ABSTRACT While transposable elements (TEs) are recognized as major drivers of fungal genome structure, evidence of their direct involvement in the interaction with host plants and the environment is only beginning to emerge. Retrotransposons can generate small RNA (sRNA) that act through cross-kingdom RNA interference, while giant DNA TEs called Starships carry dozens of cargo genes that enrich the accessory gene compartment of fungal genomes. In the polyphagous pathogen Botrytis cinerea, the Vv3 strain and other strains specialized on grapevine display a specific repertoire of TEs, including the retrotransposons BcCopia4, BcGypsy6 , and BcGypsy7 . This study first explored the putative role of sRNA generated from these retrotransposons in the interaction between the Vv3 strain and its host of origin, grapevine. Putative targets were identified among the host mRNAs, but predicted cleavage sites could not be experimentally validated. Moreover, Dicer mutants unable to produce retrotransposons-derived sRNA remained fully pathogenic on grapevine, indicating that these sRNAs do not act as virulence factors on this host. In parallel, this study provides an updated RNA-seq-based annotation of the accessory genes of the Vv3 strain, which revealed a new 93 kb- Starship harboring 43 cargo genes, some of which are related to arsenic resistance. A formal genetic approach confirmed that this locus confers resistance to this metalloid. This giant TE, named Ariane, was also detected in additional grapevine-specialized strains resistant to arsenic but not in strains isolated from other hosts such as tomato. In conclusion, this study highlights how a Starship giant transposon shaped the accessory genes compartment of the polyphagous fungus B. cinerea and may have contributed to its adaptation to vine cultivation by conferring resistance to arsenic, a compound widely used in vineyards during the last century. IMPACT STATEMENT Fungal genomes contain many families of transposons whose functional role in adaptation to the environment and in biotic interactions remained hidden for a long time. In the grey mold fungus Botrytis cinerea , strains specialized on grapevine, such as Vv3, carry a specific repertoire of transposons which provides a valuable opportunity to investigate their role in niche adaptation. In this study, we first investigated retrotransposon-derived small RNA, previously described as effectors capable of manipulating the immunity of the model plant Arabidopsis thaliana . Although in silico analysis of the specific repertoire of small RNAs of the Vv3 strain suggested that some could target the expression of grapevine genes, a genetic approach demonstrated that they do not play a significant role in virulence on this host. In contrast, this study identified a new transposon, named Ariane, that carries 43 cargo genes and confers a selective advantage to the Vv3 strain. Ariane belongs to a family of giant transposons called Starships , recently discovered in fungi and considered to be responsible for horizontal genes transfers between unrelated species. Ariane was detected only in some B. cinerea strains isolated from grapevine, and a genetic cross showed that it provides these strains with the ability to grow in presence of arsenic. Arsenic was used in vineyards until the beginning of the 21 st century to control fungal trunk diseases and insect pests. Therefore, Ariane appears to have played an important role in the adaptation of B. cinerea strains to cultivated grapevine. Overall, these results underline the importance of considering Starships when predicting emergence of resistance to antifungal compounds. DATA SUMMARY The novel data described in this study, i.e., RNA-Seq data and the Starship element are accessible under NCBI GEO accession GSE327899 and at https://doi.org/10.57745/HYWRNM , respectively. All information related to Botrytis cinerea genomes used in this study are centralized and kept up to date at the Bioinfo Bioger genomic web portal: https://bioinfo.bioger.inrae.fr/portal/genome-portal/ . Direct links to individual portals are respectively https://bioinfo.bioger.inrae.fr/portal/genome-portal/3/ for B. cinerea Vv3 genome, https://bioinfo.bioger.inrae.fr/portal/genome-portal/2/ for B. cinerea Sl3 genome, and https://bioinfo.bioger.inrae.fr/portal/genome-portal/4/ for B. cinerea populations isolated on tomato or grapevine. Each portal provides: (i) a centralized access to public genomic resources, including the genome, transposon, and RNA repositories; (ii) a data browser to download the genomic files; (iii) a genome browser that enables visualization of features within their genomic context, along with associated expression data. As a summary, the prior main public B. cinerea genomic accessions and resources used in this study are: GCA_039644125 for VV3 genome, GCA_022560135 for Sl3 genome, GCA_000143535 for B05.10 genome, PRJNA624742 for populations, https://doi.org/10.57745/HYWRNM for transposons, and GSE181592 for small RNAs. Furthermore, table S1 summarizes the list and characteristics of the 64 B. cinerea genomes publicly available to date. The genomic data for Vitis vinifera genome PN40024.v4 used in this study are available at: https://integrape.eu/resources/genes-genomes/genome-accessions/ .

Traditional cultivated grapevine (Vitis vinifera) is susceptible to many fungal and oomycete pathogens causing devastating diseases including powdery mildew, downy mildew, grey mould, black rot and trunk diseases. These infections trigger various defence mechanisms such as reinforcement of the cell wall structure, production of phytoalexins and pathogenesis-related proteins, and localized cell death. In V. vinifera susceptible varieties, these defences are not effective, while in resistant grapevine, recognition of the pathogen induces effective mechanisms that stop the infection. Breeding programmes are conducted to take advantage of this genetic resistance. Moreover, a range of exogenous defence stimulators can be used to obtain a so-called "induced resistance" in susceptible varieties. This chapter presents the recently acquired knowledge on the molecular mechanisms involved in genetic and induced resistances, and further consider other mechanisms such as ontogenic resistance. It also suggests how to exploit these resistances to durably protect vineyards against the different fungal diseases.

Transcriptomic studies have revealed that fungal pathogens of plants activate the expression of numerous biosynthetic gene clusters (BGC) exclusively when in presence of a living host plant. The identification and structural elucidation of the corresponding secondary metabolites remain challenging. The aim was to develop a polycistronic system for heterologous expression of fungal BGCs in Saccharomyces cerevisiae. Here we adapted a polycistronic vector for efficient, seamless and cost-effective cloning of biosynthetic genes using in vivo assembly (also called transformation-assisted recombination) directly in Escherichia coli followed by heterologous expression in S. cerevisiae. Two vectors were generated with different auto-inducible yeast promoters and selection markers. The effectiveness of these vectors was validated with fluorescent proteins. As a proof-of-principle, we applied our approach to the Colletochlorin family of molecules. These polyketide secondary metabolites were known from the phytopathogenic fungus Colletotrichum higginsianum but had never been linked to their biosynthetic genes. Considering the requirement for a halogenase, and by applying comparative genomics, we identified a BGC putatively involved in the biosynthesis of Colletochlorins in C. higginsianum. Following the expression of those genes in S. cerevisiae, we could identify the presence of the precursor Orsellinic acid, Colletochlorins and their non-chlorinated counterparts, the Colletorins. In conclusion, the polycistronic vectors described herein were adapted for the host S. cerevisiae and allowed to link the Colletochlorin compound family to their corresponding biosynthetic genes. This system will now enable the production and purification of infection-specific secondary metabolites of fungal phytopathogens. More widely, this system could be applied to any fungal BGC of interest.

Fungal pathogens represent a serious threat towards agriculture, health, and environment.Control of fungal diseases on crops necessitates a global understanding of fungal pathogenicity determinants and their expression during infection.Genomes of phytopathogenic fungi are often compartmentalized: the core genome contains housekeeping genes whereas the fast-evolving genome mainly contains transposable elements and species-specific genes.In this study, we analysed nucleosome landscapes of four phytopathogenic fungi with contrasted genome organizations to describe and compare nucleosome repartition patterns in relation with genome structure and gene expression level.We combined MNase-seq and RNA-seq analyses to concomitantly map nucleosome-rich and transcriptionally active regions during fungal growth in axenic culture; we developed the MNase-seq Tool Suite (MSTS) to analyse and visualise data obtained from MNase-seq experiments in combination with other genomic data and notably RNA-seq expression data.We observed different characteristics of nucleosome profiles between species, as well as between genomic regions within the same species.We further linked nucleosome repartition and gene expression.Our findings support that nucleosome positioning and occupancies are subjected to evolution, in relation with underlying genome sequence modifications.Understanding genomic organization and its role in expression regulation is the next gear to understand complex cellular mechanisms and their evolution.

We report here a new application, CustomProteinSearch (CusProSe), whose purpose is to help users to search for proteins of interest based on their domain composition. The application is customizable. It consists of two independent tools, IterHMMBuild and ProSeCDA. IterHMMBuild allows the iterative construction of Hidden Markov Model (HMM) profiles for conserved domains of selected protein sequences, while ProSeCDA scans a proteome of interest against an HMM profile database, and annotates identified proteins using user-defined rules. CusProSe was successfully used to identify, in fungal genomes, genes encoding key enzyme families involved in secondary metabolism, such as polyketide synthases (PKS), non-ribosomal peptide synthetases (NRPS), hybrid PKS-NRPS and dimethylallyl tryptophan synthases (DMATS), as well as to characterize distinct terpene synthases (TS) sub-families. The highly configurable characteristics of this application makes it a generic tool, which allows the user to refine the function of predicted proteins, to extend detection to new enzymes families, and may also be applied to biological systems other than fungi and to other proteins than those involved in secondary metabolism.

The fungus Botrytis cinerea is a polyphagous pathogen that encompasses multiple host-specialized lineages. While several secreted proteins, secondary metabolites and retrotransposons-derived small RNAs have been characterized as virulence factors, their roles in host specialization remain unknown. The aim of this study was to identify the genomic correlates of host-specialization in populations of B. cinerea associated with grapevine and tomato. Using PacBio sequencing, we produced complete assemblies of the genomes of strains Sl3 and Vv3 that represent the French populations T and G1 of B. cinerea, specialized on tomato and grapevine, respectively. Both assemblies revealed 16 core chromosomes that were highly syntenic with chromosomes of the reference strain B05.10. The main sources of variation in gene content were the subtelomeric regions and the accessory chromosomes, especially the chromosome BCIN19 of Vv3 that was absent in Sl3 and B05.10. The repertoires and density of transposable elements were clearly different between the genomes of Sl3 and Vv3 with a larger number of subfamilies (26) and a greater genome coverage in Vv3 (7.7%) than in Sl3 (14 subfamilies, 4.5% coverage). An Helitron-like element was found in almost all subtelomeric regions of the Vv3 genome, in particular in the flanking regions of a highly duplicated gene encoding a Telomere-Linked Helicase, while both features were absent from the Sl3 and B05.10 genomes. Different retrotransposons in the Sl3 and the Vv3 strains resulted in the synthesis of distinct sets of small RNAs. Finally, extending the study to additional strains indicated that the accessory chromosome BCIN19 and the small RNAs producing retrotransposons Copia_4 and Gypsy_7 are common features of the G1 population that are scarcely if ever found in strains isolated from other populations. This research reveals that accessory chromosomes, repertoires of transposons and their derived small RNAs differ between populations of B. cinerea specialized on different hosts. The genomic data characterized in our study pave the way for further studies aiming at investigating the molecular mechanisms underpinning host specialization in a polyphagous pathogen.

Botrytis cinerea is a necrotrophic fungal pathogen that affects a total of 586 genera representing approximately 1400 plant species. This pathogen produces two families of phytotoxins involved in its infection process i.e. botrydial and its relatives, and botcinic and botcineric acids and their relatives, botcinins. The botrydial biosynthetic cluster consists of seven genes, where the gene BcBOT4 encodes a cytochrome P450 monooxygenase that was shown to catalyse regio- and stereospecific hydroxylation at position C-4 of the presilphiperfolan-8-β-ol skeleton. The null mutant bcbot4Δ halted the production of botrydial and its derivatives, and instead accumulated tricyclic presilphiperfolane alcohol and overproduced a significant number of polyketides. A detailed study of the bcbot4Δ mutant led us to the isolation and characterization of five undescribed polyketides, three derived from botcinic and botcineric acids (botcinins H, I, J), one derived from the initial pentaketide (botcinin K), and one cinbotolide derivative (cinbotolide D). Botcinins are tetra-methylated tetraketides biosynthesized by the sequential assembly of a pentaketide (C10) based on an acetate primer unit which is lost through a retro-Claisen type C–C bond cleavage. The structural characterization of botcinin K showed a basic chemical structure corresponding to a botcinin (C14) derivative obtained directly from the original per-methylated pentaketide leading to the biosynthesis of botrylactone and other botcinins, confirming the previously proposed biosynthetic route.

Many fungal plant pathogens encompass multiple populations specialized on different plant species. Understanding the factors underlying pathogen adaptation to their hosts is a major challenge of evolutionary microbiology, and it should help to prevent the emergence of new specialized pathogens on novel hosts. Previous studies have shown that French populations of the gray mold pathogen Botrytis cinerea parasitizing tomato and grapevine are differentiated from each other, and have higher aggressiveness on their host of origin than on other hosts, indicating some degree of host specialization in this polyphagous pathogen. Here, we aimed at identifying the genomic features underlying the specialization of B. cinerea populations to tomato and grapevine. Based on whole genome sequences of 32 isolates, we confirmed the subdivision of B. cinerea pathogens into two genetic clusters on grapevine and another, single cluster on tomato. Levels of genetic variation in the different clusters were similar, suggesting that the tomato-specific cluster has not recently emerged following a bottleneck. Using genome scans for selective sweeps and divergent selection, tests of positive selection based on polymorphism and divergence at synonymous and nonsynonymous sites, and analyses of presence and absence variation, we identified several candidate genes that represent possible determinants of host specialization in the tomato-associated population. This work deepens our understanding of the genomic changes underlying the specialization of fungal pathogen populations.

Cultivation of the phytopathogenic fungus Botrytis cinerea using sublethal amounts of copper sulfate yielded a cryptic sesquiterpenoids family, which displayed the basic chemical structure of (+)-4-epi-eremophil-9-ene. The biosynthesis pathway was established, and the route involved the likely transformation of the diphosphate of farnesyl (FDP), to give a cis-fused eudesmane cation, through (S)-hedycaryol, finally yielding the (+)-4-epi-eremophil-9-enol derivatives. An expression study of genes that code for the sesquiterpene cydases (STC), including the recently reported gene Bcstc7 present in the B. cinerea genome, was performed in order to establish the STC involved in this biosynthesis. The results showed a higher expression level for the Bcstc7 gene with respect to the other stc1-5 genes in both wild-type strains, B05.10 and Botrytis cinerea UCA992. Deletion of the Bcstc7 gene eliminated (+)-4-epi-eremophilenol biosynthesis, which could be re-established by complementing the null mutant with the Bcstc7 gene. Chemical analysis suggested that STC7 is the principal enzyme responsible for the key step of cyclization of FDP to eremophil-9-en-11-ols. Furthermore, a thorough study of the two wild-types and the complemented mutant revealed four new eremophilenol derivatives whose structures are reported here.

The role of the sesquiterpene botrydial in the interaction of the phytopathogenic fungus Botrytis cinerea and plant-associated bacteria was analyzed. From a collection of soil and phyllospheric bacteria, nine strains sensitive to growth-inhibition by B. cinerea were identified. B. cinerea mutants unable to produce botrydial caused no bacterial inhibition, thus demonstrating the inhibitory role of botrydial. A taxonomic analysis showed that these bacteria corresponded to different Bacillus species (six strains), Pseudomonas yamanorum (two strains) and Erwinia aphidicola (one strain). Inoculation of WT and botrydial non-producing mutants of B. cinerea along with Bacillusamyloliquefaciens strain MEP218 in soil demonstrated that both microorganisms exert reciprocal inhibitory effects; the inhibition caused by B. cinerea being dependent on botrydial production. Moreover, botrydial production was modulated by the presence of B. amyloliquefaciens MEP218 in confrontation assays in vitro. Purified botrydial in turn, inhibited growth of Bacillus strains in vitro and cyclic lipopeptide (surfactin) production by B. amyloliquefaciens MEP218. As a whole, results demonstrate that botrydial confers B. cinerea the ability to inhibit potential biocontrol bacteria of the genus Bacillus. We propose that resistance to botrydial could be used as an additional criterion for the selection of biocontrol agents of plant diseases caused by B. cinerea.

Botcinic acid is a phytotoxic polyketide involved in the virulence of the gray mold fungus Botrytis cinerea. Here, we aimed to investigate the specific regulation of the cluster of Bcboa genes that is responsible for its biosynthesis. Our analysis showed that this cluster is located in a subtelomeric genomic region containing alternating G + C/A + T-balanced regions, and A + T-rich regions made from transposable elements that underwent RIP (Repeat-Induced Point mutation). Genetic analyses demonstrated that BcBoa13, a putative Zn2Cys6 transcription factor, is a nuclear protein with a major positive regulatory role on the expression of other Bcboa1-to-Bcboa12 genes, and botcinic acid production. In conclusion, the structure and the regulation of the botcinic acid gene cluster show similar features with the cluster responsible for the biosynthesis of the other known phytotoxin produced by B. cinerea, i.e., the sesquiterpene botrydial. Both clusters contain a gene encoding a pathway-specific Zn2Cys6 positive regulator, and both are surrounded by relics of transposons which raise some questions about the role of these repeated elements in the evolution and regulation of the secondary metabolism gene clusters in Botrytis.

The gray mold fungus Botrytis cinerea is a necrotrophic pathogen able to infect hundreds of host plants, including high-value crops such as grapevine, strawberry and tomato. In order to decipher its infectious strategy, a library of 2,144 mutants was generated by random insertional mutagenesis using Agrobacterium tumefaciens-mediated transformation (ATMT). Twelve mutants exhibiting total loss of virulence toward different host plants were chosen for detailed analyses. Their molecular characterization revealed a single T-DNA insertion in different loci. Using a proteomics approach, the secretome of four of these strains was compared to that of the parental strain and a common profile of reduced lytic enzymes was recorded. Significant variations in this profile, notably deficiencies in the secretion of proteases and hemicellulases, were observed and validated by biochemical tests. They were also a hallmark of the remaining eight non-pathogenic strains, suggesting the importance of these secreted proteins in the infection process. In the twelve non-pathogenic mutants, the differentiation of infection cushions was also impaired, suggesting a link between the penetration structures and the secretion of proteins involved in the virulence of the pathogen.

The host plant is often the main variable explaining population structure in fungal plant pathogens, because specialization contributes to reduce gene flow between populations associated with different hosts. Previous population genetic analysis revealed that French populations of the grey mould pathogen Botrytis cinerea were structured by hosts tomato and grapevine, suggesting host specialization in this highly polyphagous pathogen. However, these findings raised questions about the magnitude of this specialization and the possibility of specialization to other hosts. Here we report specialization of B. cinerea populations to tomato and grapevine hosts but not to other tested plants. Population genetic analysis revealed two pathogen clusters associated with tomato and grapevine, while the other clusters co-occurred on hydrangea, strawberry and bramble. Measurements of quantitative pathogenicity were consistent with host specialization of populations found on tomato, and to a lesser extent, populations found on grapevine. Pathogen populations from hydrangea and strawberry appeared to be generalist, while populations from bramble may be weakly specialized. Our results suggest that the polyphagous B. cinerea is more accurately described as a collection of generalist and specialist individuals in populations. This work opens new perspectives for grey mold management, while suggesting spatial optimization of crop organization within agricultural landscapes.

While abscisic acid (ABA) is known as a hormone produced by plants through the carotenoid pathway, a small number of phytopathogenic fungi are also able to produce this sesquiterpene but they use a distinct pathway that starts with the cyclization of farnesyl diphosphate (FPP) into 2Z,4E-α-ionylideneethane which is then subjected to several oxidation steps. To identify the sesquiterpene cyclase (STC) responsible for the biosynthesis of ABA in fungi, we conducted a genomic approach in Botrytis cinerea. The genome of the ABA-overproducing strain ATCC58025 was fully sequenced and five STC-coding genes were identified. Among them, Bcstc5 exhibits an expression profile concomitant with ABA production. Gene inactivation, complementation and chemical analysis demonstrated that BcStc5/BcAba5 is the key enzyme responsible for the key step of ABA biosynthesis in fungi. Unlike what is observed for most of the fungal secondary metabolism genes, the key enzyme-coding gene Bcstc5/Bcaba5 is not clustered with the other biosynthetic genes, i.e., Bcaba1 to Bcaba4 that are responsible for the oxidative transformation of 2Z,4E-α-ionylideneethane. Finally, our study revealed that the presence of the Bcaba genes among Botrytis species is rare and that the majority of them do not possess the ability to produce ABA.