Lichen Biosynthetic Gene Clusters Part II: Homology Mapping

Article Cite This: J. Nat. Prod. 2018, 81, 732−748

pubs.acs.org/jnp

Lichen Biosynthetic Gene Clusters Part II: Homology Mapping Suggests a Functional Diversity Robert L. Bertrand, Mona Abdel-Hameed, and John L. Sorensen* Department of Chemistry, University of Manitoba, Winnipeg, Manitoba Canada, R3T 2N2 S Supporting Information *

ABSTRACT: Lichens are renowned for their diverse natural products though little is known of the genetic programming dictating lichen natural product biosynthesis. We sequenced the genome of Cladonia uncialis and profiled its secondary metabolite biosynthetic gene clusters. Through a homology searching approach, we can now propose specific functions for gene products as well as the biosynthetic pathways that are encoded in several of these gene clusters. This analysis revealed that the lichen genome encodes the required enzymes for patulin and betaenones A−C biosynthesis, fungal toxins not known to be produced by lichens. Within several gene clusters, some (but not all) genes are genetically similar to genes devoted to secondary metabolite biosynthesis in Fungi. These lichen clusters also contain accessory tailoring genes without such genetic similarity, suggesting that the encoded tailoring enzymes perform distinct chemical transformations. We hypothesize that C. uncialis gene clusters have evolved by shuffling components of ancestral fungal clusters to create new series of chemical steps, leading to the production of hitherto undiscovered derivatives of fungal secondary metabolites.

L

biosynthesis of usnic acid, a dibenzofuran polyketide produced exclusively by lichening fungi.7 C. uncialis was chosen because it was a readily accessible species of lichen that only produced usnic acid.7 The observation of only one secondary metabolite in C. uncialis compared to the presence of up to 48 gene clusters is a surprising observation that motivated us to attempt to propose functions for the remaining gene clusters.9 We have adopted a homology mapping approach to propose specific functions for these gene clusters. Version 4.0 of FungiSMASH (Antibiotics and Secondary Metabolites Analysis Shell)10 contains a ClusterBLAST module that provides cluster similarity statistics to experimentally characterized gene clusters. The development of this technology allows us to rapidly expand our understanding of lichen natural product biosynthesis by proposing specific functions for lichen clusters based on reference sequences. The C. uncialis gene clusters were uploaded to FungiSMASH (v. 4.0).10 If ClusterBLAST indicates genetic similarity between a gene cluster of C. uncialis and another characterized fungal cluster, a pBLAST analysis was performed for each C. uncialis gene to determine similarity to the gene of the reference organism. The first line of evidence that two genes plausibly share a common biosynthetic function is therefore the observation of genetic similarity. A phylogenetic analysis of

ichens are symbionts of algae and fungi that have fascinated the natural products community because of their diverse secondary metabolites.1 Little is known as to how lichens biosynthesize secondary metabolites. In non-lichenized fungi, the function of each gene in several biosynthetic gene clusters and the sequential chemical steps in the biosynthetic pathway have been described in detail, providing textbook understanding of how natural products are produced.2 Experimental procedures for the functional elucidation of unassigned gene clusters are well established.3 In lichenized fungi, no secondary metabolite biosynthetic gene cluster has been linked to a metabolite through heterologous expression or gene knockout experiments. This is because of the frustrating challenges of working with lichen specimens including slow rate of growth and obligate symbioses.4 Although in three instances transcriptional heterologous expression has been achieved, de novo production of a lichen metabolite was not observed in any study.5 Grayanic acid,6 usnic acid,7 and 6-hydroxymellein8 are the only three lichen metabolites that have been putatively linked to a gene cluster through nondefinitive experiments. We have recently reported genome sequencing of Cladonia uncialis and profiled its secondary metabolome.9 This analysis revealed up to 48 putative secondary metabolite biosynthetic gene clusters. These included type I and type III polyketide synthases (PKS), nonribosomal peptide synthetases (NRPS), hybrid PKS-NRPS, and terpene synthases, many replete with postsynthetic tailoring genes.9 The original objective of genome sequencing was to identify the gene cluster responsible for the © 2018 American Chemical Society and American Society of Pharmacognosy

Received: September 8, 2017 Published: February 27, 2018 732

DOI: 10.1021/acs.jnatprod.7b00770 J. Nat. Prod. 2018, 81, 732−748

Journal of Natural Products

Article

Figure 1. C. grayi grayanic acid gene cluster (accession no. GU930713) and C. uncialis uncharacterized gene cluster (accession no. MG777495). Color coding indicates homology pairing. Domain abbreviations: Starter acyltransferase (SAT), ketosynthase (KS), acyltransferase (AT), product template domain (PT), acetyl carrier protein (ACP), thioesterase (TE). Gene abbreviations: Polyketide synthase (PKS).

Scheme 1. Experimentally Supported Pathway for Grayanic Acid Biosynthesis in C. grayi and, Juxtaposed, the Biosynthetic Pathway That Is Proposed to be Encoded in C. uncialis

shared functions is therefore the ability to assign functions to multiple genes, resulting in coherent biosynthetic pathways. The strength of proposals can be weighted based on these three lines of evidence: A lichen cluster with many genetically similar genes to the reference organism, with close evolutionary relationships, and producing a coherent and uninterrupted series of chemical transformations is more likely to be a correct proposal, as compared to a proposal generated from a lichen cluster with few genetic similarities, few homologies, and a biosynthetic pathway lacking requisite genes. These three lines of evidence do not provide definitive assignment of function but are intended to be used as a tool to prioritize and guide future characterization experiments. The development of a reliable method to heterologously express candidate gene clusters will be required to bridge the gap between predictive and definitive assignments of function. This homology mapping approach was used to propose a halogenated derivative of 6-hydroxymellein as the product of a C. uncialis cluster,8 based on the reference cluster devoted to terrein biosynthesis in Aspergillus terreus.13 By applying this same approach to the approximately 48 biosynthetic gene clusters recently described in C. uncialis9 we can now propose that C. uncialis encodes the requisite enzymes for patulin and betaenone biosynthesis, two fungal mycotoxins not known to be produced by lichens. Proposals of function involving

each gene pair was then performed. It is a basic premise of phylogenetics that genes with close evolutionary relationships are more likely to share similar functions than genes with distant common origin, a premise equally applied to the study of secondary metabolite biosynthesis in lichenizing fungi.11 Ancient horizontal gene transfer events have played a role in the evolution of both lichening and non-lichenizing fungi including the evolution of secondary metabolite gene clusters.12 Vertical gene transfer nonetheless predominates the evolutionary history of multicellular species of fungi and particularly so among fungal species of related and recent evolutionary origin. A phylogenetic approach among species of fungi was therefore adopted to determine whether genetically similar pairs of genes were likely to share recent evolutionary heritage. If pairs of lichen and reference genes were observed to cluster together within a common clade, genetic homology was inferred. The second line of evidence that two genes share similar biosynthetic functions is therefore the observation of genetic homology. We then attempted to map a coherent series of chemical transformations based on the characterized biosynthetic pathway of the reference organism, using those lichen genes presenting genetic similarity and/or homology to genes from the reference organism. Lichen genes presenting genetic similarity and/or homology are proposed to share the function of the reference gene. The third line of evidence of 733



Article

Table 1. BLAST Statistics and Proposed Function of C. uncialis Genes That Are Genetically Similar to Genes within the Grayanic Acid Cluster of C. grayi C. uncialis gene

accession no.

C. grayi gene

accession no.

proposed function of C. uncialis gene

similarity (%)/coverage (%)

Nr-pks-7 2 3

AUW31177 AUW31178 AUW31179

CgrPKS16 2 3

ADM79459 ADM79460 ADM79461

4-O-demethyl-sphaerophorin synthase 4-O-demethyl-sphaerophorin oxidase 4-O-demethylgrayanic acid O-methyltransferase

91/97 86/88 74/99

Figure 2. Truncated phylogenetic trees illustrating degree of relationship between genes of C. uncialis and C. grayi pertinent to grayanic acid biosynthesis: (A) 4-O-demethylsphaerophorin synthase; (B) 4-O-demethylsphaerophorin oxidase; (C) 4-O-demethylgrayanic acid O-methyltransferase. Complete phylogenetic trees (Figures S1−S3) are found in the Supporting Information. Bar signifies number of amino acid substitutions per site.

Figure 3. P. betae betaenone gene cluster (accession no. LC011911) and C. uncialis uncharacterized gene cluster (accession no. MG777513). Genes with matching colors are genetically similar. Domain abbreviations: Enoylreductase (ER), ketosynthase (KS), acyltransferase (AT), dehydratase (DH), C-methyltransferase (MT), ketoreductase (KR), acetyl carrier protein (ACP), reductase (R). Gene abbreviations: Polyketide synthase (PKS), short-chain dehydrogenase/reductase (SDR), flavin adenine dinucleotide-dependent oxidase (FAD oxidase).

pathway of grayanic acid. The grayanic acid gene cluster and proposed biosynthetic pathway are shown in Figure 1 and Scheme 1. One gene cluster in C. uncialis was identified as a possible functional homologue of the C. grayi grayanic acid cluster (Figure 1). A pBLAST search on each of the three genes within the C. uncialis cluster revealed robust similarity scores for each of the three genes in the grayanic acid gene cluster of C. grayi (Table 1). The C. uncialis cluster is presented with color coding to indicate similarity pairings (Figure 1). The domain architectures of the PKS are identical (Figure 1). Clustering of gene pairs in the resultant phylogenetic trees infers recent evolutionary heritage (Figure 2), an unsurprising result because C. uncialis and C. grayi are members of the Cladonia genus.

derivations of biosynthetic pathways associated with cytochalasin, azaphilone, fusarubin, pestheic acid, and mycophenolic acid will also be presented. These proposals are presented in order of greatest confidence to most speculative.

■

RESULTS AND DISCUSSION Grayanic Acid. Grayanic acid is a polyketide metabolite produced exclusively by lichenizing fungi. Reproducible induction of grayanic acid biosynthesis in C. grayi combined with transcriptional profiling of candidate polyketide synthase genes led to the putative identification of the grayanic acid gene cluster.6 The PKS gene was found next to genes encoding a cytochrome P-450 oxidase and an O-methyltransferase, an observation that was consistent with the expected biosynthetic 734



Article

Scheme 2. Experimentally Supported Pathway for Betaenone A−C Biosynthesis in P. betae and, Juxtaposed, the Biosynthetic Pathway That Is Proposed to be Encoded in C. uncialis

Table 2. BLAST Statistics and Proposed Function of the C. uncialis Genes That Are Genetically Similar to Genes of the Betaenone Pathway of P. betae C. uncialis gene

accession no.

P. betae gene

accession no.



1 2 3 4 R-pks-15

AUW31407 AUW31408 AUW31409 AUW31410 AUW31411

Bet4 Bet5 Bet2 Bet3 Bet1

BAQ25463 BAQ25462 BAQ25465 BAQ25464 BAQ25466

dehydroprobetaenone reductase putative betaenone C oxidase dehydroprobetaenone oxidase trans-enoylreductase dehydroprobetaenone synthase

58/98 56/98 60/96 64/98 61/99

These findings are interpreted as evidence that a grayanic acid gene cluster is encoded in C. uncialis, and a biosynthetic pathway identical to that of C. grayi has been proposed (Scheme 1). Specifically, the C. uncialis PKS named CU-NRPKS-7 is proposed to possess an equivalent function of CgrPKS16 such that its purpose is to synthesize 4-Odemethylsphaerophorin.6 The C. uncialis cytochrome P-450 and O-methyltransferase are then proposed to oxidize and methylate this intermediate, producing 4-O-demethylgrayanic acid and grayanic acid, respectively (Scheme 1). Betaenone. Betaenones A−C are phytotoxic polyketides produced by the fungus Phoma betae. Heterologous expression experiments demonstrated that the PKS (Bet1) and a transenoylreductase (Bet3) produce dehydroprobetaenone I in P. betae.14 The gene product of Bet4 reduces dehydroprobetaenone I to probetaenone I. The gene product of Bet2 is a biofunctional oxidase that may convert probetaenone I to betaenone B, and dehyroprobetaenone I to betaenone C. The function of Bet5 could not be conclusively determined but is hypothesized to convert betaenone C to betaenone A.14 The gene cluster and biosynthetic pathway of betaenones A−C are shown in Figure 3 and Scheme 2.

The ClusterBLAST module embedded within FungiSMASH (v. 4.0)10 identified a C. uncialis cluster with genetic similarities to the betaenone cluster of P. betae. A pBLAST analysis demonstrated that each of the five genes in the P. betae cluster are genetically similar to five genes in C. uncialis. Similarity scores and pairings are shown in Table 2 and Figure 3. The domain architecture of the lichen PKS is identical to Bet1, including the presence of a trans-enoylreductase (Figure 3). Phylogenetic analyses of each of these gene pairs support close genetic homology (Figure 4). These data suggest that the C. uncialis gene cluster is a complete betaenone gene cluster. Functional roles of each of the five genes can be proposed from the homology pairings, and in this case, an identical biosynthetic pathway is proposed (Scheme 2). Lichens are not known to produce betaenones A−C. Patulin. Patulin is a mycotoxin subject to strict regulation in the food industry because of its acute toxicity. The study of patulin biosynthesis was formative in the Birch acetate hypothesis and our understanding of polyketide biosynthesis in Fungi.15 Patulin biosynthesis requires 10 chemical steps facilitated by 15 gene products.16 The functions of five of these genes have been experimentally characterized. These include PatK, encoding 6-methylsalicylic acid synthase, PatG, encoding 735



Article

encoding a putative acetate transporter, and PatD, encoding a putative alcohol dehydrogenase. The role of PatA and PatD in patulin biosynthesis is unknown. Five genes within the lichen cluster were not similar to any gene member of the patulin cluster. These include a putative MFS transporter (1), a dioxygenase (2), an oxidoreductase (12), a short-chain dehydrogenase/reductase (14), and a gene ambiguously assigned as a reductase or an MFS transporter (17). Phylogenetic trees of the 13 gene pairings were generated to assess genetic homology. Functional homologues of the patulin genes of Penicillium expansum are also included in the phylogenetic trees. Of the 13 genetically similar gene pairs, 10 are present in a common clade. In most cases, the corresponding P. expansum homologue was also present (Figure 6). A close phylogenetic relationship could not be established for the lichen PKS (Cu-r-pks-9), a gene of unknown function (16), and a GMC oxidoreductase (18). It is possible that close ancestry could not be established for the PKS because its ancestry is distinct from other patulin-related genes, because the phylogenetic resolution is low, or because it is not a 6-methylsalicylic acid synthase. Complete phylogenetic trees, including the three nonsupportive trees, are displayed (Figures S9−S21, Supporting Information). Although the function of every gene product in the patulin biosynthetic pathway in A. clavatus has yet to be elucidated, the high number of homology pairings lends to the conclusion that C. uncialis likely has the requisite biosynthetic programming to produce patulin or a close derivative of patulin. We have therefore proposed a scheme by which the gene products of C. uncialis can plausibly produce patulin in a manner similar to A. clavatus and P. expansum (Scheme 3).16 Patulin has not been isolated from any species of lichen. 6-Hydroxymellein. A homology mapping approach similar to that adopted in this study was previously used to propose a biosynthetic pathway encoded in a C. uncialis gene cluster.8 This gene cluster, possessing a nonreducing PKS and four tailoring genes, was proposed to biosynthesize a halogenated derivative of 6-hydroxymellein, an intermediate metabolite of terrein biosynthesis of Aspergillus terreus.13 Plausible final products of this pathway were proposed based on known lichen metabolites catalogued in the Dictionary of Natural Products.8 Azaphilone. Azaphilones are a class of fungal polyketide pigments with pyrone−quinone structures. Azaphilones display antimicrobial, antifungal, antiviral, cytotoxic, nematicidal, and anti-inflammatory properties.18 In Monascus pilosus, the gene cluster of an azaphilone pigment was identified through TDNA random mutagenesis and selection of pigment-defective strains.19 Inactivation experiments demonstrated that the gene product of MpPKS5 is responsible for the biosynthesis of a substituted 3-methylorcinaldehyde. The authors hypothesize that the SAT domain of MpPKS5 selects crotonyl-CoA in lieu of acetyl-CoA.19 Tailoring enzymes MppB, MppD, and MppF are then proposed to facilitate hydroxylation, esterification, and acylation, yielding rubropunctatin.19 The function of the remaining genes in the M. pilosus cluster are unknown. The azaphilone cluster and biosynthetic pathway are shown in Figure 7 and Scheme 4. The ClusterBLAST module of FungiSMASH (v. 4.0)10 suggested that a C. uncialis gene cluster was genetically similar to the azaphilone pathway of M. pilosus. A total of 10 similarities were observed (Table 4; Figure 7). Phylogenetic support of genetic homology was only limited because four of

Figure 4. Truncated phylogenetic trees illustrating degree of relationship between genes of C. uncialis and P. betae pertinent to betaenone biosynthesis: (A) dehydroprobetaenone reductase; (B) betaenone C oxidase; (C) dehydroprobetaenone oxidase; (D) enoylreductase; (E) dehydroprobetaenone synthase. Complete phylogenetic trees (Figures S4−S8) are found in the Supporting Information. Bar signifies number of amino acid substitutions per site.

6-methylsalicylic acid decarboxylase, PatH, encoding m-cresol hydroxylase, PatI, encoding m-hydroxybenzyl alcohol hydroxylase, and PatN, encoding isoepoxydon dehydrogenase.17 The patulin gene cluster and biosynthetic pathway are displayed in Figure 5 and Scheme 3. The ClusterBLAST module of FungiSMASH (V. 4.0)10 indicated that a C. uncialis gene cluster was genetically similar to the patulin cluster of Aspergillus clavatus. Of the 18 genes present in the lichen cluster, pBLAST analysis revealed 13 similarity pairings (Figure 5; Table 3). The domain architectures of the C. uncialis and A. clavatus PKS are identical (Figure 5). Similarity pairings were not established for PatA, 736



Article

Figure 5. A. clavatus patulin gene cluster (accession no. A1CFL8) and C. uncialis uncharacterized gene cluster (accession no. MG777507). Genes with matching colors are genetically similar. Domain abbreviations: Ketosynthase (KS), acyltransferase (AT), dehydratase (DH), ketoreductase (KR), acetyl carrier protein (ACP). Gene abbreviations: Polyketide synthase (PKS), major facilitator superfamily transporter (MFS transporter), 6methylsalicylic acid (6-MSA), ATP-binding cassette transporter (ABC transporter), flavin adenine dinucleotide-dependent oxidase (FAD oxidase), short-chain dehydrogenase/reductase (SDR), glucose-methanol-choline oxidoreductase (GMC oxidoreductase).

evolutionarily distant and therefore are more likely to have divergent functions. Nonetheless a plausible biosynthetic pathway will be proposed. With the exception of a fatty acid synthase not found in the lichen cluster, all requisite genes for rubropunctatin biosynthesis are present. A plausible biosynthetic pathway can be proposed leading to rubropunctatin (Scheme 4). The fatty acid could be provided by a nondedicated fatty acid synthase. Four genes in C. uncialis were found to be genetically similar to genes of M. pilosus, although their product roles in M. pilosus have not been determined. It is possible that the products of these gene pairings function similarly to produce a derivative of rubropunctatin that is common to both M. pilosus and C. uncialis (Scheme 4). Cytochalasin. The cytochalasins are a group of fungal polyketide−amino acid chimeric natural products produced by a hybrid PKS-NRPS (polyketide synthase−nonribosomal peptide synthetase). The macrocyclic ring and isoindolone moieties characteristic of this group of natural products are derivatized from a highly reduced polyketide backbone fused to a phenylalanine.20 The gene clusters responsible for the biosynthesis of cytochalasins E and K have been identified in A. clavatus through genome surveying and gene inactivation.21 Biosynthesis is proposed to proceed through seven rounds of Claisen condensation between an acetyl-CoA primer and seven units of malonyl-CoA by the PKS module to form a C16polyketide backbone that is methylated three times. The polyketide chain is fused to a phenylalanine by the NRPS module and released by an NADPH-dependent terminal reductase (R) domain. The NRPS-PKS (CcsA) lacks a functional ER domain required for complete reduction of βketo sites. A trans-enoylreductase (CcsC) encoded downstream of CcsA is proposed to fulfill this role. The released aminoaldehyde intermediate is readily cyclized to form a pentaene intermediate. This intermediate then undergoes a series of modifications performed by five tailoring enzymes to produce cytochalasins E and K.21 The cytochalasin gene cluster and an abbreviated cytochalasin pathway illustrating the role of CcsA and CcsC are displayed in Figure 9 and Scheme 5.

Scheme 3. Experimentally Supported Pathway for Patulin Biosynthesis in A. clavatus and, Juxtaposed, the Biosynthetic Pathway That Is Proposed to be Encoded in C. uncialis

10 gene pairings were found to cluster within a common clade (Figure 8). The four genetically homologous genes include an amino-oxidase, an oxidoreductase, an efflux transporter, and a gene of unknown function. Complete phylogenetic trees, including trees that do not support genetic homology, are displayed (Figures S22−S31, Supporting Information). Although the lack of phylogenetic support does not automatically invalidate predictions of function, the viability of this prediction is weakened by inference that these genes are 737



Article

Table 3. BLAST Statistics and Proposed Function of the C. uncialis Genes That Are Genetically Similar to Genes of the Patulin Cluster of A. clavatus. C. uncialis gene

C. uncialis accession no.

A. clavatus gene

A. clavatus accession no.



3 4 5 6 7 8 R-pks-9 10 11 13 15 16 18

AUW31349 AUW31350 AUW31351 AUW31352 AUW31353 AUW31354 AUW31355 AUW31356 AUW31357 AUW31359 AUW31361 AUW31362 AUW31364

PatM PatN PatH PatO PatG PatL PatK PatJ PatI PatB PatC PatF PatE

XM_001273094 XM_001273095 XM_001273089 XM_001273096 XM_001273088 XM_001273093 XM_001273092 XM_001273091 XM_001273090 XM_001273083 XM_001273084 XM_001273087 XM_001273086

putative ABC transporter isoepoxydon dehydrogenase m-cresol hydroxylase isoamyl alcohol oxidase 6-methylsalicylic acid decarboxylase putative C6 transcription factor 6-methylsalicylic acid synthase unknown function m-hydroxybenzyl alcohol hydroxylase putative carboxylesterase putative MFS transporter unknown function putative GMC oxidoreductase

66/94 68/99 66/94 65/93 63/100 46/97 60/99 77/87 65/89 49/91 52/92 59/74 29/97

Four genes in C. uncialis presented similarity pairings to the F. f ujikuroi fusarubin pathway, suggesting that a possible function of the lichen cluster involves biosynthesis of a naphthoquinone. BLAST similarity scores and pairings are shown (Figure 11; Table 6). The identification of the lichen PKS, Cu-nr-pks-10, as a possible genetic homologue of the F. f ujikuroi PKS fsr1 is supported by a pBLAST similarity score of 40% (Table 6). However, the suggestion that Cu-nr-pks-10 is a plausible functional homologue of fsr1 is not supported by the domain architecture of the lichen PKS. Whereas the lichen PKS possesses a terminal thioesterase (TE), the F. fujikuroi PKS possesses a terminal reductase (R) domain (Figure 11). One gene in the lichen cluster is unusual because it appears to have acyl carrier protein (ACP) and terminal R domain-like features (Figure 11). It is possible that this gene product serves as a trans-reductase and that the terminal thioesterase of CU-NRPKS-10 is inoperative. A biosynthetic route to the first intermediate in fusarubin biosynthesis would then be feasible. Genetic similarities extend to an O-methyltransferase (3 | fsr2), a monooxygenase (8 | fsr3), and a short-chain dehydrogenase/ reductase (12 | fsr5), extending possible chemical transformations in C. uncialis to those known to occur in F. f ujikuroi (Scheme 7). Three lichen genes, a hydrolase (1), a cytochrome P-450 oxidase (6), and a monooxygenase (11), remain unassigned. Phylogenetic analysis suggests that these gene clusters have distant relationships, and therefore close genetic homology cannot be inferred (Figures S34−S37, Supporting Information). Querying the Dictionary of Natural Products for naphthoquinone polyketides returned hemoventosin, isolated from the lichen Haematomma ventosum, and coronataquinone, isolated from the lichen Pseudocyphellaria coronata.24 A biosynthetic proposal leading to both naphthoquinones from the fusarubin intermediate 8-O-methylanhydrofusarubinlactol can be proposed. One plausible route involves a two-electron reduction of 8-O-methylanhydrofusarubinlactol followed by hydroxylation, yielding O-methylcoronatoquinone (Scheme 8). Alternatively, two-electron reduction followed by two-electron rearrangement produces O-methylhemoventosin (Scheme 8). Available tailoring genes in C. uncialis include a hydrolase (1), a cytochrome P-450 (6), a monooxygenase (11), an SDR/ oxidoreductase (12), and two genes of unknown function (5, 9) (Figure 11). These gene products could plausibly perform the transformations leading to both metabolites. The presence or absence of the O-methyl residue (indicated by CH3*) on O-

A pBLAST analysis suggests that two genes in C. uncialis are similar to cluster members of the A. clavatus cytochalasin pathway. These include a C. uncialis NRPS-PKS (Cu-pks-nrps-3) and a trans-enoylreductase. BLAST statistics and similarity pairings are shown in Table 5 and Figure 9. The domain architecture of Cu-pks-nrps-3 is identical to CcsA including the requisite “R” terminal domain and the presence of a transenoylreductase encoded near the NRPS-PKS (Figure 9). Phylogenetic analyses support close genetic homology because both gene pairs were found to cluster within common clades (Figure 10). Four genes in C. uncialis did not have genetic similarity. These include an MFS transporter (2), an FAD oxidase (3), and two genes of unknown function (4, 5). It is proposed that the first two chemical steps in cytochalasin biosynthesis are encoded in C. uncialis. Derivatization of this common intermediate may then occur to produce an unknown final product (Scheme 5). It is interesting that the phylogenetic analysis of both cytochalasin-associated genes of C. uncialis also clustered with genes originating from Talaromyces sp., Penicillium sp., and Aspergillus sp. (Figure 8). The cyclized phenylalanine moiety of cytochalasin is also present in metabolites isolated from these species. Talaroconvolutin A, citridone C, carbonarin C, and preaspyridone A are four examples.22 Although no prediction could be offered of the final metabolite arising from the lichen cluster, the existence of these fungal metabolites lends a posteriori support for the proposition that the encoded NRPSPKS in C. uncialis produces a metabolite possessing the core phenylalanine moiety (Scheme 6). Fusarubin. Fusarubin is a naphthoquinone pigment produced by the ascomycetic fungus Fusarium f ujikuroi. Deletion of fsr1 abated fusarubin production, confirming Fsr1 as the nonreducing PKS devoted to fusarubin biosynthesis.23 Northern blot analyses demonstrated coregulation of Fsr1 with Fsr2 to Fsr6, suggesting that these six genes constitute the fusarubin gene cluster. Deletion experiments on each of the tailoring genes and chemical identification of the intermediate compounds elucidated the biosynthetic pathway. Deletion of fsr4 (putative short-chain dehydrogenase/reductase) and fsr5 (putative oxidoreductase) did not disrupt fusarubin biosynthesis. The authors hypothesize that these gene products have regulatory roles or are responsible for the derivatization of fusarubin to other products.23 The fusarubin gene cluster and biosynthetic pathway are shown in Figure 11 and Scheme 7. 738



Article

Figure 6. Truncated phylogenetic trees illustrating degree of relationship between genes of C. uncialis and A. clavatus pertinent to patulin biosynthesis. Functional homologues of A. clavatus genes in P. expansum are included: (A) ABC transporter; (B) isoepoxydon dehydrogenase; (C) m-cresol hydroxylase; (D) isoamyl alcohol oxidase; (E) 6-methylsalicylic acid decarboxylase; (F) C6 transcription factor; (G) unknown function; (H) m-hydroxybenzyl alcohol hydroxylase; (I) carboxylesterase; (J) MFS transporter. Complete phylogenetic trees (Figures S9−S21) are found in the Supporting Information. Bar signifies number of amino acid substitutions per site.

genes (R1 to R3). The domain architecture of PtaA is notable for lacking a terminal domain. The β-lactamase (PtaB) is proposed to replace the function of the absent terminal domain of PtaA by facilitating hydrolysis and Claisen cyclization of the polyketide, yielding atrochrysone carboxylic acid. The spontaneous dehydration of the unstable intermediate produces endocrocin anthrone. An oxygenase (PtaC) is then proposed to oxidize endocrocin anthrone, yielding endocrocin. Spontaneous decarboxylation forms emodin. A series of additional

methylhemoventosin and O-methylcoronataquinone is variable. We propose that the end-product of the C. uncialis gene cluster is derived from fusarubin pathway intermediates and may also be structurally similar to coronatoquinone or hemoventosin. Pestheic Acid. Pestalotiopsis f ici is an endophytic fungus that produces the plant growth regulator pestheic acid. The polyketide gene cluster for pestheic acid was recently identified in P. fici via gene deletion.25 The gene cluster consists of a PKS (PtaA), 11 tailoring genes (PtaB to PtaM), and three regulatory 739



Article

Figure 7. M. pilosus azaphilone gene cluster (KC148521) and C. uncialis uncharacterized gene cluster (MG777491). Genes with matching colors are genetically similar. Domain abbreviations: Starter acyltransferase (SAT), ketosynthase (KS), acyltransferase (AT), product template domain (PT), acetyl carrier protein (ACP), C-methyltransferase (MT), reductase (R). Gene abbreviations: Polyketide synthase (PKS), flavin adenine dinucleotidedependent oxidase (FAD oxidase), major facilitator superfamily transporter (MFS transporter), short-chain dehydrogenase/reductase (SDR).

Scheme 4. Experimentally Supported Pathway for Azaphilone Biosynthesis in M. pilosus and, Juxtaposed, the Biosynthetic Pathway That Is Proposed to be Encoded in C. uncialis

Table 4. BLAST Statistics and Proposed Function of the C. uncialis Genes That Are Genetically Similar to Genes of the Azaphilone Cluster of M. pilosus C. uncialis gene

accession no.

M. pilosus gene

accession no.



Nr-pks-3 3 5 6 7 8 9 11 13 14

AUW31097 AUW31098 AUW31100 AUW31101 AUW31102 AUW31103 AUW31104 AUW31106 AUW31108 AUW31109

MpPKS5 E 7 8 10 R1 F B D A

AGN71604 AGN71610 AGN71622 AGN71624 AGN71625 AGN71605 AGN71623 AGN71607 AGN71609 AGN71606

polyketide synthase oxidoreductase unknown unknown efflux transporter transcriptional regulator hydroxylase acyltransferase amino oxidase/esterase oxidoreductase

65/100 53/100 54/96 59/99 54/94 50/90 64/98 52/100 62/97 64/96

740



Article

cluster and an abridged pathway illustrating the roles of PtaA to PtaC in pestheic acid biosynthesis are shown in Figure 12 and Scheme 9. A pBLAST analysis suggested that the first three genes involved in the P. f ici pestheic acid pathway (PtaA to PtaC) are genetically similar to three genes in C. uncialis. The putative assignment of the C. uncialis PKS (Cu-nr-pks-13) as atrochrysone carboxylic acid synthase is strengthened by a robust similarity score of 71% (Table 7). This putative assignment is also supported by similar PKS domain architecture, including the absent terminal domain, and the proximity of a metallo-β-lactamase/hydrolase similar to that of P. f ici (Figure 12). The putative assignment is however weakened by the absence of convincing phylogenetic support from any of the three gene pairings (Figures S38−S40 in the Supporting Information). A metallo-β-lactamase/hydrolase, homologous to PtaB, is proposed to facilitate chain cleavage from CU-NR-PKS-13, yielding atrochrysone carboxylic acid (Scheme 9). The gene product of 1, similar to PtaC, is proposed to oxidize endocrocin to emodin. Two lichen genes are not genetically similar to P. fici genes. These include a gene of unknown function (2) and a gene ambiguously identified as an O-methyltransferase or a toxin regulator (5). A plausible biosynthetic pathway is proposed in Scheme 9. Rather than derivatizing emodin to some final product, this gene cluster could be devoted to producing emodin itself. Emodin is an intermediate in the biosynthesis of many lichen polyketides, and it is possible that the purpose of this gene cluster is to supply emodin as a starting substrate in other pathways. Several lichen metabolites possessing an emodin scaffold could be produced with a minimal number of tailoring steps from emodin. Examples include 5-chloroemodin, isolated from the lichen Nephroma laevigatum, 7-chloro-4-hydroxyemodin, isolated from Lasallia papulosa, 7-chloro-1-O-methyl-whydroxyemodin, isolated from the lichen Nephroma laevigatum, and skyrin, isolated from the lichen Rinodina peloleuca (Scheme 10).26 Mycophenolic Acid. Mycophenolic acid is a fungal metabolite consisting of a phthalide derivatized from 5methylorsellinic acid and a sesquiterpene. It was first discovered

Figure 8. Truncated phylogenetic trees illustrating degree of relationship between genes of C. uncialis and M. pilosus pertinent to azaphilone biosynthesis: (A) unknown function; (B) efflux transporter; (C) amino-oxidase; (D) oxidoreductase. Complete phylogenetic trees are shown (Figures S22−S31) in the Supporting Information. Bar signifies number of amino acid substitutions per site.

modifications by the remaining tailoring enzymes on endocrocin produces pestheic acid.25 The pestheic acid gene

Figure 9. A. clavatus cytochalasin gene cluster (accession no. DS027057) and C. uncialis uncharacterized gene cluster (accession no. MG777482). Genes with matching colors are genetically similar. Domain abbreviations: Enoylreductase (ER), ketosynthase (KS), acyltransferase (AT), dehydratase (DH), C-methyltransferase (MT), ketoreductase (KR), acetyl carrier protein (ACP), condensation (C) adenylation (A), peptidyl carrier protein (PCP), reductase (R). Gene abbreviations: Hybrid polyketide synthase-nonribosomal peptide synthetase (PKS-NRPS), major facilitator superfamily transporter (MFS transporter), flavin adenine dinucleotide-dependent oxidase (FAD oxidase). 741



Article

Scheme 5. Experimentally Supported Pathway for Cytochalasin Biosynthesis in A. clavatus and, Juxtaposed, the Biosynthetic Pathway That Is Proposed to be Encoded in C. uncialis

Table 5. BLAST Statistics and Proposed Function of the C. uncialis Genes That Are Genetically Similar to Genes of the Cytochalasin Cluster of A. clavatus C. uncialis gene

accession no.

A. clavatus gene

accession no.



1 Pks-nrps-3

AUW30907 AUW30912

CcsC CcsA

XP_001270547 XP_001270543

trans-enoyl reductase cytochalasin synthase

42/92 52/95

Scheme 6. Examples of Secondary Metabolites Possessing a Cyclized Phenylalanine Moiety from Talaromyces sp., Penicillium sp., and Aspergillus sp.

Figure 10. Truncated phylogenetic trees illustrating degree of relationship between genes of C. uncialis and A. clavatus pertinent to cytochalasin biosynthesis: (A) trans-enoyltreductase (B) PKS-NRPS. Complete phylogenetic trees (Figures S32, S33) are shown in the Supporting Information. Bar signifies number of amino acid substitutions per site.

for its effects against Bacillus anthracis and has since been noted for its antiviral, antifungal, and antitumor activities.27 The gene cluster devoted to mycophenolic acid biosynthesis was discovered in Penicillium brevicompactum.28 Deletion of the PKS (MpaC) conclusively demonstrated that biosynthesis of mycophenolic acid begins with 5-methylorsellinic acid as the first intermediate.28 MpaD is hypothesized to hydroxylate 5methylorsellinic acid because MpaD is the only candidate gene with similarity to cytochrome P-450 monooxygenases.28 Formation of the lactone moiety is likely performed by the gene product of MpaE because MpaD and MpaE form a heterodimer that facilitates lactonization in vivo.29 The

remaining tailoring enzymes are proposed to fixate the farnesyl pyrophosphate to the phthalide intermediate, methylate a hydroxy group, and perform an oxidative cleavage, yielding mycophenolic acid. As no terpene synthase is located in the mycophenolic acid gene cluster of P. brevicompactum, the farnesyl pyrophosphate is hypothesized to be produced by a nondedicated terpene synthase. The mycophenolic acid gene cluster and an abridged biosynthetic pathway are shown in Figure 13 and Scheme 11. 742



Article

Figure 11. F. f ujikuroi fusarubin gene cluster (accession no. HE613440) and C. uncialis uncharacterized gene cluster (accession no. MG777498). Genes with matching colors are genetically similar. Domain abbreviations: Starter acyltransferase (SAT), ketosynthase (KS), acyltransferase (AT), product template domain (PT), acetyl carrier protein (ACP), reductase (R), thioesterase (TE). Gene abbreviations: Polyketide synthase (PKS), short-chain dehydrogenase/reductase (SDR), major facilitator superfamily transporter (MFS transporter).

Scheme 7. Experimentally Supported Pathway for Fusarubin Biosynthesis in F. fujikuroi and, Juxtaposed, the Biosynthetic Pathway That Is Proposed to be Encoded in C. uncialis.

Table 6. BLAST Statistics and Proposed Function of the C. uncialis Genes That Are Genetically Similar to Genes of the Fusarubin Pathway of F. f ujikuroi C. uncialis gene

accession no.

F. f ujikuroi gene

accession no.



Nr-pks-10 3 8 12 13

AUW31210 AUW31211 AUW31216 AUW31220 AUW31221

Fsr1 Fsr2 Fsr3 Fsr5 Fsr1

CCE67070 CCE67071 CCE67072 CCE67074 CCE67070

6-O-demethylfusarubinaldehyde synthase 6-O-demethylfusarubinaldehyde O-methyltransferase multifunctional monooxygenase putative SDR putative trans-reductase domain

40/80 43/85 52/93 36/97 22/97

743



Article

Scheme 8. Plausible Biosynthetic Pathway Leading from the Fusarubin Intermediate 8-O-Methylanhydrofusarubinlactol to the Lichen Metabolites Haemoventosin and Coronataquinonea

a

The presence or absence of a methyl group (indicated as CH3*) is variable.

Figure 12. P. f ici pestheic acid gene cluster (accession no. KC145148) and C. uncialis uncharacterized gene cluster (accession no. MG777501). Genes with matching colors are genetically similar. Domain abbreviations: Starter acyltransferase (SAT), ketosynthase (KS), acyltransferase (AT), product template domain (PT), acetyl carrier protein (ACP). Gene abbreviations: Polyketide synthase (PKS), ATP-binding cassette transporters (ABC transporter).

Three similarity pairings were observed in C. uncialis, suggesting that functional homologues of the P. brevicompactum mycophenolic acid pathway could be encoded in the lichen. BLAST similarity scores and pairings are shown in Table 8 and Figure 13. The lichen PKS possesses identical domain architecture to MpaC (Figure 12) and is proposed to produce 5-methylorsellinic acid. C. uncialis genes, 2 and 7, respectively similar to MpaD and MpaE, are proposed to facilitate lactonization, yielding the last common intermediate, 5,7dihydroxy-4-methylphthalide. The lichen cluster also contains two GMC oxidoreductases (4,9), a cytochrome P-450 (5), two short-chain dehydrogenases/reductases (6,11), an O-methyltransferase (8), and two genes of unknown function (3, 10). A

role of these gene products may include derivatization of 5,7dihydroxy-4-methylphthalide to form a distinct final product. Phylogenetic analyses of the three gene pairings did not support genetically homologous relationships, weakening the predictive utility of the proposal (Figures S41−S43, Supporting Information).

■

CONCLUSION In summary, up to 48 secondary metabolite gene clusters were identified in the lichenizing fungus C. uncialis.9 Using characterized gene clusters in non-lichenizing fungi as reference templates we evaluated genetic similarity, genetic homology, and the emergence of logical chemical steps to propose 744



Article

Scheme 9. Experimentally Confirmed Pathway for Pestheic Acid Biosynthesis in P. fici and, Juxtaposed, the Biosynthetic Pathway That Is Proposed to be Encoded in C. uncialis

Table 7. BLAST Statistics and Proposed Function of the C. uncialis Genes That Are Genetically Similar to Genes of the Pestheic Acid Cluster of P. f ici C. uncialis gene

accession no.

P. f ici gene

accession no.



1 3 Nr-pks-13

AUW31243 AUW31245 AUW31246

PtaC PtaB PtaA

AGO59043 AGO59041 AGO59040

endocrocin anthrone monooxygenase atrochrysone carboxylic acid thioesterase atrochrysone carboxylic acid synthase

47/99 55/96 71/88

Scheme 10. Proposed Biosynthesis of Anthraquinone Lichen Natural Products from Emodin

745



Article

Figure 13. P. brevicompactum mycophenolic acid gene cluster (accession no. HQ731031) and C. uncialis unassigned gene cluster (accession no. MG777499). Color coding indicates similarity pairings. Domain abbreviations: Starter acyltransferase (SAT), ketosynthase (KS), acyltransferase (AT), product template domain (PT), acetyl carrier protein (ACP), C-methyltransferase (MT), thioesterase (TE). Gene abbreviations: Polyketide synthase (PKS), glucose-methanol-choline oxidoreductase (GMC oxidoreductase), short-chain dehydrogenase/reductase (SDR).

Scheme 11. Experimentally Confirmed Pathway for Mycophenolic Acid Biosynthesis in P. brevicompactum and, Juxtaposed, the Proposed Biosynthetic Pathway Encoded in C. uncialis

Table 8. BLAST Statistics and Proposed Function of the C. uncialis Genes That Are Genetically Similar to Genes of the Mycophenolic Acid Cluster of P. brevicompactum C. uncialis gene

accession no.

P. brevicom-pactum gene

accession no.

Nr-pks-11 2 7

AUW31224 AUW31225 AUW31230

MpaC MpaD MpaE

ADY00130 ADY00131 ADY00132

proposed function of C. uncialis gene 5-methylorsellinic acid synthase 5-methylorsellinic acid oxidase 4,6-dihydroxy-2-hydroxymethyl-3-methylbenzoic acid lactamase

biosynthetic functions for several lichen clusters. Through this approach we have identified three clusters that appear to be responsible for the biosynthesis of grayanic acid, patulin, and betaenones A−C. Some gene clusters appear only partially related to a reference cluster. In these cases, we have proposed that C. uncialis and the reference fungi share two or more common biochemical steps leading to a last common intermediate. Whereas these reference organisms derivatize the last common intermediate to produce terrein,8 azaphilone, cytochalasin, fusarubin, pestheic acid, and mycophenolic acid, C. uncialis is hypothesized to produce different and unknown final products. Lichen cultures are difficult to establish and grow. For these reasons our understanding of how lichens produce natural products is poorly understood. This homology mapping approach was adopted to provide putative assignments of function. The application of gene knockout or

similarity (%)/coverage (%) 41/98 60/67 50/97

heterologous expression will be required to confirm these putative assignments. We anticipate that the present work may help us move toward a deeper understanding of the rich biosynthetic potential of lichens.

■

EXPERIMENTAL SECTION

Identification of Genetically Similar Gene Clusters in Fungi. Collection and taxonomic conformation of C. uncialis, subculturing of axenic mycobiont culture, whole-genome sequencing, and annotation of secondary metabolite gene clusters have been recently reported.7,9 C. uncialis mycobiont gene clusters were uploaded to FungiSMASH (v. 4.0) (Antibiotics and Secondary Metabolites Analysis Shell),10 available at http://www.secondarymetabolites.org/. The ClusterBLAST module embedded within FungiSMASH (v. 4.0) provided cluster similarity statistics to characterized gene clusters. A pBLAST analysis was performed on each genetically similar gene to determine percent similarity scores.30 Lichen genes not identified as genetically 746



Article

similar to the fungal cluster constituents were manually analyzed using comparative pBLAST with each gene in the fungal cluster to confirm an absence of similarity. The reporting of “no significant similarity found” was interpreted as a negative result. Phylogenetic Analysis. Genes were analyzed via pBLAST30 and the 38 most genetically similar entries in GenBank were compiled for phylogenetic analysis. Out-group genes were selected from organisms known to be more distantly related than organisms of the in-group genes that also share the general function of in-group genes (e.g., Omethyltransferase). The hypothesized fungal homologue found via the ClusterBLAST module of FungiSMASH,10 and one out-group gene, was included for a total of 40 entries per phylogenetic tree. Multiple sequence alignments were performed with MEGA (v. 7.0),31 freely available for download at http://www.megasoftware.net/. Phylogenetic trees were constructed in MEGA (v. 7.0)31 using the neighborjoining method.32 Branch confidence was estimated using 1000 replicates of the interior branch test.33 Nodes with confidence of 70% or greater are shown on each tree. The Poisson correction method was used to estimate evolutionary distance34 and is reported on each tree in units of amino acid substitutions per site.

■

(5) (a) Gagunashvili, A. N.; Davídsson, S. P.; Jónsson, Z. O.; Andrésson, Ó .S. Mycol. Res. 2009, 113, 354−363. (b) Chooi, Y. H.; Stalker, D. M.; Davis, M. A.; Fujii, I.; Elix, J. A.; Louwhoff, S. H. J. J.; Lawrie, A. C. Mycol. Res. 2008, 112, 147−161. (c) Wang, Y.; Liu, B.; Zhang, X. Y.; Zhou, Q. M.; Zhang, T.; Li, J.; Yu, Y. F.; Zhang, X. L.; Hao, X. Y.; Wang, M.; Wang, L.; Wei, J. C. BMC Genomics 2014, 15, 34. (6) Armaleo, D.; Sun, X.; Culberson, C. Mycologia 2011, 103, 741− 754. (7) Abdel-Hameed, M.; Bertrand, R.; Piercey-Normore, M.; Sorensen, J. Fungal Biol. 2016, 120, 306−316. (8) Abdel-Hameed, M.; Bertrand, R.; Piercey-Normore, M. D.; Sorensen, J. J. Nat. Prod. 2016, 79, 1645−1650. (9) Bertrand, R. L.; Abdel-Hameed, M.; Sorensen, J. L. J. Nat. Prod. 2018, 10.1021/acs.jnatprod.7b00769 (10) Blin, K.; Wolf, T.; Chevrette, M. G.; Lu, X.; Schwalen, C. J.; Kautsar, S. A.; Suarez-Duran, H. G.; de Los Santos, E. L. C.; Kim, H. U.; Nave, M.; Dickschat, J. S.; Mitchell, D. A.; Shelest, E.; Breitling, R.; Takano, E.; Lee, S. Y.; Weber, T.; Medema, M. H. Nucleic Acids Res. 2017, 45, W36−W41. (11) (a) Schmitt, I.; Kautz, S.; Lumbsch, T. H. Mycol. Res. 2008, 112, 289−296. (b) Timsina, B. A.; Hausner, G.; Piercey-Normore, M. D. Fungal Biol. 2014, 118, 896−909. (c) Timsina, B. A.; StockerWörgötter, E.; Piercey-Normore, M. D. Botany 2012, 90, 1295−1307. (12) (a) Schmitt, I.; Lumbsch, H. T. PLoS One 2009, 4, e4437. (b) Beck, A.; Divakar, P. K.; Zhang, N.; Molina, M. C.; Struwe, L. Org. Divers. Evol. 2015, 15, 235−248. (c) Tunjić, M.; Korać, P. Period. Biol. 2013, 115, 321−329. (13) Zaehle, C.; Gressler, M.; Shelest, E.; Geib, E.; Hertweck, C.; Brock, M. Chem. Biol. 2014, 21, 719−731. (14) Ugai, T.; Minami, A.; Fujii, R.; Tanaka, M.; Oguri, H.; Gomi, K.; Oikawa, H. Chem. Commun. 2015, 51, 1878−1881. (15) (a) Birch, A. J.; Massy-Westropp, R. A.; Moye, C. J. Aust. J. Chem. 1955, 8, 539−544. (b) Lynen, F.; Tada, M. Angew. Chem. 1961, 73, 513−519. (16) (a) Puel, O.; Galtier, P.; Oswald, I. P. Toxins 2010, 2, 613−631. (b) Tannous, J.; El Khoury, R.; Snini, S. P.; Lippi, Y.; El Khoury, A.; Atoui, A.; Lteif, R.; Oswald, I. P.; Puel, O. Int. J. Food Microbiol. 2014, 189, 51−60. (17) (a) Puel, O.; Tadrist, S.; Delaforge, M.; Oswald, I. P.; Lebrihi, A. Int. J. Food Microbiol. 2007, 115, 131−139. (b) Snini, S. P.; Tadrist, S.; Laffitte, J.; Jamin, E. L.; Oswald, I. P.; Puel, O. Int. J. Food Microbiol. 2014, 171, 77−83. (c) Artigot, M. P.; Loiseau, N.; Laffitte, J.; MasReguieg, L.; Tadrist, S.; Oswald, I. P.; Puel, O. Microbiology 2009, 155, 1738−1747. (d) Dombrink-Kurtzman, M. A.; Engberg, A. E. Mycol. Res. 2006, 110, 1111−1118. (18) Osmanova, N.; Schultze, W.; Ayoub, N. Phytochem. Rev. 2010, 9, 315−342. (19) Balakrishnan, B.; Karki, S.; Chiu, S. H.; Kim, H. J.; Suh, J. W.; Nam, B.; Yoon, Y. M.; Chen, C. C.; Kwon, H. J. Appl. Microbiol. Biotechnol. 2013, 97, 6337−6345. (20) Scherlach, K.; Boettger, D.; Remme, N.; Hertweck, C. Nat. Prod. Rep. 2010, 27, 869−886. (21) Qiao, K.; Chooi, Y. H.; Tang, Y. Metab. Eng. 2011, 13, 723−732. (22) (a) Suzuki, S.; Hosoe, T.; Nozawa, K.; Kawai, K.; Yaguchi, T.; Udagawa, S. J. Nat. Prod. 2000, 63, 768−772. (b) Fukuda, T.; Tomoda, H.; Omura, S. J. Antibiot. 2005, 58, 315−321. (c) Alfatafta, A. A.; Dowd, P. F.; Gloer, J. B.; Wicklow, D. T. U.S. Patent 5,519,052, 1998. (d) Xu, W.; Cai, X.; Jung, M. E.; Tang, Y. J. Am. Chem. Soc. 2010, 132, 13604−13607. (23) Studt, L.; Wiemann, P.; Kleigrewe, K.; Humpf, H. U.; Tudzynski, B. Appl. Environ. Microbiol. 2012, 78, 4468−4480. (24) (a) Bruun, T.; Lamvik, A. Acta Chem. Scand. 1971, 25, 483−486. (b) Rycroft, D. S.; Connolly, J. D.; Huneck, S.; Himmelreich, U. Z. Naturforsch., B: J. Chem. Sci. 1995, 50, 1557−1563. (c) Ernst-Russell, M. A.; Elix, J. A.; Chai, C. L. L.; Rive, M. J.; Wardlaw, J. H. Aust. J. Chem. 2000, 53, 303−306. (25) Xu, X.; Liu, L.; Zhang, F.; Wang, W.; Li, J.; Guo, L.; Che, Y.; Liu, G. ChemBioChem 2014, 15, 284−292.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jnatprod.7b00770. Complete phylogenetic trees (Figures S1 to S43) (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel: +1-204-4749504. ORCID

John L. Sorensen: 0000-0002-1648-2310 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS J.L.S. acknowledges a Discovery Grant provided by the National Sciences and Engineering Research Council of Canada (NSERC). R.L.B. acknowledges a postgraduate scholarship provided by NSERC. All authors thank Michele PierceyNormore (Department of Biology, University of Manitoba) for providing the Cladonia uncialis specimen and for valuable technical advice. All authors also thank two anonymous reviewers for constructive comments during manuscript review.

■

REFERENCES

(1) (a) Shrestha, G.; St. Clair, L. L. Phytochem. Rev. 2013, 12, 229− 224. (b) Zambare, V. P.; Christopher, L. P. Pharm. Biol. 2012, 50, 778−798. (c) Yousuf, S.; Choudhary, I.; Rahman, A. Stud. Nat. Prod. Chem. 2014, 43, 223−259. (d) Calcott, M. J.; Ackerley, D. F.; Knight, A.; Keyzers, R. A.; Owen, J. G.; Chem. Soc. Rev. 2018 10.1039/ C7CS00431A. (2) (a) Schmidt-Dannert, C. Adv. Biochem. Eng./Biotechnol. 2014, 148, 19−61. (b) Boettger, D.; Hertweck, C. ChemBioChem 2013, 14, 28−42. (c) Chooi, Y. H.; Tang, Y. J. Org. Chem. 2012, 77, 9933−9953. (3) (a) Yaegashi, J.; Oakley, B. R.; Wang, C. C. C. J. Ind. Microbiol. Biotechnol. 2014, 41, 433−442. (b) Wiemann, P.; Keller, N. P. J. Ind. Microbiol. Biotechnol. 2014, 41, 301−313. (c) Anyaogu, D. C.; Mortensen, U. H. Front. Microbiol. 2015, 6, 77. (4) Grube, M.; Berg, G.; Andrésson, Ó .S.; Vilhelmsson, O.; Dyer, P. S.; Miao, V. P. W. In The Ecological Genomics of Fungi; Martin, F., Ed.; John Wiley & Sons: New York, 2014; pp 191−212. 747



Article

(26) (a) Cohen, P. A.; Towers, G. H. N. Phytochemistry 1996, 42, 1325−1329. (b) Matzer, M.; Mayrhofer, H.; Elix, J. A. N. Z. J. Bot. 1998, 32, 175−188. (c) Bohman, G. Acta Chem. Scand. 1969, 23, 2241−2244. (27) Bentley, R. Chem. Rev. 2000, 100, 3801−3826. (28) Regueira, T. B.; Kildegaard, K. R.; Hansen, B. G.; Mortensen, U. H.; Hertweck, C.; Nielsen, J. Appl. Environ. Microbiol. 2011, 77, 3035− 3043. (29) Hansen, B. G.; Mnich, E.; Nielsen, K. F.; Nielsen, J. B.; Nielsen, M. T.; Mortensen, U. H.; Larsen, T. O.; Patil, K. R. Appl. Environ. Microbiol. 2012, 78, 4908−4913. (30) Altschul, S.; Gish, W.; Miller, W.; Myers, E. W.; Lipman, D. J. J. Mol. Biol. 1990, 215, 403−410. (31) Kumar, S.; Stecher, G.; Tamura, K. Mol. Biol. Evol. 2016, 33, 1870−1874. (32) Saitou, N.; Nei, M. Mol. Biol. Evol. 1987, 4, 406−425. (33) (a) Dopazo, J. J. Mol. Evol. 1994, 38, 300−304. (b) Rzhetsky, A.; Nei, M. Mol. Biol. Evol. 1992, 9, 945−967. (34) Zuckerkandl, E.; Pauling, L. In Evolving Genes and Proteins; Bryson, V.; Vogel, H. J., Eds.; Academic Press: New York, 1965; pp 97−166.

748


Lichen Biosynthetic Gene Clusters Part II: Homology Mapping

Recommend Documents