3.2.3. Meroterpenoids

Meroterpenoids are secondary metabolites with structures consisting of at least two parts: a terpenoid fragment (mainly mevalonate pathway) and a nonterpenoid fragment [83]. The different nonterpenoid moiety based on the biosynthetic pathway, various terpenoid (the length of the terpenoid chain and its cyclization mode), and the tailoring reactions make the chemical diversity of meroterpenoids.

Chermebilaene A (**75**) (Figure 20), an unprecedented acorane-type sesquiterpene hybridized with an octadecadienoic acid skeleton, together with an unusual orthoester meroterpenoid, chermebilaene B (**76**) were isolated from the co-culture extract of *P. bilaiae* MA-267 (from the rhizosphere of the mangrove *Lumnitzera racemosa*) and *P. chermesinum* EN-480 (from the fresh tissue of marine red algal *Pterocladiella tenuis*) [84]. Compound **75** showed potent inhibitory activities against *Ceratobasidium cornigerum* and *Edwardsiella tarda*, and may prove helpful as an antibiotic against aquatic or plant pathogens.

**Scheme 11.** (**a**) Proposed biosynthetic pathway for compounds **70**–**72** [80,81]; (**b**) Proposed biosynthetic pathway for compounds **73**–**74** [82].

**Figure 20.** Structures of compounds **75** and **76**.

Simpterpenoid A (**77**) (Figure 21), an unconventional meroterpenoid containing a highly functionalized cyclohexadiene moiety with *gem-*propane-1,2-dione and methylformate groups, was isolated from the fungal strain *Penicillium simplicissimum* MA-332, obtained from the rhizospheric soil of the mangrove plant *Bruguiera sexangular* var. *rhynchopetala* [85]. The intricate polycyclic skeleton is unique in natural sources. Compound **77** exhibited potent inhibitory activity against influenza neuraminidase with an IC50 value of 8.1 nM.

**Figure 21.** Structures of compounds **76**–**79**.

Two new meroterpenoids, penicianstinoids A and B (**78** and **79**, Figure 21), were obtained from the mangrove-derived fungus *Penicillium* sp. TGM112 isolated from the mangrove *Bruguiera sexangula* var. *rhynchopetala* [86]. Compared with **79**, compound **78** represents an austinoid-like meroterpenoid that is reported for the second time [87], in which a carbon−carbon double bond at C-1-−C-2- was oxidized to a carbonyl group at

C-1-−C-2-. Compounds **78** and **79** showed growth inhibition activity against newly hatched larvae of *Helicoverpa armigera* (Hubner) with IC50 values of 200 *μ*g/mL. In addition, **78** and **79** displayed insecticidal activity against *Caenorhabditis elegans* with EC50 values of 9.4 (±1.0) and 9.9 (±0.0) *μ*g/mL, respectively. Biogenetically, compounds **76**–**79** are derived from the same intermediate **S2**, which is produced by the combination of a polyketide intermediate 3,5-dimethylorsellinicacid (DMOA) and the terpenoid precursor farnesyl pyrophosphate (FPP), following by a series of further modifications to generate a profile of meroterpenoids with diverse skeletons bearing polycyclic cores. DMOA-based meroterpenoids exhibit diverse structures due to the cyclization of the terpenoid moiety, divergence of postcyclization modification reactions, and various tailoring reactions (Scheme 12) [83].

**Scheme 12.** Proposed biosynthetic pathway for compounds **76**–**79** [83].

Two hybrid sesquiterpene-cyclopaldic acid metabolites with an unusual carbon scaffold, namely pestalotiopens A and B (**80** and **81**) (Figure 22), were obtained from the endophytic fungus *Pestalotiopsis* sp. (from the leaves of the Chinese mangrove *Rhizophora mucronate*), together with the known phytotoxin altiloxin B [88]. A plausible biosynthetic pathway of **80** and **81** is proposed (Scheme 13). The cyclopaldic acid and altiloxin B were deduced as precursors.

**Figure 22.** Structures of compounds **80** and **81**.

**Scheme 13.** Proposed biosynthetic pathway for compounds **80** and **81** [88].

Indole terpenoids are structurally diverse meroterpenoids containing an indole ring from tryptophan and cyclic sesquiterpenes or diterpene backbone moiety [83].

Three indole sesquiterpenes, indotertine A (**82**) [89] and indotertine B (**83a**/**83b**) [90] (Figure 23) were discovered from actinomycete *Streptomyces* sp. CHQ-64 (derived from the rhizosphere soil of reeds). They possess an unusual skeleton with a condensed ring system made up of a tryptophan-derived indole moiety and a sesquiterpene unit, which represents a new subgroup of indole terpenoids combining amino acid and mevalonate pathways. Indotertine B (**83a**/**83b**) exists as a pair of rotamers about the N−C(O) bond with a 2:1 ratio, inseparable by HPLC because of the dynamic interconversion. The analysis of the NOESY spectrum implied that the formyl−N-1 amide bond was *S-trans* in **83a** and *S-cis* in **83b**. Compound **83** displays cytotoxic activities against HCT-8 and A549 tumor cell lines

with IC50 values of 6.96 and 4.88 *μ*M. Further chemical investigation of this fungal strain led to the isolation of drimentine I (**84**) [91], containing a rare heptacyclic skeleton formed via two bridging linkages. The pentacyclic product indotertine A (**82**) was hypothetically synthesized by iminium-olefin cyclization. In contrast, tetracyclic product drimentine F could take place from amidic nitrogen by nucleophilic addition to the *α*-position of the indole moiety (Scheme 14). However, cyclization of **84** happened on indol-NH to afford the linkage between C-14 and N-6 of drimentine F. Compound **84** was found to have weak activity against human cervical carcinoma cell line HeLa, with IC50 values of 16.73 *μ*M.

**Figure 23.** Structures of compounds **82**–**84**.

**Scheme 14.** Proposed biosynthetic pathway for compounds **82**–**84** [89].

Secopaxilline A (**85**) [92] (Figure 24) is the first example of indole diterpenoid derivatives possessing a carbon-nitrogen bond cleavage skeleton, which was isolated from metabolites of the aciduric fungus *Penicillium camemberti* OUCMDZ-1492 (separated from the soil and mud around the roots of *Rhizophora apiculata*). A plausible biosynthetic pathway for secopaxilline A (**85**) from paxilline was postulated, (Scheme 15), and the process has been conducted by chemical reactions with a 45% overall yield. Paxilline was derived from the common indole-diterpenoid precursor 3-geranylgeranylindole (GGI) derived from geranylgeranyl pyrophosphate (GGPP) and indole-3-glycerol phosphate [93] (Scheme 15).

**Figure 24.** Structures of compounds **85**–**89**.

**Scheme 15.** Proposed biosynthetic pathway for compounds **86**–**89** [93].

The fungus *Mucor irregularis*, isolated from the fresh inner tissue of the mangrove *Rhizophora stylosa*, yields three unusual indole-diterpenes, rhizovarin A–C (**86**–**88**, Figure 24) [94], which represent the most complex members of the reported indole-diterpenes. Even though the main structural elements resemble those of other reported indole diterpenes, the presence of an unusual acetal linked to a hemiketal (**86**) or a ketal (**87** and **88**) unit in an unprecedented 4,6,6,8,5,6,6,6,6-fused indole-diterpene ring system makes them chemically unique. Their structures and absolute configurations were elucidated by spectroscopic analysis, modified Mosher's method, and chemical calculations. For rhizovarin A (**86**), the biosynthetic pathway may involve more oxidative steps than penitrem A, a known indole-diterpene derived from a paxilline and two isopentenyl-diphosphate units. (Scheme 15) The biosynthetic pathway has been elucidated by reconstitution of the biosynthetic genes in *Aspergillus oryzae* [95]. Another unusual indole-diterpene, containing a complex 6,8,6,6,6-fused ring system, rhizovarin D (**89**), was also obtained in this study. NOESY experiments determined the relative configuration for the stereogenic centers of **89**. Each isolated compound was evaluated for antitumor activity against HL-60 and A-549 cell lines. Compounds **86** and **87** showed activities against the human A-549 and HL-60 cancer cell lines (IC50 < 10 *μ*M).

Bioassay-guided fractionation of the bacterial strain *Erythrobacter* sp. SNB-035 (from mangrove sediments) led to the isolation of erythrazoles A and B (**90** and **91**) [96] (Figure 25). Structurally, **90** and **91** possess an abenzothiazole moiety, which is rare among NPs. Furthermore, **91** arises from four biosynthetic pathways: NRPS, terpene, shikimate, and polyketide. Although combinations of two of the four pathways are common among NPs, four biosynthetic pathways simultaneously involved are extremely rare (Scheme 16).

**Figure 25.** Structures of compounds **90** and **91**.

**Scheme 16.** Proposed biosynthetic pathway for compounds **90** and **91** [96].

*3.3. Alkaloids and Other Nitrogen-Containing Metabolites*

3.3.1. Diketopiperazines

Diketopiperazines (DKPs) are an essential group of structurally diverse cyclic dipeptides with significant biological properties [97].

Effusin A (**92**) (Figure 26) is a spirobicyclic *N*,*O*-acetal derivative with an unprecedented 3-,3a-,5-,6--tetrahydrospiro-[piperazine-2,2--pyrano[2,3,4-*de*]chromene] ring system. Besides this, a spiro-polyketide-diketopiperazine hybrid dihydrocryptoechinulin D (**93**) were isolated from a mangrove rhizosphere soil-derived fungus *Aspergillus effuses* H1- 1 [98]. Compounds **92** and **93** occurred as racemates. The enantiomers were separated and characterized by online HPLC-ECD analysis, and their absolute configurations were determined by the TDDFT ECD calculation approach. The spirobicyclic *N*,*O*-acetal moiety of **92** could be obtained through a domino ring-closure reaction between the substituted salicylaldehyde moiety in aspergin and the eneamide moiety of the diketopiperazine unit in neoechinulin B [98]. On the contrary, the spirobicycle of **93** is produced by an enzymecatalyzed regiospecific [4 + 2] Diels Alder reaction (Scheme 17). The cytotoxic effects of **92**–**93** were evaluated, **93** showed potent activity on P388 cells with an IC50 value of 1.83 *μ*M. The target of racemic **93** was also evaluated, and the (12 *R*,28*S*,31*S*)-**93** enantiomer (**93a**) showed selectivity against topoisomerase I.

**Figure 26.** Structures of compounds **92** and **93**.

**Scheme 17.** Proposed biosynthetic pathway for compounds **92** and **93** [98].

Using the OSMAC (one strain many compounds) approach, a metabolically rich strain of *Penicillium brocae* MA-231 (isolated from mangrove *Avicennia marina*) could produce two

new diketopiperazines, spirobrocazines A–B (**94**–**95**) (Figure 27), which had a 6/5/6/5/6 cyclic system with a rare spirocyclic center at C-2 [99]. In addition, a deep-sea sedimentderived fungus *Eutypella* sp. Also yields three new spirocyclic DKPs, eutypellazines N–P (**96**–**98**) [100]. Compound **96** was determined as the C-2- isomer of spirobrocazine A (**91**). Notably, **97** and **98** are the first compounds isolated from a wild-type fungus to contain a spirocyclic tetrahydrobenzothiophene motif. Furthermore, eight new dioxopiperazines **99**–**106** (penispirozines A-H) were discovered from the mangrove-derived fungus *Penicillium janthinellum* HDN13-309 [101]. The structures of **99**–**104** were similar to eutypellazines O–P (**97**–**88**). They were distinguished by not only the existence of a spiro-thiophane or spiro-furan ring system but also the chirality of the pentacyclic moiety. Moreover, penispirozine A (**99**) had an unusual pyrazino[1,2]oxazadecaline coupled with a thiophane ring system, while penispirozine B (**100**) possessed a 6/5/6/5/6 pentacyclic ring system with two rare spirocyclic centers. Biosynthetically, the precursor to these structurally diverse penispirozines was considered to be the diketopiperazine cyclo-L-Phe-L-Phe (Scheme 18). In addition, compounds **101** and **102** increased the expression of the two relevant phase-II detoxifying enzymes, SOD2 and HO-1, at 10 *μ*M.

; 25 0H*RS*

**Figure 27.** Structures of compounds **94**–**106**.

**Scheme 18.** Proposed biosynthetic pathway for compounds **94**–**98** [99,100].

A pair of unusual enantiomeric indole diketopiperazine alkaloid dimers, ( ±)-asperginulin A (**107a**/**b**) (Figure 28), with an unprecedented 6/5/4/5/6 pentacyclic skeleton, were −

−

discovered from the mangrove endophytic fungus *Aspergillus* sp. SK-28, guided by UPLC-HRMS [102]. Chiral-phase HPLC separated the enantiomeric dimers. Their structures, including the absolute configurations, were elucidated by spectroscopic analysis, X-ray diffraction, and quantum chemical calculation. (+)-Asperginulin A (**107b**) exhibited antifouling activity against the barnacle *Balanus reticulatus*. **107** was possibly derived, in vivo, from intermolecular [2 + 2] cycloaddition of its monomer precursor by nonenzymatic processes (Scheme 19).

**Figure 28.** Structures of compounds **107a/b**.

**Scheme 19.** Proposed biosynthetic pathway for compounds **107a/b** [102].

A class of pyrazinopyrimidine-type alkaloids, namely pyrasplorines A–C (**108**–**110**) (Figure 29) were discovered from the fungus *Aspergillus versicolor* HDN11-84 [103]. Pyrasplorine A (**105**) represents the first compound with spiro-cyclopentane in pyrazinopyrimidinetype alkaloids. The cyclopentane moiety is common in terpenes but rare in alkaloids and diketopiperazines, and it is only found in maremycins [104]. The structure is probably constructed by the condensation of anthranilic acid with diketopiperazine and followed by successive steps to yield the key intermediate **S3**. Then, compound **108** was derived from the **S3** via a series of reactions [105] (Scheme 20).

**Figure 29.** Structures of compounds **108**–**110**.

**Scheme 20.** Proposed biosynthetic pathway for compound **108** [105].

#### 3.3.2. Indole and Isoindole Alkaloids Derivatives

Various mangrove fungi produce indole and isoindole alkaloids with a plethora of biologically active. The indole-terpenes which also belong to meroterpenes have been described in Section 3.2.3.

Cytochalasan alkaloid usually consists of a 10-(indol-3-yl) group, a macrocyclic ring, and a perhydroisoindolone moiety. Chaetoglobosin is one class of cytochalasan alkaloid. The mangrove endophytic fungus *Penicillium chrysogenum* V11 afforded two unusual new Chaetoglobosins, penochalasin I and K (**111** and **112**) [106,107] (Figure 30), with an unprecedented six-cyclic 6/5/6/5/6/13 fused ring system formed by the connection of C-5 and C-2- of the chaetoglobosin class. Additionally, the biomimetic semi-synthesis of **111** and **112** was successfully carried out from the corresponding co-occurrence analogue chaetoglobosin C and chaetoglobosin A, respectively [107]. Compound **112** displayed significant inhibitory activities against *Colletotrichum gloeosporioides* and *Rhizoctonia solani* (MICs = 6.13 *μ*M, 12.26 *μ*M, respectively), which was better than those of control carbendazim. It also exhibited potent cytotoxicity against MDA-MB-435, SGC-7901, and A549 cells (IC50 < 10 *μ*M). In addition, compound **111** exhibited significant cytotoxicity against MDA-MB-435 and SGC-7901 cells (IC50< 10 *μ*M).

**Figure 30.** Structures of compounds **111** and **112**, chaetoglobosin A and C.

The typical paraherquamides (PHQs) are prenylated indole alkaloids with diverse ring systems. PHQs are derived from three building blocks: *L*-tryptophan, acyclic amino acid (either proline, *β*-methyl proline, or pipecolic acid), and one or two isoprenyl units. Interestingly, compounds **113**–**115** (Figure 31) (mangrovamides A–C, isolated from the *Penicillium* sp. Separated from a mangrove sediment sample of the South China Sea) feature a bicyclo [2.2.2] diazaoctane core and contain the first documented examples of isoprene derived dimethyl *γ*-pyrone and *γ*-methyl proline, instead of the usual *β*-methyl proline in the PHQ family [108]. A plausible biosynthetic pathway starting from L-ornithine to account for the formation of the observed *γ*-methyl proline is outlined (Scheme 21). Moreover, the X-ray data determined the absolute configuration of all chiral centers in **113**. In an activity assay, **115** showed a moderate acetylcholinesterase inhibitory effect with an IC50 value of 58.0 *μ*M.

**Figure 31.** Structures of compounds **113**–**115**.

**Scheme 21.** Proposed biosynthetic pathway for compounds **113**–**115** [108].

Diaporisoindoles A and B (**116** and **117**) [109], and D and E (**118** and **119**) [52] (Figure 32), isolated from the mangrove endophytic fungus *Diaporthe* sp. SYSU-HQ3 (from a fresh branch of the mangrove plant *Excoecaria agallocha*) and could be derived from tenellone B, are the first reported examples of isoprenylisoindole alkaloids with a rare 1,4-benzodioxan moiety. In addition, siaporisoindole A (**116**) showed significant inhibitory activity against *Mycobacterium tuberculosis* protein tyrosine phosphatase B with IC50 4.2 *μ*M compared to 22.1 *μ*M for the positive control (oleanolic acid,). Furthermore, **116** and **117** exhibited potent inhibitory activity against NO production in RAW 264.7 cells with IC50 values less than 10 *μ*M. Then She et al. continued an extensive study of isolating an unusual diisoprenylisoindole dimer diaporisoindole C (**120**). It was presumed to be derived from compounds **116** or **117** via addition reaction, dehydration, and aromatization (Scheme 22).

**Figure 32.** Structures of compounds **116**–**120**.

**Scheme 22.** Proposed biosynthetic pathway for compounds **116**–**120** [109].

Quinazoline containing indole alkaloids have pyrimidine [2, 1-b] quinazoline and imidazole [1, 2-a] indole groups linked by methylene (and, in some cases, further linked by additional helical Bridges). Two unusual quinazoline-containing indole alkaloids neosartoryadins A and B (**121** and **122**) (Figure 33) along with fiscalin C (a known compound to be related to biosynthesis) were identified from the mangrove endophytic fungus *Neosartorya udagawae* HDN13-313 [110]. Compounds **121** and **122** is a quinazoline-containing indole alkaloid featuring a unique 6/6/6/5 quinazoline ring directly linked to the 6/5/5 imidazolinone ring. **121** and **122** differs from conventional fumiquinazoline alkaloids such as fiscalin C by the unprecedented pyrido[2, 1-b]- quinazoline moiety, which binds to a pyridine (C ring) rather than a pyrimidine ring, in addition to the presence of a unique tetrahydrofuran ring (D ring). It is speculated that **121** and **122** are biosynthesized from *L*-tryptophan, anthranilic acid (ATA), L-valine, and 2-aminoisobutyric acid (Aib). The unprecedented C ring was formed by the key intermediate fiscalin C through further modification by oxidation, hydrolysis, water nucleophilic attack, dehydration, deprotonation, and subsequent aldol reaction (Scheme 23).

**Figure 33.** Structures of compounds **121** and **122**.

**Scheme 23.** Proposed biosynthetic pathway for compounds **121** and **122 [110]**.

Streptocarbazoles, the staurosporine analogues with extraordinary cyclic N-glycosidic connections between 1,3-carbon atoms of the glycosyl moiety and two indole nitrogen atoms of the indolocarbazole core, have also been produced by mangrove actinomycetes.

*Streptomyces* sp. FMA, isolated from mangrove soil collected in Sanya, Hainan Province of China provided streptocarbazoles A (**123**) and B (**124**) [111] (Figure 34). Compound **123** was cytotoxic to HL60, A549, P338, and HeLa cells with IC50 values of 1.4, 5.0, 18.9, and 34.5 *μ*M, respectively, while compound **124** was active against P388 and HeLa cells with IC50 values of 12.8 and 22.5 *μ*M, respectively. In addition, it was demonstrated that streptocarbazoles A arrest the HeLa cells in the G2/M phase at 10 *μ*M. A plausible biogenetic pathway of **123** and **124** was postulated (Scheme 24). The indolocarbazole unit (K252c) was derived from tryptophan, while the glycosyl moiety was probably developed from 2-deoxy-D-pyranoglucose. Subsequently, the first cloning and characterization of an indolocarbazole gene cluster isolated from *Streptomyces sanyensis* FMA were reported. Indolocarbazole biosynthesis was confirmed by gene inactivation and heterologous expression in *Streptomyces coelicolor* M1152 [112].

**Figure 34.** Structures of compounds **123** and **124**.

**Scheme 24.** Proposed biosynthetic pathway for compounds **123** and **124** [111].
