability and diverse applications.10,53 This fungus is commonly found in forests worldwide and easily recognised as it grows on dead wood which is stained blue. Historically, xylindein-stained wood has been used in wood art for centuries and those pieces of art still exhibit the vibrant colour nowadays (Figure 6B).53 Chapter 1 20 Figure 6. The stability and optoelectronic properties of xylindein make it promising for organic solar cells. (A) Chlorociboria aeruginascens grows on wood and stains it blue due to the production of the blue-green organic natural pigment xylindein. (B) Xylindein-stained wood remained coloured in wood art from the 15th century (Adapted from Gutierrez and Robinson, 2017).53 (C) Photobleaching of xylindein in solution. Xylindein shows great resistance to light (Adapted from Harrison et al., 2017).72 (D) The absorption of xylindein film after having been annealed at successively higher temperatures in air. Xylindein shows resistance to heat up to 180 ¡C (Adapted from Harrison et al., 2017).72 Table 1. Electron mobility (μ; cm2 V-1 s-1) of common organic semiconductors in organic solar cells Semi-conductor μ Origin iso-BBT-4T:PCBM 1.3 synthetic Xylindein 0.4 natural PCPDTBT 10-2 synthetic P3HT:PCBM 10-4 synthetic PM6:Y5 10 -4 synthetic PPDT2FBT:Y6 10-4 synthetic PBDB-TF:BTP-4Cl 10-4 synthetic Adapted from Chen et al., 2020; Vollbrecht et al., 202376,77 General Introduction 21 Xylindein was reported to be very stable under light exposure and resistant to temperatures up to 180 oC (Figure 6C, 6D).72,73 This stability contributes to its protective role in C. aeruginascens, likely functioning as a UV-protectant to enhance the survival of this slow-growing fungus.74 Additionally, toxicity studies have demonstrated that purified xylindein exhibits no toxicity to zebrafish embryos at concentrations ranging from 0 to 50 mM75, supporting its potential application in textiles, food products, and artworks.72,73 Recent studies have also highlighted the optoelectronic properties of xylindein, suggesting promising applications as a semiconductor in organic solar cells (OSCs), which represent emerging thirdgeneration solar cell technologies.73 A major parameter that determines the energy conversion efficiency is electron mobility ( µ ). Organic semiconductors typically exhibit values in the 10-2 to 10-4 cm2 V-1 s-1 range (Table 1), with the PCPDTBT compound being considered to have excellent mobility. Electron mobility for xylindein was found to be up to 1000-fold higher and reached 0.4 cm2 V-1 s-1 (Table 1), making this fungal compound particularly promising for efficient OSCs.76,77 Xylindein is thus a natural compound that seems to be superior to most organic semiconductors which are synthetic compounds. However, a major limitation of xylindein development for OSCs and other applications is its low levels of production. C. aeruginascens grows slowly and efforts to increase xylindein production in fermenters are yet not compatible with high-yield production. Xylindein cannot be easily produced by chemical synthesis78, and the extraction and purification of xylindein is cumbersome because of its low solubility in most organic solvents.79,80 Various attempts have been made to optimise the growth of C. aeruginascens and improve xylindein production. Robinson and colleagues discovered that supplementing malt agar media with specific wood types, including Acer saccharum, Populus tremuloides, spalted P. tremuloides, and Ailanthus altissima, could stimulate the growth of C. aeruginascens and enhance xylindein production.80 Further studies have compared the effects of various nutrients and environmental growth factors. While high nutrient concentrations promoted biomass production without pigmentation, low nitrogen concentrations favoured pigmentation but resulted in reduced biomass.81 Thus, nitrogen content appears to be a critical factor for optimal xylindein production.81 The optimised growth conditions for C. aeruginascens have been determined as a temperature of 20 °C, pH 4, dark conditions, and aerobic cultivation.74 Recent advances have successfully scaled up xylindein production from 3 to 70 L bioreactor cultivations, identifying key bioprocess parameters such as low shear stress (150 rpm, tip speed < 0.5 m s-1) to achieve optimal pigment yield (4.8 mg L⁻¹ d⁻¹).82 Chapter 1 22 Despite these efforts, current production levels remain insufficient to meet the demand for xylindein as a functional material in OSCs and other potential applications. Therefore, finding a cheap, simple and efficient way to produce xylindein is needed to develop this promising pigment for diverse applications. Elucidating the biosynthetic pathway of xylindein and mastering its biosynthesis in a fermentation process using a heterologous host offers a highly efficient alternative to the low production and slow growth of Chlorociboria species. Aims and outline of this thesis This thesis explores the evolutionary and functional aspects of the biosynthesis of fungal secondary metabolites, focusing on the xylindein biosynthetic pathway and its related pyranone compounds. We investigate these pathways by combining bioinformatics and heterologous expression approaches to discover and understand the origin, evolution and function of enzymes involved in the biosynthesis of xylindein and related pyranones. In Chapter 2, we focused on the identification of the BGC involved in the production of the blue-green pigment xylindein and on the elucidation of the first biosynthetic steps. We re-sequenced the genome of C. aeruginascens as well as that of the related C. aeruginosa species. Genome mining and phylogenetic dereplication identified a candidate BGC containing a non-reducing polyketide synthase (XLNpks) and a pair of fatty acid synthase (XLNfas1 and XLNfas2) genes. RNA sequencing established the co-regulation of genes within the candidate BGC. We performed heterologous expression in Aspergillus oryzae of XLNpks, XLNfas1 and XLNfas2, as well as of the related viriditoxin nrPKS. Altogether, it allowed us to propose a putative biosynthetic pathway for xylindein in Chlorociboria species. In Chapter 3, we explored the tailoring steps in the biosynthesis of xylindein and related compounds, with an emphasis on the unique dimerisation pattern that distinguishes xylindein from talaroderxine A, a predicted xylindein intermediate produced by Talaromyces derxii. Comparative genomic analyses identified homologous BGCs confirming that the candidate gene XLNcnh is unique and potentially responsible for the distinctive dimerisation of xylindein. Phylogenetic analysis of the candidate xylindein tailoring enzymes was performed to investigate their origin and evolution. Using heterologous expression in A. oryzae, we coexpressed the viriditoxin nrPKS (vdtA) with different tailoring enzymes, characterising the tailoring steps that lead to the production of intermediates related to talaroderxine A and viriditoxin. This chapter advanced our understanding of General Introduction 23 xylindein biosynthesis and highlighted the potential of tailoring enzymes for engineering applications. In Chapter 4, we investigated further the evolutionary origin and divergence of fungal fatty acid synthases (FASs) and the mechanism of their recruitment from primary metabolism to secondary metabolism. Using phylogenetic and comparative genomics analyses of FAS1 and FAS2 across the Pezizomycotina, we identified two independent evolutionary lineages: one highly conserved lineage with FASs involved in primary metabolism, and a second one, more diverse, with FASs involved in secondary metabolism. This chapter also examined how FASs coevolved with other core enzymes in diverse secondary metabolite pathways. Our findings provided a foundation for understanding the evolutionary mechanisms that drive BGC divergence and fungal metabolic diversity. In Chapter 5, we discussed the key findings of the thesis, including the co-evolution of the xylindein BGC, the dimerisation mechanism, and proposed future research directions and potential applications of xylindein in organic solar cells. Chapter 1 30 75. Almurshidi BH, Van Court RC, Gutierrez SMV, Harper S, Harper B, Robinson SC. Preliminary examination of the toxicity of spalting fungal pigments: A comparison between extraction methods. Journal of Fungi. 2021;7(2):1-16. doi:10.3390/jof7020155 76. Chen L, Chen Q, Wang C, Li Y. Interfacial dipole in organic and perovskite solar cells. J Am Chem Soc. 2020;142(43):18281-18292. doi:10.1021/jacs.0c07439 77. Vollbrecht J, Tokmoldin N, Sun B, et al. On the relationship of the effective mobility and photoconductance mobility in organic solar cells. Energy Advances. 2023;2(9):1390-1398. doi:10.1039/d3ya00125c 78. Donner CD, Cuzzupe AN, Falzon CL, Gill M. Investigations towards the synthesis of xylindein, a blue-green pigment from the fungus Chlorociboria aeruginosa. Tetrahedron. 2012;68(13):2799-2805. doi:10.1016/J.TET.2012.02.009 79. Boonloed A, Weber GL, Ramzy KM, Dias VR, Remcho VT. Centrifugal partition chromatography: A preparative tool for isolation and purification of xylindein from Chlorociboria aeruginosa. J Chromatogr A. 2016;1478:19-25. 80. Robinson SC, Tudor D, Snider H, Cooper PA. Stimulating growth and xylindein production of Chlorociboria aeruginascens in agar-based systems. AMB Express. 2012;2(1):1-7. doi:10.1186/2191-0855-2-15 81. Stange S, Steudler S, Delenk H, Werner A, Walther T, Wagenführ A. Influence of the nutrients on the biomass and pigment production of Chlorociboria aeruginascens. Journal of Fungi. 2019;5(2):1-14. doi:10.3390/jof5020040 82. Zschätzsch M, Steudler S, Reinhardt O, et al. Production of natural colorants by liquid fermentation with Chlorociboria aeruginascens and Laetiporus sulphureus and prospective applications. Eng Life Sci. 2021;21(3-4):270-282. doi:10.1002/elsc.202000079 General Introduction 31 32 33 Chapter 2 Identification of a biosynthetic gene cluster for the production of the blue-green pigment xylindein by the fungus Chlorociboria aeruginascens Yanfang Guo1,2, Jorge C. Navarro-Muñoz3, Caroline Rodenbach1,2,Elske Dwars1,2, Chendo Dieleman1, Bart van den Hout4, Bazante Sanders4, Miaomiao Zhou4,Ayodele Arogunjo5, Russell J. Cox5, Arnold J.M. Driessen*2, Jérôme Collemare*1 1 Fungal Natural Products Group, Westerdijk Fungal Biodiversity Institute, 3584 CT Utrecht, Netherlands. 2 Department of Molecular Microbiology, University of Groningen, 9747 AG Groningen, Netherlands. 3 Bioinformatics Group, Wageningen University and Research, 6708 PB Wageningen, Netherlands. 4 School of Life Sciences and Technology, Avans University of Applied Sciences, 4818 AJ Breda, Netherlands. 5 Institute for Organic Chemistry and BMWZ, Leibniz Universität Hannover, 30167, Hannover, Germany. Published as: Journal of Natural Products. 2025. DOI: 10.1021/acs.jnatprod.4c00350 Chapter 2 34 Abstract Xylindein is a blue-green pigment produced by the fungi Chlorociboria aeruginascens and Chlorociboria aeruginosa. Its stunning colour and optoelectronic properties make xylindein valuable for textile and as a natural semiconductor material. However, producing xylindein from culture broths remains challenging because of the slow growth of Chlorociboria species and the poor solubility of xylindein in organic solvents. An alternative production route for obtaining pure xylindein is heterologous expression of the xylindein biosynthetic genes. Here, we re-sequenced the genomes of C. aeruginascens and C. aeruginosa, and subsequent genome mining and phylogenetic dereplication identified a unique candidate biosynthetic gene cluster with a non-reducing polyketide synthase (nrPKS). RNA sequencing during xylindein production revealed that the core gene XLNpks is coregulated with eight other genes at the locus. Among those, XLNfas1 and XLNfas2 encode a putative fatty acid synthase which likely provides the starter unit to XLNpks. Attempts to heterologously express in Aspergillus oryzae XLNpks alone or in combination with XLNfas1 and XLNfas2 did not yield any intermediate, but expression of the closely related viriditoxin nrPKS (VdtA) produced the expected intermediate. Based on our results, we propose a biosynthetic route to xylindein and suggest that the obtained A. oryzae transformants open ways to further study xylindein biosynthesis. Key words: genome mining, RNA sequencing, heterologous expression, viriditoxin, biosynthetic pathway Identification of xylindein BGC 35 Introduction Blue-green-stained wood is commonly found in humid areas of temperate forests worldwide. This colour is usually due to wood colonisation by the Leotiomycetes fungi Chlorociboria aeruginascens or Chlorociboria aeruginosa which produce the blue-green pigment xylindein. This compound was one of the first isolated fungal secondary metabolites (SMs), extracted from 20 kg of stained wood from the Fontainebleau forest near Paris in 1868.1 The stunning colour and resistance against UV light make xylindein valuable in textile colouration and the decorative wood industry. Xylindein-spalted wooden artefacts produced 500 years ago still exhibit this bright cyan colour.2 More recently, xylindein has been investigated for its optoelectronic performance as a natural semiconductor material.3 Because of these properties, production of xylindein for industrial applications has been investigated.4,5 Xylindein cannot be easily produced by chemical synthesis,6 and its extraction and purification from culture broths are cumbersome and inefficient due to the slow growth of Chlorociboria species and its poor solubility in common organic solvents.7 Heterologous expression has been widely used to elucidate biosynthetic pathways of fungal SMs,8 but it is also a promising alternative strategy for the production of valuable metabolites from fungi that are difficult to manipulate, such as C. aeruginascens.9 For example, titers up to 1.3 g L-1 of the hybrid cyclodepsipeptide hexa-bassianolide were produced in an engineered Aspergillus niger strain.10 A seven-gene cluster for producing the antibiotic pleuromutilin was heterologously expressed in Aspergillus. oryzae and gave a significant increase in production by more than three orders of magnitude, while no improvement in yield was achieved by using various targeted approaches in the natural basidiomycete producer Clitopilus passeckerianus.11 Heterologous expression in A. oryzae also benefits from fast growth under controlled laboratory conditions, genetic tractability for further engineering, and mycotoxin-free production.12 Xylindein is a dimeric naphtho-α-pyranone polyketide that resembles talaroderxine A produced by Talaromyces derxii (Figure 1).13 The production of xylindein is expected to involve a non-reducing polyketide synthase (nrPKS).14,15 Several fungal nrPKSs have been reported to produce similar naphtho-α-pyranone intermediates, including VdtA and Ctb1 which are involved in the biosynthesis of viriditoxin and cercosporin, respectively (Figure 1).16-18 The chemical structures of xylindein and viriditoxin monomers share similarities that suggest common biosynthetic steps. The viriditoxin biosynthetic pathway from the fungus Paecilomyces variotii was Chapter 2 36 elucidated using targeted gene deletion and heterologous expression in Aspergillus nidulans.16,17 The nrPKS VdtA produces the pyranone backbone, which is methylated by the O-methyltransferase VdtC and reduced by the short-chain dehydrogenase VdtF.17 The phenol-coupling dimerisation of viriditoxin involves the laccase VdtB and the catalytically inactive hydrolase VdtD.17 Homologous biosynthetic gene clusters (BGCs) to viriditoxin were also reported in diverse fungal species known to produce naphthopyrones like vioxanthin and xanthoepocin.19 Although the nrPKSs found in these BGCs were not characterised, the laccase AvVirL found in the Aspergillus viridinutans predicted naphtho-α-pyranone BGC exhibited the same phenol-coupling activity as VdtB.19 The genome sequence of C. aeruginasens was recently reported14, and its analysis indicated the presence of 32 BGCs, including 14 for the production of polyketide compounds. Yet, no BGC has been linked to xylindein. In this study, we resequenced the genome of C. aeruginascens, as well as that of the related C. aeruginosa species, and performed genome mining and phylogenetic dereplication to identify a candidate BGC for the production of xylindein. We have initiated the elucidation of this biosynthetic pathway using heterologous expression in Aspergillus oryzae. Figure 1. Xylindein and related compounds. The common naphtho-α-pyranone backbone is highlighted in grey. Nor-toralactone is the biosynthetic precursor of cercosporin. Identification of xylindein BGC 37 Results and discussion Phylogenetic dereplication of C. aeruginascens identifies a single candidate nrPKS for xylindein production Linking biosynthetic genes to an already characterised molecule can be done by exploring the genome of the fungal producer, based on hypotheses according to the predicted enzymatic requirements to yield the chemical structure of interest. Retrobiosynthesis of xylindein indicates that the monomeric precursor released by the nrPKS is likely related to viriditoxin monomer and nor-toralactone, the precursor of cercosporin (Figure 1). Based on this observation, the nrPKS controlling xylindein production is likely related to the corresponding VdtA and Ctb1 enzymes, which both belong to group IV of nrPKSs.20 We thus embarked on mining the C. aeruginascens genome, searching for group IV nrPKS genes. The available genome assembly of C. aeruginasens DSM 107184 is highly fragmented, with 588 contigs in total.14 To obtain a better assembly for genome mining, we sequenced the C. aeruginascens CBS 122017 strain, as well as two C. aeruginosa strains, using Oxford Nanopore long-read technology. The newly obtained assembly of C. aeruginascens is slightly larger with a size of 38.3 Mb in a total of 13 assembled contigs, and with 9,457 predicted genes (Table S1). The assembly of C. aeruginosa CBS 139.28 is similar in size (38.9 Mb) and has a similar number of predicted genes (9,630). In contrast, the assembly of C. aeruginosa CBS 123.57 is larger, with a size of 45.9 Mb and encodes more predicted genes (12,210, Table S1). The low number of large contigs over 1 Mb in the C. aeruginascens assembly and detection of telomeric repeats in all three assemblies suggest that Chlorociboria species comprise between four and six chromosomes (Figure S1). We identified regions with predicted BGCs using antiSMASH (Table S2).21 Out of the 14 predicted PKSs in C. aeruginascens, only two share the conserved domain organisation of nrPKSs, which are encoded by g4260 and g423 genes, respectively (Table S3). Similarly, the C. aeruginosa CBS 139.28 assembly contains only two nrPKS genes, while the C. aeruginosa CBS 123.57 assembly encodes five of them (Table S2 and Table S3). To determine if one of the two nrPKSs encoded in C. aeruginascens genome is related to group IV nrPKSs, we performed a phylogenetic dereplication that included 92 characterised nrPKSs from the Minimum Information about a Biosynthetic Gene Cluster (MIBiG) database and from literature20 (Supplementary files S1, S2 and S3). The obtained maximum likelihood phylogenetic tree showed that the nrPKS Chapter 2 38 encoded by g4260 belongs to group VI which produces methylorsellinic acid or 3,5dimethylorsellinic acid (Figure 2).20 These molecules are tetraketides derived from an acetate starter unit which is not consistent with the structure of xylindein which is likely to be a heptaketide derived from a C4 starter unit. In contrast, the other nrPKS encoded by g423 belongs to the expected group IV, together with the nrPKSs involved in the biosynthesis of fusarubin, viriditoxin, cercosporin and aflatoxins with a strong branch support (Figure 2). However, g423 and VdtA do not form a monophyletic clade. Instead, g423 forms a strongly supported outgroup to the dothistromin, sterigmatocystin and aflatoxin nrPKSs (Figure 2). These later enzymes were previously separated into group IVa, consistent with the production of an anthraquinone intermediate rather than a pyranone backbone.20 Figure 2. Phylogenetic dereplication identifies a candidate non-reducing polyketide synthase (nrPKS) for the production of xylindein. A maximum likelihood phylogenetic tree was built with characterised nrPKSs from the MIBiG database and literature. Ultrafast bootstrap values over 95 and likelihood ratio test above 80 are indicated with black dots at the nodes. The tree is midpoint rooted. Colours and numbers indicate the nrPKS phylogenetic clades. The phylogenetic tree presented here confirms that the ancestor of group IV nrPKSs likely released a naphthopyrone intermediate. The divergence of group IVa is likely Identification of xylindein BGC 39 linked to the modification of the starter unit as nrPKSs from this group use hexanoylCoA produced by fatty acid synthases encoded in the BGC instead of acetyl-CoA as the polyketide starter unit.22 Phylogenetic dereplication with all nrPKSs encoded in the four Chlorociboria assemblies included in this study confirms that g423 is the only nrPKS conserved in all four strains (Figures S2, S4, S5 and S6). Altogether, the phylogenetic dereplication identified a single candidate nrPKS for xylindein production, which we hereafter name XLNpks. Prediction of a unique biosynthetic gene cluster for the production of xylindein Although xylindein is a unique pigment, so far only reported to have been isolated from the Chlorociboria genus, we searched for orthologs of XLNpks in publicly available fungal genomes. This search identified only four orthologous nrPKSs from distant Dothideomycetes species, and one close homologue in the Leotiomycetes Polyphilus sieberi (Figure 3A). While there is no report of naphtho-α-pyranones produced by these Dothideomycetes species, P. sieberi was recently found to produce talaroderxine C, a compound closely related to xylindein and talaroderxine A, but exhibiting an hexaketide starter unit.13 This phylogenetic analysis including close homologues confirms that the XLNpks share a common ancestry with the aflatoxin nrPKS (Figure 3A), which diverged from a common ancestor with other naphtho-α-pyranone producing nrPKSs like VdtA and Ctb1. To determine a potential BGC for xylindein production, we then compared the XLNpks genomic locus to those of all its close homologs (Figure 3B; Table S4, S5). The gene component at the nrPKS loci in the different species is consistent with the phylogenetic relationships. A clear conserved BGC is found within the XLNpks monophyletic clade, as well as in P. sieberi, consisting next to the nrPKS gene of a common set of six genes that encode putative tailoring enzymes or transcription factors based on their functionally conserved domains (Figure 3B; Table 1). XLNtf1 (g430) and XLNtf2 (g429) encode transcription factors, the latter being related to the aflatoxin aflR regulator (Table 1), which might be involved in the regulation of these BGCs. Reminiscent of the aflatoxin BGC, two genes encode a pair of fatty acid synthase (FAS) sub-units, XLNfas1 (g432) and XLNfas2 (g431). The presence of this pair of genes is consistent with the biosynthesis of a fatty acyl-CoA starter unit (Figure 1). The other two conserved genes, XLNsdh (g424) and XLNlac (g425), encode a shortchain dehydrogenase and a laccase, respectively (Figure 3B; Table 1), consistent with redox processes required for the formation of xylindein. Homologs of those Chapter 2 46 S4 and S5). Thus, product 1 is likely produced by an endogenous enzyme of A. oryzae that reduced the ketone to an alcohol on the side chain. For example, undesired oxidations of biosynthetic intermediates were detected during heterologous biosynthesis of the polyketide solanapyrone in A. oryzae.23 Our results show that A. oryzae was able to express VdtA and produce the expected pyranone product 2. A putative biosynthetic route to xylindein. Based on the genomic predictions and heterologous expression, we propose a biosynthetic pathway to yield xylindein. Several hypotheses could explain the absence of product when XLNfas1, XLNfas2 and XLNpks are co-expressed. First, the release mechanism of the polyketide from the nrPKS may require an extra enzyme as exemplified with group V nrPKSs that need a b-lactamase enzyme encoded within their BGCs.24 However, in the cercosporin biosynthetic pathway, it was shown that the release mechanism from the nrPKS Ctb1 is catalysed by its TE domain that forms the pyrone.25 Thus, the observed pyrone ring in the structure of nor-toralactone, viriditoxin and xylindein is expected to be the result of the releasing mechanism. A second hypothesis is a codon usage bias incompatible with A. oryzae. To evaluate this hypothesis, we calculated the codon adaptation index (CAI)26 between vdtA, XLNpks, XLNfas1 and XLNfas2 genes with the host organism A. oryzae. The CAI scores range from 0.0 to 1.0, and genes with scores above 0.8 are considered to be suitable for heterologous expression.27 CAI scores are very similar for the three XLN and vdtA genes and all are greater than 0.8 (Table S7), indicating that C. aeruginascens codon bias is unlikely to explain the absence of product. In the aflatoxin pathway, the starter unit hexanoyl-CoA is directly transferred from the FAS complex to the nrPKS.22 Thus, no other enzyme is expected to be required to produce the first intermediate from butanoyl-CoA and malonyl-CoA in the xylindein pathway. Based on these considerations, the first steps of the pathways are expected to be very similar to the aflatoxin and viriditoxin pathways (Figure 7). However, we cannot exclude that tailoring enzymes are required in these early steps in the biosynthetic pathway as found for sporothriolides whose biosynthesis involves FASs (FasA and FasB) and a citrate synthase (SpoE).28 Co-expression of the three genes in A. oryzae did not yield any product, and the first stable intermediate was obtained only when the decarboxylase (spoK) and citrate dehydratase (spoL) were also co-expressed.28 There is no obvious candidate gene in the xylindein BGC that would encode a tailoring enzyme involved in the production of the expected first stable intermediate. Finally, the predicted starter unit butanoyl-CoA could be rapidly degraded by b- Identification of xylindein BGC 47 oxidation,29 preventing the XLNpks to use it, while the acetyl-CoA starter unit is easily available for VdtA and explains the successful expression in A. oryzae. Similarities between xylindein and viriditoxin (Figure 1) suggest common tailoring biosynthetic steps. XLNsdh likely exhibits a similar function as VdtF to reduce the pyrone ring17 and we propose that XLNsdh will catalyse the first tailoring step in the xylindein pathway (Figure 7). It was demonstrated that VdtB catalyses the dimerisation step to yield viriditoxin.17 XLNlac is a multicopper oxidase homologous to VdtB and is thus expected to catalyse a similar phenol coupling in xylindein biosynthesis (Figure 7).17 The dimerisation of xylindein is, however unique because of additional C-O coupling (Figure 1). This feature suggests the xylindein BGC should comprise a gene that is not found outside of the Chlorociboria genus (Figure 3). Although the predicted function is not compatible with the required oxidative chemistry, the only candidate is the putative carbonic anhydrase XLNcnh which we propose to contribute to the final step of the xylindein biosynthesis (Figure 7). The exact enzymatic activity of XLNcnh remains to be determined with functional analyses. Figure 7. Proposed biosynthetic route of xylindein. XLNfas1 and XLNfas2 provide the precursor butanoyl-CoA to XLNpks. Along with 6 malonyl-CoA, XLNpks releases the first intermediate, which is subsequently reduced by XLNsdh. XLNlac dimerises the monomer with XLNcnh and produces the final product xylindein. Chapter 2 48 In conclusion, through a combination of genome mining, phylogenetic dereplication, and expression analysis, this study identified the BGC responsible for the production of xylindein, a valuable blue-green pigment derived from wood colonisation by Chlorociboria species. Elucidation of the biosynthetic pathway requires further effort, especially to understand the early biosynthetic steps. Most importantly, the A. oryzae transformant expressing the viriditoxin nrPKS VdtA opens avenues to characterise the xylindein tailoring enzymes, especially those responsible for the unique dimerisation of xylindein. Material and methods General experimental procedures High Resolution Mass Spectrometry (HRMS) was performed on a Q-Tof Premier mass spectrometer (Waters) coupled to an Acquity UPLC system (Waters). Electron spray Ionisation (ESI) mass spectroscopy was measured in positive or negative mode depending on the compound. All solvents and chemicals used for HR ESI-MS and chromatography were LC-MS grade, while the solvents for metabolite extraction were HPLC grade. Water was purified by using a Milli-Q ultrapure water system. Fungal strains and growth conditions C. aeruginascens CBS 122017, C. aeruginosa CBS 139.28 and C. aeruginosa CBS 123.57 strains were grown and maintained on MEA plates at 21 °C. P. variotii CBS 101075 was grown and maintained on MEA plates at 30 °C. For genomic DNA extraction, strains were grown in 50 mL MB (Table S8) liquid medium for 14 days at 21 °C under constant agitation at 200 rpm. For total RNA extraction, C. aeruginascens was grown in 50 mL 5% OJ liquid medium (Table S8) at 21 °C under constant agitation at 200 rpm; mycelium was harvested at 6-8 days, 10-12 days and 14-18 days when exhibiting different white, light blue and dark blue pigmentation, respectively. A. oryzae NSAR1 was grown and maintained on MEA plates at 30 °C. Saccharomyces cerevisiae BMA 64 Δura3 used for homologous recombination was grown and maintained on YPD plates at 30 °C. Nucleic acid extraction and RT-PCR. The mycelium of Chlorociboria species and A. oryzae transformants from liquid cultures was filtered through a paper filter, frozen in liquid nitrogen, and ground using a mortar and pestle. Genomic DNA of transformants was isolated using the DNeasy plant minikit (Qiagen, Hilden, Germany) according to the manufacturer’s recommendations. For genomic DNA for genome sequencing, 100 mg of the ground Identification of xylindein BGC 49 mycelium was mixed with 0.6 mL of warm cetyltrimethylammonium bromide (CTAB) buffer in a 1.5 mL microcentrifuge tube and incubated for 15 min at 65 °C. The resulting lysate was mixed with 0.6 mL chloroform isoamyl alcohol (24:1), violently mixed by vortex for 5 min. The sample was centrifuged at 12,000 g for 10 min at room temperature. The aqueous phase was transferred into a new microcentrifuge tube, incubated on ice for 10 min then mixed with 0.5 mL of cold isopropanol. The sample was centrifuged at 12,000 g for 15 min to obtain the pellet, followed by washing with 0.5 mL of cold ethanol and centrifugation for 10 min to remove the supernatant. The air-dried sample was eluted with 50 µL of water and mixed with 5 µL of RNase A (Thermo Fisher Scientific, Waltham, MA), incubated at 37 °C for 30 min. For total RNA extraction, 100 mg of the ground mycelium was mixed with 1 mL of Invitrogen TRIzol reagent (Thermo Fisher Scientific, Waltham, MA) in a 1.5 mL microcentrifuge tube and incubated for 5 min at 25 °C. The resulting lysate was mixed with 0.2 mL chloroform, gently mixed by hand, and incubated for 5 min. Samples were centrifuged at 12,000 g for 15 min at room temperature. The aqueous phase was transferred into a new microcentrifuge tube, mixed with 0.5 volumes of 100% ethanol, and loaded into a column from the NucleoSpin RNA extraction kit (Macherey Nagel, Allentown, PA). Downstream steps were performed according to the manufacturer’s protocol. Five micrograms of RNA were used to synthesise cDNA using oligo(dT) primers and GoScript reverse transcription (RT) mix (Promega, Madison, WI) according to the manufacturer’s protocol. PCR was performed for XLNpks, XLNfas1, XLNfas2, vdtA genes and the housekeeping control gene H2B using GoTaq DNA polymerase (Promega, Madison, WI) (Table 8). Genome sequencing, assembly and gene prediction The genomic DNA was purified using 0.4 volume of AMPure XP Beads (Beckman Coulter, Brea, CA). After mixing the genomic DNA with AMPure XP beads, the suspension was incubated on a rotator mixer for 10 min at room temperature. The sample was spun down shortly and placed on a magnet for 5 min to pellet the beads. The supernatant was discarded and the pellet was washed twice with 80% ethanol on the magnetic stand. Air-dried beads were resuspended in 35 µL of water and incubated at 37 °C for 10 minutes, gently flicking the sample every 2 minutes to encourage DNA elution. After spinning down and pelleting the beads on the magnet, the eluate was placed into a new tube. Purified genomic DNA samples were prepared for long-read sequencing with the Oxford Nanopore Technologies native barcoding kit SQK-NBD114.24 (ONT, Oxford, United Kingdom) according to the Chapter 2 50 manufacturer’s protocol (version NBE_9769_v114_revI_15Sep2022, latest update 12/07/2023) using approximately 1 µg DNA per sample. The final prepared libraries were sequenced on R10.4.1 flow cells (ONT, Oxford, United Kingdom) with the GridION sequencer (ONT, Oxford, United Kingdom). Basecalling of the reads was performed by Guppy with the high-accuracy basecalling model in Minknow. The first and last 50 bp of all “pass” reads were then chopped, using CHOPPER.30 The quality of the chopped reads was then checked using FastQC version 0.12.131 and Flye 2.9.2-b178632 was used to assemble the chopped reads. The quality of the assembly was then checked using Quast version 5.2.0.33 Completeness of the assemblies was determined by a BUSCO v5 analysis using gVolante34 with the Leotiomycetes_odb10 dataset. Genes were predicted using webAugustus35 with the Botrytis cinerea training set. Assemblies were visualised and telomeric regions predicted using Telovision (https://github.com/WesterdijkInstitute/TeloVision). Phylogenetic trees and comparative genomics Close homologues of C. aeruginascens XLNpks were retrieved from the Joint Genome Institute (JGI) MycoCosm repository36 using BLASTp against Pezizomycotina. Regions containing BGCs were retrieved using fungiSMASH 7.0 with default parameters.21 BGC comparison was performed using Clinker.37 Characterised nrPKSs were retrieved from the Minimum Information about a Biosynthetic Gene Cluster (MIBiG) database.38 Protein alignments were performed using MAFFT version 7.49039 (parameters --reorder) with nrPKS sequences. Poorly aligned regions were removed using trimaL version 1.440 (build 2013-12-17; parameter -automated1). Maximum likelihood trees were built with IQ-TREE version 2.2.0-beta with model finder and ultrafast bootstrapping as well as an approximate likelihood-ratio test41 (parameters-mset LG -bb 1000 -alrt 1000 -T AUTO). The substitution model used for trees of nrPKS phylogenetic dereplication, XLNpks homologue phylogeny, and phylogeny of Chlorociboria nrPKSs are LG+F+R6, LG+R6, and LG+F+R6, respectively. The resulting trees were visualised using iTOL.42 All curated alignments and phylogenetic tree files are provided in the supplementary (Supplementary files S1-S6). Gene amplification and plasmid digestion (XLNpks, XLNfas1, XLNfas2, vdtA) XLNpks and vdtA were amplified from cDNA of C. aeruginascens and genomic DNA of P. variotii, respectively, using primers which harbour 30-bp sequences homologous to the pEYA2 plasmid43 (Table S9). XLNfas1 and XLNfas2 were amplified from cDNA of C. aeruginascens using primers that contain 30-bp Identification of xylindein BGC 51 sequences homologous to pTYGSade plasmid (behind promoters Padh and Pgpd; Table S9). All PCR fragments were amplified using Phusion high-fidelity DNA polymerase (Thermo Fisher Scientific, Waltham, MA) according to the manufacturer’s protocol. One microgram of the pEYA2 plasmid or pTYGSsde plasmid was digested for 1h with 10 U NotI or AscI (Promega, Madison, WI), respectively, at 37 °C. Fragments of the expected size and the linearised plasmids were purified from a 0.8% agarose gel or directly from the PCR mix using a Geneclean II kit (MP Biomedicals, Santa Ana, CA). Transformation-associated recombination in Saccharomyces cerevisiae (pEYA2::PKS, pTYGSade::FAS1::FAS2, pEYA2::vdtA) A strain of S. cerevisiae BMA 64 with a ura3-auxotrophic marker was used for transformation-associated recombination according to a protocol adapted from the one described previously.44 S. cerevisiae was grown overnight at 30 °C in 3 mL yeast YPD medium (Table S8). Two millilitres of culture containing 108 cells was transferred into 50 mL YPD medium and incubated at 30 °C under agitation at 200 rpm for about 5 h until reaching an optical density at 600 nm (OD600) of 1 to 1.5. Yeast cells were centrifuged for 5 min at 2,500 rpm at 4 °C, and the supernatant was discarded. Fifty microliters of cells were mixed with 250 µL DTT (100mM), and incubated for 10 minutes at room temperature. The mixture was centrifuged for 15 seconds and the supernatant was discarded. The pellet was gently mixed with 500 microliters PLTE buffer (800 µL 50% PEG, 100 µL 1M LiAc, 20 µL 50mM, 10 µL 1M Tris HCl pH 7.5, 70 µL H2O), 4 µL of each DNA fragments and digested plasmid, 50 µL boiled salmon sperm DNA, and incubated for one hour at 30 °C with inverting the tube once every 20 minutes. After one hour of incubation, cells were heat shocked at 45 °C for 15 minutes. The pellet was collected by centrifuging for 15 seconds and the supernatant was discarded. Cells were gently resuspended in 1 mL sterile water and centrifuged for 15 seconds to remove the water. The pellet was resuspended in 1 mL YPD liquid medium and incubated at 30 °C for 30 minutes without shaking. Cells were collected after 15-second centrifugation. The pellet was reconstituted in 200 µL sterile water and plated on SDM plates (Table S8). Plates were incubated for 3 to 7 days at 30 °C. Yeast transformants were transferred to a new SDM plate and grown overnight. Single colonies were transferred into a microcentrifuge tube in 30 µL of 25 mM NaOH and boiled for 10 min at 100 °C. Next, 1 mL was used for PCR screening with GoTaq DNA polymerase (Promega, Madison, WI) and corresponding cloning primers (Table S9). Positive transformants were grown overnight in liquid SDM to isolate the plasmid using the Zymoprep yeast Chapter 2 52 plasmid miniprep kit (Zymo Research, Irvine, CA). The obtained plasmids were subsequently introduced into electrocompetent Escherichia coli DH5a cells (for pEYA2::PKS and pEYA2::vdtA; Thermo Fisher Scientific, Waltham, MA) and 2T1 cells (for pTYGSade::FAS1::FASs2; Thermo Fisher Scientific, Waltham, MA) using an electroporation method according to the manufacturer’s protocol (electroporator conditions: 2.0 kV, 200 Ω, 25 µF). PCR screening was performed by transferring individual colonies into the PCR mixture with GoTaq DNA polymerase (Promega, Madison, WI). The plasmid was isolated from confirmed positive clones using the Zyppy plasmid miniprep kit (Zymo Research, Irvine, CA), and the plasmid was validated by sequencing (Macrogen, Seoul, South Korea) with sequencing primers (Table S9). Plasmid pTYGSade::FAS1::FASs2 was subsequently isolated from sequence sequence-confirmed clone by the HiPure Plasmid Midiprep Kit (Thermo Fisher Scientific, Waltham, MA). Construction of the expression vector (pTYGSarg::PKS, pTYGSade::vdtA) Seventy nanograms of the pEYA2::PKS or pEYA2::vdtA entry vector and 100 ng of the pTYGSarg or pTYGSade destination vector38 were mixed with 1 µL of the Gateway® LR Clonase II enzyme (Thermo Fisher Scientific, Waltham, MA) in a 5µL final volume, and the reaction mixture was incubated at 25 °C for 2 h. The total reaction mixture was introduced into chemically competent E. coli DH5a cells (Thermo Fisher Scientific, Waltham, MA) using a heat shock protocol. The pTYGSarg::PKS and pTYGSade::vdtA expression vectors were isolated from positive colonies using the Zyppy plasmid miniprep kit (Zymo Research, Irvine, CA) and the HiPure Plasmid Midiprep Kit (Thermo Fisher Scientific, Waltham, MA) for A. oryzae transformation. Transformation of A. oryzae NSAR1 Spores from A. oryzae NSAR1 were harvested from MEA plates in 5 mL of sterile water, and 1 mL of this spore suspension was inoculated into 50 mL of MB liquid medium and grown overnight at 28 °C with shaking at 200 rpm. Germinating spores were collected by centrifugation at room temperature for 10 min at 3,500 rpm and resuspended in 25 mL of 0.8 M NaCl. After spin down for 10 min at 3,500 rpm at room temperature, germinated spores were resuspended in 10 mL of a freshly made filter-sterilized protoplasting solution (200 mg Trichoderma lysing enzyme (Thermo Fisher Scientific, Waltham, MA) and 50 mg Driselase (Thermo Fisher Scientifific, Waltham, MA) in 0.8 M NaCl) and incubated at 30 °C for 2 to 2.5 h with shaking at 100 rpm. Protoplasts were filtered through sterile Miracloth and then centrifuged for Identification of xylindein BGC 53 5 min at 3,000 rpm at 4 °C. Protoplasts were resuspended in 200 µL of solution 1 (0.8 M NaCl, 10 mM CaCl2, and 50 mM Tris-HCl, pH 7.5) and aliquoted to 100 µL in 2-mL microcentrifuge tubes. Ten micrograms of the pTYGSarg::PKS, pTYGSade::FAS1::FAS2, pTYGSade::vdtA expression plasmids or pTYGSarg and pTYGSade empty vectors were added to protoplasts, and the mixture was incubated on ice for 2 min. One millilitre of solution 2 (60% (wt/vol) polyethylene glycol 3350, 0.8 M NaCl, 10 mM CaCl2, and 50 mM TrisHCl, pH 7.5) was added, and the tubes were gently inverted before incubation at room temperature for 20 min. Protoplasts were then mixed with 25 mL of cooled Top CZD agar A, B or C (Table S8) and immediately plated onto the corresponding Bottom CZD agar plates (Table S8). Transformation plates were incubated at 30 °C for 3 to 10 days. Transformants were transferred onto new selection CZD agar plates individually, and transferred onto DPY agar plates (Table S8) for induction. Secondary-metabolite extraction, HPLC and HRMS analyses The 4-day-old A. oryzae transformant on DPY agar plates was cut and collected into the 50 mL tubes and isolated with ethyl acetate (VWR Chemicals, Radnor, PA) for screening the metabolic profile. After shaking on an orbital shaker for at least 1 h, the organic phase was transferred to a new 50 mL tube and evaporated under nitrogen flow. For a 50 mL DPY broth culture, secondary metabolites from 4-day-old transformant liquid culture filtrates were isolated with a 1:1 volume of ethyl acetate, and followed by the same shaking and evaporating steps. The resulting solid was dissolved in acetonitrile. Organic extracts were analysed with a Shimadzu LC-2030 3D Prominence-i PDA system coupled to a Shimadzu LCMS-2020 mass spectrometer and equipped with a Shimadzu Shim-pack GIST C18-HP reversedphase column (3 mm, 4.6 by 100 mm). The following method was used: a linear gradient of buffer B (5% to 95%) for 10 min, 2 min of 95% buffer B, gradient of buffer B (95% to 5%) for 1 min and then 5% buffer B for 5 min. Water with 0.1% trifluoroacetic acid (TFA) for high-performance liquid chromatography (HPLC), or 0.05% formic acid for mass spectrometry (MS)-coupled analyses was used as buffer A, and acetonitrile (LC-MS grade) with 0.1% TFA for HPLC, or 0.05% formic acid for MS-coupled analyses was used as buffer B. The flow rate was 1 mL min or 0.5 mL min-1 for UV-HPLC or MS-coupled analyses, respectively. The equipment was controlled and results were analysed using Shimadzu LabSolutions LC-MS software. High Resolution Mass Spectrometry (HRMS) was performed on a Q-Tof Premier mass spectrometer (Waters) coupled to an Acquity UPLC system (Waters). Electron spray Ionisation (ESI) mass spectroscopy was measured in positive or negative mode Chapter 2 54 depending on the compound. The chemical formula of product 1 and product 2 was determined using the measured exact mass on the ChemCalc server.45 RNA sequencing Library preparation was performed using a custom barcoding protocol as previously described,46 before continuing with the protocol of the Oxford Nanopore direct RNA sequencing kit (SQK-RNA002). The concentration of each individually barcoded sample was determined using the Qubit RNA high-sensitivity kit (Invitrogen, Q32852). Samples were then pooled and the final concentration was determined using Qubit (Invitrogen, Q32851). RNA of S. cerevisiae Enolase II was used as a spike-in. Between 50 and 300 ng of the libraries were sequenced on the Oxford Nanopore MinION Mk1C with R9 flow cells, and the sequencing runs were monitored using the accompanying MinKNOW 23.07.12 software.47 Base-calling was performed with Guppy 7.1.4. Demultiplexing was performed with the ‘resnet20final.h5’ barcode prediction model according to the custom barcoding protocol with the CPU method described in Deeplexicon.46 The spike-in reads were aligned using Minimap2 (2.24).48 The spike-in and sample reads were separated using Samtools (1.18) view (-F4 -hS and -f4 -hS).49 The latter was converted to fastq format using the Samtools bam2fq command. The fastq files were combined to generate a single file according to experimental condition, biological and technical replications. The total reads were then mapped to the reference genome of C. aeruginascens CBS 122017 using Minimap2.48 Expression was quantified using Featurecounts (Subread 2.0.6)50 with C. aeruginascens gene prediction. Codon adaptation index (CAI) calculation To calculate CAI, the codon usage table for A. oryzae was acquired from the Codon Usage Database (https://www.kazusa.or.jp/codon/). This table was then fed into two calculators, OPTIMIZER and E-CAI from CAIcal.27 CAI of the genes was evaluated by submitting exonic sequences only. Supporting information Amino acid sequences, alignments, and tree files are mentioned in the main text. Additional experimental details, methods and analysis, including supplementary figures and tables. The published sequence and assembly of C. aeruginasens DSM 107184 are on NCBI with the accession of NCSK00000000.2. Other sequencing data and assemblies are available on NCBI under the umbrella PRJNA1062589 bioproject. This Whole Genome Shotgun project has been deposited at DDBJ/ENA/GenBank under the accessions JBAWJB000000000, Identification of xylindein BGC 55 JBAWJC000000000 and JBAWJD000000000. The version described in this paper is version JBAWJB010000000, JBAWJC010000000 and JBAWJD010000000. The xylindein BGC locus has been submitted to GenBank. Supplementary files Supplementary file S1. nrPKS sequences used in the phylogenetic dereplication. Supplementary file S2. Trimmed alignment of nrPKSs used in the phylogenetic dereplication. Supplementary file S3. Tree file of the phylogenetic dereplication. Supplementary file S4. nrPKS sequences used to build the phylogeny of Chlorociboria nrPKSs. Supplementary file S5. Trimmed alignment of Chlorociboria nrPKSs. Supplementary file S6. Tree file of Chlorociboria nrPKSs. Supplementary file S7. RNAseq quantification of gene expression in Chlorociboria aeruginascens. Supplementary materials can be accessed and downloaded via the following link: https://pubs.acs.org/doi/10.1021/acs.jnatprod.4c00350 Chapter 2 62 Table S3. Conserved domain organization of predicted non-reducing polyketides synthases. Strain antiSMASH region Gene ID Domain architecture C. Aeruginasce ns CBS 122017 fungismash.contig_4.reg ion007 g4260 SAT | KS | KS_C | AT | DH | T/ACP | T/ACP | Abhydrolase_3 | Peptidase_S9 fungismash.contig_16.re gion003 g423 SAT | KS | KS_C | AT | PT | T/ACP | T/ACP | TE C. Aeruginosa CBS 139.28 fungismash.contig_101.r egion001 g486 SAT | KS | KS_C | AT | PT | T/ACP | T/ACP | TE fungismash.contig_11.re gion001 g1253 SAT | KS | KS_C | AT | DH | T/ACP | T/ACP | Abhydrolase_3 | Peptidase_S9 C. Aeruginosa CBS 123.57 fungismash.contig_16.re gion004 g2144 SAT | KS | KS_C | AT | PT | T/ACP | TE fungismash.contig_16.re gion009 g2763 SAT | KS | KS_C | AT | PT | T/ACP | TE fungismash.contig_17.re gion002 g3060 adh_short | SAT | KS | KS_C | AT | PT | T/ACP | T/ACP | TE fungismash.contig_19.re gion004 g3704 SAT | KS | KS_C | AT | T/ACP | T/ACP | HTH_51 | CMeT | NAD_binding_4 | FAD_binding_3 fungismash.contig_25.re gion003 g7034 SAT | KS | KS_C | AT | PT | T/ACP | TE SAT: Starter Acyl-carrier protein Transacylase; KS: Keto-Synthase: AT: AcylTransferase; PT: Product Template; T: Thiolation; ACP: Acyl-Carrier Protein; TE: ThioEsterase; HTH: Helix-turn-helix; CMeT: C-methyl transferase; NAD: Nicotinamide Adenine Dinucleotide; FAD: Flavin Adenine Dinucleotide. Identification of xylindein BGC 63 Table S4. BlastP search results for proteins encoded at the XLNpks locus. Protein query Best BlastP hit in Polyphilus sieberi (evalue|%identit y|%coverage) Best BlastP hit in Hysterium pulicare (evalue|%identity| %coverage) Best BlastP hit in Stagonospora sp. (evalue|%identity| %coverage) Best BlastP hit in Flavomyces fulophazii (evalue|%identity| %coverage) Best BlastP hit in Periconia macrospinosa (evalue|%identity|% coverage) g421 788440 (0.0e0|79|92) scaffold 15 115568 (1.80e78|58|99) scaffold 239 122177 (3.51e78|57|96) scaffold 5 658531 (2.52e79|58|99) scaffold 104 619275 (1.42e77|57|99) scaffold 290 g422 726361 (3.76e180|61|93) scaffold 15 120174 (2.01e34|42|59) scaffold 456 290194 (1.16e35|43|59) scaffold 2 94634 (8.47e38|40|59) scaffold 15 667290 (2.68e36|37|64) scaffold 8 g423 pks 872537 (0.0e0|64|100) scaffold 16 121385 (0.0e0|70|100) scaffold 572 234140 (0.0e0|67|97) scaffold 21 739674 (0.0e0|65|100) scaffold 47 698930 (0.0e0|65|100) scaffold 36 g424 sdh 789957 (3.62e108|71|100) scaffold 16 121383 (6.34e163|80|100) scaffold 572 298559 (1.10e156|76|100) scaffold 21 710374 (2.33e159|77|100) scaffold 47 521457 (3.54e157|76|100) scaffold 36 g425 lac 658105 (1.57e109|64|89) scaffold 16 121384 (1.40e12-|67|93) scaffold 572 321074 (8.01e101|63|87) scaffold 21 616058 (7.19e101|64|86) scaffold 47 521543 (6.71e101|64|97) scaffold 36 g426 cnh 303359 (4.20e26|42|82) scaffold 25 121984 (1.32e33|45|26) scaffold 583 333352 (1.36e6|40|23) scaffold 3 No hit 635458 (7.17e10|42|27) scaffold 396 g427 tf3 638301 (4.32e73|60|90) scaffold 4 121983 (1.51e16|49|49) scaffold 583 38731 (1.21e7|25|22) scaffold 13 730149 (3.22e19|37|23) scaffold 16 695211 (5.01e12|36|28) scaffold 272 g428 met 748822 (3.04e31|31|60) scaffold 16 111424 (6.08e16|32|20) scaffold 106 292291 (3.09e15|29|50) scaffold 5 631741 (2.57e22|41|38) scaffold 7 670136 (9.22e23|43|37) scaffold 37 g429 tf2 166291 (3.25e25|60|49) scaffold 16 121387 (4.97e23|61|58) scaffold 572 382208 (9.93e20|54|58) scaffold 21 299616 (7.17e21|60|45) scaffold 47 612484 (1.48e38|57|41) scaffold 36 g430 tf1 618621 (4.77e98|59|87) scaffold 16 121388 (4.88e54|56|90) scaffold 572 86029 (3.27e79|57|67) scaffold 21 562513 (8.81e81|54|83) scaffold 47 268080 (5.56e80|56|78) scaffold 36 g431 fas2 891520 (0.0e0|60|91) scaffold 16 121381 (0.0e0|62|81) scaffold 572 286687 (9.42e165|59|82) scaffold 21 299705 (2.73e174|60|83) scaffold 47 268045 (6.05e173|59|84) scaffold 36 g432 fas1 693732 (0.0e0|56|86) scaffold 16 121382 (0.0e0|56|89) scaffold 572 353560 (0.0e0|53|86) scaffold 21 689350 (0.0e0|54|89) scaffold 47 558682 (0.0e0|55|94) scaffold 36 g433 739970 (1.20e102|82|91) scaffold 4 112596 (5.09e51|68|65) scaffold 14 288481 (9.52e75|70|64) scaffold 1 709703 (9.23e76|67|59) scaffold 37 569438 (1.09e74|71|60) scaffold 163 g434 765476 (1.32e108|73|85) scaffold 4 112597 (1.61e17|60|60) scaffold 14 311761 (9.55e15|54|61) scaffold 1 737218 (1.75e15|56|57) scaffold 37 117331 (8.62e16|56|57) scaffold 163 Hits in green correspond to conserved homologous loci. Hits in red are distant homologues. Chapter 2 64 Table S5. Level of conservation between conserved enzymes encoded at the XLNpks locus in different species. C. aeruginas cens C. aerugin osa CBS 139.28 C. aerugin osa CBS 123.57 P. macrospi nosa F. fuloph azii Stagon ospora sp. H. pulicar e P. siebe ri A. flavus P. vario tii F. fujiku roi C. betico la g432 fas1 1.00 0.85 0.48 0.49 0.48 0.51 0.49 0.38 < 0.3 < 0.3 < 0.3 g431 fas2 0.99 0.89 0.51 0.5 0.51 0.54 0.56 0.44 < 0.3 < 0.3 < 0.3 g430 tf1 0.96 0.79 0.49 0.5 0.46 < 0.3 0.47 < 0.3 < 0.3 < 0.3 < 0.3 g429 tf2 1.00 0.61 0.42 0.4 0.4 0.43 0.39 < 0.3 < 0.3 < 0.3 < 0.3 g428 met 1.00 0.88 < 0.3 < 0.3 < 0.3 < 0.3 < 0.3 < 0.3 0.4 < 0.3 < 0.3 g426 cnh 0.98 0.70 < 0.3 < 0.3 < 0.3 < 0.3 < 0.3 < 0.3 < 0.3 < 0.3 < 0.3 g425 lac 0.97 0.82 0.54 0.57 0.57 0.62 0.57 < 0.3 0.47 < 0.3 0.33 g424 sdh 0.99 0.98 0.77 0.77 0.76 0.8 0.71 < 0.3 0.38 < 0.3 < 0.3 g423 pks 0.98 0.96 0.62 0.61 0.61 0.66 0.59 0.46 0.45 0.35 0.41 TheAmino acid identity percentage compared to C. aeruginascens was determined using Clinker. Table S6. Number of transformants obtained in this study. Name Total number of obtained transformants Number of transformants producing additional compounds / Number of transformants screened by HPLC Number of transformants expressing introduced genes / Number of transformants tested by RT-PCR AoARG::PKS 44 0/44 7/11 AoADE::FAS1::FAS2 17 0/17 6/10 AoADEARG::FAS1::F AS2::PKS 49 0/49 4/15 AoARGADE::PKS::F AS1::FAS2 35 0/35 4/12 AoADE::vdtA 55 14/30 4/4 Table S7. Codon adaptation index (CAI) scores of VdtA, XLNpks, XLNfas1 and XLNfas2 genes in Aspergillus oryzae. Gene Original organism OPTIMIZER E-CAI vdtA P. variotii 0.838 0.855 XLNpks(g423) C. aeruginascens 0.843 0.835 XLNfas1(g432) C. aeruginascens 0.814 0.846 XLNfas2(g431) C. aeruginascens 0.809 0.822 CAI scores were determined using Optimizer and E-CAI calculators from CAI-cal. Identification of xylindein BGC 65 Table S8. Media used in this study. Media Recipe Brand MEA 50 g/L Malt extract agar, pH 5.4 ± 0.2. oxoid cm59 MB 30 g/L malt extract, pH 5.4 ± 0.2. oxoid L39 5% OJ broth 5 ml/L orange juice 5% OJ agar 5 ml/L orange juice, 15 g/L agar. Top CZD agar (A, B, D) Base: 35 g/L Czapek-Dox, 1 g/L ammonium sulfate, 182.1g/L1M Sorbitol, 8 g/L selective agar. A: 0.5 g/L adenine, 1.5 g/L methionine Sigma-Aldrich, St. Louis, MO. B: 1 g/L arginine, 1.5 g/L methionine D: 1.5 g/L methionine Bottom CZD agar (A, B, D) Base: 35 g/L Czapek-Dox, 1 g/L ammonium sulfate, 182.1g/L1M Sorbitol, 15 g/L selective agar. A: 0.5 g/L adenine, 1.5 g/L methionine Sigma-Aldrich, St. Louis, MO. B: 1 g/L arginine, 1.5 g/L methionine D: 1.5 g/L methionine Re-selection CZD agar (A, B, D) Base: 35 g/L Czapek Dox, 1 g/L ammonium sulfate, 15 g/L selective agar. A: 0.5 g/L adenine, 1.5 g/L methionine Sigma Aldrich, St. Louis, MO. B: 1 g/L arginine, 1.5 g/L methionine D: 1.5 g/L methionine DPY broth 20 g/L dextrine from potato starch, 10 g/L polypeptone, 5 g/L yeast extract, 5 g/L KH2PO4, 0.5 g/L MgSO4 ∙ H2O. Sigma Aldrich, St. Louis, MO. DPY agar 20 g/L dextrine from potato starch, 10 g/L polypeptone, 5 g/L yeast extract, 5 g/L KH2PO4, 0.5 g/L MgSO4 ∙ H2O, 25 g/L agar. Sigma Aldrich, St. Louis, MO. YPD 10 g/L Yeast extrac, 20 g/L Bacto peptone, 20 g/L D (+)-Glucose. Difco 212759; Difco 211677; Merck 1.08337. SDM 6.7 g/L Yeast nitrogen base without amino, 1.92 g/L Yeast dropout supplements without uracil, 20 g/L D-glucose, 20 g/L selective agar. Sigma Y0626, Sigma Y1501, Merck 1.08337, Invitrogen 30391049. LB 10 g/L Tryptone, 5 g/L Yeast extract, 5 g/L Sodium Chloride, 0.02 g/L Thymine. oxoid LP0042; difco 212750; baker 0278; sigma T-0376. LA 10 g/L Tryptone, 5 g/L Yeast extract, 5 g/L Sodium Chloride, 0.02 g/L Thymine, 15 g/L Agar Bacteriological. oxoid LP0042; difco 212750; baker 0278; sigma T-0376; oxoid L11. Chapter 2 66 Table S9. Primers used in this study. Primer name Sequence (5’-3’) Product size Ta Description PKS RNA-1F taatgccaactttgtacaaaaaagcaggctATGGCTCAA TCACTGCAAGTCTATCTTTTTGG 3030 bp 60°C XLNpks cloning; lower case sequence is homologous to pEYA2 vector PKS RNA-1R TGCCATCTGCATGCAGGTAG PKS RNA-2F TCGCAGTCACCGGTTCAAATC 2458 bp 60°C PKS RNA-2R TCAGCCACTGCAATTGCTGC PKS RNA-3F TGACGGTGATAAGATCGTGG 1944 bp 60°C PKS RNA-3R taatgccaactttgtacaagaaagctgggtTCAAGCCAG AGCCTTCGCAATAAAAGCGCCCAC PadhChlaerFAS1-Fw tttcaacacaagatcccaaagtcaaaATGTCTTCCTAC CTTGATGTAGAGTATCTTG 2582 bp 60°C XLNfas1 cloning; lower case sequence is homologous to pTYGSade vector FASChlaer1-1R AGCCATCATACGGCTTCCAAG FASChlaer1-2F ATTGTCCTGATTGCCGGTAG 2275 bp 60°C FASChlaer1-2R TGACCGGATTTCCATGGCATG FASChlaer1-3F AGGACCAGAACTTATCGGCG 1558 bp 60°C TadhChlaerFAS1-Rv ttcattctatgcgttatgaacatgttccctCTAGTAGTCCA AGCCTTCCAGCGCCCGACT PgpdChlaerFAS2-Fw aacagctaccccgcttgagcagacatcaccATGCGTGG AGACGCTAGTAAATCCAGAGATC 2752 bp 60°C XLNfas2 cloning; lower case sequence is homologous to pTYGSade vector FASChlaer2-1R ATCCAATTCGCAGCGTGGAG FASChlaer2-2F ATCTCCTGACGCTATTGACG 1406 bp 60°C FASChlaer2-2R ACTCTCAGTAGTCGGACGAG FASChlaer2-3F TGTACAAGGATCAGTACCTCG 1991 bp 60°C TgpdChlaerFAS2-Rv acgacaatgtccatatcatcaatcatgaccCTAATAGCTT CTATCACCCTTCCCATGCAA vdtA_f1_F actttgtacaaaaaagcaggctccgcATGGCGCAAAA GCTTCGT 2626 bp 60°C vdtA cloning; lower case sequence is homologous to pEYA2 vector vdtA_f1_R GGTCCTCTACTAGAACTGAGCTA vdtA_f2_F GAACAACACTATACCTCCCCATC 2648 bp 60°C vdtA_f2_R CCACATCACCCTTCCACTGA vdtA_f3_F GTTCGTTAATCATGGGTGGGA 2392 bp 60°C vdtA_f3_R tcggcgcgcctgtttaaactgcggccCTAGCCCTTATA ACTGC H2B_Chlaer_F TCTCCAACCGTGCTATGTC gDNA: 290 bp; cDNA: 193 bp 60°C Gene expression checking H2B_Chlaer_R ACAGCCTTGGTACCTTCG H2B_Aspory_F GCTGCTGCCTCTGGTGAC gDNA: 532 bp; cDNA: 381 bp 60°C Gene expression checking H2B_Aspory_R GCTGCTGCCTCTGGTGAC PamyB-Fw ACCGACAACATCACATCAAGC XLNpks whole gene sequencing after cloning xln1seq-2F TGACCTCGATCTCAGATTGG xln1seq-3F TGGATGGAGACCAACTCTGC xln1seq-4F TGCATTTGAGCCTACGGAATG xln1seq-5F TGTCCTAACAGACTCCGATAC xln1seq-6F TGCCGTTGTAAAGGCCATTG xln1seq-7F TGTCCTTACGCAAACAGCTG xln1seq-8F TGGATCCACCCTCTGATGCTG xln1seq-9F TGAAATGCCACTACGACGAC xln1seq-1R AGAGGCTGCATCAGAATCTC xln1seq-3R TGCAGTAATGATGACCACGTG Identification of xylindein BGC 67 xln1seq-5R TGAAGTCCGACTCTGCAATG xln1seq-7R AGGGCTGAGTCGATCAACTC TamyB-Rev ACTCACTGTCCAATGCCAG Padh-Fw CCTAACTCCACCGCAACCTC XLNfas1 whole gene sequencing after cloning fas1seq-1R TGGATGCTACAACGTTGACC fas1seq-2F AGGAGACTCGCGGCTATATG fas1seq-2R TCACATCCTGTGATCGGAAG fas1seq-3F AGTCCTCATCTTAAAGGCGC fas1seq-3R TGCGAATCTTTCCGTACGCC fas1seq-4F TGATTGACGGGCTGACTATTG fas1seq-4R TGGAGCAACTCAGTTTCTGC fas1seq-5F AGCGGTGGATAGATTCTTCC fas1seq-5R TCCTGTACTCCATCTCCTCG fas1seq-6F TAAGGCCATCAAGGACTCTC fas1seq-6R AGGAGAACCGTATCGAGTCAG fas1seq-7F TCATGACCCAGGTCACTTAC fas1seq-7R TGCTGGTTGCGCGAATTGAG fas1seq-8F TCCAACTACCTAGCCATGAC Tadh-Rv CGATGGAATCTCATAATACTC Pgpd-Fw CGAGCTTTCCCACTTCATCG XLNfas2 whole gene sequencing after cloning fas2seq-1R TACTGTGGAAGCTTGCTGAG fas2seq-2F TCTAGGCAGTCTGCAGTCTTG fas2seq-2R TGCGCCTGTGACTAGAACATC fas2seq-3F ACCTAGATTGGTAAGAGGCC fas2seq-3R TCATCACGCCTTTAGACTCG fas2seq-4F TATCTCACCGTATGTGGAGC fas2seq-4R AAAGCCTGTCGGCAGTTGAC fas2seq-5F ATCCGACAGGTCGGAAATTC fas2seq-5R ACTCCTCTGAATCTGTTGAGC fas2seq-6F ACTCAACATAGACTACCGTCG fas2seq-6R AGGATCCATAAGTACTGCCG fas2seq-7F ACGCAACGTAAGAGACTTGC Tgpd-Rv CTCTTCATTTCTTTCATTATC pEYA2-Fw TCGAACTGGATCTCAACAGCG vdtA whole gene sequencing after cloning vdtA_seq_F1 TGCAGGATGTACTCCTTCG vdtA_f2_F GAACAACACTATACCTCCCCATC vdtA_seq_F2 ATGACAGCAGCGTAGTCATG vdtA_f3_F GTTCGTTAATCATGGGTGGGA vdtA_seq_F3 ATAGACTCGCTGCTGTCCTTG vdtA_seq_R1 TCCGTCTTCCAGAGTCACTG vdtA_seq_R2 ACACGTAGTTGATTCGGCCTG vdtA_seq_R3 TGCAAGCTGCATGCACGTATTG vdtA_seq_R4 AAGCCGTGACTCGCAGTAATTG vdtA_seq_R5 ATTGCTTTGGCGCGTCTGATAC vdtA_seq_R6 TAAGCGAGAATGCCACCG pEYA2-Rev ACAGTACTGCGATGAGTGGC Chapter 2 68 References 1. Thenard, P.; Rommier, A. (France) Sur un nouvelle matière colorante appelée xylindeine et extraite de certains bois morts. Comptes rendus hebdomadaires des séances de l'Académie des Sciences: Paris, 1868; Vol. t.66, 108–109. 2. Gutierrez, P.T.V.; Robinson, S.C. Determining the presence of spaltedwood in Spanish marquetry woodworks of the 1500s through the 1800s. Coatings 2017, 7, 1–14, doi:10.3390/coatings7110188. 3. Giesbers, G.; Van Schenck, J.; Quinn, A.; Van Court, R.; Vega Gutierrez, S.M.; Robinson, S.C.; Ostroverkhova, O. Xylindein: Naturally Produced Fungal Compound for Sustainable (Opto)electronics. ACS Omega 2019, 4, 13309–13318, doi:10.1021/acsomega.9b01490. 4. Krueger, T.D.; Tang, L.; Giesbers, G.; Van Court, R.C.; Zhu, L.; Robinson, S.C.; Ostroverkhova, O.; Fang, C. Ultrafast Triplet State Formation in a Methylated Fungi-Derived Pigment: Toward Rational Molecular Design for Sustainable Optoelectronics. J. Phys. Chem. C 2021, 125, 17565–17572, doi:10.1021/ACS.JPCC.1C06260. 5. Robinson, S.C.; Tudor, D.; Snider, H.; Cooper, P.A. Stimulating growth and xylindein production of Chlorociboria aeruginascens in agar-based systems. AMB Express 2012, 2, 1–7, doi:10.1186/2191-0855-2-15. 6. Donner, C.D.; Cuzzupe, A.N.; Falzon, C.L.; Gill, M. Investigations towards the synthesis of xylindein, a blue-green pigment from the fungus Chlorociboria aeruginosa. Tetrahedron 2012, 68, 2799–2805, doi:10.1016/J.TET.2012.02.009. 7. Boonloed, A.; Weber, G.L.; Ramzy, K.M.; Dias, V.R.; Remcho, V.T. Centrifugal partition chromatography: A preparative tool for isolation and purification of xylindein from Chlorociboria aeruginosa. J. Chromatogr. A 2016, 1478, 19–25, doi:10.1016/j.chroma.2016.11.026. 8. Mosunova, O.; Navarro-Muñoz, J.C.; Collemare, J. The Biosynthesis of Fungal Secondary Metabolites: From Fundamentals to Biotechnological Applications. Encycl. Mycol. 2021, 458–476, doi:10.1016/B978-0-12-8096338.21072-8. 9. Skellam, E. Strategies for Engineering Natural Product Biosynthesis in Fungi. Trends Biotechnol. 2019, 37, 416–427, doi:10.1016/j.tibtech.2018.09.003. 10. Steiniger, C.; Hoffmann, S.; Mainz, A.; Kaiser, M.; Voigt, K.; Meyer, V.; Süssmuth, R.D. Harnessing fungal nonribosomal cyclodepsipeptide synthetases for mechanistic insights and tailored engineering. Chem. Sci. 2017, 8, 7834–7843, doi:10.1039/c7sc03093b. 11. Bailey, A.M.; Alberti, F.; Kilaru, S.; Collins, C.M.; De Mattos-Shipley, K.; Hartley, A.J.; Hayes, P.; Griffin, A.; Lazarus, C.M.; Cox, R.J.; et al. Identification Identification of xylindein BGC 69 and manipulation of the pleuromutilin gene cluster from Clitopilus passeckerianus for increased rapid antibiotic production. Sci. Rep. 2016, 6, 1–11, doi:10.1038/srep25202. 12. Meng, X.; Fang, Y.; Ding, M.; Zhang, Y.; Jia, K.; Li, Z.; Collemare, J.; Liu, W. Developing fungal heterologous expression platforms to explore and improve the production of natural products from fungal biodiversity. Biotechnol. Adv. 2022, 54, 107866, doi:10.1016/j.biotechadv.2021.107866. 13. Wennrich, J.P.; Sepanian, E.; Ebada, S.S.; Llanos-Lopez, N.A.; Ashrafi, S.; Maier, W.; Kurtán, T.; Stadler, M. Bioactive Naphtho-α-Pyranones from Two Endophytic Fungi of the Genus Polyphilus. Antibiotics 2023, 12, 1–9, doi:10.3390/antibiotics12081273. 14. Büttner, E.; Liers, C.; Gebauer, A. M.; Collemare, J.; Navarro-Muñoz, J. C.; Hofrichter, M.; Kellner, H. Draft Genome Sequence of the Wood-Staining Ascomycete Chlorociboria aeruginascens DSM 107184. Microbiol Resour. Announc. 2019, 8, 17, e00249-19. doi: 10.1128/MRA.00249-19. 15. Cox, R.J.; Simpson, T.J. Chapter 3 Fungal Type I Polyketide Synthases; 1st ed.; Elsevier Inc., 2009; Vol. 459; ISBN 9780123745910. 16. Urquhart, A.S.; Hu, J.; Chooi, Y.-H.; Idnurm, A. The fungal gene cluster for biosynthesis of the antibacterial agent viriditoxin. Fungal Biol. Biotechnol. 2019, 6, 2, doi:10.1186/s40694-019-0072-y. 17. Hu, J.; Li, H.; Chooi, Y.-H. Fungal Dirigent Protein Controls the Stereoselectivity of Multicopper Oxidase-Catalyzed Phenol Coupling in Viriditoxin Biosynthesis. J. Am. Chem. Soc. 2019, 141, 8068–8072, doi:10.1021/jacs.9b03354. 18. Newman, A.G.; Townsend, C.A. Molecular Characterization of the Cercosporin Biosynthetic Pathway in the Fungal Plant Pathogen Cercospora nicotianae. J. Am. Chem. Soc. 2016, 138, 4219–4228, doi:10.1021/jacs.6b00633. 19. Fürtges, L.; Obermaier, S.; Thiele, W.; Foegen, S.; Müller, M. Diversity in Fungal Intermolecular Phenol Coupling of Polyketides: Regioselective LaccaseBased Systems. ChemBioChem 2019, 20, 1928–1932, doi:10.1002/cbic.201900041. 20. Mosunova, O. V; Navarro-Muñoz, J.C.; Haksar, D.; van Neer, J.; Hoeksma, J.; den Hertog, J.; Collemare, J. Evolution-Informed Discovery of the Naphthalenone Biosynthetic Pathway in Fungi. MBio 2022, 13, e0022322, doi:10.1128/mbio.0022322. 21. Blin, K.; Shaw, S.; Augustijn, H.E.; Reitz, Z.L.; Biermann, F.; Alanjary, M.; Fetter, A.; Terlouw, B.R.; Metcalf, W.W.; Helfrich, E.J.N.; et al. antiSMASH 7.0: new and improved predictions for detection, regulation, chemical structures and visualisation. Nucleic Acids Res. 2023, doi:10.1093/nar/gkad344. 22. Caceres, I.; Khoury, A. A.; Khoury, R. E.; Lorber, S.; Oswald, I. P.; Khoury, A. E.; Atoui, A.; Puel, O.; Bailly, J. D. Aflatoxin Biosynthesis and Genetic Chapter 2 70 Regulation: A Review. Toxins 2020, 12(3), 150. https://doi.org/10.3390/toxins12030150 23. Fujii, R.; Ugai, T.; Ichinose, H.; Hatakeyama, M.; Kosaki, T.; Gomi, K.; Fujii, I.; Minami, A.; Oikawa, H. Reconstitution of biosynthetic machinery of fungal polyketides: Unexpected oxidations of biosynthetic intermediates by expression host. Biosci. Biotechnol. Biochem. 2016, 80, 426–431, doi:10.1080/09168451.2015.1104234. 24. Griffiths, S.; Mesarich, C.H.; Saccomanno, B.; Vaisberg, A.; De Wit, P.J.G.M.; Cox, R.; Collemare, J. Elucidation of cladofulvin biosynthesis reveals a cytochrome P450 monooxygenase required for anthraquinone dimerization. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, 6851–6856, doi:10.1073/pnas.1603528113. 25. Newman, A.G.; Vagstad, A.L.; Belecki, K.; Scheerer, J.R.; Townsend, C.A. Analysis of the cercosporin polyketide synthase CTB1 reveals a new fungal thioesterase function. Chem. Commun. (Camb). 2012, 48, 11772–11774, doi:10.1039/c2cc36010a. 26. Sharp, P. M.; Li, W. H. The codon Adaptation Index--a measure of directional synonymous codon usage bias, and its potential applications. Nucleic Acids Res. 1987, 15(3), 1281–1295. https://doi.org/10.1093/nar/15.3.1281. 27. Puigbò, P.; Bravo, I.G.; Garcia-Vallve, S. CAIcal: A combined set of tools to assess codon usage adaptation. Biol. Direct 2008, 3, 1–8, doi:10.1186/1745-61503-38. 28. Tian, D.-S.; Kuhnert, E.; Ouazzani, J.; Wibberg, D.; Kalinowski, J.; Cox, R.J. The sporothriolides. A new biosynthetic family of fungal secondary metabolites. Chem. Sci. 2020, 11, 12477–12484, doi:10.1039/d0sc04886k. 29. Hynes, M.J.; Murray, S.L.; Duncan, A.; Khew, G.S.; Davis, M.A. Regulatory genes controlling fatty acid catabolism and peroxisomal functions in the filamentous fungus Aspergillus nidulans. Eukaryot. Cell 2006, 5, 794-805, doi:10.1128/EC.5.5.794-805.2006. 30. De Coster, W.; Rademakers, R. NanoPack2: population-scale evaluation of long-read sequencing data. Bioinformatics 2023, 39, btad311. doi:10.1093/bioinformatics/btad311. 31. Andrews, S. FastQC: A Quality Control Tool for High Throughput Sequence Data. 2010 [Online]. Available online at: http://www.bioinformatics.babraham.ac.uk/projects/fastqc/ 32. Kolmogorov, M.; Yuan, J.; Lin, Y.; Pevzner, P.A. Assembly of long, errorprone reads using repeat graphs. Nat. Biotechnol. 2019, 37, 540-546. doi:10.1038/s41587-019-0072-8. Identification of xylindein BGC 71 33. Mikheenko, A.; Prjibelski, A.; Saveliev, V.; Antipov, D.; Gurevich, A. Versatile genome assembly evaluation with QUAST-LG. Bioinformatics 2018, 34, i142-i150. doi:10.1093/bioinformatics/bty266. 34. Nishimura, O.; Hara, Y.; Kuraku, S. Evaluating Genome Assemblies and Gene Models Using gVolante. Methods Mol. Biol. 2019, 1962, 247–256, doi:10.1007/978-1-4939-9173-0_15. 35. Hoff, K.J.; Stanke, M. WebAUGUSTUS--a web service for training AUGUSTUS and predicting genes in eukaryotes. Nucleic Acids Res. 2013, 41, 123– 128, doi:10.1093/nar/gkt418. 36. Grigoriev, I. V; Nikitin, R.; Haridas, S.; Kuo, A.; Ohm, R.; Otillar, R.; Riley, R.; Salamov, A.; Zhao, X.; Korzeniewski, F.; et al. MycoCosm portal: gearing up for 1000 fungal genomes. Nucleic Acids Res. 2014, 42, D699-704, doi:10.1093/nar/gkt1183. 37. Gilchrist, C.L.M.; Chooi, Y.H. Clinker & clustermap.js: Automatic generation of gene cluster comparison figures. Bioinformatics 2021, 37, 2473–2475, doi:10.1093/bioinformatics/btab007. 38. Terlouw, B.R.; Blin, K.; Navarro-Muñoz, J.C.; Avalon, N.E.; Chevrette, M.G.; Egbert, S.; Lee, S.; Meijer, D.; Recchia, M.J.J.; Reitz, Z.L.; et al. MIBiG 3.0: a community-driven effort to annotate experimentally validated biosynthetic gene clusters. Nucleic Acids Res. 2023, 51, D603–D610, doi:10.1093/nar/gkac1049. 39. Katoh, K.; Misawa, K.; Kuma, K.; Miyata, T. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res. 2002, 30, 3059–3066, doi:10.1093/nar/gkf436. 40. Capella-Gutiérrez, S.; Silla-Martínez, J.M.; Gabaldón, T. trimAl: a tool for automated alignment trimming in large-scale phylogenetic analyses. Bioinformatics 2009, 25, 1972–1973, doi:10.1093/bioinformatics/btp348. 41. Guindon, S.; Dufayard, J.-F.; Lefort, V.; Anisimova, M.; Hordijk, W.; Gascuel, O. New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0. Syst. Biol. 2010, 59, 307– 321, doi:10.1093/sysbio/syq010. 42. Letunic, I.; Bork, P. Interactive Tree Of Life (iTOL) v5: an online tool for phylogenetic tree display and annotation. Nucleic Acids Res. 2021, 49, W293–W296, doi:10.1093/nar/gkab301. 43. Lazarus, C.M.; Williams, K.; Bailey, A.M. Reconstructing fungal natural product biosynthetic pathways. Nat. Prod. Rep. 2014, 31, 1339–1347, doi:10.1039/c4np00084f. 44. van Leeuwen, J.; Andrews, B.; Boone, C.; Tan, G. Rapid and Efficient Plasmid Construction by Homologous Recombination in Yeast. Cold Spring Harb. Protoc. 2015, 2015, 853–862, doi:10.1101/pdb.prot085100. Chapter 3 78 and a laccase (XLNlac and vdtB) (Figure 1A).11,14 Based on those similarities, a biosynthetic pathway for xylindein was proposed, which includes talaroderxine A as an intermediate (Figure 1B).14 However, a hallmark in the xylindein structure is that the dimerisation mechanism not only involves a traditional phenol coupling like observed in viriditoxin and talaroderxines, but it also requires C-O bonds uncommonly observed for this kind of polyketide. Such bonds are found in depsidone compounds in which a P450 is involved, like DepG from Aspergillus sp. SCSIO SX7S7 and MollD from Ovatospora sp. SCSIO SY280D that catalyse ether bonds to form nornidulin and mollicellins, respectively.15 Similarly, DpeB from Preussia isomera catalyses the formation of the C-O bond by oxidative conversion of lecanoric acid to depsidone.16 However, no cytochrome P450 is found encoded in the predicted xylindein BGC, and it was hypothesised that the putative carbonic anhydrase (CNH) XLNcnh could be involved in this dimerisation step (Figure 1B).14 Figure 1. Related biosynthetic gene clusters (BGCs) and shared biosynthetic steps in xylindein and viriditoxin production. (A) The xylindein BGC shares three homologous genes with the viriditoxin BGC, while g428 is not part of the xylindein BGC. Grey arrows represent genes included in the BGCs; white arrows represent genes located outside the BGCs. Numbers on the connecting lines indicate percent identity between homologous genes. (B) Overview of similar biosynthetic steps in xylindein and viriditoxin production. (XLNpks and VdtA: non-reducing polyketide synthases; XLNsdh and VdtF: short-chain dehydrogenases; XLNlac and VdtB: laccases; XLNfas1: fatty acid synthase β-subunit; XLNfas2: fatty acid synthase α-subunit; XLNcnh: carbonic anhydrase; XLNtf1, XLNtf2, XLNtf3 and VdtR: transcription factors; VdtC: O-methyltransferase; VdtD: α/β hydrolaselike protein; VdtE: Baeyer–Villiger monooxygenase; VdtG: transporter.) Adapted from Guo et al., 2025; Hu et al., 2019. Dimerisation of xylindein 79 In this study, we investigated the tailoring steps in xylindein biosynthesis. We refined the hypotheses about the biosynthetic steps using comparative genomics and phylogenetic analysis of the tailoring enzymes, and we performed heterologous expression in Aspergillus oryzae to characterise candidate enzymes. These efforts aim to advance our understanding of xylindein biosynthesis and pave the way for the scalable production of this promising natural product. Results and discussion Comparative genomics found a unique candidate gene for xylindein dimerisation Our previous research revealed that xylindein BGC is also present in a few other species, including the Leotiomycetes fungus Polyphilus sieberi, which is known to produce the xylindein-related compound talaroderxine C.14,17 In addition, talaroderxine A, the expected intermediate possibly produced by XLNlac in the xylindein pathway, is known to be produced by T. derxii.18 Thus, we expect that the genes involved in the early steps of xylindein biosynthesis in Chlorociboria species are also present in T. derxii, but the gene responsible for the uncommon dimerisation of xylindein must be absent. We performed de novo genome sequencing and genome mining of T. derxii (strain CBS 412.89), searching for the genes involved in the production of talaroderxine A. The obtained assembly of T. derxii is 31.1 Mb in size with a total of 11 assembled contigs and comprising 8229 predicted genes (Table S1). The low number of large contigs over 1 Mb in the T. derxii assembly and detection of telomeric repeats suggest that T. derxii very likely comprises 8 chromosomes (Figure S1). We then identified regions with predicted BGCs using antiSMASH.19 In total, 58 BGCs and 66 core biosynthetic proteins (CBPs) were found encoded in T. derxii genome, including 5 nrPKSs, 10 reducing PKSs, 5 nonribosomal peptide synthetases (NRPSs), 2 hybrid PKS-NRPSs, 15 NRPS-like enzymes, 11 terpene cyclases and 18 other CBPs (Table S2). As we hypothesise that the talaroderxine A BGC shares similarity with the xylindein BGC, especially for genes involved in the early biosynthetic steps, the single locus that harbours an nrPKS gene and a pair of fatty acid synthase (FAS) genes from contig 8 was identified as the candidate BGC for producing talaroderxine A (Figure 2). In addition, phylogenetic dereplication confirmed that the nrPKS in this BGC is the only one belonging to the nrPKS group IV like C. aeruginascens XLNpks (Figure S2). This BGC in T. derxii consists of genes encoding an nrPKS, a FAS, a LAC, an SDH, a Major Facilitator Superfamily (MFS) transporter, two transcription factors and two hypothetical proteins with no conserved domain (Figure 2). It corresponds to the conserved orthologous xylindein BGCs in other fungal species, Chapter 3 80 including P. sieberi (Figure 2).14 Because of this high conservation among all species, the predicted BGC for talaroderxines comprises three additional genes, encoding the transporter and two hypothetical proteins, compared to the predicted xylindein BGC (Figure 2). A Blastp search with homologues from Hysterium pulicare (g121378 and g121379) was conducted to investigate if these three proteins were present in the predicted proteome of C. aeruginascens (Table S3). The two hypothetical proteins showed no significant hit (C. aeruginascens g11112 shared 52% identity with H. pulicare g121379, but with a coverage of 69.4% only) or no hit at all (Table S3). Four significant hits were obtained for the MFS transporter (C. aeruginascens g2980, g4622, g3313, g9668) (Table S3), but they all share the same level of conservation (52 to 63%) and they are all located on different contigs. Thus, we conclude that true orthologues of these three genes are not present in C. aeruginascens genome. Figure 2. Comparative genomics of homologous xylindein biosynthetic gene clusters (BGCs). Conserved loci of the species with XLNpks homologous predicted by antiSMASH were compared using Clinker with an identity threshold of 0.3. Numbers on the links indicate the identity between the two linked genes. Phylogenetic analysis of tailoring enzymes putatively involved in xylindein biosynthesis Comparative genomic analysis shows a clear rearrangement of the xylindein genes in the genome of C. aeruginascens compared to the homologous locus in other fungi. Compared to T. derxii locus, the location of XLNfas1 and XLNfas2 was swapped and moved upstream to XLNtf1 and XLNtf2. XLNlac and XLNsdh were also swapped but remained downstream to XLNpks (Figure 2). Importantly, three additional genes were observed in the xylindein BGC compared to orthologous BGCs, including XLNtf3, XLNcnh, and a gene encoding an O-methyltransferase, which is also present in the viriditoxin BGC (Figure 1 and Figure 2). The gene encoding an Omethyltransferase is not expected to be involved in the putative xylindein pathway, Dimerisation of xylindein 81 and XLNtf3 is likely involved in the regulation of xylindein production (Figure 1).14 Thus, the observation that XLNcnh is the only gene at the locus in Chlorociboria species that is absent from the talaroderxine BGC strengthens the hypothesis that XLNcnh is likely involved in the formation of the C-O bonds during xylindein dimerisation. Although the conserved domain predicted for XLNcnh suggests it encodes a carbonic anhydrase (CNH), it is unlikely that a CNH plays a direct role in the biosynthesis of xylindein. CNHs are primarily involved in facilitating the interconversion between CO₂, H₂O, carbonic acid (H₂CO₃), and its dissociative ions (HCO₃⁻ and H⁺).20 In contrast, the biosynthesis of xylindein likely requires an oxidative enzyme capable of forming the C–O bonds essential for its structure. For instance, cytochrome P450 enzymes are known to catalyse the formation of ether linkages in depsidones.15,16 However, no P450 enzyme has been identified within the xylindein BGC14, suggesting that XLNcnh may encode an enzyme with a function distinct from typical CNHs. Alternatively, it is also possible that a cytochrome P450 gene located outside the xylindein BGC contributes to its biosynthesis. Our previous research proposed a putative biosynthetic pathway for xylindein with three candidate tailoring enzymes, XLNsdh, XLNlac, and XLNcnh (Figure 1).14 The comparative genomic analysis showed that XLNsdh and XLNlac genes are conserved among species with an XLNpks ortholog, including T. derxii, P. sieberi, H. pulicare, Stagonospora sp., F. fulophazii and P. macrospinosa (Figure 2).14 Such conservation suggests that XLNsdh and XLNlac co-evolved with XLNpks. In contrast, the origin of XLNcnh is uncertain because this gene was only found in the xylindein BGC.14 Phylogenetic analyses of these three tailoring enzymes were performed to explore their origin and evolutionary history and gain insight into their functions. In total, using XLNlac as a query, 1426 laccase sequences were retrieved from Pezizomycotina species, including 10 characterised enzymes that catalyse the oxidative dimerisation of viriditoxin (VdtB from P. variotii)11, aurofusarin (GIP1 from Gibberella zeae)9, dinapinone A (MCE from Talaromyces pinophilus)13, ustilaginoidin A (UstL from Ustilaginoidea virens)12, cercosporin (CTB12 from Cercospora beticola)21, elsinochrome C (ElcG from Phaeosphaeria nodorum)22 and four DHN melanin-related laccases (YA from Aspergillus nidulans and Aspergillus oryzae; Abr2 from Aspergillus fumigatus; Laccase from A. oryzae)23–26, respectively (Supplementary files S1-S3). The laccase tree is divided into well-supported phylogenetic clades and five of them contain the characterised laccases involved in SM biosynthesis (Figure 3A). Clade A exhibits high diversity, with most sequences Chapter 3 82 identified in Eurotiomycetes and Sordariomycetes (Figure 3A). XLNlac belongs to this clade, which also includes the laccase MCE, GIP1, VdtB and UstL involved in coupling dinapinone A, aurofusarin, viriditoxin, and ustilaginoidin A, respectively (Figure 3A). The monomers of xylindein, viriditoxin, and dinapinone share similar pyranone ring backbones; however, the regioselectivity of MCE to produce dinapinone (C8-C8’ coupling) is different compared to the other two (C6-C6’ coupling). Although XLNlac and VdtB share the same regioselectivity, XLNlac is phylogenetically closer to MCE (Figure 3A), suggesting that the phylogenetic relationship of laccases does not reflect their regiospecificity. Similarly, Gip1 and UstL belong to clade A, but differ in their regioselectivity (C8-C8’ and C9-C9’ couplings, respectively). The distant homology between XLNlac and VdtB is consistent with the phylogeny of nrPKSs as XLNpks was found to be a distant homolog of VdtA.14 Although many laccases function independently, like MCE and UstL, several laccases operate in collaboration with other enzymes. For example, the viriditoxin laccase VdtB works alongside VdtD, an α/β hydrolase-like protein that controls the stereoselectivity of VdtB’s catalysis.11 Similarly, Gip1 can only function in the form of a protein complex with AurS, a fasciclin-domain-containing protein, to facilitate the C-C bond formation in aurofusarin.9 Because laccases from the same clade function with different other tailoring enzymes, such enzymatic cooperation likely evolved independently. Clades B, C and D are restricted to Eurotiomycetes only and include DHN melaninrelated laccases (Figure 3A). In Aspergillus species, DHN melanin is produced by the PKS wA which releases YWA1 as the first intermediate.24 This PKS belongs to a BGC which encodes two laccases, Abr1 and Abr2, in A. fumigatus.25 Clade D is represented by Abr2 (Figure 3A), however, Abr1 was not retrieved in the sequence similarity search and thus is not included in this tree. It was shown that two laccases are actually involved in the production of a green conidial pigment in A. oryzae, YA (AO090011000755) and Laccase (AO090102000546).27 Two YA sequences from A. nidulans and A. oryzae define clade B, while A. oryzae Laccase is present in clade C (Figure 3A). As these three clades all contain DHN melanin-related laccases and are found in the same species, although there is no strong support to show they share a common ancestor, they likely originate from duplication events of a common ancestor. Clade E is more distantly related and harbours laccases involved in the production of perylene quinones (Figure 3A). The formation of two embedded C-C bridges between the naphthol units in elsinochromes is catalysed by the laccase ElcG and a berberine bridge enzyme-like oxidase ElcE. The laccase catalyses the first radical coupling reaction between the two naphthol units (Figure 3A).6 However, Dimerisation of xylindein 83 the coupling in cercosporin is synthesised by the protein complex of laccase CTB12 and fasciclin-domain-containing protein CTB11, which is similar to that of aurofusarin biosynthesis.21 Besides the five characterised clades, there are also large clades with no characterised enzyme and it remains to be determined whether they are involved in SM biosynthesis or other biological processes. Each clade in the phylogenetic tree appears to represent different BGCs, but the relationships between clades are not fully resolved. Yet, XLNlac likely shares a common ancestor with laccases involved in the production of viriditoxin, aurofusarin, dinapinone A and ustilaginoidin A. Within clade A, Chlorociboria XLNlac clearly shares a common ancestor with laccases from T. derxii, P. sieberi, P. macrospinosa, F. fulophazii, Stagonospora sp., and H. pulicare (Figure 3B). Unexpectedly, both PKS and laccase in the Leotiomycetes P. sieberi seem to have diverged more rapidly as they form an outgroup in both trees (Figure 3B and 3C). The topology of this clade in the laccase tree is identical to the topology of the phylogenetic tree of XLNpks and its orthologues in these species (Figure 3C; Supplementary file S4-S6)14 suggesting co-evolution between XLNlac and XLNpks. Using XLNsdh as a query, 2820 SDH sequences were retrieved from Pezizomycotina species to perform the phylogenetic analysis (Supplementary file S7-S9), including 8 characterised sequences (Figure 4A). Two characterised SDH sequences from Agaricomycetes were used as an outgroup to root the tree. EriH from Hericium erinaceus is part of the gene cluster that mediates the biosynthesis of erinacines28 29, and ARMGADRAFT_1018421 from Armillaria gallica is involved in the biosynthesis of melleolides (Figure 4A).30 Five SDH characterised sequences, ClaC from Fulvia fulva (cladofulvin BGC)31, 10, AacuN from Aspergillus aculeatus (secalonic acid D BGC) 32, 33, PenD from Penicillium thymicola (penigequinolones BGC) 34, NsrJ from Aspergillus novofumigatus (neosartorin BGC)35, and MdpC from Aspergillus nidulans (monodictyphenone BGC)36 belong to a single clade that forms an outgroup to the rest of the sequences (Figure 4A). The last characterised sequence is the viriditoxin SDH, VdtF from P. variotii.11 Similarly to the nrPKS and laccase trees, VdtF is only distantly related to XLNsdh (Figure 4A). The topology of the SDH tree shows that XLNsdh shares a common ancestor with SDH from P. macrospinosa, F. fulophazii, Stagonospora sp., H. pulicare and T. derxii. However, the SDH encoded in the BGC in P. sieberi is not closely related (Figure 4A), consistent with the observed faster divergence of the PKS and laccase (Figure 3B and 3C). Yet, because the topology of XLNsdh orthologs is similar to both nrPKS and laccase trees, we conclude that XLNsdh, XLNlac and XLNpks have co-evolved (Figure 3B, 3C and 4B). Chapter 3 84 Figure 3. Phylogenetic analysis of fungal laccases. (A) The maximum-likelihood phylogenetic tree was built with homologous sequences to XLNlac. Branch colours indicate fungal taxonomic classes. Black dots indicate significant branch support (likelihood ratio test >80 and ultrafast bootstrap >95). The tree is midpoint rooted. Chemical structures out of the tree present the characterised examples catalysed by laccases. Different C-C coupling types were highlighted with the same colour as the corresponding clade. (B) XLNlac and its orthologs clade extracted from (A). (C) Maximum-likelihood phylogenetic tree of XLNpks and its orthologs. The tree is rooted with the Paecilomyces variotii VdtA sequence. Protein identifiers from JGI Mycocosm are indicated next to the species names. Dimerisation of xylindein 85 Figure 4. Phylogenetic analysis of fungal short-chain dehydrogenases (SDH). (A) The maximum-likelihood phylogenetic tree was built with homologous sequences to XLNsdh. Branch colours indicate fungal taxonomic classes. Black dots indicate significant branch support (likelihood ratio test >80 and ultrafast bootstrap >95). The tree is rooted with Basidiomycota SDHs. (B) XLNsdh and its orthologs clade extracted from (A). Protein identifiers from JGI Mycocosm are indicated next to the species names. Chapter 3 86 Figure 5. Phylogenetic analysis of carbonic anhydrases (CNH). (A) The maximumlikelihood phylogenetic tree was built with homologous sequences to XLNcnh. Branch colours indicate fungal taxonomic classes. Black dots indicate significant branch support (likelihood ratio test >80 and ultrafast bootstrap >95). The tree is rooted with mammal CNHs. (B) XLNcnh and closest homologs clade extracted from (A). Protein identifiers from JGI Mycocosm are indicated next to the species names. Because XLNcnh is unique in the BGC of Chlorociboria species, we investigated how this gene could have been recruited in the pathway. CNHs catalyse the reversible hydration of CO2 to H2CO3 to maintain intracellular pH homeostasis.20,37 The only characterised fungal CNH, NCE103 from Saccharomyces cerevisiae, is involved in fungal growth and its main role appears to provide inorganic carbon for the bicarbonate-dependent carboxylation reactions in primary metabolic Dimerisation of xylindein 87 pathways.38,39 To understand the origin of XLNcnh, we retrieved 4210 CNH sequences from Pezizomycotina species (Supplementary file S10-S12) and included four mammal CNHs to root the phylogenetic tree (Figure 5A). NCE103 is found in a large well-supported clade with diverse species (Figure 5A). XLNcnh does not belong to this clade and is not found in a strongly supported clade, with the most closely homologous found in other Leotiomycetes and one Eurotiomycetes (Figure 5B). This phylogenetic tree does not allow for determining the origin of XLNcnh. Previously, it was suggested that the enzymatic activity of XLNcnh is unlikely to be the same as other CNHs if it would be involved in xylindein dimerisation14. The position of XLNcnh in a distant clade to NCE103 is coherent with this assumption and XLNcnh may have acquired a different enzymatic activity for the dimerisation of xylindein. Indeed, engineered CNHs could be obtained to catalyse non-natural reactions like styrene epoxidation and hydrogenation of imines, but also to exhibit an efficient peroxidase activity that would be compatible with the expected oxidative chemistry.20 Functional characterisation of XLNsdh as a pyranone ring reducing enzyme In the proposed biosynthetic pathway of xylindein14, the step after releasing the intermediate from XLNpks is catalysed by XLNsdh. Although previous attempts to obtain the XLNfas and XLNpks precursor failed, heterologous expression of the homologous nrPKS VdtA from the viriditoxin pathway yielded the expected intermediate.14 This molecule is very similar to the predicted xylindein monomer and thus we speculated that XLNsdh would be able to use it as a substrate. The A. oryzae transformant expressing vdtA produces two related main compounds, product 1 and product 2, which only differ by the reduction of the side acyl chain (Figure 6A and 6B).14 Heterologous co-expression was conducted by randomly inserting XLNsdh cloned from cDNA in the genome of the A. oryzae AoADE::VDTA transformant. A total of 30 AoADEARG::VDTA::SDH transformants were obtained, of which 26 resulted in a difference in compound composition compared to the vdtA control (Figure 6A, Table S4). RT-PCR confirmed the gene expression of vdtA and XLNsdh in the selected transformants (Figure S3). HPLC profiling shows they all produced two additional compounds that were not observed in the vdtA control (Figure 6A). Product 3 [retention time (RT) = 4.941 min; UV maximum = 220, 259, 312 and 367 nm; m/z (electrospray; ES−) 303 [M−H]−] and product 4 [RT = 5.067 min; UV maximum = 220, 259, 313 and 368 nm; m/z (electrospray; ES−) 301 [M−H]−] are related based on their UV spectra (Figure 6B-6E). High-resolution mass spectrometry (HRMS) determined the exact mass of product 3 [m/z (ES+) 305,1020 Chapter 3 94 Nanopore Technologies native barcoding kit SQK-NBD114.24 (ONT, Oxford, United Kingdom) according to the manufacturer’s protocol (version NBE_9769_v114_revI_15Sep2022, latest update 12/07/2023) using approximately 1 µg DNA per sample. The final prepared libraries were sequenced on R10.4.1 flow cells (ONT, Oxford, United Kingdom) with the GridION sequencer (ONT, Oxford, United Kingdom). Base-calling of the reads was performed by Guppy with the high accuracy base-calling model in Minknow. The first and last 50 bp of all “pass” reads were then chopped using CHOPPER.40 The quality of the chopped reads was then checked using FastQC version 0.12.141 and Flye 2.9.2-b178642 (time flye-- nano) was used to assemble the chopped reads. The quality of the assembly was checked using Quast version 5.2.0.43 Completeness of the assemblies was determined by a BUSCO v5 analysis using gVolante44 with the Eurotiomycetes lineage dataset. Genes were predicted by Augustus v3.5.045 with the Aspergillus oryzae species model. Assemblies were visualised and telomeric regions were predicted using Telovision (https://github.com/WesterdijkInstitute/TeloVision). Phylogenetic trees and comparative genomics Close homologues of C. aeruginascens XLNlac, XLNsdh and XLNcnh were retrieved from the Joint Genome Institute (JGI) MycoCosm repository46 using BLASTp against Pezizomycotina. Regions containing BGCs were retrieved using fungiSMASH 7.0 with default parameters.19 BGC comparison was performed using Clinker.47 Characterised laccases were retrieved from UniProt.48 Protein alignments were performed using MAFFT version 7.49049 (parameters --reorder). Poorly aligned regions were removed using trimaL version 1.450 (build 2013-12-17; parameter -automated1). Maximum likelihood trees were built with IQ-TREE version 2.2.0-beta with model finder and ultrafast bootstrapping, as well as an approximate likelihood-ratio test (parameters-mset LG -bb 1000 -alrt 1000 -T AUTO).51 The substitution model used for trees of XLNlac, XLNpks, XLNsdh and XLNcnh homolog phylogeny are LG+I+R9, LG+F+G4, Q. insect+R10 and LG+R9, respectively. The resulting trees were visualised using iTOL.52 All curated alignments and phylogenetic tree files are provided in the supplementary (Supplementary files S1-S12). Gene amplification (XLNsdh, XLNlac, XLNcnh, vdtB) and expression vector construction (pTYGSarg::XLNsdh, pTYGSbar::XLNlac, pTYGSbar::vdtB, pTYGSbar::XLNcnh, pTYGSbar::XLNlac::XLNcnh, pTYGSbar::vdtB::XLNcnh) Dimerisation of xylindein 95 XLNsdh, XLNlac, and XLNcnh genes were amplified from the cDNA of C. aeruginascens; vdtA was amplified from the genomic DNA of P. variotii. Primers containing attB1 and attB2 tails were used to create the DNA template for BP Gateway® cloning (Table S6). XLNcnh was also amplified with the primers containing Padh and Tadh tails for yeast homologous recombination (Table S6). All the genes were amplified using Phusion high-fidelity DNA polymerase (Thermo Fisher Scientific, Waltham, MA) according to the manufacturer’s protocol. Amplicons were purified using a Wizard SV Gel and purification kit (Promega, Madison, U.S.A.). Purified attB-tailed XLNsdh, XLNlac, and XLNcnh were cloned in a pDONR207 plasmid through BP-gateway cloning. For this, 70 ng of these PCR products were mixed with 70 ng empty pDONR207 plasmid, 1 µL BP Clonase II (Invitrogen, Waltham, U.S.A.), and MilliQ water to 5 μL per reaction. After 3 hours of incubation at 25℃, the reaction was mixed with 50 μL chemically competent DH5α E. coli cells, and heat-shock transformation was performed. Hereafter, the transformed E. coli was incubated at 37℃ overnight using Lysogeny agar (LA) containing gentamycin. Positive colonies were confirmed by colony-PCR using pDONOR207 primers (Table S6) and sequencing (Macrogen, Amsterdam, The Netherlands). Correct plasmids were selected for the LR reaction, which followed a similar protocol to the BP reaction. Seventy nanograms of the gene-containing pDONOR207 entry vector and 70 ng of the pTYGSarg or pTYGSbar destination vector were mixed with 1 µL of the Gateway LR Clonase II enzyme (Thermo Fisher Scientific, Waltham, MA) in a 5-µL final volume, and the reaction mixture was incubated at 25 °C for three hours. The total reaction mixture was introduced into chemically competent E. coli DH5a cells (Thermo Fisher Scientific, Waltham, MA) using a heat shock protocol. The gene-containing pTYGSarg and pTYGSbar expression vectors were isolated from positive colonies using the Zyppy plasmid miniprep kit (Zymo Research, Irvine, CA) and HiPure Plasmid Midiprep Kit (Thermo Fisher Scientific, Waltham, MA) for A. oryzae transformation. The yeast homologous recombination protocol was adapted from our previous research.14 XLNcnh was assembled under the adh promoter on the pTYGSbar::XLNlac and pTYGSbar::vdtB plasmids, respectively. A. oryzae transformation The A. oryzae transformant AoADE::VDTA was used as a host strain and transformed with expression vectors of different combinations. Spores from AoADE::VDTA were harvested from selection CZD agar B (Table S5) plates in 5 mL of sterile water, and 1 mL of this spore suspension was inoculated into 50 mL of MB liquid medium and Chapter 3 96 grown overnight at 28 °C with shaking at 200 rpm. Germinating spores were collected by centrifugation at room temperature for 10 min at 3,500 rpm and resuspended in 25 mL of 0.8 M NaCl. After spin down for 10 min at 3,500 rpm at room temperature, germinated spores were resuspended in 10 mL of a freshly made filter-sterilized protoplasting solution (200 mg Trichoderma lysing enzyme and 50 mg Driselase (both from ThermoFisher Scientific, Waltham, MA) in 0.8 M NaCl) and incubated at 30 °C for 2 to 2.5 h with shaking at 100 rpm. Protoplasts were filtered through sterile Miracloth and then centrifuged for 5 min at 3,000 rpm at 4 °C. Protoplasts were resuspended in 200 µL of solution 1 (0.8 M NaCl, 10 mM CaCl2, and 50 mM Tris-HCl pH 7.5) and aliquoted to 100 µL in 2-mL microcentrifuge tubes. Ten micrograms of each expression plasmid or pTYGSarg and pTYGSbar empty vectors were added to protoplasts, and the mixture was incubated on ice for 2 min. One millilitre of solution 2 (60% (wt/vol) polyethylene glycol 3350, 0.8 M NaCl, 10 mM CaCl2, and 50 mM TrisHCl pH 7.5) was added, and the tubes were gently inverted before incubation at room temperature for 20 min. Protoplasts were then mixed with 25 mL of cooled Top CZD agar (Table S5) and immediately plated onto the corresponding Bottom CZD agar plates (Table S5). Transformation plates were incubated at 30 °C for 3 to 10 days. Transformants were transferred onto new selection CZD agar plates individually and transferred onto DPY agar plates (Table S5) for induction. Secondary metabolite extraction, HPLC and HRMS analyses. The 4-day-old A. oryzae transformants on DPY agar plates (Table S5) were cut and collected into the 50 mL tubes and isolated with ethyl acetate (VWR Chemicals, Radnor, PA) for screening the metabolic profile. After shaking on an orbital shaker for at least 1 h, the organic phase was transferred to a new 50 mL tube and evaporated under nitrogen flow. For 50 mL DPY broth culture (Table S5), secondary metabolites from 4-day-old transformant liquid culture filtrates were isolated with a 1:1 volume of ethyl acetate and followed by the same shaking and evaporating steps. The resulting solid was dissolved in acetonitrile. Organic extracts were analysed with a Shimadzu LC-2030 3D Prominence-i PDA system coupled to a Shimadzu LCMS2020 mass spectrometer and equipped with a Shimadzu Shim-pack GIST C18-HP reversed-phase column (3 mm, 4.6 by 100 mm). The following method was used: a linear gradient of buffer B (5% to 95%) for 10 min, 2 min of 95% buffer B, the gradient of buffer B (95% to 5%) for 1 min, and then 5% buffer B for 5 min. Water with 0.1% trifluoroacetic acid (TFA) for high-performance liquid chromatography (HPLC), or 0.05% formic acid for mass spectrometry (MS)-coupled analyses were Dimerisation of xylindein 97 used as buffer A, and acetonitrile (LCMS grade) with 0.1% TFA for HPLC, or 0.05% formic acid for MS-coupled analyses was used as buffer B. The flow rate was 1 mL min-1 or 0.5 mL min-1 for UV-HPLC or MS-coupled analyses, respectively. The results were analysed using Shimadzu LabSolutions LC-MS software. High Resolution Mass Spectrometry (HRMS) was performed on a Q-Tof Premier mass spectrometer (Waters) coupled to an Acquity UPLC system (Waters). Electron spray Ionisation (ESI) mass spectroscopy was measured in positive or negative mode depending on the compound. Supplementary Files Supplementary file S1. Laccase sequences used in the phylogenetic analysis. Supplementary file S2. Trimmed alignment of laccase sequences. Supplementary file S3. Tree file of laccase phylogeny. Supplementary file S4. Non-reducing polyketide synthase (nrPKS) sequences used in the phylogenetic analysis. Supplementary file S5. Trimmed alignment of nrPKS sequences. Supplementary file S6. Tree file of nrPKS phylogeny. Supplementary file S7. Short-chain dehydrogenase sequences used in the phylogenetic analysis. Supplementary file S8. Trimmed alignment of short-chain dehydrogenases. Supplementary file S9. Tree file of short-chain dehydrogenase phylogeny. Supplementary file S10. Carbonic anhydrase sequences used in the phylogenetic analysis. Supplementary file S11. Trimmed alignment of carbonic anhydrases. Supplementary file S12. Tree file of carbonic anhydrase phylogeny. Supplementary materials can be accessed and downloaded via the following link: https://doi.org/10.5281/zenodo.16631399 Chapter 3 98 Supplementary figures Figure S1. Chromosome-level assembly of Talaromyces derxii CBS 412.89 visualized by TeloVision. Each scaffold is drawn to scale, and the GC content is indicated with the color scale. Detected telomeric regions are visualized as black boxes at the ends of scaffolds, while the absence of telomeres is indicated with a white box at the end of scaffolds. T. derxii very likely comprises 8 chromosomes. The detected consensus telomeric repeat is CCTAAC. Dimerisation of xylindein 99 Figure S2. Phylogenetic dereplication of polyketide synthases (PKSs) of Talaromyces derxii CBS 412.89. (A) Phylogenetic tree of non-reducing PKSs. Clades are defined according to Mosunova et al. (2022). (B) Phylogenetic tree of reducing PKSs. Phylogeny trees were built with characterised enzymes from the MIBiG 2.0 database. Red triangle indicates the candidate nrPKS for talaroderxine A biosynthesis and red dots indicate other PKSs from T. derxii. (C) Ochratoxin, (D) communesin and (E) fumagillin BGC comparison using Clinker. group V group VII group I group IVa group III group IIa group VI group IVb group VIII group IX group IIb group X eurotiomycetes|Aspfla|BGC0000008|AAS90022.1||aflatoxin eurotiomycetes|Aspfla|BGC0000006|AAS90093.1|pksA|aflatoxin eurotiomycetes|Aspory|BGC0000004|BAE71314.1|pksA|aflatoxin eurotiomycetes|Aspfla|BGC0000007|AAS89999.1||aflatoxin eurotiomycetes|Aspnom|BGC0000009|AAS90047.1||aflatoxin eurotiomycetes|Aspoch|BGC0000011|ACH72912.1||sterigmatocystin eurotiomycetes|Aspnid|BGC0000152|AAC49191.1|stcA|sterigmatocystin dothideomycetes|Dotsep|BGC0000048|AAZ95017.1|pksA|dothistromin CBS 412.89.contig 8.region002~L0+CDS5 dothideomycetes|Cerbet|BGC0001541|PIB02405.1||cercosporin dothideomycetes|Cerzei|BGC0001542|ARU80380.1|CTB1|cercosporin dothideomycetes|Parnod|BGC0001865|XP 001798923.1||elsinochrome sordariomycetes|Fusfuj|BGC0001242|CCE67070.1|fsr1|oxyjavanicin eurotiomycetes|Bysspe|BGC0001866|XP 028481820.1||viriditoxin eurotiomycetes|Aspfla|BGC0001446|EED57518.1||asparasone eurotiomycetes|Aspory|FNP030|XP 001822700.1||YWA1 eurotiomycetes|Aspnid|BGC0000107|CBF74114.1|wA|naphthopyrone eurotiomycetes|Paevar|FNP129|P.|variotii|Ywa1 CBS 412.89.contig 9.region007~L0+CDS6 eurotiomycetes|Aspory|FNP123|XP 023094012.1||8-methyldiaporthin sordariomycetes|Fusfuj|BGC0000030|CCT67991.1||bikaverin eurotiomycetes|Endpus|FNP095|ERF77221.1|| eurotiomycetes|Talsti|BGC0001578|EED21099.1||duclauxin sordariomycetes|Nodsp.|BGC0001257|AAD38786.1|pks1|1 sordariomycetes|Dalesc|BGC0001906|||naphthalene sordariomycetes|Pesfic|FNP011|XP 007833873.1||DHN sordariomycetes|Pyrory|FNP001|XP 003715434.1||1 leotiomycetes|Glaloz|BGC0001258|AAN59953.1|pks1|1 dothideomycetes|Bipory|BGC0001265|BAD22832.1|PKS1|melanin eurotiomycetes|Aspter|BGC0000161|EAU38791.1||isoterrein lecanoromycetes|Claunc|BGC0001489|AUW31184.1||6-hydroxymellein eurotiomycetes|Aspfum|BGC0001403|EAL89339.1||trypacidin CBS 412.89.contig 2.region002~L0+CDS8 sordariomycetes|Pesfic|BGC0000121|AGO59040.1|ptaA|RES-1214-2 eurotiomycetes|Aspnid|BGC0000101|CBF90097.1||emodin eurotiomycetes|Aspnov|BGC0001988|PKX92308.1||neosartorin sordariomycetes|Escweb|BGC0001583|DAB41653.1|EmoG|emodin dothideomycetes|Pasful|FNP054|184395|184395|cladofulvin sordariomycetes|Clapur|BGC0001886|CCE31584.1||secalonic eurotiomycetes|Aspfum|BGC0001118|EAL84397.1||endocrocin eurotiomycetes|Aspfla|BGC0001304|EED53479.1||aflavarin eurotiomycetes|Penaet|BGC0000070|ADI24953.1||griseofulvin dothideomycetes|Parnod|BGC0001284|AKN45693.1||alternariol eurotiomycetes|Aspnig|BGC0000156|AEN83889.1|adaA|TAN-1612 eurotiomycetes|Aspnid|BGC0000684|CBF70387.1||asperthecin eurotiomycetes|Penaet|BGC0000168|ADI24926.1||viridicatumtoxin eurotiomycetes|Triton|BGC0001144|EGD99348.1||neosartoricin eurotiomycetes|Aspory|FNP124|XP 001823362.1||dichlorodiaporthin eurotiomycetes|Aspnid|BGC0000013|CBF79143.1||alternariol lecanoromycetes|Clagra|BGC0001266|ADM79459.1|PKS16|grayanic sordariomycetes|Hypsub|BGC0000076|ACD39753.1|hpm3|hypothemycin sordariomycetes|Hypsub|BGC0000077|ACD39762.1|hpm3|hypothemycin dothideomycetes|Lasthe|BGC0001245|AHV78247.1|lasS2|lasiodiplodin sordariomycetes|Fusgra|BGC0001057|ABB90282.1|PKS13|zearalenone sordariomycetes|Pocchl|BGC0000134|ACD39770.1|Rdc1|radicicol sordariomycetes|Sarzea|BGC0001246|AHV78253.1|resS2|trans-resorcylide eurotiomycetes|Aspter|BGC0000045|AGC95321.1|curS2|dehydrocurvularin eurotiomycetes|Aspnid|BGC0000057|EAA59563.1||F9775A ||BGC0001121|CBF73505.1|| sordariomycetes|Trihar|BGC0001854|KKP00966.1||tricholignan agaricomycetes|Monror|FNP092|ESK96613.1|| ||BGC0001401|AFL91703.1|| sordariomycetes|Beabas|BGC0001720|EJP62792.1||oosporein CBS 412.89.contig 13.region007~L0+CDS9 sordariomycetes|Sarsp.|BGC0001827|AWM95789.1|aspks1|xenovulene dothideomycetes|Phosp.|BGC0001976|QCO93110.1|eupA|eupenifeldin eurotiomycetes|Aspnid|FNP117|||derivative CBS 412.89.contig 8.region005~L0+CDS20 eurotiomycetes|Talsti|BGC0000154|EED18001.1||stipitatic eurotiomycetes|Monpil|BGC0000027|AGN71604.1||ankaflavin eurotiomycetes|Talmar|BGC0000099|ADH01663.1||monascorubrin eurotiomycetes|Aspnig|BGC0001143|EHA28237.1||azanigerone sordariomycetes|Acrchr|FNP051|KFH44362.1||Sorbicillinoid eurotiomycetes|Penrub|BGC0001404|CAP95404.1|Pc21g05070|sorbicillin eurotiomycetes|Aspnid|BGC0000022|EAA65602.1||asperfuranone sordariomycetes|Chaglo|BGC0001405|EAQ86392.1||chaetoviridin sordariomycetes|Trigui|FNP085|OPB37950.1||trigazaphilones eurotiomycetes|Aspnid|BGC0001722|CBF83139.1||aspernidine dothideomycetes|Ascfab|FNP050|QEN17973.1||ascochitine eurotiomycetes|Monrub|BGC0001338|ALI92655.1|citS|citrinin eurotiomycetes|Aspnid|BGC0000037|CBF69451.1||cichorine eurotiomycetes|Penroq|BGC0001360|CDM36726.1||mycophenolic eurotiomycetes|Penroq|BGC0001677|API82664.1|mpaC|mycophenolic eurotiomycetes|Penbre|BGC0000104|ADY00130.1|mpaC|mycophenolic eurotiomycetes|Penroq|BGC0001508|ART41209.1|adrD|andrastin eurotiomycetes|Aspnov|BGC0001708|PKX88487.1||novofumigatonin sordariomycetes|Chaglo|BGC0001219|EAQ84779.1||shanorellin sordariomycetes|Stabis|BGC0001390|BAV19379.1|stbA|LL-Z1272beta sordariomycetes|Stachl|BGC0001626|KFA69335.1||isoindolinone sordariomycetes|Acregy|BGC0001923|BBF25315.1|ascC|ascochlorin s A group macrolide group MSA group alkyl citrates group lipopeptide group basidiomycota sordariomycetes|Hypsub|BGC0000076|ACD39758.1|hpm8|hypothemycin sordariomycetes|Hypsub|BGC0000077|ACD39767.1|hpm8|hypothemycin dothideomycetes|Lasthe|BGC0001245|AHV78245.1|lasS1|lasiodiplodin sordariomycetes|Fusgra|BGC0001057|ABB90283.1|PKS4|zearalenone eurotiomycetes|Aspter|BGC0000045|AGC95324.1|curS1|dehydrocurvularin sordariomycetes|Sarzea|BGC0001246|AHV78252.1|resS1|trans-resorcylide sordariomycetes|Pocchl|BGC0000134|ACD39774.1|Rdc5|radicicol sordariomycetes|Trihar|BGC0001901|KKP00963.1||tricholignan sordariomycetes|Fusver|BGC0001190|EWG54266.1||fusaric CBS 412.89.contig 8.region007~L0+CDS14 dothideomycetes|Epinig|FNP068|OSS48297.1||Epipyrone dothideomycetes|Altsol|BGC0000012|BAD83684.1|alt5|alternapyrone eurotiomycetes|Talver|BGC0001926|BBG28484.1|cdmE|chrodrimanin CBS 412.89.contig 13.region014~L0+CDS3 eurotiomycetes|Pennor|BGC0001030|AUS29495.1||ochratoxin CBS 412.89.contig 8.region006~L0+CDS9 sordariomycetes|Pyrory|FNP002|XP 003718178.1||nectriapyrone sordariomycetes|Zopcur|FNP102|BBU42026.1|zopA|zopfiellin sordariomycetes|Difcur|FNP104|QTE75992.1|zopPKS|zopfiellin leotiomycetes|Scyalb|FNP103|QTE76000.1|scyPKS|scytalidin eurotiomycetes|Paediv|BGC0001436|ASK38717.1|pvpks1|cornexistin eurotiomycetes|Paediv|BGC0001557|ASK38717.1|pvpks1|cornexistin eurotiomycetes|Bysful|BGC0001340|ANF07288.1|bfpks1|byssochlamic CBS 412.89.contig 9.region001~L0+CDS7 |funsp.|BGC0001913|BBG28498.1|phiA|phomoidride dothideomycetes|Altalt|BGC0000003|BAD97694.1|AFT9-1|AF-toxin sordariomycetes|Sorara|BGC0001373|BAV32159.1|sdnO|sordarin sordariomycetes|Metani|FNP128|KFG78606.1||BAA eurotiomycetes|Penexp|BGC0001205|KGO40478.1||communesin CBS 412.89.contig 9.region005~L0+CDS7 eurotiomycetes|Aspnid|BGC0000022|EAA65604.1||asperfuranone sordariomycetes|Chaglo|BGC0001405|EAQ86385.1||chaetoviridin sordariomycetes|Trigui|FNP085|OPB37944.1||trigazaphilones sordariomycetes|Acrchr|FNP051|KFH44396.1||Sorbicillinoid eurotiomycetes|Penrub|BGC0001404|CAP95405.1|Pc21g05080|sorbicillin CBS 412.89.contig 8.region005~L0+CDS26 eurotiomycetes|Aspter|BGC0000088|AAD34559.1|lovF|monacolin eurotiomycetes|Monpil|BGC0000098|ABA02240.1|mkB|monacolin eurotiomycetes|Pencit|BGC0000039|BAC20566.1|mlcB|compactin eurotiomycetes|Aspnig|BGC0001143|EHA28244.1||azanigerone dothideomycetes|Phosp.|BGC0001339|AMY15057.1|mfpks1|squalestatin dothideomycetes|Lepmac|BGC0001899|CBX99534.1||abscisic dothideomycetes|Altalt|BGC0001252|BAN19720.1|ACRTS2|UNII-YC2Q1O94PT |Shuvec|BGC0001816|AQM58285.1|lptB|valactamide CBS 412.89.contig 1.region001~L0+CDS11 sordariomycetes|Fusgra|FNP046|XP 011320599.1||Fusaristatin sordariomycetes|Fuspse|FNP049|||Fusaristatin eurotiomycetes|Aspnid|BGC0001290|CBF87072.1||emericellamide sordariomycetes|Fuspse|FNP048|XP 009260579.1|PKS6|W493 dothideomycetes|Phosp.|BGC0001966|QCC63000.1||BII-rafflesfungin sordariomycetes|Fusver|BGC0000062|AAD43562.2|FUM1|fumonisin sordariomycetes|Fusoxy|BGC0000063|ACB12550.1|fum1|fumonisin leotiomycetes|Glaloz|BGC0001035|EPE34340.1||pneumocandin sordariomycetes|Purlil|BGC0001358|OAQ83765.1||leucinostatin CBS 412.89.contig 11.region002~L0+CDS6 eurotiomycetes|Bysspe|BGC0001866|XP 028481819.1||viriditoxin sordariomycetes|Trigui|FNP085|QHD56300.1|polyketide|trigazaphilones sordariomycetes|Dalesc|BGC0001907|||chromane dothideomycetes|Phosp.|BGC0001339|AMY15068.1|mfpks2|squalestatin |Pensop|FNP066|||soppiline sordariomycetes|Purlil|BGC0001358|OAQ83760.1||leucinostatin sordariomycetes|Beafel|BGC0001565|ATQ39432.1||cyclosporin sordariomycetes|Neucra|FNP065|XP 965600.1|pks-6|sordarial sordariomycetes|Neucra|BGC0001697|EAA36364.1|pks-6|neurosporin sordariomycetes|Pyrory|BGC0001749|EHA52508.1||epipyriculol sordariomycetes|Trivir|FNP119|EHK18438.1||trichoxide dothideomycetes|Altsol|BGC0000146|BAJ09789.1|sol1|solanapyrone leotiomycetes|Botcin|BGC0001892|ATZ45185.1|Bcboa9|botcinic dothideomycetes|Altbra|BGC0000046|ACZ57548.1|DEP5|depudecin CBS 412.89.contig 2.region001~L0+CDS6 |Mac|FNP109|mpmlA|ProteinId eurotiomycetes|Penbre|BGC0001141|AIA58899.1||4-epi-15-epi-brefeldin sordariomycetes|Artpha|FNP026|BBJ21455.1|apmlA|phaeospelide eurotiomycetes|Aspfum|BGC0001067|EAL85129.1||fumagillin dothideomycetes|Parspo|FNP091|OAG05545.1|| CBS 412.89.contig 7.region002~L0+CDS8 eurotiomycetes|Aspluc|FNP107|GAA85575.1||AKML sordariomycetes|Colinc|FNP108|KZL86691.1||CIML sordariomycetes|Fusfuj|BGC0001606|CCT75967.1||gibepyrone-A sordariomycetes|Calarb|FNP074|ALD83627.1|polyketide|Aurovertin eurotiomycetes|Aspter|BGC0001400|EAU29808.1||citreoviridin eurotiomycetes|Aspnid|BGC0001998|XP 659388.1||aspernidgulene eurotiomycetes|Aspory|FNP125|XP 023091701.1||2 eurotiomycetes|Aspfum|BGC0000129|EAL89230.2||pyripyropene |Phobet|BGC0001264|BAQ25466.1|bet1|betaenone |Phobet|BGC0001280|BAQ25466.1|bet1|betaenone dothideomycetes|Parnod|FNP015|XP 001798193.1||stemphyloxin sordariomycetes|Fusgra|BGC0001600|CEF75886.1||fusarielin leotiomycetes|Botcin|BGC0001892|ATZ45182.1|Bcboa6|botcinic sordariomycetes|Fusfuj|BGC0001305|CCT72377.1||fujikurin eurotiomycetes|Aspacu|FNP019|XP 020054826.1||acurin dothideomycetes|Altalt|BGC0001254|BAJ14522.1|ACTTS3|ACT-Toxin agaricomycetes|Agasp.|FNP078|ART89046.1|pps1|18methyl19oxoicosaoctaenoic |Leosul|FNP079|QNJ99675.1|lpaA|laetiporic agaricomycetes|Hydpin|FNP094|KIJ60886.1|PKS2| agaricomycetes|Punstr|FNP093|EIN09536.1|pks1| agaricomycetes|Strten|BGC0001909|ATV82110.1|stpks1|strobilurin eurotiomycetes|Aspter|BGC0000160|EAU32819.1||terreic eurotiomycetes|Aspter|BGC0001276|BAA20102.2|atX|6-methylsalicyclic dothideomycetes|Parnod|BGC0001244|AIW00670.1|mlnS| eurotiomycetes|Aspacu|FNP110|XP 020053918.1||6-MSA eurotiomycetes|Penexp|BGC0000120|AIG62146.1|patK|patulin eurotiomycetes|Aspoch|BGC0001273|AAS98200.1||asperlactone leotiomycetes|Glaloz|BGC0001275|AAX35547.1|pks2|6-methylsalicyclic eurotiomycetes|Aspnig|BGC0000170|EHA22196.1||yanuthone eurotiomycetes|Aspust|FNP088|KIA75596.1||Ustethylin eurotiomycetes|Penbra|BGC0001729|BAV69313.1|prhL|paraherquonin ota s B C D E Chapter 3 100 Figure S3. Gene expression of transformants AoADEARG::VDTA::SDH and AoADEARGBAR::VDTA::SDH::VDTB as determined by RT-PCR. The H2B housekeeping gene was used as an expression control. Numbers on the gel represent three different clones. Dimerisation of xylindein 101 1 2 Chapter 3 102 Figure S4. HRMS data of products 1 to 4 from transformant AoADEARG::vdtA::XLNsdh. Figure S5. HPLC profile of co-expression transformants. Additional peaks produced by AoADEARGBAR::VDTA::SDH::VDTB transformant are highlighted in the red square. 3 4 Dimerisation of xylindein 103 Figure S6. Possible dimers potentially synthesised by AoADEARGBAR::VDTA::SDH::VDTB transformant. Atropisomers are not considered. Chapter 3 110 24. Strycker BD, Han Z, Bahari A, et al. Raman characterization of fungal dhn and dopa melanin biosynthesis pathways. Journal of Fungi. 2021;7(10). doi:10.3390/jof7100841 25. Tamano K, Takayama H, Yasokawa S, Sano M, Baker SE. Major involvement of two laccase genes in conidial pigment biosynthesis in Aspergillus oryzae. Appl Microbiol Biotechnol. 2022;106(1):287-300. doi:10.1007/s00253-02111669-1 26. Perez-Cuesta U, Aparicio-Fernandez L, Guruceaga X, et al. Melanin and pyomelanin in Aspergillus fumigatus: from its genetics to host interaction. International Microbiology. 2020;23(1):55-63. doi:10.1007/s10123-019-00078-0 27. Tamano K, Takayama H, Yasokawa S, Sano M, Baker SE. Major involvement of two laccase genes in conidial pigment biosynthesis in Aspergillus oryzae. Appl Microbiol Biotechnol. 2022;106(1):287-300. doi:10.1007/s00253-02111669-1 28. Yang Y, Zhang S, Ma K, et al. Discovery and Characterization of a New Family of Diterpene Cyclases in Bacteria and Fungi. Angewandte Chemie. 2017;129(17):4827-4830. doi:10.1002/ange.201700565 29. Liu C, Minami A, Ozaki T, et al. Efficient Reconstitution of Basidiomycota Diterpene Erinacine Gene Cluster in Ascomycota Host Aspergillus oryzae Based on Genomic DNA Sequences. J Am Chem Soc. 2019;141(39):15519-15523. doi:10.1021/jacs.9b08935 30. Engels B, Heinig U, Grothe T, Stadler M, Jennewein S. Cloning and characterization of an Armillaria gallica cDNA encoding protoilludene synthase, which catalyzes the first committed step in the synthesis of antimicrobial melleolides. Journal of Biological Chemistry. 2011;286(9):6871-6878. doi:10.1074/jbc.M110.165845 31. Collemare J, Griffiths S, Iida Y, et al. Secondary metabolism and biotrophic lifestyle in the tomato pathogen Cladosporium fulvum. PLoS One. 2014;9(1). doi:10.1371/journal.pone.0085877 32. Greco C, De Mattos-Shipley K, Bailey AM, et al. Structure revision of cryptosporioptides and determination of the genetic basis for dimeric xanthone biosynthesis in fungi. Chem Sci. 2019;10(10):2930-2939. doi:10.1039/c8sc05126g 33. Wei X, Chen X, Chen L, Yan D, Wang WG, Matsuda Y. Heterologous Biosynthesis of Tetrahydroxanthone Dimers: Determination of Key Factors for Selective or Divergent Synthesis. J Nat Prod. 2021;84(5):1544-1549. doi:10.1021/acs.jnatprod.1c00022 34. Zou Y, Zhan Z, Li D, et al. Tandem prenyltransferases catalyze isoprenoid elongation and complexity generation in biosynthesis of quinolone alkaloids. J Am Chem Soc. 2015;137(15):4980-4983. doi:10.1021/jacs.5b03022 Dimerisation of xylindein 111 35. Wei X, Matsuda Y. Unraveling the Fungal Strategy for Tetrahydroxanthone Biosynthesis and Diversification. Org Lett. 2020;22(5):1919-1923. doi:10.1021/acs.orglett.0c00285 36. Schätzle MA, Husain SM, Ferlaino S, Müller M. Tautomers of anthrahydroquinones: Enzymatic reduction and implications for chrysophanol, monodictyphenone, and related xanthone biosyntheses. J Am Chem Soc. 2012;134(36):14742-14745. doi:10.1021/ja307151x 37. Temperini C, Innocenti A, Guerri A, Scozzafava A, Rusconi S, Supuran CT. Phosph(on)ate as a zinc-binding group in metalloenzyme inhibitors: X-ray crystal structure of the antiviral drug foscarnet complexed to human carbonic anhydrase I. Bioorg Med Chem Lett. 2007;17(8):2210-2215. doi:10.1016/j.bmcl.2007.01.113 38. Isik S, Kockar F, Arslan O, Guler OO, Innocenti A, Supuran CT. Carbonic anhydrase inhibitors. Inhibition of the β-class enzyme from the yeast Saccharomyces cerevisiae with anions. Bioorg Med Chem Lett. 2008;18(24):6327-6331. doi:10.1016/j.bmcl.2008.10.100 39. Aguilera J, Van Dijken JP, De Winde JH, Pronk JT. Carbonic anhydrase (Nce103p): An essential biosynthetic enzyme for growth of Saccharomyces cerevisiae at atmospheric carbon dioxide pressure. Biochemical Journal. 2005;391(2):311-316. doi:10.1042/BJ20050556 40. De Coster W, Rademakers R. NanoPack2: population-scale evaluation of long-read sequencing data. Bioinformatics. 2023;39(5). doi:10.1093/bioinformatics/btad311 41. Andrews, S. FastQC: A Quality Control Tool for High Throughput Sequence Data. 2010 [Online]. Available online at: http://www.bioinformatics.babraham.ac.uk/projects/fastqc/ 42. Kolmogorov M, Yuan J, Lin Y, Pevzner PA. Assembly of long, error-prone reads using repeat graphs. Nat Biotechnol. 2019;37(5):540-546. doi:10.1038/s41587-019-0072-8 43. Mikheenko A, Prjibelski A, Saveliev V, Antipov D, Gurevich A. Versatile genome assembly evaluation with QUAST-LG. In: Bioinformatics. Vol 34. Oxford University Press; 2018:i142-i150. doi:10.1093/bioinformatics/bty266 44. Nishimura O, Hara Y, Kuraku S. Evaluating Genome Assemblies and Gene Models Using gVolante. Methods Mol Biol. 2019;1962:247-256. doi:10.1007/9781-4939-9173-0_15 45. Augustus Web https://github.com/Gaius-Augustus/Augustus 46. Grigoriev I V, Nikitin R, Haridas S, et al. MycoCosm portal: gearing up for 1000 fungal genomes. Nucleic Acids Res. 2014;42(Database issue):D699-704. doi:10.1093/nar/gkt1183 Chapter 3 112 47. Gilchrist CLM, Chooi YH. Clinker & clustermap.js: Automatic generation of gene cluster comparison figures. Bioinformatics. 2021;37(16):2473-2475. doi:10.1093/bioinformatics/btab007 48. Coudert E, Gehant S, de Castro E, et al. Annotation of biologically relevant ligands in UniProtKB using ChEBI. Bioinformatics. 2023;39(1). doi:10.1093/bioinformatics/btac793 49. Katoh K, Misawa K, Kuma K ichi, Miyata T. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res. 2002;30(14):3059-3066. doi:10.1093/nar/gkf436 50. Capella-Gutiérrez S, Silla-Martínez JM, Gabaldón T. trimAl: a tool for automated alignment trimming in large-scale phylogenetic analyses. Bioinformatics. 2009;25(15):1972-1973. doi:10.1093/bioinformatics/btp348 51. Minh BQ, Schmidt HA, Chernomor O, et al. IQ-TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era. Mol Biol Evol. 2020;37(5):1530-1534. doi:10.1093/molbev/msaa015 52. Letunic I, Bork P. Interactive Tree Of Life (iTOL) v5: an online tool for phylogenetic tree display and annotation. Nucleic Acids Res. 2021;49(W1):W293W296. doi:10.1093/nar/gkab301 Dimerisation of xylindein 113 114 115 Chapter 4 Evolution of fungal fatty acid synthases from primary to secondary metabolism Yanfang Guo1,2, Jorge C. Navarro-Muñoz3, Arnold J.M. Driessen2, Jérôme Collemare*1 1 Fungal Natural Products Group, Westerdijk Fungal Biodiversity Institute, 3584 CT Utrecht, Netherlands. 2 Department of Molecular Microbiology, University of Groningen, 9747 AG Groningen, Netherlands. 3 Bioinformatics Group, Wageningen University and Research, 6708 PB Wageningen, Netherlands. Chapter 4 116 Abstract Fatty acid synthases (FASs) are essential multi-domain enzymes responsible for the synthesis of fatty acids in all living organisms. While type I FASs are well known for their role in producing long-chain fatty acids for membrane biosynthesis (primary metabolism, PM), distinct FASs in fungi are implicated in secondary metabolism (SM) biosynthetic pathways in which they provide specific acyl-CoA precursors to produce metabolites such as aflatoxins, HC-toxin and sporothriolide. In fungi, both PM and SM FASs comprise two complementary α and β subunits which are encoded by a pair of genes (FAS1 and FAS2) located at the same genomic locus. SM FAS genes are also co-localising with biosynthetic genes involved in SM biosynthetic pathways, being part of SM biosynthetic gene clusters (BGCs). How SM FAS genes have been recruited in SM BGCs in fungi and how they have evolved from PM FASs is not known. In this study, we investigated the evolutionary origin and divergence of fungal FASs, focusing on their functional specialisation in PM and SM. A comprehensive phylogenetic and comparative genomics analysis of FAS1 and FAS2 gene pairs from 1,324 Pezizomycotina genomes revealed the existence of two distinct evolutionary lineages: a conserved clade associated with PM, and a taxonomically widespread, functionally heterogeneous clade associated with SM. Our data support the notion that SM FASs originated from a single ancestral duplication event of PM FASs, followed by extensive diversification and recruitment into BGCs. Interestingly, we also identified recent PM FAS duplications, including one instance where the duplicated genes are embedded as self-resistance genes in the BGC for production of the FAS-inhibitor cerulenin. SM FASs were found to belong to BGCs that comprise diverse biosynthetic enzymes, including polyketide synthases, alkyl citrate synthases and non-ribosomal peptide synthetases. The genomic loci of SM FASs are consistent with the phylogenetic clades, showing evidence of coevolution between SM FASs and their associated biosynthetic partners. Altogether, our results point at gene duplication as the major evolutionary mechanism behind the recruitment and diversification of fungal FASs and their associated SM pathways. Such a mechanism is likely to explain the evolution of SM BGCs more generally, and to play a major role in expanding the chemical diversity of fungal natural products. Keywords: evolution history, phylogenetic analysis, genomic comparison, duplication events, fungal natural products Evolution of FAS in fungi 117 Introduction Fatty acids are essential across all domains of life, contributing fundamentally to both primary metabolism (PM) and secondary metabolism (SM). In PM, fatty acids serve structural, energetic, and signalling roles, synthesising key components of cellular membranes and storage lipids.1 In fungi, as in other organisms, the precise regulation and diversity of fatty acids allow them to adapt to changing environments. This adaptive capacity is achieved through modulation of fatty acid chain length, degree of unsaturation, and incorporation into specialised lipid classes, enabling fungi to maintain membrane fluidity, energy homeostasis, and stress resistance under fluctuating environmental conditions.2,3 Beyond their roles in PM, fatty acids are also involved in secondary metabolic pathways, acting as building blocks for complex and diverse bioactive compounds such as pigments, toxins, and antimicrobial agents. For example, aflatoxins produced by the genus Aspergillus are one of the major mycotoxins in food and feed4, while xylindein is a complex cyan pigment produced by Cholociboria species (Chapter 2).5 The diversity of secondary metabolites that contain a fatty acid component encompasses three different major chemical families: polyketides5–7, non-ribosomal peptides8–11 and alkyl citrates12–14 (Figure 1). Figure 1. Examples of secondary metabolites synthesised by fungal fatty acid synthases (FASs). Chemical backbones synthesised by fungal FASs are highlighted in bold. The biosynthesis of fatty acids is catalysed by fatty acid synthases (FASs), which are multifunctional enzymatic complexes. The conserved domains which compose Chapter 4 118 FASs include acetyl transferase (AT), enoylreductase (ER), dehydratase (DH), malonyl-palmitoyl transferase (MPT), acyl-carrier protein (ACP), ketoreductase (KR), ketosynthase (KS), thioesterase (TE) and phosphopantetheinyltransferase (PPT) domains. The PPT domain does not strictly play a role in fatty acid biosynthesis, but it catalyses the post-translational modification of ACP domains that is essential for the ACP to be activated and to function as a “swinging arm” that carries intermediates between enzymatic domains.15 Fungal FASs, as well as FASs from animals and CMN-group bacteria (Corynebacterium, Mycobacterium, and Nocardia)16, are type I FASs, which are large and highly conserved multi-enzyme complexes that orchestrate the stepwise assembly of fatty acids from simple acetyland malonyl-CoA building blocks.16 In contrast, Archaea, most bacteria and plants use type II FASs, characterised by discrete, monofunctional enzymes for fatty acid synthesis.17,18 The type I systems can be further subdivided based on the domain organisation of the multifunctional proteins and their quaternary structure. Microbial type I FASs are hexamers with a domain sequence of AT-ER-DH-MPT/ACP-KRKS (Figure 2), forming either α6β6 (fungi) or α6 (bacteria) oligomers (type Ia), with Saccharomyces cerevisiae and Mycobacterium tuberculosis as examples, respectively.16 In contrast, animal FASs, such as human FASs, are α2 dimers with the domain sequence KS-AT-DH-ER-KR-ACP-TE (type Ib; Figure 2).19 Figure 2. Domain organisation and gene splitting of type I fatty acid synthases (FASs). This schematic illustrates the domains and gene splitting patterns in type I FASs. The fungal β-chain (Fas1) typically harbours domains such as acetyl transferase (AT), enoyl reductase (ER), dehydratase (DH), and the majority of the split malonyl-palmitoyl transferase (MPT) domain. The α-chain (Fas2) contains a smaller portion of the MPT domain along with the acyl carrier protein (ACP), ketoreductase (KR), ketosynthase (KS), phosphopantetheinyl transferase (PPT), and thioesterase (TE) domains. Methyltransferase (Me) domains are found in humans. Evolution of FAS in fungi 119 In contrast to CMN-bacteria FASs that are encoded by a single gene, the catalytic domains of type I FASs in most fungal species are distributed over two distinct protein chains, α-chain and β-chain, which are encoded by separate genes FAS2 and FAS1, respectively (Figure 2).20 However, the FAS in the Basidiomycota fungus Coprinopsis cinerea is encoded by a single gene and resembles the bacterial ones, suggesting this configuration is the ancestral state.3 The PPT domain is encoded as a separate protein in bacteria and animals, but it is integrated into the FAS genes in fungi (Figure 2), suggesting a fusion event in fungi.20 Consistent with this single gene ancestor, the domain organisation and the splitting position of FAS1 and FAS2 vary among different fungal species, indicative of independent splitting events during fungal evolution.20 In S. cerevisiae, the β-chain Fas1 consists of the AT, ER, DH and the major part of the split MPT domains, while the α-chain Fas2 encompasses a smaller portion of the MPT domain as well as the ACP, KR, KS, and PPT domains.20 In the oleaginous yeast Rhodosporidium toruloides, the splitting site is within a FAS-meander domain that connects the ER and DH domains. ACP duplication also occurs in R. toruloides FAS, which likely increases substrate shuttling capacity and is beneficial for lipid overproduction.21 Finally, a third configuration is observed in the Basidomycota Cryptococcus neoformans in which the splitting site is between the ACP and KR domains.20 The SM FAS in Aspergillus parasiticus, which produces hexanoyl-CoA delivered to the non-reducing polyketide synthase (nrPKS) for aflatoxin production, exhibits a similar splitting pattern to S. cerevisiae, though it shows a sequence loss between the ACP and KR domains, and was suggested to form a distinct α₂β₂γ₂ complex together with the nrPKS, contrasting with the hexameric α₆β₆ structure of yeast PM FAS (Figure 2).22 In fungi, PM FASs typically generate long-chain fatty acids like palmitate and stearate.16 In contrast, SM FASs produce specific CoA-activated intermediates that are often short fatty acyl-CoAs, which are further processed by other enzymes in the biosynthesis of complex secondary metabolites.16,23 For example, the FAS-produced hexanoyl-CoA in A. parasiticus and suggested FAS-produced butanoyl-CoA from Chlorociboria aeruginascens act as precursors for the nrPKS in the biosynthesis of aflatoxin and xylindein, respectively (Figure 1) (Chapter 2).5,7 Similarly, in other pathways, FASs provide precursors such as octanoyl-CoA in the biosynthesis of hexylitaconic acid by alkyl citrate synthase (ACS) in Aspergillus niger24. They also support non-ribosomal peptide synthetase (NRPS)-mediated synthesis of HC toxin and apicidin F in Cochliobolus carbonum and Fusarium fujikuroi, respectively, by synthesising decanoic acid (Figure 1).8,11 Chapter 4 126 the branch support, the topology of the Fas2 tree is consistent with the species tree, indicating ancestral origin and differential retention in these few species (Figure S4). Comparison of the corresponding genomic loci revealed that the Fas-containing BGC is well conserved across A. laciniosus, A. ambiguus, A. appendiculatus, A. umbrosus, A. cumulatus, and A. cibarius, though the BGC in the last two species is found on a different contig, likely because of fragmented assemblies. This BGC comprises genes encoding a thioesterase, an oxidoreductase, an amidotransferase, two cytochrome P450 monooxygenases and one cytochrome b5 (Figure 5C). Genes downstream of FAS2 are also conserved in these species and encode an FAD-linked oxidase, a transporter, and a transcription factor, which could also belong to the same BGC (Figure 5C). This BGC shows striking synteny with the recently characterised cerulenin BGC from Sarocladium oryzae and Sarocladium attenuatum (Figure S6).42 Such conservation suggests these Aspergillus species may produce cerulenin, which remains to be determined. The phylogeny of the Sarocladium Fas2 protein from the cerulenin BGC is consistent with the species tree (Figure S7), suggesting an old origin of the cerulenin BGC and retention in a few species only. In A. cumulatus and A. cibarius, the conserved BGC is located on a separate contig in both genomes (Figure S5). Notably, A. laciniosus and A. ambiguous lack the FAS1 gene (Figure 5C). Combined with synteny patterns and the ancestral origin of this BGC (Figure 5C), we hypothesise that the duplication of FAS1 and FAS2 first occurred with both paralogs being recruited into the cerulenin BGC. During species divergence, the FAS1 paralog was lost in several species and only the FAS2 paralog has remained in the BGC of A. laciniosus and A. ambiguous. In contrast, in A. alliaceus and A. albertensis, which no longer harbour this BGC, the FAS2 gene is located at a distant genomic locus lacking any recognisable core biosynthetic genes or typical tailoring enzymes, suggesting that FAS2 was translocated after the BGC was lost in these species. Finally, the potential more recent recruitment of functional FAS paralogs from PM to SM pathways may be difficult. Indeed, SM Fas2 proteins lack a conserved sequence located between the FAS_I_H and KR domains (Figure 3 and Figure S3). Although the functional importance of this sequence remains unknown, its absence in SM FAS may be required for compatibility with noncanonical enzymatic partners in SM pathways. Consequently, only after sequence divergence, such as deletion of the insertion or accumulation of adaptive mutations, could these paralogs be effectively integrated into functional SM BGCs. Evolution of FAS in fungi 127 Overall, we detected three ancestral duplications of FAS1 and FAS2 genes in Dothideomycetes, four in Sordariomycetes, and one in Eurotiomycetes, and we linked one of these duplication events to the origin of the FAS inhibitor cerulenin BGC in fungi. Multiple lineage-specific duplication and recruitment events of fatty acid synthases in secondary metabolism To investigate the evolutionary history and functional diversification of FAS genes in fungal SM, we examined the phylogeny of the assigned SM Fas1 and Fas2 homologs, which presumably originated from ancestral duplications of PM FAS genes. Both trees show a similar topology and are divided into several wellsupported phylogenetic clades (Figure 6). In contrast to their PM counterparts, these SM FAS genes are consistently co-localised with core biosynthetic proteins (CBPs), such as nrPKSs, NRPSs, or ACSs, indicating a non-random association and probable co-evolution between FAS paralogs and their biosynthetic partners. By combining the phylogeny of SM Fas and CBP types, seven clades were defined and dubbed according to the main CBP type (Figure 6). Clades P1 and P2 are associated with nrPKS genes. P1 includes enzymes involved in the biosynthesis of compounds such as 2,4-dihydroxy-3-methyl-6-(2-oxoundecyl)benzaldehyde, and is largely restricted to Eurotiomycetes; P2 includes Fas proteins from aflatoxin, sterigmatocystin, xylindein, and dothistromin pathways, and is more taxonomically widespread (Figures 6 and 7; Figure S8 ).5–7 Clades C1 and C2 are associated with ACSs, involved in alkyl citrate pathways. C1 includes functionally validated BGCs such as sporothriolide and hexylitaconic acid; C2 lacks characterised examples but shows consistent co-localisation with ACSs (Figure 6; Figures S9 and S10).13,14 Loci in the P1 and a C1 subclade appear to comprise a conserved triterpene synthase and sesquiterpene synthase, respectively, though the corresponding characterised molecules, 2,4-dihydroxy-3-methyl-6-(2-oxoundecyl)benzaldehyde and oryzine B, do not harbour any terpene moiety (Figures 6 and 7; Figure S9).12 Clades N1, N2, and N3 contain FAS genes associated with NRPSs. These include HC toxin, apicidin F (N1), aspercryptin (N2), and gramillin A (N3).8–10 Notably, N1 does not form a strongly supported monophyletic group, possibly reflecting under-sampling or functional diversity (Figure 6; Figures S11, S12 and S13). Importantly, the phylogenetic placement of SM FAS genes closely correlates with their associated CBP type, suggesting a history of co-evolution and functional recruitment, rather than random clustering. Chapter 4 128 The presence of SM Fas homologs from different clades within a single fungal genome (e.g., Aspergillus flavus, A. pseudotamarii, A. mottae, A. ibericus; Figure S14) supports the idea that these clades represent paralogous lineages, each arising from distinct ancestral gene duplication events. This diversification likely enabled the evolution of specialised roles in different secondary metabolic pathways. Phylogenetic analysis suggests that clades P1, P2, and C1 form a larger monophyletic group, with P1 and C1 sharing a more recent common ancestor that diverged from the P2 lineage (Figure 6). The P1 clade is restricted to Eurotiomycetes and a few Leotiomycetes species and shows strong bootstrap support, while C1 and P2 include homologs from multiple fungal classes, including Eurotiomycetes, Sordariomycetes, Dothideomycetes, Leotiomycetes, and Lecanoromycetes (Figure 6). The ancestral FAS gene of these clades was likely positioned adjacent to an nrPKS, consistent with the widespread association of P1 and P2 with nrPKS genes. After duplication, the C1 paralog appears to have been recruited into alkyl citrate biosynthesis by relocating next to an ACS gene. This recruitment likely occurred before the divergence of Sordariomycetes and Eurotiomycetes (Figure S1), after which the P1 clade was lost in most lineages except Eurotiomycetes, suggesting lineage-specific retention and loss. Further duplications shaped the evolution of other clades. N2 and N3, both associated with NRPSs, appear to have arisen from a recent duplication event (Figure 6). Moreover, some species (e.g., Aspergillus carbonarius, Xylaria curta) harbour two C1 paralogs, indicating another duplication after recruitment into alkyl citrate pathways (Figure 6 and Figure S14). The C2 clade is distant from the C1 clade and groups with N1, N2 and N3 clades. The most plausible scenario is that the common ancestor of the Nx clades and C2 was recruited next to an NRPS gene, and later the C2 paralog was independently recruited to an alkyl citrate pathway, implying convergent recruitment of FAS genes into similar biosynthetic contexts. In contrast, the evolutionary history of the N1 clade remains unclear, likely due to under-sampling or rapid sequence divergence. Nevertheless, its consistent association with NRPSs and its placement apart from other NRPS-associated clades support a distinct functional specialisation. Especially, this clade comprises enzymes encoded in the HC-toxin and apicidin BGCs. While two FAS genes are found in the case of HC-toxin, the conserved apicidin BGC in Fusarium semitectum and Fusarium fujikuroi contains the Fas2 gene only8, similar to the cerulenin BGC. The deletion of APS5 in F. semitectum was shown to abolish apicidin production45, however, the orthologue APF5 in F. fujikuroi is not co-regulated with other genes involved in the pathway8, and the role of this unique FAS gene in the production of Evolution of FAS in fungi 129 apicidin is questionable because both αand β-subunits are required for FAS activity. Alternatively, the production of apicidins could involve a reducing PKS located at another locus following the biosynthetic mechanism of other lipopeptides like emericellamide.46 The singleton position of this FAS protein in the phylogeny suggests that potentially non-functional FAS genes may remain in fungal genomes and BGCs. Collectively, these observations indicate that fungal SM FAS genes have undergone multiple rounds of duplication, divergent recruitment, and clade-specific adaptation, resulting in the diverse biosynthetic roles observed across fungal lineages today. Figure 5. Fatty acid synthase genes experience duplication events with localisation of the duplicated copies at different genomic loci. A, Inverted maximum-likelihood phylogenetic tree of Fas1 and Fas2 proteins involved in primary metabolism. Branch colours indicate fungal taxonomic classes. Internal lines connect intraspecies paralogues that resulted from duplication events at the species level. Black lines indicate only FAS1 or FAS2 duplication, while pink lines indicate duplication events from both genes were detected. The middle dots indicate the presence of core biosynthetic proteins (CBPs) in the corresponding FAS loci. Black dots indicate significant branch support (likelihood ratio test >80 and ultrafast bootstrap >95). Grey arrows indicate the whole genome duplication (WGD). Pink arrows indicate the duplicated homologues next to a polyketide synthase (PKS) within the loci. Red boxes on the trees indicate the characterised sequences. B, Maximum-likelihood phylogenetic tree of duplicated Fas2 homologues next to a PKS. Black dots indicate significant branch support (likelihood ratio test >80 and ultrafast bootstrap >95). C, Conserved genomic locus of the duplicated FAS1 and FAS2 paralogues located next to a PKS gene. Figure 6. Divergent evolution of fatty acid synthase genes in secondary metabolism. Maximum-likelihood phylogenetic tree of Fas1 and Fas2 proteins involved in secondary metabolism (SM). Branch colours indicate fungal taxonomic classes. Black dots indicate significant branch support (likelihood ratio test >80 and ultrafast bootstrap >95). The trees are mid-point rooted. The outside dots indicate the presence of core biosynthetic proteins (CBPs) in the corresponding FAS loci. Trees were grouped into several SM clades with corresponding secondary metabolites (P, polyketides; C, citrates; N, non-ribosomal peptides). Chemical structures out of the trees present the examples synthesised by SM Fas within the corresponding clades. Backbones synthesised by Fas are highlighted in bold. Green stars indicate the characterised sequences, while blue triangles represent the Chlorociboria Fas sequences. Chapter 4 130 Figure 5. Evolution of FAS in fungi 131 Figure 6. Chapter 4 132 Clade-specific conservation and diversification of fatty acid synthase genomic loci in secondary metabolism To further understand the FAS evolution within each clade, we analysed the genomic loci in detail. Overall, synteny patterns in the P1 clade largely mirrored the FAS phylogeny. Each subclade shows strong conservation, whereas extensive rearrangements are observed between subclades (Figure 7). The characterised BGC responsible for 2,4-dihydroxy-3-methyl-6-(2-oxoundecyl)benzaldehyde production in Aspergillus shows strong synteny with several uncharacterised loci in the same subclade (Figure 7), consistently retaining the FAS gene pair, a PKS, and tailoring enzymes such as oxidoreductases and P450s. This suggests these loci may produce related alkylresorcinol-type compounds with structural variation arising from modifications to side chains or redox state. In the two basal subclades of P1, the genomic regions downstream of the FAS genes are highly conserved. In contrast, their upstream regions exhibit considerable variability, reflecting either lineagespecific recruitment of auxiliary genes into the BGCs or rearrangements in the genomic localisation. In the upper P1 subclades, FAS genes appear to have been independently recruited into distinct BGCs. Although downstream regions still show moderate conservation, the upstream loci display at least three distinct rearrangement patterns, indicative of flexible modular integration of tailoring genes (Figure 7). Additional loci in the upper subclades further illustrate that these P1 FAS genes, which originated from a common duplication event, have undergone independent recruitment into different biosynthetic contexts across lineages, leading to loci with minimal shared synteny (Figure 7). These observations support the hypothesis that after duplication, FAS genes in the P1 clade were co-opted into divergent secondary metabolic pathways. Figure 7. Comparative genomic analysis of P1 clade FAS loci across fungal lineages. Phylogenetic analysis and comparative genomics of the P1 clade reveal conserved synteny among FAS loci across species within Eurotiomycetes. The left panel shows a phylogenetic tree of P1 clade FAS proteins with branch support and taxonomic annotations. The green star on the tree marks the characterised Fas1 that is involved in the biosynthesis of 2,4-dihydroxy3-methyl-6-(2-oxoundecyl)benzaldehyde. The right panel depicts the corresponding genomic loci, highlighting conserved gene arrangements. Red boxes indicate regions with the highest synteny conservation, particularly downstream of the FAS genes. While the downstream gene content is generally stable, upstream regions show greater variability, suggesting differential recruitment of auxiliary genes. Evolution of FAS in fungi 133 Figure 7. Chapter 4 134 The P2 clade loci also display conserved synteny, but with greater structural diversification. Sterigmatocystin-producing species such as Aspergillus nidulans maintain conserved gene order downstream of the FAS genes, including a cytochrome P450 and several biosynthetic genes (Figure S8). In A. ochraceoroseus, however, these genes are rearranged upstream of the FAS genes, although the overall BGC architecture remains largely preserved, suggesting localised inversions or gene transpositions. Further divergence is observed in aflatoxin-producing species, where the genomic region between the FAS and PKS genes has undergone significant rearrangement (Figure S8). Key biosynthetic elements, including P450s and shortchain dehydrogenases are relocated to the downstream region of the FAS gene, resulting in a different BGC organisation (Figure S8). Beyond Eurotiomycetes, P2 clade loci become increasingly diverse and often lack conserved BGC architecture (Figure S8). However, in the xylindein-producing fungus C. aeruginascens, XLNfas1 and XLNfas2 are embedded within a conserved BGC (Figure 8).5 These genes are co-localised with XLNpks, XLNlac, and XLNsdh, encoding a non-reducing polyketide synthase, a laccase, and a short-chain dehydrogenase, respectively (Figures S8 and Figure 8) (Chapter 2). The consistent presence of these genes across multiple distinct species indicates a co-evolved biosynthetic module. Comparative genomic analyses of the C1, C2, N1, N2, and N3 clades (Figures S9– S13) further support the role of FAS gene duplication and subsequent diversification in shaping fungal SM. In the C1 clade, FAS1 and FAS2 do not share a bidirectional promoter, but FAS2 is consistently located next to an ACS gene, and the two genes share a common promoter (Figure S9). The position of FAS2 relative to FAS1 varies across different pathways, being located either upstream or downstream depending on the species (Figure S9). Characterised pathways like sporothriolide and hexylitaconic acid share a conserved FAS–ACS core, and many uncharacterised loci exhibit similar architectures (Figure S9). While this implies a common biosynthetic framework for alkyl citrate-like molecules, variations in associated tailoring genes point to potential structural diversity within the clade. The C2 clade, largely restricted to Sordariomycetes, displays a highly conserved genomic locus. In nearly all examined loci, the ACS gene is positioned downstream of the FAS1 and FAS2 gene pair, forming a stable functional unit (Figure S10). However, upstream of FAS2, lineage-specific rearrangements are observed, suggesting partial modular reshuffling while maintaining the FAS and ACS cores (Figure S10). In contrast, the N1 clade presents a more heterogeneous genomic landscape, consistent with the lower phylogenetic support (Figure S11). Nonetheless, a subset of N1 FAS genes shares a conserved locus, which includes cytochrome P450 and transporter genes, suggesting Evolution of FAS in fungi 135 a conserved pathway which remains to be characterised. The N2 and N3 clades show a higher degree of synteny conservation, consistent with the stronger phylogenetic signal. In the N2 clade (e.g., aspercryptin), a consistent set of biosynthetic genes, including P450, short-chain dehydrogenase, and transporter, is present across multiple species (Figure S12). Similarly, in the N3 clade (e.g., gramillin A), the FAS loci are conserved with genes encoding a P450 monooxygenase, dehydratase, hydrolase, and transporter (Figure S13). While minor gene rearrangements are observed within subclades, the overall syntenic architecture remains well conserved. Together, these findings suggest that the FAS duplication event that gave rise to the seven clades was followed by independent recruitment into functionally diverse biosynthetic pathways. The co-localisation of FAS genes with lineage-specific tailoring enzymes across species reflects a co-evolutionary process that underlies the emergence of complex fungal metabolites. Figure 8. Xylindein fatty acid synthases (FASs) and the orthologs. (A) XLNfas1 (g6389) and its orthologs clade extracted from the Fas1 tree. (B) XLNfas2 (g6390) and its orthologs clade extracted from the Fas2 tree. (C) Conserved genomic locus of XLNfas1 and its orthologs. Chapter 4 142 Figure S5. BLAST analysis reveals the presence of PKS and contiguous tailoring genes in fragmented genome assemblies of Aspergillus cumulans and Aspergillus cibarius. The BLAST results show high-confidence hits for PKS homologs located on contigs separate from the main FAS-containing clusters in A. cumulates and A. cibarius. In both species, the top hits show >85% identity and 100% coverage, supporting the presence of the PKS gene despite its absence in the syntenic cluster visualised in Figure 5C. AntiSMASH analysis further revealed that these PKS loci are positioned adjacent to conserved tailoring genes, consistent with the expected cluster structure. These findings indicate that the apparent absence of the PKS gene in the cluster is likely due to incomplete or fragmented genome assemblies rather than true gene loss. Evolution of FAS in fungi 143 Figure S6. Comparative analysis of fatty acid synthase (FAS)-containing biosynthetic gene clusters (BGCs) across Aspergillus species and Sarocladium oryzae. Synteny and homology relationships of BGCs are shown, highlighting the conservation of core biosynthetic genes including FAS1, FAS2, reducing polyketide synthase (rPKS), and tailoring enzymes such as cytochrome P450, oxidoreductase, and FAD-linked oxidase. Homologous genes are connected by shaded ribbons, with percent protein identity indicated. The BGCs from Aspergillus umbrosus, A. appendiculatus, and related Eurotiomycetes display strong synteny and sequence similarity with the cerulenin BGC from the Sordariomycete S. oryzae, suggesting a potential shared evolutionary origin through vertical descent or horizontal gene transfer (HGT). Colour-coded arrows represent predicted gene functions as indicated in the legend. Chapter 4 144 Figure S7. The phylogeny of the Sarocladium oryzae CerI protein from the cerulenin BGC is consistent with the species tree. Branch colours indicate fungal taxonomic classes. Black dots indicate significant branch support (likelihood ratio test >80 and ultrafast bootstrap >95). The tree is mid-point rooted. Evolution of FAS in fungi 145 Figure S8. Comparative genomic analysis of P2 clade FAS loci across fungal lineages. The figure illustrates the genomic organisation of loci containing P2 clade Fas1 genes, highlighting both conserved and divergent features across species. The left panel shows a phylogenetic tree of P1 clade Fas1 proteins with branch support and taxonomic annotations. The right panel depicts the corresponding genomic loci, highlighting conserved gene arrangements. Red boxes indicate regions with the highest synteny conservation. The preservation of this module across diverse taxa suggests a co-evolved, functionally integrated BGC characteristic of the P2 clade. Chapter 4 146 Figure S9. Comparative genomic analysis of C1 clade FAS loci across fungal lineages. Maximum likelihood tree showing the evolutionary relationships among FAS1 in the C1 clade. The left panel shows a phylogenetic tree of C1 clade Fas1 proteins with branch support and taxonomic annotations. The right panel depicts the corresponding genomic loci, highlighting conserved gene arrangements. Red boxes indicate regions with the highest synteny conservation, and blue boxes show the conjunction of FAS2 genes and alkyl citrate synthase (ACS) genes. Gene cluster synteny indicates shared ancestry and limited rearrangement across species. Evolution of FAS in fungi 147 Figure S10. Comparative genomic analysis of C2 clade FAS loci across fungal lineages. This figure shows the phylogeny and associated gene cluster structures of FAS1-containing loci within the C2 clade. The left panel shows a phylogenetic tree of C2 clade Fas1 proteins with branch support and taxonomic annotations. The right panel depicts the corresponding genomic loci, highlighting conserved gene arrangements. Red boxes indicate regions with the highest synteny conservation. Chapter 4 148 Figure S11. Comparative genomic analysis of N1 clade FAS loci across fungal lineages. The phylogenetic tree and synteny maps illustrate the structure of the N1 clade loci. The left panel shows a phylogenetic tree of N1 clade Fas1 proteins with branch support and taxonomic annotations. The right panel depicts the corresponding genomic loci, highlighting conserved gene arrangements. Red boxes indicate regions with the highest synteny conservation. Evolution of FAS in fungi 149 Figure S12. Comparative genomic analysis of N2 clade FAS loci across fungal lineages. The left panel shows a phylogenetic tree of N2 clade Fas1 proteins with branch support and taxonomic annotations. The right panel depicts the corresponding genomic loci, highlighting conserved gene arrangements. Red boxes indicate regions with the highest synteny conservation. Chapter 4 150 Figure S13. Comparative genomic analysis of N3 clade FAS loci across fungal lineages. The left panel shows a phylogenetic tree of N3 clade Fas1 proteins with branch support and taxonomic annotations. The right panel depicts the corresponding genomic loci, highlighting conserved gene arrangements. Red boxes indicate regions with the highest synteny conservation. Figure S14. Frequent duplication events of fatty acid synthase genes happened in secondary metabolism. Inverted maximum-likelihood phylogenetic tree of Fas1 and Fas2 proteins involved in secondary metabolism. Internal lines connect intraspecies paralogues that resulted from duplication events at the species level. Green stars indicate the characterised sequences, while blue triangles represent the Chlorociboria FAS sequences. Evolution of FAS in fungi 151 Figure S14.