Phylogenetic Comparison of Swainsonine Biosynthetic Gene Clusters among Fungi

Swainsonine is a cytotoxic alkaloid produced by fungi. Genome sequence analyses revealed that these fungi share an orthologous gene cluster, SWN, necessary for swainsonine biosynthesis. To investigate the SWN cluster, the gene sequences and intergenic regions were assessed in organisms containing swnK, which is conserved across all fungi that produce swainsonine. The orders of fungi which contained orthologous swainsonine genes included Pleosporales, Onygenales, Hypocreales, Chaetothyriales, Xylariales, Capnodiales, Microthyriales, Caliciales, Patellariales, Eurotiales, and a species of the Leotiomycetes. SwnK and swnH2 genes were conserved across all fungi containing the SWN cluster; in contrast, swnT and swnA were found in a limited number of fungi containing the SWN cluster. The phylogenetic data suggest that in some orders that the SWN cluster was gained once from a common ancestor while in other orders it was likely gained several times from one or more common ancestors. The data also show that rearrangements and inversions of the SWN cluster happened within a genus as species diverged. Analysis of the intergenic regions revealed different combinations and inversions of open reading frames, as well as absence of genes. These results provide evidence of a complex evolutionary history of the SWN cluster in fungi.


Introduction
Secondary metabolites are organic molecules produced by plants, bacteria, and fungi and are critical for virulence, defense, and communication. The secondary metabolite swainsonine is a toxic indolizidine alkaloid that inhibits α-mannosidase and disrupts the endomembrane system of the animal cells causing a lysosomal storage disease and inhibiting mannosidase II in the Golgi apparatus altering glycoprotein synthesis [1][2][3]. Swainsonine was first isolated from the plant Swainsona canescens (Fabaceae) in Australia [4] and subsequently other plant genera within the Convolvulaceae, Fabaceae, and Malvaceae [5].

Materials and Methods
Protein sequences of SWN gene clusters (SwnA-EXU97977, SwnH1-EXU97978, SwnH2-EXU97983, SwnK-EXU97982, SwnN-EXU97980, SwnR-EXU97981, and SwnT-EXU97979) of the fungus Metarhizium robertsii order Hypocreales were obtained from NCBI as reference sequences. Homology of each SWN protein component was compared using the pblast program of NCBI. SwnK was used as the reference gene to initiate all analyses. Protein sequences for SwnK were identified first; all 36 organisms that contained SwnK were used for subsequent analyses. The sequences for each SWN protein (if available) were obtained from NCBI, downloaded in fasta file format ( Table 1).
Sequences of each protein were compared with Geneious Prime ® 2020.2.2 and aligned with MUSCLE with the following parameters: sequences grouped by similarity, anchor optimization, distance measure kmer6_6, clustering method UPGMB, tree rooting method pseudo, and sequence weighting scheme CLUSTALW. Sequences were manually edited to remove any low-quality bases. Maximum parsimony trees were used to construct the phylogenetic protein trees for the seven swainsonine genes and the five intergenic-region trees using PAUP* plugin with heuristic search strategy, fastStep search type and 1000 replications. The outgroup for each SWN gene represented the closest non-SWN-containing taxon of the Hypocreales. The binary tree was constructed using winclada and nona, max trees 100, and 10 replications. The ITS tree was constructed using nucleotides sequences from NCBI for the 14

), and
Tothia was the outgroup. Slafractonia was not included in this analysis because the SWN cluster is not contiguous. Trees were built using PHYML 3.3.20180621 [29], substitution model HKY85, and X1000 bootstrap values.

Swn Genes
All swainsonine gene matches were to fungal species, representing 11 orders within the Ascomycota, with the exception of Quercus suber. Quercus suber belongs to the kingdom Plantae. SwnK and swnH2 genes were identified from all fungi assessed in this study, while swnA and swnT were the identified for the fewest fungi ( Table 1). The swnA gene was the least frequently identified among the fungi assessed and was not identified among all Pleosporales and Xylariales members, as well as the Fusarium sp., Tothia fuscella, Pseudogymnoascus sp., and Rhizodiscinia lignyota. Quercus suber contained all the SWN genes with the exception of swnR, swnT, and swnA.
As expected, the swnK gene was identified in all the fungi assessed (Table 1, Figure 1) since it was used as the basis for inclusion in the study. Protein similarity, as shown in the SwnK tree, was generally shared by fungi within the same order, with the exception of the Fusarium sp. from the Hypocreales and the taxa from the Pleosporales which formed three different clades. Notably, there was a clade of four taxa representing at least three orders with bootstrap support of 81%. All taxa that have been reported to contain swainsonine contained swnK consistent with reports showing it to be essential for swainsonine biosynthesis [12] (Table 1).  All taxa that have been reported to contain swainsonine contained swnH2 as did all taxa that contained swnK ( Table 1). The SwnH2 tree was similar to SwnK for the fungal orders, Hypocreales and Onygenales, with the exception of Microsporum canis not grouping together with other Onygenales taxa. Other members from other orders diverged in the SwnH2 tree as well (Figure 2), including some taxa of the Xylariales and Pleosporales. For example, in the Xylariales two taxa grouped together with 94% bootstrap support while two others grouped with a taxon from the Microthyriales with 100% bootstrap support. All taxa that have been reported to contain swainsonine contained swnH2 as did all taxa that contained swnK ( Table 1). The SwnH2 tree was similar to SwnK for the fungal orders, Hypocreales and Onygenales, with the exception of Microsporum canis not grouping together with other Onygenales taxa. Other members from other orders diverged in the SwnH2 tree as well (Figure 2), including some taxa of the Xylariales and Pleosporales. For example, in the Xylariales two taxa grouped together with 94% bootstrap support while two others grouped with a taxon from the Microthyriales with 100% bootstrap support. The swnN gene was found in all taxa of interest except Rosellinia necatrix and Xylaria polymorpha ( Table 1). All taxa that have been reported to contain swainsonine contain swnN ( Table 1). The SwnN tree was similar to the SwnK tree for the order Onygenales and Hypocreales. The SwnN tree was unique in that all of the Pleosporales members with the exception of Periconia macrospinosa grouped together with 52% bootsrap support, which was not observed in the other trees ( Figure 3). Similar to the SwnK tree, there was the clade of four taxa representing at least three orders with a bootstrap support of 82%.  (Table 1). All taxa that have been reported to contain swainsonine contain swnN ( Table 1). The SwnN tree was similar to the SwnK tree for the order Onygenales and Hypocreales. The SwnN tree was unique in that all of the Pleosporales members with the exception of Periconia macrospinosa grouped together with 52% bootsrap support, which was not observed in the other trees ( Figure 3). Similar to the SwnK tree, there was the clade of four taxa representing at least three orders with a bootstrap support of 82%.
The swnH1 gene was found in all fungi except for Rosellinia necatrix, Xylaria polymorpha, and Pseudovirgaria hyperparasitica ( Table 1). All taxa that have been reported to contain swainsonine contain swnH1 ( Table 1). The SwnH1 tree showed a similar pattern to the SwnK tree where similar members of the Hypocreales, Onygenales, and Xylariales, and some Pleosporales grouped together ( Figure 4).
The swnR gene was found in all fungi examined except Trichophyton mentagrophytes ( Table 1). All taxa that have been reported to contain swainsonine contain swnR (Table 1). In the SwnR tree, only the Hypocreales and Onygenales members formed their own clades. Members of the Pleosporales showed a similar grouping pattern to the SwnK tree in which the Pleosporaceae species (P. seminiperda and A. oxytropis) grouped into one clade, and the C. aquaticus and S. leguminicola grouped into another clade ( Figure 5). The swnH1 gene was found in all fungi except for Rosellinia necatrix, Xylaria polymorpha, and Pseudovirgaria hyperparasitica ( Table 1). All taxa that have been reported to contain swainsonine contain swnH1 ( Table 1). The SwnH1 tree showed a similar pattern to the SwnK tree where similar members of the Hypocreales, Onygenales, and Xylariales, and some Pleosporales grouped together ( Figure 4).  The swnR gene was found in all fungi examined except Trichophyton mentagrophytes ( Table 1). All taxa that have been reported to contain swainsonine contain swnR (Table 1). In the SwnR tree, only the Hypocreales and Onygenales members formed their own clades. Members of the Pleosporales showed a similar grouping pattern to the SwnK tree in which the Pleosporaceae species (P. seminiperda and A. oxytropis) grouped into one clade, and the C. aquaticus and S. leguminicola grouped into another clade ( Figure 5).  The swnA gene was detected in the Onygenales, all Hypocreales with the exception of the Fusarium sp., and a single taxon of four other orders ( Table 1). Taxa that have been reported to produce and not produce swainsonine were among the taxa that contained swnA ( Table 1). Members of the Hypocreales and Onygenales grouped together in the SwnA tree, respectively, and were part of a larger clade that included the two taxa belonging to the Eurotiales and Caliciales, while the two taxa belonging to the Capnodiales and the Chaetothyriales grouped together into a single clade with 88% bootstrap values (Figure 6). Notably, the grouping together of the two taxa belonging to the Capnodiales and the Chaetothyriales had not been observed in other trees. The swnA gene was detected in the Onygenales, all Hypocreales with the exception of the Fusarium sp., and a single taxon of four other orders ( Table 1). Taxa that have been reported to produce and not produce swainsonine were among the taxa that contained swnA ( Table 1). Members of the Hypocreales and Onygenales grouped together in the SwnA tree, respectively, and were part of a larger clade that included the two taxa belonging to the Eurotiales and Caliciales, while the two taxa belonging to the Capnodiales and the Chaetothyriales grouped together into a single clade with 88% bootstrap values ( Figure 6). Notably, the grouping together of the two taxa belonging to the Capnodiales and the Chaetothyriales had not been observed in other trees.
The swnT gene was detected in the Hypocreales, Onygenales, two taxa of the Xylariales, and single taxa of three other orders. Taxa that have been reported to produce and not produce swainsonine were among the taxa that contained swnT (Table 1). Interestingly Slafractonia leguminicola was the only taxon of the Pleosporales found to contain swnT (Table 1, Figure 7). Similar to other trees, in the SwnT tree, members of the Onygenales, Hypocreales, and Xylariales grouped together, respectively (Figure 7). The swnT gene was detected in the Hypocreales, Onygenales, two taxa of the Xylariales, and single taxa of three other orders. Taxa that have been reported to produce and not produce swainsonine were among the taxa that contained swnT (Table 1). Interestingly Slafractonia leguminicola was the only taxon of the Pleosporales found to contain swnT (Table 1, Figure 7). Similar to other trees, in the SwnT tree, members of the Onygenales, Hypocreales, and Xylariales grouped together, respectively (Figure 7).
The swnH2 gene appeared to be the most variable position in that it was found joined with all the other genes, except with swnN. The swnN and swnH1 genes were found in four combinations, and swnR, swnT, and swnA genes were found in three combinations. The swnK was the least variable and it joined only with swnH2 and swnR.
The sizes of the intergenic regions were greater than 500 bp for most of the fungi with exception to the intergenic region between swnH1 and swnH2, which was under 50 bp in the Pleosporales (Table 2). Intergenic regions greater than 1000 bp were recorded with the hydrogenases: swnH2-swnK, swnH1-swnT, swnT-swnH2, swnH2-swnR, swnA-swnH2, and swnR-swnH1. The intergenic region between swnH2-swnK was conserved in all fungi expect for Clohesyomyces aquaticus, and the region between swnR-swnN was the second most conserved. The Chaetothyriaceae sp. fungus and Xylaria hypoxylon possessed the largest intergenic region of swnH2-swnK and swnH1-swnT. Alternaria oxytropis contained a large intergenic region of 2007 bp between swnN-swnH1.

Intergenic Regions
The intergenic regions between the SWN genes for 14 representative fungi were assembled and analyzed (Table 2). Six fungi with the following 13 characteristics (combinations) were denoted as Type A: swnH2-swnK, swnR-swnN, swnK-swnR, swnH1-swnT, swnN-swnH1, swnT-swnN, swnT-swnH2, swnH2-swnR, swnH1-swnH2, swnA-swnH2, swnA-swnH1, swnN-swnA, and swnR-swnH1 (Table 2, Figure 8). Inverted intergenic regions were detected in two fungi (except for swnT-swnH2 and swnR-swnH1 that had no inversions) were denoted as Type-B. Six fungi had a mix of both types and were classified as Type A/B.  In the phylogenetic tree of the intergenic regions, the 14 fungi separated into two major groups (Figure 9). One group consisted of all Pleosporales members, Pseudogymnoascus sp., and Rosellinia necatrix, and the other group consisted of all other fungi. Alternaria oxytropis, Pyrenophora seminiperda, and Clohesyomyces aquaticus were separated from Periconia macrospinosa by characters 7 (swnH2-swR), 8 (swnH1-swnH2), and 12 (swnR-swnH1). The black circle on the branches indicates the characters only changed at that spot on the tree, while number 1 indicates the presence of that character and 0 indicates absence. Xylaria hypoxylon shared more similarities with the bottom group, and the Onygenales shared the highest number of similarities. ascus sp., and Rosellinia necatrix, and the other group consisted of all other fungi. Alternaria oxytropis, Pyrenophora seminiperda, and Clohesyomyces aquaticus were separated from Periconia macrospinosa by characters 7 (swnH2-swR), 8 (swnH1-swnH2), and 12 (swnR-swnH1). The black circle on the branches indicates the characters only changed at that spot on the tree, while number 1 indicates the presence of that character and 0 indicates absence. Xylaria hypoxylon shared more similarities with the bottom group, and the Onygenales shared the highest number of similarities.  Table 2). The 0 and 1 under each circle represent  Table 2). The 0 and 1 under each circle represent absence (0) or presence (1) of that character. The closed circles indicate that the character only changes at that one spot on the tree, and the open circles indicate changes elsewhere on the tree also.
Five trees of the intergenic regions were informative among some of the taxa investigated: swnH2-swnK, swnH1-swnT, swnK-swnR, swnN-swnH1, and swnT-swnN. In the swnH2-swnK tree, two members of the Pleosporales, Alternaria oxytropis, and Pyrenophora seminiperda, and two members of the Onygenales, Nannizzia gypsea and Trichophyton mentagrophytes, grouped together, respectively, with high boots strap support ( Figure 10). These same taxa grouped together with high bootstrap support in other intergenic trees when they were present (Figures 11-14). In some trees these respective taxa grouped with other members of the same order, respectively, such as Microsporum canis and Clohesyomyces aquaticus (Figures 11-14). Five trees of the intergenic regions were informative among some of the taxa inves- tigated: swnH2-swnK, swnH1-swnT, swnK-swnR, swnN-swnH1, and swnT-swnN. In the swnH2-swnK tree, two members of the Pleosporales, Alternaria oxytropis, and Pyrenophora seminiperda, and two members of the Onygenales, Nannizzia gypsea and Trichophyton mentagrophytes, grouped together, respectively, with high boots strap support ( Figure 10). These same taxa grouped together with high bootstrap support in other intergenic trees when they were present (Figures 11-14). In some trees these respective taxa grouped with other members of the same order, respectively, such as Microsporum canis and Clohesyomyces aquaticus (Figures 11-14).

ITS Phylogeny
A phylogenetic tree of the internal transcribed sequences (ITS) for the 14 representative fungi was constructed to help compare the relationships between the fungi for this set of noncoding regions with those from the SWN intergenic regions and SWN proteins. Taxa belonging to the same orders, Onygenales, Pleosporales, and Hypocreales grouped together with good boot strap support as one would expect (Figure 15). Metarhizium Figure 14. Maximum parsimony DNA tree resulting from the analysis of the swnT-swnN intergenic region sequence data. Bootstrap confidence values from 1000 bootstrap replicates are presented at each corresponding node. Fungal orders are indicated next to each organism/clade.

ITS Phylogeny
A phylogenetic tree of the internal transcribed sequences (ITS) for the 14 representative fungi was constructed to help compare the relationships between the fungi for this set of noncoding regions with those from the SWN intergenic regions and SWN proteins. Taxa belonging to the same orders, Onygenales, Pleosporales, and Hypocreales grouped together with good boot strap support as one would expect (Figure 15). Metarhizium acridum (Hypocreales) grouped with the Xylariales with high bootstrap support consistent with the fact that they both belong to the same fungal class, Sordariomycetes.

Open Reading Frames within the Intergenic Regions
Many of the intergenic regions contained open reading frames (ORFS) ( Table 2, Supplementary Tables S2-S14). However, generally only ORFS under 30 amino acids (aa) showed matching identity with proteins in databases. Most matches were to fungi, with much fewer matches to bacteria, and amoebae. There were also matches to plants, which may be due to errors in the databases. The swnN-swnH1, swnT-swnH2, and swnR-swnH1 intergenic regions matched only fungi, while the other intergenic regions had matched with bacteria or amoebae also. Overall, 5 of the 14 fungi assessed only matched fungal ORFS.

Open Reading Frames within the Intergenic Regions
Many of the intergenic regions contained open reading frames (ORFS) ( Table 2, Supplementary Tables S2-S14). However, generally only ORFS under 30 amino acids (aa) showed matching identity with proteins in databases. Most matches were to fungi, with much fewer matches to bacteria, and amoebae. There were also matches to plants, which may be due to errors in the databases. The swnN-swnH1, swnT-swnH2, and swnR-swnH1 intergenic regions matched only fungi, while the other intergenic regions had matched with bacteria or amoebae also. Overall, 5 of the 14 fungi assessed only matched fungal ORFS.
Interestingly, some of the ORFS were matches to bacterial species. Nannizzia gypsea recorded the highest number of bacterial similarities in the following intergenic regions and their bacterial matches: swnK-swnR (Euplotes aediculatus), swnH1-swnT (Shewanella piezotolerans), swnT-swnN (Escherichia coli), and swnN-swnA (Escherichia coli and Agrobacterium tumefaciens) and percent identity and query cover were high. Alternaria oxytropis recorded bacterial similarities as well, in swnH2-swnK (Geobacillus thermodenitrificans). Other bacterial similarities were homofermentative species Lacticaseibacillus paracasei and Leuconostoc mesenteroides were matches to Rosellinia necatrix in the swnH2-swnK with 100% identity and 67% query cover. The Gram-negative fish bacteria Yersinia ruckeri was a match to Clohesomyces aquaticus in the swnN-swnR with 81% query cover and 69.23% identity.
The amoeboid slime mold Dictyostelium discoideumn was recorded in swnH2-swnK as a match to Pyrenophora seminiperda, Periconia macrospinosa, and Pseudogymnoascus sp., in swnR-swnN as a match to Chaetothyriaceae sp., in swnK-swnR as a match to Nannizzia gypsea, and in swnH2-swnA as a match to Microsporum canis. The Xylariales member Xylaria hypoxylon shared similarities with the protozoan pathogen Trypanosoma cruzi with 80% query cover and 67% identity in the swnK-swnR. Four different accessions of Ipomoea carnea (Kingdom: Planta) were matches to the seed transmitted fungus Chaetothyriaceae with 100% query covers in swnH2-swnK.
Some fungal matches shared predicted functions for their orfs, while others were unrelated to secondary metabolites. The matches to Fusarium, Alternaria, Aschochyta, Aspergillus, Omphalotus, Metarhizium, and Trichophyton, showed a function of non-ribosomal peptide synthetase, highly reducing polyketide synthase, non-reducing reducing polyketide synthase, and swainsonine genes. Other fungal matches showed dissimilarities in function, for example, Saccharomyces cerevisiae match in the swnH2-swnK was a ribonuclease P protein component in the mitochondria. Bacterial matches were more dissimilar in their function.

Swainsonine
Swainsonine was detected in the isolate of Pyrenophora seminiperda (Table 1). Swainsonine was previously detected in several of the species herein [12].

Discussion
The SWN gene clusters were identified from fungi within the orders Hypocreales, Chaetothyriales, Onygenales, Pleosporales, and a Leotiomycetes sp., as previously reported [12], and in additional species of these orders. Additional orders of fungi were also found to contain the SWN gene cluster including the Xylariales, Capnodiales, Microthyriales, Caliciales, Patellariales, and Eurotiales. The SWN cluster was also detected in the higher plant Quercus suber which is not expected as swainsonine has not been shown to be a plant product. We suspect that the Quercus suber plant material that was the source of the DNA sequenced was likely associated with a fungal symbiont of some unknown genus that contained the SWN genes. Notably, the morning glory plant Ipomoea carnea was matched with the fungus Chaetothyriaceae sp. intergenic region; Chaetothyriaceae sp. is a seed transmitted fungus of the plant Ipomoea carnea [7,32]. In both cases, we suspect that these genes belonged to fungi living on or within the plants.
The swnK gene was identified in all fungi. This was expected since swnK is required for swainsonine production and is highly conserved among swainsonine-producing fungi [12,20]. The swnH2 gene was found in all taxa also suggesting the important role of this gene for the synthesis of swainsonine. Deletion of the swnH2 or swnH1 gene in Metarhizium robertsii resulted in the inability of the fungus to produce swainsonine while deletion of swnN, swnR, swnT, and swnA reduced swainsonine production to varying amounts in M. robertsii but did not eliminate it [19]. These results may suggest that species lacking the SwnH1 gene may not produce swainsonine. Furthermore, not all SWN genes are likely required for swainsonine biosynthesis in all taxa as Slafractonia leguminicola, Pyrenophora seminiperda, and Alternaria oxytropis were missing some of the SWN genes and still produced swainsonine.
In general, the SWN proteins from different fungi that clustered together with the most confidence were also fungi that are the most closely related. For example, all Trichophyton and Metarhizium species grouped together with greater than 90% percent bootstrap support for each respective protein within the SWN cluster. These results suggest that the SWN gene cluster was present in a recent common ancestor from each of these respective genera. Subsequently as the individual species in these two genera have diverged there have been rearrangements and recombination events that have resulted in the order of the respective SWN genes, as shown herein.
In contrast, the SWN proteins in the different taxa of the Pleosporales clustered with one other Pleosporales species but not others or clustered with taxa representing other orders. Specifically, Alternaria oxytropis and Pyrenophora seminiperda grouped together in all trees with bootstrap confidence values over 91%, while Slafractonia leguminicola and Clohesyomyces aquaticus grouped together in the SwnK, SwnR, and SwnN trees. Pyrenophora macrospinosa did not group with any member of the Pleosporales, but instead grouped with Tothia fuscella (Microthyriales), Pseudogymnoascus sp. (Leotiomycetes), and Rhizodiscinia lignyota (Patellariales). These results suggest that Alternaria oxytropis and Pyrenophora seminiperda likely shared the same common ancestor containing the SWN cluster. These results are consistent with a recent report of Creamer et al. [21] demonstrating that several polyketide synthases in Alternaria oxytropis were most closely related to paralogs in Pyrenophora seminiperda rather than other Alternaria species. Further supporting this observation is the fact that none of the closely related Alternaria species that have been sequenced contain the SWN cluster. The three groups among the different species of the Pleosporales belong to different families, which might explain the high variability within this group. The patterns observed in this study, combinations and inversions, and absence of genes, suggest that the swainsonine gene cluster was not transferred as a whole-cluster gain to a common ancestor within the Pleosporales but may have originated several times within the order.
Slafractonia leguminicola was the only member of the Pleosporales which contained swnT. It is likely that SwnT serves a role in transferring swainsonine within or out of the cell as was suggested in Cook et al. [12]. Slafractonia was also separated from other fungi in the swnT tree. The presence and absence of different SWN genes among different members of the Pleosporales suggests that the SWN genes may have been inherited from different common ancestors, and that the SWN genes may still be undergoing evolutionary changes within the Pleosporales.
In summary, the divergence of the secondary metabolites gene clusters between closely related Ascomycota spp. is hypothesized to be the result of functional diversity, de novo assembly, and/or horizontal gene transfer [33]. Manning et al. [34] reported a horizontal gene transfer and gene duplications events in NRPSs of Pyrenophora triticirepentis. Horizontal gene transfer and orthologous functional diversity were also discussed as contributing factors to the variability found in Alternaria oxytropis SWN genes [21]. In this study, the diversity found within fungal members and the bacterial matches in the ORFs could be due to a combination of horizontal gene transfer, gene duplications, and functional diversity. The high percentage identity recorded for bacterial matches in the open reading frames could suggest bacterial ontogeny. The gene transfer of secondary metabolites from bacteria to ascomycetes was suggested by Lawrence et al. [28], which reported that the hybrid NPS/PKS could have been acquired by Ascomycota via HGT from a bacterial donor in the Burkholderiales early in the evolution of Pezizomycotina. The work here reinforces the diversity and the complex evolution of the swainsonine gene cluster in fungi.

Conclusions
Few fungal species are known for their production of the toxic alkaloid swainsonine. Despite their diversity, the SWN gene cluster was identified among several fungi. The diversity of the SWN cluster and the intergenic regions within the cluster mirrors the diversity of the various fungal orders; Pleosporales are highly diverse, while the Onygenales and Hypocreales are extremely conserved. The open reading frames in the intergenic regions matched primarily fungi, however the high percentage identity recorded for bacterial matches could suggest bacterial ontogeny for portions of the cluster. These SWN cluster analyses provide a basis for understanding the evolution of secondary metabolites and the mechanisms responsible for the complexity of this cluster in fungi.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jof8040359/s1. Table S1A: SWN gene arrangements in each fungus; Table S1B: The genomic positions. swnN intergenic region was not found in Periconia macrospinosa as it is in different scaffold; Table S2: Open reading frames ORFs in the intergenic region swnH2-swnK information; Table S3: Open reading frames ORFs in the intergenic region swnR-swnN information; Table S4: Open reading frames ORFs in the intergenic region swnK-swnR information; Table S5: Open reading frames ORFs in the intergenic region swnH1-swnT information; Table S6: Open reading frames ORFs in the intergenic region swnN-swnH1 information; Table S7: Open reading frames ORFs in the intergenic region swnT-swnN information; Table S8: Open reading frames ORFs in the intergenic region swnT-swnH2 information; Table S9: Open reading frames ORFs in the intergenic region swnH2-swnR information; Table S10: Open reading frames ORFs in the intergenic region swnH1-swnH2 information; Table S11: Open reading frames ORFs in the intergenic region swnA-swnH2 information; Table S12: Open reading frames ORFs in the intergenic region swnA-swnH1 information; Table S13: Open reading frames ORFs in the intergenic region swnN-swnA information; Table S14: Open reading frames ORFs in the intergenic region swnR-swnH1 information.