**1. Introduction**

**\***

As well as we know, microbial metabolites are an important source of drug discovery and development [1]. However, with the deepening of research, many strains in the conventional environment have been repeatedly studied, resulting in the increase of the recurrence rate of known compounds and the decrease of the occurrence rate of new bioactive compounds [2]. Mangrove fungi have attracted much attention because of their special growth environment and unique metabolic mechanism, resulting in the diversity of the structure and bioactivity of their secondary metabolites, which has become a new hotspot in drug development [3,4]. Our previous work reported 9 drimane sesquiterpenoids, 8 benzofurans [4], and 18 ophiobolins [5] from mangrove-derived fungus *Aspergillus ustus* 094102, among which ustusorane E and ustusolate E exhibited cytotoxic activity against the HL-60 cells with IC50 values of 0.13 and 9.0 μM, respectively [4]. In addition, more than 50 drimane sesquiterpenoids have been reported from fungi, including cytotoxic

**Citation:** Gui, P.; Fan, J.; Zhu, T.; Fu, P.; Hong, K.; Zhu, W. Sesquiterpenoids from the Mangrove-Derived *Aspergillus ustus* 094102. *Mar. Drugs* **2022**, *20*, 408. https://doi.org/10.3390/md20070408

Academic Editors: Wenhan Lin, Guoqiang Li and Jing Xu

Received: 23 May 2022 Accepted: 20 June 2022 Published: 22 June 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

strobilactones A and B from *Strobilurus ohshimae* [6], (6-strobilactone-B) ester of (*E*,*E*)- 6-oxo-2,4-hexadienoic acid from marine sponge-derived *A*. *ustus* [7], (6-strobilactone-B) ester of (*E*,*<sup>E</sup>*)-6-carbon-7-hydroxy-2,4-octadienoic acid from mangrove-derived *A*. *ustus* [8], synergistic antibacterial ustusoic acid B from *A*. *ustus* [9], and ET-1 binding inhibitory (2-*E*,4-*E*,6-*E*)-6-(l'-carboxyocta-2-,4-,6--triene)-9-hydroxydrim-7-ene-l l-al from *A*. *ustus* var. *pseudodeflectus* [10], etc. In order to further explore the new drimane sesquiterpenoids produced by *A*. *ustus* strain 094102, we continued to study its secondary metabolites. As a result, we isolated and identified four new drimane sesquiterpenoids (**1**–**4**), as well as three known analogues, (strobilactone A-6-yl) (2*E*,4*E*)-6,7-dihydroxyocta-2,4-dienoate (**5**) [7] that were further isolated as four isomers **5a**–**5d**, mono(6-strobilactone A) ester of (*E*,*E*)-2,4- hexadienedioic acid (**6**) [7], and <sup>2</sup>α,9<sup>α</sup>,11-trihydroxy-6-oxodrim-7-ene (**7**) [7]. The structures elucidation including absolute configurations and the antiproliferative activity will be discussed here.

#### **2. Results and Discussion**

The bioactive EtOAc extract of the fermentation broth of the mangrove-derived fungus *Aspergillus ustus* 094102 was chromatographed on silica gel, Sephadex LH-20, and preparative HPLC columns to give compounds **1**–**7** (Figure 1).

**Figure 1.** Structures of compounds **1**–**7** from *Aspergillus ustus* 094102.

Compound **1** was obtained as a colorless oily solid. Its molecular formula was determined as C15H24O5 based on the high-resolution mass spectrometry (HRMS, ESI-Orbitrap) peak at *m*/*z* 285.1694 [M+H]+ or 283.1547 [M–H]− (Figure S1), indicating 4 index of hydrogen deficiency (IHD). The IR spectrum at νmax 3399 and 1663 cm<sup>−</sup><sup>1</sup> (Figure S2), corresponded to a hydroxy and an *<sup>α</sup>*,*β*-unsaturated carbonyl group, respectively. The 1H-NMR data (Table 1, Figure S3) of **1** revealed four tertiary methyl groups at *δ*H 1.04 (s, H-13/15), 1.14 (s, H-14) and 1.96 (s, H-12), an oxymethylene signal at *δ*H 3.53/3.64 (d/d, H-11), a methylene signal at *δ*H 1.85/1.69 (dd/t, H-1), one olefinic proton signal at *δ*H 5.61 (d, H-7), three methine signals at *δ*H 3.47 (dt, H-2), 2.67 (d, H-3) and 2.81 (s, H-5), as well as four exchangeable proton signals at *δ*H 4.48 (HO-3/2), 4.91 (HO-11) and 5.06 (HO-9). The 13C-NMR and DEPT data (Table 1, Figures S4 and S5) of **1** revealed 15 carbon signals, including a ketone carbonyl signal at *δ*C 199.2 (C-6), two olefinic carbons at *δ*C 128.2/157.4 (C-7/C-8), four methyl signals at *δ*C 16.7/19.1/19.2/29.3 (C-15/C-13/C-12/C-14), two methylenes at *δ*C 38.6/61.9 (C-1/C-11), three methines at *δ*C 55.0/66.4/81.6 (C-5/C-2/C-3) and three nonhydrogenated carbons at *δ*C 37.8/45.4/74.6 (C-4/C-10/C-9). These NMR data (Table 1) were closely related to those of 3*β*,9*<sup>α</sup>*,11-trihydroxydrim-7- en-6-one (that is 3*β*,9*<sup>α</sup>*,11-trihydroxy-6-oxodrim-7-ene [7]), indicating the presence of a drimane sesquiterpene skeleton. The key difference was that compound **1** possessed an additional hydroxy group that resided at C-2 of ring A. On the basis of correlations in the COSY experiments between HO-3/H-3, H-3/H-2/H-1 and HO-11/H-11, as well

as the key HMBC correlations from H-1 to C-5/C-10/C-13, H-3 to C-4/C-14/C-15, H-5 to C-4/C-6/C-9/C-10/C-13/C-14/C-15, H-7 to C-5/C-9/C-12, H-11 to C-8/C-9/C-10, H-12 to C-7/C-8/C-9, H-13 to C-5/C-9/C-10, H-14 to C-3/C-4/C-5/C-15, and H-15 to C-3/C-4/C-14 (Figures 2 and S6–S8) further confirmed the constitution of **1** (Figure 1). The relative configuration was deduced from the NOESY spectrum (Figures 3 and S9), which showed correlations of H-1*α* to H-3/H-5/HO-9, H-11 to H-1*β*/H-2/H-13, and H-2 to H-13 indicated *cis*-orientation of HO-2/H-5/HO-9, and H-2/HO-3/CH3-13/CH2-11, and a *trans*-fused decalin nucleus. The absolute configuration of **1** was determined by its ECD spectrum. On the basis of the octant rule for cyclohexenones [11–13], the positive Cotton effect at λmax 336 nm (Δε + 8.4) and the negative Cotton effect at λmax 240 nm (Δε − 41.3) (Figures 3 and S10) indicated the (2*R*,3*R*,5*S*,9*R*,10*S*)-configuration, consistent with the core configuration of the drimane sesquiterpene, 9*<sup>α</sup>*,11-dihydroxydrim-7-en-6-one (that is 6- oxo-7-drimen-9*<sup>α</sup>*,11-diol [14]), whose absolute configurations have been established by chemical synthesis. Therefore, compound **1,** which we named ustusol F, was determined as (2*R*,3*R*,5*S*,9*R*,10*S*)-2,3,9,11-tetrahydroxydrim-7-en-6-one.

**Table 1.** 1H (500 MHz) and 13C (125 MHz) NMR Data for Compounds **1**–**3** and **7** (DMSO-*d*6, TMS, *δ* ppm).


**Figure 2.** Key COSY and HMBC correlations of compounds **1**–**4** and **5a**–**5e**.

**Figure 3.** NOESY correlations of compounds **1**–**4** & **5e** and the octant rule for **1** and **3**.

Compound **2** was obtained as a light-yellow oil. Its molecular formula was determined as C15H24O4 based on the HRESIMS peak at *m*/*z* 269.1751 [M+H]+ (Figure S11). The similar IR and UV absorptions to those of **1** implied that they shared the same molecular skeleton (Figure S12). The 1D NMR data (Table 1, Figures S13–S15) were also similar to **1** except for a methine signal at *δ*C/H 57.1/2.29 which replaced the nonhydrogenated oxycarbon at *δ*C 74.6, the disappearance of a hydroxy signal at *δ*H 5.06 (HO-9), and the changes of chemical shifts around C-9. These data combined with the 16 amu less of molecular weight than **1** revealed compound **2** as the 9-deoxy derivative of compound **1**. Key COSY of H-9/H-11/HO-11 and HMBC of H-11 to C-8 and C-10 and H-9 to C-10 (Figures 2, S16 and S18) supported the inference. The same relative configuration to **1** was deduced from the NOESY correlations of H-2 (*δ*H 3.45) to H-13 (*δ*H 0.88), H-15 (*δ*H 1.03) and H-1*β* (*δ*H 2.09), H-1*α* (*δ*H 1.33) to H-3 (*δ*H 2.75), H-5 (*δ*H 2.22) and H-9 (*δ*H 2.29), and H-3 to H-14 (*δ*H 1.12) (Figures 3, S16 and S18). The absolute configuration of the *threo*-2,3-diol in **2** was assigned

by a dimolybdenum-induced ECD method [15,16]. Upon addition of Mo2(OAc)4 to a DMSO solution of compound **2**, a chiral dimolybdenum complex was generated in situ as an auxiliary chromophore. Because the contribution from the inherent ECD was subtracted to give the induced ECD of the complex, the observed sign of the Cotton effect in the induced spectrum originates solely from the chirality of the *ortho*-diol moiety expressed by the sign of the O–C–C–O torsion angle. The positive Cotton effect at λmax 332 (Δ*ε* + 6.8) nm (Figure S20) permitted us to assign the (2*R*,3*<sup>R</sup>*)-configuration on the basis of Snatzke's empirical rule [15]. In addition, compounds **1** and **2** also showed a similar ECD Cotton effect, indicating the same absolute configuration. Thus, compound **2,** which we named 9-deoxyustusol F, was determined as (2*R*,3*R*,5*R*,9*S*,10*R*)-2,3,11-trihydroxydrim-7-en-6-one.

Compound **3** was obtained as a colorless oily solid. Its molecular formula was determined as C15H24O4 based on the HRESIMS peak at *m*/*z* 269.1750 [M+H]+ (Figure S21), indicating an isomer of **2**. Similar 1D NMR data (Table 1, Figures S23–S25) with **2** were observed. In addition, a methylene signal (*δ*H/C 1.47/26.7) and an oxymethylene signal (*δ*H/C 4.19/4.26/61.3) replaced the methyl signal (*δ*H/C 1.98/21.5) and oxymethine signal (*δ*H/C 3.45/66.1). Key COSY of H-1/H2-2/H-3 as well as the HMBC of H2-12 (*δ*H 4.19/4.26) to C-8 (*δ*C 162.3), H-7 (*δ*H 5.96) to C-12 (*δ*C 61.3) and H2-2 (*δ*H 1.47) to C-4 (*δ*C 37.5) and C-10 (*δ*C 41.7) revealed that 2-OH was moved to C-12 to form 2-CH2 and 12-CH2OH, respectively (Figures 2, S26 and S28). The relative configuration of compound **3** was deduced from the NOE difference (NOEdiff) experiment. NOEdiff of **3** showed that H-5 (*δ*H 2.15) and H-1a (*δ*H 1.44) were enhanced after the irradiation of H-9 (*δ*H 2.31), while H-3 (*δ*H 3.02) and H-9 (*δ*H 2.31) were enhanced after the irradiation of H-5. The NOE enhancements of H-3 (*δ*H 3.02) and H-5 (*δ*H 2.15) were also observed after H-14 (*δ*H 1.10) was irradiated, while H-13 (*δ*H 0.80) was enhanced after the irradiation of H-15 (*δ*H 0.99). H-1b (*δ*H 1.88) and H-15 was enhanced after the irradiation of H-13 (Figure S29). These NOE data indicated the *cis*-orientation of H-3, H-5, H-9 and H-14 as well as H-13 and H-15, indicating the same relative configuration of **3** to **2** in the chiral centers of C-3, C-5, C-9, and C-10. The similar ECD spectrum to that of **2** implied the same absolute configuration, which was confirmed by octant rule for cyclohexanone [11–13], the positive Cotton effect at λmax 335 nm (Δε + 10.6) and the negative Cotton effect at λmax 241 nm (Δε – 19.1) (Figures 4 and S30). Accordingly, compound **3**, which we named ustusol G, was elucidated as (3*S*,5*R*,9*R*,10*R*)-3,11,12-trihydroxydrim-7-en-6-one.

**Figure 4.** The preparation of acetonide **5e** from **5a**.

Compound **4** was obtained as a colorless solid. Its molecular formula was determined as C21H28O7 based on the HRESIMS peak at *m*/*z* 391.1762 [M–H]–, indicating 8 HIDs (Figure S32). The IR spectrum showed absorption bands of hydroxyl and conjugated carbonyl at νmax 3434 and 1696 cm<sup>−</sup><sup>1</sup> (Figure S33), respectively. The 1D NMR spectra of **4** (Table 2, Figures S34–S36) were very similar to those of (2*E*,4*E*)-(strobilactone A-6-yl)-5- carboxypenta-2,4-dienoate (that is mono(6-strobilactone B) ester of (*E*,*E*)-2,4-hexadienedioic acid [7]), which we named ustusolate J (**6**) for convenience, suggesting that they shared the same molecular scaffold. The only difference was a replacement of the lactone carbonyl signal (*δ*C 174.6 in **6**) by the hemiacetal methine group (*δ*C/H 97.4/5.20 in **4**). In addition, the chemical shifts for C-9 and C-7 have a grea<sup>t</sup> increase and decrease, respectively (Table 2 and Figure S35). These data combined with a 2 amu more than **6** suggested that the γlactone of **6** was reduced to the corresponding hemiacetal in **4**. The key HMBC correlations from hemiacetal proton (*δ*H-11 5.20) to C-9 (*δ*C 76.4)/C-10 (*δ*C 38.0)/C-12 (*δ*C 65.8), from H-12 (*δ*H 4.08/4.38) to C-7 (*δ*C 117.0)/C-8 (*δ*C 143.2)/C-9/C-11 (*δ*C 97.3), and from H-7 (*δ*H 5.49) to C-5 (*δ*C 45.1)/C-9 verified the deduction (Figures 2 and S39). Compound **4** displayed the key NOESY correlations of H-6 (*δ*H 5.58) with H-5 (*δ*H 2.07) and H-14 (*δ*H 0.91), H-5 with H-1b (*δ*H 1.86) and H-2a (*δ*H 1.42), H-11 (*δ*H 5.20) with H-1a (*δ*H 1.22), as well as H-13 (*δ*H 1.12) with H-2b (*δ*H 1.58) (Figures 3 and S40), indicating *cis*-orientation of H-5 with H-6 and *trans*-orientation of H-5 with H-11 and H-13 which is the same relative configuration of **4** to **1** and **6** in the decalin (decahydronaphthalene) nucleus. The same relative configuration of HO-9 was deduced from the same biosynthetic pathway to those of compounds **1** and **5**–**7**. Subsequently, the same ECD pattern of **4**–**6** (Figure S78) implied the same absolute configuration of the drimane nucleus. Compound **4,** which we named ustusolate H, was thus elucidated as (5*S*,6*R*,9*S*,10*S*,11*R*,2-*E*,4-*E*)-(11-deoxy-11-hydroxystrobilactone A-6-yl)-5- carboxypenta-2,4-dienoate.

**Table 2.** 1H (500 MHz) and 13C (125 MHz) NMR Data for Compounds **4** and **6** (DMSO-*d*6, TMS, *δ* ppm).


a Overlapping signals of H-2- with H-5-; b Overlapping signals of H-3- with H-4-

Compound **5** was obtained as a yellow oil. Its molecular formula was determined as C21H28O7 based on the ESIMS peak at *m*/*z* 419.1 for [M–H]– and *m*/*z* 464.9 for [M + HCO2]–(Figure S42), indicating 8 HIDs. A literature search verified that the constitution (planar structure) of compound **5** was the same as the (strobilactone A-6-yl) (2*E*,4*E*)-6,7- dihydroxyocta-2,4-dienoate (that is (6-strobilactone-B) esters of (*E*,*<sup>E</sup>*)-6,7-dihydroxy-2,4- octadienoic acid [7]), for almost the same NMR data. However, four sets of 13C NMR signals of compound **5** (Figure S40) for the side chain at *δ*C 165.51/165.50/165.49/165.47 (C-1-), 120.03/120.99/119.95/119.90 (C-2-), 145.41/145.37/145.34/145.26 (C-3-), 127.54/127.35/ 127.16/126.98 (C-4-), 146.18/146.12/145.48/145.45 (C-5-), 75.16/75.00/74.64/74.46 (C-6-), 69.64/69.62/69.33/69.32 (C-7-), and 19.34/19.26/18.26/18.24 (C-8-) were observed, indi-

.

cating four stereoisomers of **5** resulted from the *ortho*-diol chiral centers of the side chain. With the help of HPLC, compound **5**, which we named ustusolate I for convenience, was confirmed to have four baseline-separated peaks, then purified **5a**, **5b**, **5c**, and **5d** were obtained (Figure S83). The NMR differences of **5a**–**5d** were concentrated in the side chains (Tables 3 and 4, Figures S45–S60), and indicated that compounds **5a** and **5b**, **5c**, and **5d** were two pairs of enantiomers of the *ortho*-diol in the side chain. To elucidate the relative configuration of 6-,7--diol moiety, the acetonide (**5e**) was prepared from **5a** (Figure 4). The 1D and 2D NMR spectra (Tables 3 and 4, Figures S66–S70), as well as the NOESY correlations of H-5- (*δ*H 6.24)/H3-8- (*δ*H 1.01) and H3-11- (*δ*H 1.40), H-6- (*δ*H 4.62)/H-7- (*δ*H 4.34) and H3-10- (*δ*H 1.28) in **5e** (Figures 3 and S71) clearly suggested an *erythro*-6-,7--diol in **5a** and **5b**, and a *threo*-6-,7--diol in **5c** and **5d** was accordingly elucidated. This conclusion is consistent with the chemical shift rule of methyl carbon (*δ*CH3) for 1-methyl-1,2-diol by chemical synthesis, that is 18.1–18.6 and 19.1–19.6 ppm for *threo*- and *erythro*-1,2-diol, respectively [17]. The absolute configuration of the *threo*-6-,7--diol in **5c** and **5d** was assigned by a dimolybdenum-induced ECD method [15,16] in the same manner as that of compound **2**. Upon addition of Mo2(OAc)4 to a solution of compounds **5c** and **5d** in DMSO, a chiral dimolybdenum complex was generated in situ as an auxiliary chromophore. According to the negative ECD Cotton effects of **5c** at λmax 303 ( Δ*ε* – 7.9) nm and the positive ECD Cotton effects of compound **5d** at λmax 305 ( Δ*ε* +2.37) nm (Figures S73 and S74), the absolute configuration of *threo*-6-,7--diol in **5c** and **5d** were determined to be (6- *R*,7- *R*) and (6- *S*,7- *S*), respectively. Thus, the structure of **5c** and **5d** was unambiguously determined as (2- *E*,4- *E*,6- *R*,7- *R*)-ustusolate I (**5c**) and (2- *E*,4- *E*,6- *S*,7- *S*)-ustusolate I (**5d**), respectively. Unfortunately, the absolute configuration of compounds **5a** and **5b** were not determined ye<sup>t</sup> in this paper, which we tentatively named (2- *E*,4- *E*;6-,7- -*erythro*)-ustusolate I (**5a**) and (2- *E*,4- *E*; *ent*-6-,7- -*erythro*)-ustusolate I (**5b**), respectively.

**Table 3.** 1H NMR Data for Compounds **5a**–**5e** (600 MHz, DMSO-*d*6, TMS, *δ* ppm).



**Table 4.** 13C NMR Data for Compounds **5a**–**5e** (150 MHz, DMSO-*d*6, TMS, δ ppm).

Compounds **6** and **7** which could be a 3-deoxy derivative of ustusol F (**1**) were identified by respective comparison of NMR data with those of mono(6-strobilactone-B) ester of (*E*,*E*)-2,4-hexadienedioic acid [7] and <sup>2</sup>*α*,9*<sup>α</sup>*,11-trihydroxy-6-oxodrim-7-ene [7]. The same ECD pattern of compound **6** with **5** (Figure S78) and compound **7** with **1** (Figure S31) indicated they shared the same absolute configuration. Thus, compounds **6** and **7** were respectively identified as (5*S*,6 *R*,9*S*,10*S*,2- *E*,4- *E*)-(strobilactone A-6-yl)-5-carboxypenta-2,4- dienoate (ustusolate J, **6**) and (2*S*,5*S*,9 *<sup>R</sup>*,10*S*)-2,9,11-trihydroxydrim-7-en-6-one (ustusol B, **7**) in this paper. In addition, compound 7 showed almost the same NMR data as our previously reported ustusol B [4] (Table S1) and displayed the same retention times in the co-HPLC (Figure S91). Thus, the structure of ustusol B was revised as structure **7,** which was named ustusol B.

The drimane sesquiterpenoids **1**–**7** were postulated to be biosynthesized from farnesyl-PP (**I**) which generated intermediate **II**, **III** and **IV** after cyclization and oxidation. The intermediates **II** and **III** were subjected to further oxidation to form compounds **1**, **2**, **3**, and **7**. The intermediate **II** was further oxidized to intermediate **IV**, and the latter was subjected to oxidation, hemi acetalization, and esterification to form compounds **4**, **5**, and **6** (Figure 5).

The antiproliferations of compounds **1**–**7** were evaluated against 29 human cancer cell lines and a normal cell line (the names of cell lines are listed in the Supplementary Files) by the cell counting Kit-8 (CCK-8) methods [18,19]. Only compound **5**, the mixture of **5a**/**5b**/**5c**/**5d**, showed antiproliferative activity against the human thyroid cancer cells (CAL-62) and human osteosarcoma cells (MG-63) with the IC50 values of 16.28 ± 1.01 and 10.08 ± 0.04 μM, respectively, while the pure compounds **5a**–**5d** were inactive (IC50 ≥ 50 μM). The IC50 values of doxorubicin (positive control) against CAL-62 and MG-63 were 0.062 ± 0.022 and 0.096 ± 0.012 μM, respectively. The bacteriostatic activities of compounds **1**–**7** against 6 human pathogenic bacteria and 6 aquatic pathogenic bacteria (the names are listed in the Supplementary Files) were tested by the diffusion method of

filter paper, but no inhibition zone was observed at the concentration of 100 μg/mL for compounds **1**–**7**.

**Figure 5.** Proposed biosynthetic pathway for drimane sesquiterpenoids from *A*. *ustus* 094102.
