Benzyl Alcohol/Salicylaldehyde-Type Polyketide Metabolites of Fungi: Sources, Biosynthesis, Biological Activities, and Synthesis

Marine microorganisms are an important source of natural polyketides, which have become a significant reservoir of lead structures for drug design due to their diverse biological activities. In this review, we provide a summary of the resources, structures, biological activities, and proposed biosynthetic pathways of the benzyl alcohol/salicylaldehyde-type polyketides. In addition, the total syntheses of these secondary metabolites from their discoveries to the present day are presented. This review could be helpful for researchers in the total synthesis of complex natural products and the use of polyketide bioactive molecules for pharmacological purposes and applications in medicinal chemistry.


Introduction
Polyketides represent a vast group of secondary metabolites that are produced by certain living organisms to provide sustaining advantages for them. In terms of their chemical structures, Baerson and Rimando [1] described polyketide as a natural compound consisting of carbonyl and methylene groups alternating with each other. Polyketides and their derivatives represent an important group of compounds with remarkable diversity in both structure and function. Even though most polyketides are products of microbial origin (fungi and bacteria), these ubiquitous metabolites can be produced by a range of other organisms, such as plants, insects, lichens, mollusks, algae, sponges, crinoids, and dinoflagellates [1,2]. Polyketides are highly biologically active and multifaceted organic compounds. The structural diversity of polyketides allows them to have specific properties, such as pharmacological properties and biological activities. According to Dechert-Schmitt et al. [3], there is up to a five times higher probability of discovering a new polyketide-based drug compared to other groups of naturally occurring substances. Structurally smaller molecules of polyketide origin are among the top 20% of the best-selling drugs in the world [3]. This review provides a comprehensive summary of the resources, structures, biological activities, biosynthesis, and laboratory synthesis of benzyl alcohol/salicylaldehyde-type polyketide metabolites. The key criterion for the selection of the described compounds was the benzyl alcohol/salicylaldehyde structural pattern in the isolated secondary metabolites across the environment. As far as we can tell, known metabolites with the described structural characteristics are exclusively produced by fungi. A particularly significant source is fungi of marine origin. The metabolites obtained from terrestrial sources are included in order to cover all natural substances of interest produced by these species. This overview could be helpful for organic chemists in the search for new lead structures in drug design.

Biosynthesis
The authors [11] suggested several plausible biosynthetic relationships between the isolated polyketide derivatives.
Malmstrøm et al. [11] suggested that diene 6 could act as a precursor to varitriol (1) and that its biosynthesis probably proceeds via varioxirane (2). The enzymatic epoxidation of the C-5-C-6 double bond of 6 could provide epoxide 2. Subsequent enzymatically catalyzed intramolecular SN2 oxirane ring opening at C-6 by the hydroxyl group in the C-3 position in varioxirane (2) could easily yield varitriol (1) (Scheme 1).
Malmstrøm et al. [11] suggested that diene 6 could act as a precursor to varitriol (1) and that its biosynthesis probably proceeds via varioxirane (2). The enzymatic epoxidation of the C-5-C-6 double bond of 6 could provide epoxide 2. Subsequent enzymatically catalyzed intramolecular S N 2 oxirane ring opening at C-6 by the hydroxyl group in the C-3 position in varioxirane (2) could easily yield varitriol (1) (Scheme 1). 3

Bioactivity
Metabolites isolated from the Emericella variecolor M75-2 fungus were evaluated for antimicrobial activity against several Gram-positive and Gram-negative bacteria. Among the tested compounds, varixanthone (3) showed the best results. The MIC values for compound 3 were significantly lower than those for xanthone derivatives 4 and 5 [11].

Synthesis
The diversity of the biological activities of these relatively simple and structurally small secondary polyketide metabolites has attracted the interest of synthetic chemists. Among this group of natural compounds, varitriol (1) aroused the greatest interest, mainly for its biological properties. Therefore, most of the known syntheses are devoted to this metabolite.

Varitriol
Since the first synthesis [15] of the unnatural enantiomer (−)-1, accomplished by Jennings in 2006, about 22 total syntheses of varitriol have been reported in the literature. The review by Majik et al. [16] described different synthetic approaches towards 1 and its analogues in detail, covering the period up to 2013. The first two synthetic approaches commencing from D-ribose provided unnatural varitriol ((−)-1) (Scheme 2) [15]. Clemens and Jennings [15] achieved the total synthesis of (−)-1 using cross-metathesis to link the carbohydrate A and aromatic moieties of the molecule. Taylor's group reported a flexible approach to (−)-1, applying the HWE/conjugate addition/Ramberg-Bäcklund rearrangement sequence via intermediate B [17]. The aromatic part of (−)-1 was obtained, either from 2,6-dihydroxybenzoic acid or methyl 2-methoxy-5-methylbenzoate. The first total synthesis of naturally occurring varitriol ((+)-1) utilizing methyl α-D-mannopyranoside was accomplished by Shaw and Kumar [18] (Scheme 2). The approach was based on an alkene metathesis of the corresponding styrene and tetrahydrofuran subunits. The key furanoside alkene C was obtained from methyl α-D-mannopyranoside. Scheme 1. Biosynthesis of varitriol (1) by Malstrøm et al. [11].

Bioactivity
Metabolites isolated from the Emericella variecolor M75-2 fungus were evaluated for antimicrobial activity against several Gram-positive and Gram-negative bacteria. Among the tested compounds, varixanthone (3) showed the best results. The MIC values for compound 3 were significantly lower than those for xanthone derivatives 4 and 5 [11].

Synthesis
The diversity of the biological activities of these relatively simple and structurally small secondary polyketide metabolites has attracted the interest of synthetic chemists. Among this group of natural compounds, varitriol (1) aroused the greatest interest, mainly for its biological properties. Therefore, most of the known syntheses are devoted to this metabolite.

Varitriol
Since the first synthesis [15] of the unnatural enantiomer (−)-1, accomplished by Jennings in 2006, about 22 total syntheses of varitriol have been reported in the literature. The review by Majik et al. [16] described different synthetic approaches towards 1 and its analogues in detail, covering the period up to 2013. The first two synthetic approaches commencing from D-ribose provided unnatural varitriol ((−)-1) (Scheme 2) [15]. Clemens and Jennings [15] achieved the total synthesis of (−)-1 using cross-metathesis to link the carbohydrate A and aromatic moieties of the molecule. Taylor's group reported a flexible approach to (−)-1, applying the HWE/conjugate addition/Ramberg-Bäcklund rearrangement sequence via intermediate B [17]. The aromatic part of (−)-1 was obtained, either from 2,6-dihydroxybenzoic acid or methyl 2-methoxy-5-methylbenzoate. The first total synthesis of naturally occurring varitriol ((+)-1) utilizing methyl α-D-mannopyranoside was accomplished by Shaw and Kumar [18] (Scheme 2). The approach was based on an alkene metathesis of the corresponding styrene and tetrahydrofuran subunits. The key furanoside alkene C was obtained from methyl α-D-mannopyranoside.
After the definitive confirmation of the absolute configuration of natural varitriol (1), several other total syntheses were developed. These synthetic approaches usually involve multistep sequences and can be differentiated by the means of the disconnection of individual bonds in the varitriol (1) backbone (Scheme 3). The majority of the known strategies involve alkene metathesis for coupling the furanoside part and the corresponding styrene derivative (path a). The most common starting material for the synthesis of a furanosidecontaining vinyl moiety (intermediate A/B) is D-ribose [19][20][21][22][23][24][25][26] Alternatively, an olefinic intermediate C can be obtained from ethyl (S)-lactate (7) [20] via a synthetic sequence involving epoxidation, cyclisation, and dihydroxylation reactions. Njardanson and coworkers [27] have described the syntheses of the carbohydrate part of (+)-1 from 2,4-diene-heptanol (8) or but-2-enal (9) using the Cu-catalyzed vinyl oxirane ring expansion reaction. Another approach exploits the olefination of an aldehyde (path b). The Julia-Kocieński olefination was applied for the coupling of the furan-derivative-bearing sulfonyl group and the substituted benzaldehyde [28][29][30][31], which was easily available from 2,3-dimethylanisole. The required furanoside sulfone was prepared from dimethyl L-tartrate by employing the bicyclization strategy developed for the construction of 2,3-trans-substituted tetrahydrofurans [28,29]. The authors provided X-ray conformation of the absolute configuration of the natural (+)-varitriol (1) for the first time. The most efficient synthesis of natural varitriol (1) so far has been achieved from γ-D-ribonolactone and dimethylanisole (eight steps and 41% overall yield) [30]. The key steps of the route include a highly stereoselective introduction of the methyl group at C-1 and the installation of the side chain with the aromatic moiety using Julia-Kocieński olefination at C-5 of the starting carbon skeleton. This strategy was applied in the synthesis of novel varitriol analogues and a subsequent SAR study [31]. Krause et al. [32] reported the synthesis of (+)-1 and its analogues. The authors utilized the Horner-Wittig-Emmons (HWE) condensation of the aromatic phosphonate and the chiral furanoside aldehyde, which was obtained from hept-2-en-4-ynol (10) using asymmetric synthesis. The linkage of the aromatic and sugar moieties was also achieved by a Heck C-C cross-coupling reaction between an aromatic triflate and an olefinic furanoside prepared from D-mannitol [33] (path c). Liu and coworkers [23] have developed a short synthesis of (+)-1 by applying the Sonogashira reaction. The cross-coupling allowed the connection of the alkynic sugar moiety, obtained from D-ribose, and the aromatic triflate (path d). He and Qin [24] reported the formal synthesis of (+)-1 using the Zn-catalyzed cross-coupling of the corresponding furanoside acetate with a benzene-derivative-bearing terminal alkynyl group (path d). Finally, varioxirane (2) was successfully converted into varitriol ((+)-1) via the intramolecular ring opening of the epoxide (path e) [34], supporting the proposed biosynthetic pathway [11]. After the definitive confirmation of the absolute configuration of natural varitriol (1), several other total syntheses were developed. These synthetic approaches usually involve multistep sequences and can be differentiated by the means of the disconnection of individual bonds in the varitriol (1) backbone (Scheme 3). The majority of the known strategies involve alkene metathesis for coupling the furanoside part and the corresponding styrene derivative (path a). The most common starting material for the synthesis of a furanoside-containing vinyl moiety (intermediate A/B) is D-ribose [19][20][21][22][23][24][25][26] Alternatively, an olefinic intermediate C can be obtained from ethyl (S)-lactate (7) [20] via a synthetic sequence involving epoxidation, cyclisation, and dihydroxylation reactions. Njardanson and coworkers [27] have described the syntheses of the carbohydrate part of Scheme 2. Retrosynthetic analyses of varitriol ((−)-1) and ((+)-1) [15,17,18].

Varioxirane and Andytriol
Sudhakar and Raghvaiah [34] developed the first synthesis of varioxirane (2) (Scheme 4). The synthesis commenced with the Sharpless kinetic resolution of α-hydroxy ester 11, which afforded the epoxy alcohol 12 with the configuration corresponding to the target compound 2. The key intermediate for Horner-Wittig-Emmons olefination, the phosphonate 14, was obtained in two steps. A silylation of the secondary hydroxyl in 12 furnished 13, which was treated with lithiated methyl phosphonate to afford 14. The coupling of the phosphonate 14 with the aldehyde 15, prepared from dimethylanisol [21], afforded ketone 16. The desilylation of 16 with the subsequent reduction of ketone 17 using LiEt3BH in ether led to the desired anti-diol 18 with high diastereoselectivity (dr 32:1). The final deacetylation of 18 provided the target varioxirane (2) in seven reaction steps with a total yield of 15%.

Varioxirane and Andytriol
Sudhakar and Raghvaiah [34] developed the first synthesis of varioxirane (2) (Scheme 4). The synthesis commenced with the Sharpless kinetic resolution of α-hydroxy ester 11, which afforded the epoxy alcohol 12 with the configuration corresponding to the target compound 2. The key intermediate for Horner-Wittig-Emmons olefination, the phosphonate 14, was obtained in two steps. A silylation of the secondary hydroxyl in 12 furnished 13, which was treated with lithiated methyl phosphonate to afford 14. The coupling of the phosphonate 14 with the aldehyde 15, prepared from dimethylanisol [21], afforded ketone 16. The desilylation of 16 with the subsequent reduction of ketone 17 using LiEt 3 BH in ether led to the desired anti-diol 18 with high diastereoselectivity (dr 32:1). The final deacetylation of 18 provided the target varioxirane (2) in seven reaction steps with a total yield of 15%.

Biosynthesis
The authors described varioxiranols F (44) and G (45) as products derived from the coupling of tajixanthone (48) with varioxiranol A (39) and andytriol (6), respectively (Scheme 8). The epoxide ring of tajixanthone (48) is formed by the oxidation of the carboncarbon double bond at the prenyl group of 4, which is mediated by cyclooxygenase [40]. The oxirane ring of 48 is then opened at the C-16 position by a nucleophilic attack of benzyl alcohol 6 or 39 to form varioxiranols F (44) and G (45). According to the authors, the enzymatic mediation of the epoxide ring opening is necessary to achieve the formation of ethers 44 and 45 with such high regioselectivity [38].
coupling of tajixanthone (48) with varioxiranol A (39) and andytriol (6), respectively (Scheme 8). The epoxide ring of tajixanthone (48) is formed by the oxidation of the carbon-carbon double bond at the prenyl group of 4, which is mediated by cyclooxygenase [40]. The oxirane ring of 48 is then opened at the C-16 position by a nucleophilic attack of benzyl alcohol 6 or 39 to form varioxiranols F (44) and G (45). According to the authors, the enzymatic mediation of the epoxide ring opening is necessary to achieve the formation of ethers 44 and 45 with such high regioselectivity [38].

Bioactivity
Isolated metabolites 1, 2, 6, and 39-47 were tested for lipid-lowering effects against oleic acid (OA)-elicited lipid accumulation in HepG2 liver cells. Varioxiranol A (39), andytriol (6), and preshamixanthone (47) exhibited inhibitory effects at a dose of 10 μM, and they did not show any toxicity up to 100 μM towards the tested cells. Preshamixantone (47), as the metabolite with the highest activity levels, has shown an inhibitory effect comparable to the positive control simvastatin. In addition, a further evaluation revealed that compound 47 exhibited inhibition against intracellular triglyceride (TG) levels and dramatically reduced total cholesterol (TC) at a dose of 10 μM. Varioxiranol A (39) reduced TG levels with weak effects on TC, whereas andytriol (6) showed exactly opposite results, as it induced a reduction in TC, with a weak effect on TG levels. These surprising findings suggest there is a distinct mechanism for the anti-hyperlipidemic effects, since 6 differs from 39 only by having an extra double bond.

Synthesis of Varioxiranol A and 4-Epi-Varioxiranol A
Lásiková and coworkers [41] have developed the only total synthesis of varioxiranol A (39) to date (Scheme 9). This chiral pool approach employed Julia-Kocieński olefination to connect the aromatic sulfone and the carbohydrate fragment. The synthesis of alkyl fragment 54 commenced from 2,3-O-isopropylidene D-glyceraldehyde (49), employing the Grignard addition, TBS protection, and acetonide deprotection to give a diastereomeric mixture of partially protected L-erythro/D-threo triols 52 in the ratio of 67:33. The aldehyde 54 was prepared using a selective protection-deprotection sequence involving a selective tritylation of the primary alcohol in 52, followed by the acetylation of the secondary hydroxyl group and smooth tritylether hydrolysis using formic acid in ether furnishing alcohol 53. The Swern oxidation of alcohol provided the desired aldehyde 54, which was subjected to Julia-Kocieński coupling with sulfone 23. The resulting E-alkenes 55/epi-55 were separated by MPLC and the final removal of all protecting

Bioactivity
Isolated metabolites 1, 2, 6, and 39-47 were tested for lipid-lowering effects against oleic acid (OA)-elicited lipid accumulation in HepG2 liver cells. Varioxiranol A (39), andytriol (6), and preshamixanthone (47) exhibited inhibitory effects at a dose of 10 µM, and they did not show any toxicity up to 100 µM towards the tested cells. Preshamixantone (47), as the metabolite with the highest activity levels, has shown an inhibitory effect comparable to the positive control simvastatin. In addition, a further evaluation revealed that compound 47 exhibited inhibition against intracellular triglyceride (TG) levels and dramatically reduced total cholesterol (TC) at a dose of 10 µM. Varioxiranol A (39) reduced TG levels with weak effects on TC, whereas andytriol (6) showed exactly opposite results, as it induced a reduction in TC, with a weak effect on TG levels. These surprising findings suggest there is a distinct mechanism for the anti-hyperlipidemic effects, since 6 differs from 39 only by having an extra double bond.

Synthesis of Varioxiranol A and 4-Epi-Varioxiranol A
Lásiková and coworkers [41] have developed the only total synthesis of varioxiranol A (39) to date (Scheme 9). This chiral pool approach employed Julia-Kocieński olefination to connect the aromatic sulfone and the carbohydrate fragment. The synthesis of alkyl fragment 54 commenced from 2,3-O-isopropylidene D-glyceraldehyde (49), employing the Grignard addition, TBS protection, and acetonide deprotection to give a diastereomeric mixture of partially protected L-erythro/D-threo triols 52 in the ratio of 67:33. The aldehyde 54 was prepared using a selective protection-deprotection sequence involving a selective tritylation of the primary alcohol in 52, followed by the acetylation of the secondary hydroxyl group and smooth tritylether hydrolysis using formic acid in ether furnishing alcohol 53. The Swern oxidation of alcohol provided the desired aldehyde 54, which was subjected to Julia-Kocieński coupling with sulfone 23. The resulting E-alkenes 55/epi-55 were separated by MPLC and the final removal of all protecting groups, which were run in parallel with the pure diastereomers 55 and epi-55. The first basic hydrolysis of acetyl groups was followed by the treatment of the mixture with TBAF in THF, and finally flash chromatography purification furnished the target compounds 39 and epi-39 in 92% and 57% yields, respectively. The absolute stereochemistry of both natural varioxiranol A (39) and its 4-epimer epi-39 was confirmed by a single-crystal X-ray analysis.
groups, which were run in parallel with the pure diastereomers 55 and epi-55. The first basic hydrolysis of acetyl groups was followed by the treatment of the mixture with TBAF in THF, and finally flash chromatography purification furnished the target compounds 39 and epi-39 in 92% and 57% yields, respectively. The absolute stereochemistry of both natural varioxiranol A (39) and its 4-epimer epi-39 was confirmed by a single-crystal X-ray analysis.

Isolation
New strain of polyketide derivatives with benzyl alcohol scaffolds 56, 57, and 58, along with previously established analogue 2 [11,42] was obtained by the bioassay-guided isolation of metabolites from cultures of the plant-derived fungus Emericella sp. TJ29 ( Figure 3) [42,43]. During the structural evaluation of 56-58, a benzofuran motif appeared to be reoccurring, and the relative configuration of the stereogenic centers was further established based on spectroscopic experiments (HMBC, COSY, TOCSY, and NOESY) and an X-ray analysis [17,18]. The stereochemistry of the alkyl chain of polyketides 56-58 corresponds to the configuration of andytriol (6

Isolation
New strain of polyketide derivatives with benzyl alcohol scaffolds 56, 57, and 58, along with previously established analogue 2 [11,42] was obtained by the bioassay-guided isolation of metabolites from cultures of the plant-derived fungus Emericella sp. TJ29 ( Figure 3) [42,43]. During the structural evaluation of 56-58, a benzofuran motif appeared to be reoccurring, and the relative configuration of the stereogenic centers was further established based on spectroscopic experiments (HMBC, COSY, TOCSY, and NOESY) and an X-ray analysis [17,18]. The stereochemistry of the alkyl chain of polyketides 56-58 corresponds to the configuration of andytriol (6).
Mar. Drugs 2023, 21, x FOR PEER REVIEW 10 of 28 groups, which were run in parallel with the pure diastereomers 55 and epi-55. The first basic hydrolysis of acetyl groups was followed by the treatment of the mixture with TBAF in THF, and finally flash chromatography purification furnished the target compounds 39 and epi-39 in 92% and 57% yields, respectively. The absolute stereochemistry of both natural varioxiranol A (39) and its 4-epimer epi-39 was confirmed by a single-crystal X-ray analysis.

Isolation
New strain of polyketide derivatives with benzyl alcohol scaffolds 56, 57, and 58, along with previously established analogue 2 [11,42] was obtained by the bioassay-guided isolation of metabolites from cultures of the plant-derived fungus Emericella sp. TJ29 ( Figure 3) [42,43]. During the structural evaluation of 56-58, a benzofuran motif appeared to be reoccurring, and the relative configuration of the stereogenic centers was further established based on spectroscopic experiments (HMBC, COSY, TOCSY, and NOESY) and an X-ray analysis [17,18]. The stereochemistry of the alkyl chain of polyketides 56-58 corresponds to the configuration of andytriol (6).

Bioactivity
Wang, Xue, Zhang, and coworkers [43] performed an extensive biological screening against five antibiotic-resistant bacteria to try to identify promising novel antibiotics. Among the isolated compounds, varioxiranediol A (56) demonstrated the highest antibac-terial activity against ESBL-producing E. coli and P. aeruginosa, with MICs of 4 µg/mL for both strains.

Bioactivity
Wang, Xue, Zhang, and coworkers [43] performed an extensive biological screening against five antibiotic-resistant bacteria to try to identify promising novel antibiotics. Among the isolated compounds, varioxiranediol A (56) demonstrated the highest antibacterial activity against ESBL-producing E. coli and P. aeruginosa, with MICs of 4 μg/mL for both strains.

Biosynthesis
In 2011, Tanaka et al. [48] synthetized heptatrienylsalicylaldehyde (73) in a deuterium-labeled form as a plausible intermediate for the biosynthesis of metabolites 59-66. Based on a biosynthetic study, the authors proposed a biosynthetic pathway of polyketides derivatives 59-66 (Scheme 10). Their findings showed that the biosynthesis of 73 involves the condensation of acetyl-CoA with malonyl-CoA and the subsequent aromatization of polyketide 72. The resulting triene 73 appears to be biosynthetic precursor for salicylaldehyde-type metabolites [49]. An epoxidation of the C-5-C-6 double bond of 73, followed by the hydrolysis of the oxirane ring in 76a and 76b, gives pyriculariol (64). A reduction of the aromatic aldehyde 64 affords dihydropyriculariol (65). Similarly, a fermentation by a shaking culture allows the epoxidation of the 3,4 double bond of 73, which leads to the formation of epoxides 74a and/or 74b. Epoxy rearrangement in 74a

Biosynthesis
In 2011, Tanaka et al. [48] synthetized heptatrienylsalicylaldehyde (73) in a deuteriumlabeled form as a plausible intermediate for the biosynthesis of metabolites 59-66. Based on a biosynthetic study, the authors proposed a biosynthetic pathway of polyketides derivatives 59-66 (Scheme 10). Their findings showed that the biosynthesis of 73 involves the condensation of acetyl-CoA with malonyl-CoA and the subsequent aromatization of polyketide 72. The resulting triene 73 appears to be biosynthetic precursor for salicylaldehyde-type metabolites [49]. An epoxidation of the C-5-C-6 double bond of 73, followed by the hydrolysis of the oxirane ring in 76a and 76b, gives pyriculariol (64). A reduction of the aromatic aldehyde 64 affords dihydropyriculariol (65). Similarly, a fermentation by a shaking culture allows the epoxidation of the 3,4 double bond of 73, which leads to the formation of epoxides 74a and/or 74b. Epoxy rearrangement in 74a and/or 74b (routes b and b ), followed by the regiospecific reduction of the aliphatic aldehyde 75, provides pyricuol (59). The hydrolysis of epoxides 74a and 74b provides pyriculol (60) with the 3R,4S absolute configuration (routes a and a ). The authors claimed that the biosynthesis of 60 could be performed by asymmetric epoxidation followed by regioselective hydrolysis and/or nonstereoselective oxidation followed by stereospecific hydration. The oxidation and reduction of 60 provide pyriculone (66) and dihydropyriculol (62), respectively. Interestingly, the only structural difference between dihydropyriculol (62) and the secondary metabolite of Emericella species, andytriol (6), is the presence of a methyl group attached to the phenolic OH.
dihydropyriculol (62) and the secondary metabolite of Emericella species, andytriol (6), is the presence of a methyl group attached to the phenolic OH.
Recently, the gene encoding a polyketide synthase, MoPKS19, in M. oryzea was identified to be responsible for the biosynthesis of pyriculol and related compounds [47] MoPKS19 resides in a gene cluster that contains additional biosynthetic enzymes found in fungal PKS pathways, including oxidoreductases and cupin-domain-containing proteins. The roles of most of the other enzymes in this cluster are unknown. Recently, the gene encoding a polyketide synthase, MoPKS19, in M. oryzea was identified to be responsible for the biosynthesis of pyriculol and related compounds [47] MoPKS19 resides in a gene cluster that contains additional biosynthetic enzymes found in fungal PKS pathways, including oxidoreductases and cupin-domain-containing proteins. The roles of most of the other enzymes in this cluster are unknown.

Bioactivity
Magnaporthe species are a huge threat for rice plants. Their metabolites have strong phytotoxic properties and cause the inhibition of growth and leaf damage at levels as low as 60-500 ppm [50]. An investigation of the metabolites showed that pyriculol (60) is responsible for the disease. The results suggested that salicylaldehyde functionality brings about the phytotoxicity since benzyl alcohols 62 and 63 were described as nonphytotoxic [51].

Synthesis
Pyriculol Among the phytotoxic metabolites [39,[44][45][46][47] produced by Pyricularia oryzae, three were synthesized, and their definitive structures as well as the absolute configurations were determined [52][53][54][55][56][57][58]. First, pyriculol (60) [39], a compound bearing a salicylaldehyde moiety with an unsaturated side chain containing two stereogenic centers, was prepared [52,53]. The authors developed a convergent approach based on the preparation of aromatic and aliphatic fragments and their subsequent condensation (Scheme 11). The aromatic synthon was prepared from 2,3-dimethylphenol (77) in accordance with the previously reported protocol [59]. The phenol 77 was acetylated and successively treated with N-bromosuccinimide, sodium acetate, potassium hydroxide or lithium aluminum hydride, and 2,2-dimethoxypropane or acetone to give 78. The resulting hydroxyacetonide 78 was converted to a corresponding chloride 79 using triphenylphosphine-carbon tetrachlochloride, which was transformed into phosphonium salt 80. The synthesis of the side chain moiety commenced with 2,3-O-cyclohexylidene-D-glyceraldehyde (81), which afforded upon prop-1-ynyllithium addition, followed by an LAH reduction of the carbon-carbon triple bond, a separable diastereomeric mixture of E-allylic alcohols 82 (erythro/threo, 2.1:1). In addition, the threo-diastereomer was transformed into erythro-triol using a Mitsunobu inversion. The erythro product 82 was converted to the carbonate 84 through the following sequence: the removal of a cyclohexylidene group from 82 with 70% acetic acid, the tritylation of a primary hydroxyl group, the cyclic carbonation of a residual 2,3-diol with dimethyl carbonate and sodium hydride, and finally the acidic hydrolysis of the trityl protecting group, providing the alcohol 83. The Swern oxidation of 83 gave aldehyde 84, which was used without further purification in Wittig olefination with 80. A mixture of geometrical isomers 85 (E/Z, 4:1) was obtained in a 32% yield over two steps. The desired E-isomer was obtained by thin-layer chromatography. The acetonide protection on the aromatic ring was removed with p-TsOH upon refluxing in an aqueous THF solution to give a hydroxymethyl moiety. Finally, the oxidation of a primary alcohol with activated manganese dioxide, followed by the hydrolysis of a carbonate protecting group, gave pyriculol (60).

Pyricuol
In 2003, Kiyota and coworkers [54] developed a synthesis of pyricuol (59), which allowed the confirmation of its structure. However, the absolute configuration remained undetermined. Even though its following syntheses, commencing from (R)-lactate [55] and/or L-serine [56] provided the unnatural enantiomer (S)-59, they allowed the confirmation of the absolute stereochemistry of the natural pyricuol (59). Based on these results, the authors [57] successfully carried out the first synthesis of natural pyricuol (59)

Pyricuol
In 2003, Kiyota and coworkers [54] developed a synthesis of pyricuol (59), which allowed the confirmation of its structure. However, the absolute configuration remained undetermined. Even though its following syntheses, commencing from (R)-lactate [55] and/or L-serine [56] provided the unnatural enantiomer (S)-59, they allowed the confirmation of the absolute stereochemistry of the natural pyricuol (59). Based on these results, the authors [57] successfully carried out the first synthesis of natural pyricuol (59) employing Stille coupling and [2,3]-Wittig rearrangement reactions as the key steps (Scheme 12). Aromatic fragment 88, bearing a tributyltin group, was obtained from the known aldehyde 86 [60]. The treatment of 86 with Ohira-Bestmann [61,62] reagent, followed by hydrostannylation of 87 using Bu 3 SnH/CuCN/BuLi in THF, gave styryltin compound 88. The partner for the C-C cross-coupling reaction, Z-vinyl iodide 90, was prepared from the corresponding aldehyde 89 [63] via a Wittig olefination reaction with (Ph 3 P + CH 2 I)I − in the presence of NaHMDS. An undesired E-isomer was removed using column chromatography. The Stille coupling of 88 with 90 afforded aromatic diene 91. The TBAF removal of the silyl protecting group provided dienol 92 in a 20% yield over three steps. Next, the free hydroxyl group was etherified with Bu 3 SnCH 2 I, and the resulting stannylmethyl ether was treated with nBuLi, giving a [2,3]-Wittig-rearranged product 93. The acidic hydrolysis of 93, yielding 94, followed by the selective oxidation of the benzylic hydroxy group with MnO 2 , furnished natural (R)-pyricuol (59

Pyriculariol
In 2009, Kiyota and coworkers [58] reported the first total synthesis of (-)-pyriculariol (64). The synthesis of the alkenyl chain commenced with a Ferrier reaction of L-rhamnal diacetate (95) [64] providing aliphatic 96, which upon the acetylation of the free hydroxyl group provided aldehyde 97 (Scheme 13). A Corey-Fuchs reaction using PPh3/CBr4, followed by the basic hydrolysis of both acetyl protecting groups in 98, afforded diol 99. The treatment of 99 with nBuLi gave enynediol 100, which was converted into dienyl stannane 101 by radical-promoted hydrostannylation with Bu3SnH. Finally, two fragments (stannyl compound 101 and aromatic triflate 102 [64]) were coupled using the abovementioned procedure (Stille coupling) employing Pd2dba3, AsPh3, and LiCl in dimethylformamide under microwave irradiation, which allowed the formation of the desired product 64.

Pyriculariol
In 2009, Kiyota and coworkers [58] reported the first total synthesis of (-)-pyriculariol (64). The synthesis of the alkenyl chain commenced with a Ferrier reaction of L-rhamnal diacetate (95) [64] providing aliphatic 96, which upon the acetylation of the free hydroxyl group provided aldehyde 97 (Scheme 13). A Corey-Fuchs reaction using PPh 3 /CBr 4 , followed by the basic hydrolysis of both acetyl protecting groups in 98, afforded diol 99. The treatment of 99 with nBuLi gave enynediol 100, which was converted into dienyl stannane 101 by radical-promoted hydrostannylation with Bu 3 SnH. Finally, two fragments (stannyl compound 101 and aromatic triflate 102 [64]) were coupled using the abovementioned procedure (Stille coupling) employing Pd 2 dba 3 , AsPh 3 , and LiCl in dimethylformamide under microwave irradiation, which allowed the formation of the desired product 64.
PPh3/CBr4, followed by the basic hydrolysis of both acetyl protecting groups in 98, afforded diol 99. The treatment of 99 with nBuLi gave enynediol 100, which was converted into dienyl stannane 101 by radical-promoted hydrostannylation with Bu3SnH. Finally, two fragments (stannyl compound 101 and aromatic triflate 102 [64]) were coupled using the abovementioned procedure (Stille coupling) employing Pd2dba3, AsPh3, and LiCl in dimethylformamide under microwave irradiation, which allowed the formation of the desired product 64.
Yang, Kong, and coworkers [65] described the structure elucidation and antioxidant activities of polyketides 109-116 isolated from solid cultures of S. macrospora that were obtained from the bark of Ilex comuta. The hexaketides from S. macrospora are chemically related to the heptaketide pyriculol family, in which all products, except for 109, contain a 1,2,3-trisubstituted benzene ring. Another typical structural motif, an alkenyl vicinal diol attached to an aromatic ring, was observed in these metabolites. The absolute configuration 3R,4S was established for all determined structures. Moreover, the dimeric metabolites 48-51 were identified for the first time. Bisordariol A (113) and B (114) can be described as sordariol dimers connected via a new C-C bond, while bisordarial C (115) and D (116) represent novel ether-bond-linked sordariol dimers.  cluster (srdA, srdB, srdC, srdD, srdE, and srdG) is responsible for the synthesis of sordarial. The authors suggested that a polyketide chain Yang, Kong, and coworkers [65] described the structure elucidation and antioxidant activities of polyketides 109-116 isolated from solid cultures of S. macrospora that were obtained from the bark of Ilex comuta. The hexaketides from S. macrospora are chemically related to the heptaketide pyriculol family, in which all products, except for 109, contain a 1,2,3-trisubstituted benzene ring. Another typical structural motif, an alkenyl vicinal diol attached to an aromatic ring, was observed in these metabolites. The absolute configuration 3R,4S was established for all determined structures. Moreover, the dimeric metabolites 48-51 were identified for the first time. Bisordariol A (113) and B (114) can be described as sordariol dimers connected via a new C-C bond, while bisordarial C (115) and D (116) represent novel ether-bond-linked sordariol dimers.

Biosynthesis
Scheme 14 shows the proposed biosynthesis [47] of sordarial (106). The experimental results could indicate that the srd gene cluster (srdA, srdB, srdC, srdD, srdE, and srdG) is responsible for the synthesis of sordarial. The authors suggested that a polyketide chain 117 is formed by the condensation of one molecule of acetyl-CoA and five molecules of malonyl-CoA. Formed hexaketide is then reductively released from PKS as an aldehyde 117, and a corresponding hydroxyl group is enzymatically oxidized to give 118. Ketone formation enables the intramolecular aldol condensation, which is followed by dehydration to yield salicylaldehyde 119. The selective epoxidation of the 3,4 double bond in 119 and the hydrolysis of the oxirane ring in 120 affords sordarial (106).  Although the biosynthesis of 103 has not been described so far, the formation o sordariol (103) by the reduction of the aldehyde group of 106 is conceivable.

Bioactivity
Metabolites of S. macrospora were tested for their antioxidant activities using ABTS radical scavenging assays with Trolox as a positive control [65,66]. Sordariol (103) was not phytotoxic towards race roots or race shoots [50] and did not show any immuno suppressive activity [66]. In contrast to the monomeric compounds of this origin, the dimers 113-115 exhibited more potent effects. The authors suggested that the way in which the dimers are connected affects the ABTS radical scavenging abilities, considering that bisordariol D (51) showed a significantly higher value of EC50 (57.6 ± 7.5 mM) com pared to the rest of the dimers 113-115.
Additionally, the compounds were tested for their cytotoxicity against human tu mor cell lines U2OS, MCF-7, and HepG2. No activity was observed at concentrations up to 50.0 mM [31].
Although the biosynthesis of 103 has not been described so far, the formation of sordariol (103) by the reduction of the aldehyde group of 106 is conceivable.

Bioactivity
Metabolites of S. macrospora were tested for their antioxidant activities using ABTS radical scavenging assays with Trolox as a positive control [65,66]. Sordariol (103) was not phytotoxic towards race roots or race shoots [50] and did not show any immunosuppressive activity [66]. In contrast to the monomeric compounds of this origin, the dimers 113-115 exhibited more potent effects. The authors suggested that the way in which the dimers are connected affects the ABTS radical scavenging abilities, considering that bisordariol D (51) showed a significantly higher value of EC 50 (57.6 ± 7.5 mM) compared to the rest of the dimers 113-115.
Additionally, the compounds were tested for their cytotoxicity against human tumor cell lines U2OS, MCF-7, and HepG2. No activity was observed at concentrations up to 50.0 mM [31].

Bioactivity
Preliminary tests on leaves of several weed species have disclosed the relatively low phytotoxicity of agropyrenol (121) and agropyrenal (122), while agropyrenone (123) was described as inactive. None of the metabolites showed antibiotic, fungicidal, or zootoxic activity [67]. Considering the simple chemical structure of agropyrenol (121) and the identified biological properties, the authors decided to perform an SAR study of this metabolite to determine its potential as a new natural herbicide [68]. Compounds 124-129, prepared by the modification of the functional groups of 121 (Figure 6), were assayed for phytotoxic, antibacterial, and zootoxic activities, and the SAR was examined. The results of the tests on nonhost weedy and agrarian plants, fungi, Gram-positive and Gram-negative bacteria, and brine shrimp larvae are summarized in Table 1. The data show that the presence of the double bond and the diol in the aliphatic chain as well as the aldehyde group attached to the aromatic ring seems to play a key role in the biological activity of agropyrenol (121). The introduction of protecting groups (compounds 124, 125, and 129) did not cause any significant loss of the biological activity, as the authors assumed that their hydrolysis occurs at a physiological pH. These less polar derivatives of agropyrenol (121) also showed significant zootoxic and slight antimicrobial activities. The results from this biological activity evaluation could be useful in the testing of other secondary metabolites.

Bioactivity
Preliminary tests on leaves of several weed species have disclosed the relatively low phytotoxicity of agropyrenol (121) and agropyrenal (122), while agropyrenone (123) was described as inactive. None of the metabolites showed antibiotic, fungicidal, or zootoxic activity [67]. Considering the simple chemical structure of agropyrenol (121) and the identified biological properties, the authors decided to perform an SAR study of this metabolite to determine its potential as a new natural herbicide [68]. Compounds 124-129, prepared by the modification of the functional groups of 121 (Figure 6), were assayed for phytotoxic, antibacterial, and zootoxic activities, and the SAR was examined. The results of the tests on nonhost weedy and agrarian plants, fungi, Gram-positive and Gram-negative bacteria, and brine shrimp larvae are summarized in Table 1. The data show that the presence of the double bond and the diol in the aliphatic chain as well as the aldehyde group attached to the aromatic ring seems to play a key role in the biological activity of agropyrenol (121). The introduction of protecting groups (compounds 124, 125, and 129) did not cause any significant loss of the biological activity, as the authors assumed that their hydrolysis occurs at a physiological pH. These less polar derivatives of agropyrenol (121) also showed significant zootoxic and slight antimicrobial activities. The results from this biological activity evaluation could be useful in the testing of other secondary metabolites.
Data are expressed using a visual empiric scale from 0 = no symptoms to 4 = necrosis larger than 1 cm. b Data are expressed as a percentage of seed germination (SD in parentheses). c Data are expressed in mm (SD in parentheses). d Data are expressed in mg/L (SD in parentheses). e Data are expressed in mg/well (SD in parentheses). f Data are expressed as a percentage of dead larvae to the total (SD in parentheses).

Biosynthesis
Polyketides are common biosynthetic precursors in the synthesis of aromatics and macrolides in microorganisms. Based on the structural similarities of the heterocornols to sordarial (106), the biosynthesis of which is described (Scheme 14) [47] it can be assumed that the biosynthesis of these secondary metabolites proceeds very similarly. The authors [69] suggest that the identified secondary metabolites from Pestalotiopsis heterocornis were formed from complementary polyketide precursors in the presence of PKSs. Scheme 15 shows the reaction centers where the enzymatic transformations occur. The length of the chain and the structures of the final polyketide derivatives depend on the number of malonyl-CoAs that participate the condensation reaction.

Bioactivity
Metabolites of P. heterocornis were evaluated by an MTT assay for their cytotoxic activities against four human cancer cell lines, including a human gastric carcinoma (BGC-823), a human large-cell lung carcinoma (H460), a human prostate cancer (PC-3), and a human hepatocellular carcinoma (SMMC-7721) [68] . 6-Alkylsalicylaldehydes 121, 130, 131,  135, 146, and 148 and benzooxepines 132, 136, and 137 showed moderate cytotoxicity, with IC 50 values of 15-100 µM with adriamycin assayed as a positive control. In the antimicrobial assay, the MIC values for the same compounds were evaluated between 25 and 100 µg/mL for the pathogens Staphylococcus aureus and Bacillus subtilis. The antifungal properties of the isolated derivatives appeared to be dependent on the presence of a pent-4-ene-2,3-diol block. The derivatives 132, 121, 136, and 148 showed moderate antifungal activity against Candida parapsilosis and Cryptococcus neofromans at 100 µg/mL ( Table 2).

Conclusions
To date, more than 100 polyketides containing either benzyl alcohol or a salicylaldehyde aromatic moiety have been isolated and characterized. These secondary metabolites are mainly produced by the fungus Emericella variecolor, which is widely distributed in marine and terrestrial environments. Structurally related polyketides derivatives of the benzyl alcohol/salicylaldehyde-type were obtained from other genera of the ascomycetes, namely Sordaria, Magnaporthe, Pestalotipsis, and Ascochyta. The biosyntheses of these polyketide structures involve the condensation of acetyl-CoA with malonyl-CoA in the presence of the specific polyketide synthases (PKS) and the subsequent aromatization of the polyketide chain. The resulting salicylaldehyde derivative with an aliphatic polyenyl substituent appears to be key for the biosynthesis of salicylaldehyde-and benzyl alcohol-type metabolites. The reported polyketide derivatives exhibited valuable bioactivities, particularly those of marine origin. The highest cytotoxicity towards the human cell line panel was demonstrated with varitriol (1), while agropyrenol (121) and other polyketides of the salicylaldehyde-type, such as heterocornols 130-132 and 135-137, showed moderate activity. In the antimicrobial test, the MIC values ranged from 25 to 100 µg/mL in the pathogens Staphylococcus aureus and Bacillus subtilis. Interesting results were obtained in tests for lipid-lowering effects against oleic acid in HepG2 liver cells. Varioxiranol A (39), andytriol (6), and preshamixanthone (11) exhibited inhibitory effects at a dose of 10 µM and showed no toxicity up to 100 µM toward the tested cells. Moreover, preliminary tests of agropyrenol (121) and its derivatives 124-129 showed significant zootoxic and slight antimicrobial activities. This article also provides an overview of the laboratory syntheses of polyketides containing benzyl alcohol or salicylaldehyde functionalities (Table 3). We believe that a deeper understanding of the structural elements influencing biological properties offers a powerful tool for the controlled and rational design of new drugs or phytoherbicides based on natural polyketide derivatives. cancer (BGC-823), liver cancer (HepG2), and kidney cancer (7860). The tested metabolites showed activity, with IC50 values of 20.4-94.2 μM. Compounds 143 and 152 showed no cytotoxic activity [73].

Conclusions
To date, more than 100 polyketides containing either benzyl alcohol or a salicylaldehyde aromatic moiety have been isolated and characterized. These secondary metabolites are mainly produced by the fungus Emericella variecolor, which is widely distributed in marine and terrestrial environments. Structurally related polyketides derivatives of the benzyl alcohol/salicylaldehyde-type were obtained from other genera of the ascomycetes, namely Sordaria, Magnaporthe, Pestalotipsis, and Ascochyta. The biosyntheses of these polyketide structures involve the condensation of acetyl-CoA with malonyl-CoA in the presence of the specific polyketide synthases (PKS) and the subsequent aromatization of the polyketide chain. The resulting salicylaldehyde derivative with an aliphatic polyenyl substituent appears to be key for the biosynthesis of salicylaldehyde-and benzyl alcohol-type metabolites. The reported polyketide derivatives exhibited valuable bioactivities, particularly those of marine origin. The highest cytotoxicity towards the human cell line panel was demonstrated with varitriol (1), while agropyrenol (121) and other polyketides of the salicylaldehyde-type, such as heterocornols 130-132 and 135-137, showed moderate activity. In the antimicrobial test, the MIC values ranged from 25 to 100 μg/mL in the pathogens Staphylococcus aureus and Bacillus subtilis. Interesting results were obtained in tests for lipid-lowering effects against oleic acid in HepG2 liver cells. Varioxiranol A (39), andytriol (6), and preshamixanthone (11) exhibited inhibitory effects at a dose of 10 μM and showed no toxicity up to 100 μM toward the tested cells. Moreover, preliminary tests of agropyrenol (121) and its derivatives 124-129 showed significant zootoxic and slight antimicrobial activities. This article also provides an overview of the laboratory syntheses of polyketides containing benzyl alcohol or salicylaldehyde functionalities (Table 3). We believe that a deeper understanding of the structural elements influencing biological properties offers a powerful tool for the controlled and rational design of new drugs or phytoherbicides based on natural polyketide derivatives.