**2. Results**

Phomoparagin A (**1**) was obtained as a colorless amorphous powder. Its molecular formula, C28H35NO3 with 12 degrees of unsaturation, was established using the high-resolution-electrospray ionization mass spectrometry (HRESIMS) positive ion at *m*/*z* 434.2644 ([M+H]+, calcd for 434.2695). The 1H NMR data of **1** (Table 1), as well as the coupling constants of the connected protons, indicated the presence of a tertiary methyl at *δ*H (1.74, 3H, s, H3-23), two secondary methyl groups at *δ*H (0.73, 3H, d, *J* = 6.1 Hz, H3-11; 0.95, 3H, d, *J* = 6.7 Hz, H3-22), an exocyclic methylene group at *δ*H (5.13 and 4.96, 2H, both s, H2-12), three oxygenated methine groups at *δ*H (4.13, 1H, d, *J* = 9.8 Hz, H-7; 4.19, 1H, t, *J* = 9.7 Hz, H-14; 2.99, 1H, d, *J* = 2.4 Hz, H-21), an olefinic methine group at *δ*H (5.28, 1H, br s, H-19), and typical resonance of a single substituted phenyl at *δ*H (7.22–7.30, 5H). Its 13C NMR spectrum (Table 2) disclosed 28 carbon resonances, including three sp<sup>3</sup> methyls, three sp<sup>3</sup> methylenes, 10 sp<sup>3</sup> methines, one sp<sup>3</sup> quaternary carbon, one sp<sup>2</sup> exocyclic methylene, six sp<sup>2</sup> olefinic methines, and four sp<sup>2</sup> quaternary carbons (three olefinic carbons and one amide carbonyl), as supported by the DEPT and HSQC spectra. The complete structure of **1** was established by extensive analysis of its 2D NMR spectra. The 1H-1H COSY (Figure 2) and HSQC spectra suggested the presence of the fragments CH2(10)-CH(3)-CH(4)-CH(5)- CH3(11)-, CH(7)-CH(8)-CH(13)-CH(14)-CH2(15)-CH(16)-CH2(17), incorporating CH3-(22), which was coupled to CH-(16); -CH(7)-CH(8)-CH(13)-CH(20)-CH(21)-, incorporating CH- (19), which was coupled to CH-(20), and CH(2-) to CH(6-). In the HMBC spectrum (Figure 2), 13C-1H long-range correlations were observed from H-3 (*δ*H 3.30, m) and H-4 (*δ*H 2.64, m) to C-1 (*δ*C 171.7); H-4 and H-5 (*δ*H 2.65, m) to C-9 (*δ*C 53.8); H-7 (*δ*H 4.13, d, 9.8) to C-5(*δ*C 33.4), C-6(*δ*C 150.9), C-12(*δ*C 113.4), C-14(*δ*C 76.8), and C-8(*δ*C 41.9), establishing the phenylalanine moiety (rings A and B). HMBC correlations from H-8 (*δ*H 2.19, t, 10.0 Hz) to C-1, C-4 (*δ*C 47.9), C-9 (*δ*C 53.8), and H-4 (*δ*H 2.49, m) to C-21 (*δ*C 75.6) supported that the five-membered ring C was fused to ring B via C-8 and C-9. An eight-membered ring D was elucidated using the HMBC correlations from H-19 (*δ*H 5.28, br s) to C-17 (*δ*C 43.4), C-21 (*δ*C 75.6), C-18 (*δ*C 138.9), and C-23 (*δ*C 27.7), and from H-22 (*δ*H 0.95, d, 6.7) to C-15 (*δ*C 44.4) and C-17 (*δ*C 43.4). Additionally, H-10 (*δ*H 2.76, 2.69) to C-3 (*δ*C 54.5), C-4 (*δ*C 47.6), C-1- (*δ*C 138.7), and C-2-/C-6- (*δ*C 131.1) revealed the connection of the phenyl to C-3 via C-10. Comparison of the 1H and 13C NMR spectra of **1** with those of phomopchalasin A (**4**), which was previously isolated from the endophytic fungus *Phomopsis* sp. shj2 associated with *Isodon eriocalyx* var. laxif lora. [12], indicated that these two compounds possessed the same 5/6/5/8-fused tetracyclic cytochalasan ring system (rings A-D). These two compounds are different, in that a new five-membered epoxy unit (ring E) was formed by the dehydration reaction between 7-OH and 14-OH. To confirm this observation, evidence was obtained from the chemical shifts of C-8 (*δ*C 41.9) and C-13 (*δ*C 42.5) in **1**, which were significantly shifted upfield compared with C-8 (*δ*C 51.3) and C-13 (*δ*C 51.7) in **4**, along with the HRESIMS data. Thus, the connection from C-7 to C-14 via an oxygen atom was deducedfromtheaboveevidence.

**Figure 2.** Key COSY and HMBC correlations of Compounds **1**–**3**.


**Table 1.** 1H NMR data (δ) for **1**–**3** in CD3OD (400 MHz) (δ in ppm, J in Hz).

**Table 2.** 13C NMR data (δ) for **1**–**3** in CD3OD (100 MHz) (δ in ppm, J in Hz).


The relative configuration of **1** was subsequently established by analyzing the NOESY spectrum. The NOE crosspeaks (Figure 3) between H-8/H-14, H-14/H-20, and H-20/H-16 indicated that these protons have *β* orientations. H-7/H-13 and H-13/H-21 were observed, suggesting that these protons are cofacial and have α-orientations. Considering the above evidence, we assumed that the stereochemistry is the same as that reported for **4,** based on the similarity in the 2D NMR resonances and the shared biogenetic origin.

**Figure 3.** Key NOESY correlations of Compounds **1**–**3**.

The absolute configuration of **1** was determined by comparing experimental and calculated ECD spectra using time-dependent density-functional theory (TDDFT). Two feasible configurations, 3*S*, 4*R*, 5*S*, 7*S*, 8*R*, 9*R*, 13*R*, 14*R*, 16*R*, 20*S*, 21*R* and 3*R*, 4*S*, 5*R*, 7*R*, 8*S*, 9*S*, 13*S*, 14*S*, 16*S*, 20*R*, 21*S* (**1** and *ent*-**1**, respectively), were calculated at the B3LYP/6- 31+G(d,p) level with a PCM solvent model for MeOH. The calculated ECD spectrum of **1** showed an excellent fit with the experimental spectrum (Figure 4), which indicated the absolute configuration to 3*S*, 4*R*, 5*S*, 7*S*, 8*R*, 9*R*, 13*R*, 14*R*, 16*R*, 20*S*, 21*R*. Thus, the complete structure of **1** was established.

**Figure 4.** Experimental and calculated electronic circular dichroism (ECD) spectra of **1**–**3**.

Phomoparagin B (**2**) was obtained as colorless needles and has a molecular formula of C30H39NO5 based on HRESIMS (*m/z* 516.2727, calcd for [M+Na]+ 516.2726), implying 12 degrees of unsaturation. The 1H and 13C NMR spectra (Table 1 and 2) of **2** were similar to those of phomopchalasin A (**4**) [12], except that the hydroxyl group at C-21 of the latter was replaced by the acetoxy group of **2**. This was confirmed by the molecular weight difference of 42 amu observed between compounds **2** and **4**, along with the strong HMBC correlation (Figure 2) from the protons of the methyl ester group (*δ* 1.99, 3H, s) and H-21 (*δ* 3.71, 1H, *d*, 5.0) to C-24 (*δ* 173.1), which supports the presence of an acetoxy group at C-21 in **2**. The relative configuration of **2** was determined by interpreting the NOESY data (Figure 3). As expected, the experimental ECD spectrum (Figure 4) of **2** matched exactly with the calculated spectrum. Accordingly, the absolute configuration of **2** was determined to be 3*S*, 4*R*, 5*S*, 7*S*, 8*R*, 9*R*, 13*R*, 14*R*, 16*R*, 20*S*, 21*R,* and it was named phomoparagin B.

Phomoparagin C (**3**), a colorless amorphous powder, has a molecular formula of C28H35NO3, as established by HRESIMS (*m/z* 434.2678, calcd for [M+H]+ 434.2695), corresponding to 12 degrees of unsaturation. The 1H and 13C NMR data (Tables 1 and 2) of **3** were similar to those of **2,** except for the signals of the -OCOCH3 substituent and an aliphatic methine proton within **2,** which were absent in **3**. Instead, signals of an olefinic double bond at (*δ*C 137.9, s, C-20; *δ*H 5.53, 1H, s; *δ*C 120.1, d, CH-21) appeared, which

accounts for the molecular weight difference of 60 amu that was observed between the compounds. Conformational evidence was obtained from the HMBC correlations of H-21 to C-8(*δ*C 38.2)/C-9(*δ*C 52.7)/C-13(*δ*C 47.5) (Figure 2). The relative stereochemistry of compound **3** was determined using the key NOSEY cross peaks (Figure 3) of H-3 with H3-11; H-8 with H-4 and H-20; and H-14 with H-16 and H-20. The absolute configuration could be determined as 3*S*, 4 *R*, 5*S*, 7*S*, 8 *R*, 9 *R*, 13 *R*, 14 *R*, 16*R,* by comparing the experimental and calculated ECD spectra using TDDFT (Figure 4). Thus, the structure of **3** was determined, and it was named phomoparagin C.

Previous isotope labeling experiments revealed that cyctochalasans might rationally share the same biosynthetic pathway, and it most likely originates from a polyketide synthase (PKS)/nonribosomal peptide synthetase (NRPS) hybrid machinery [31,32]. The stepwise assembly is realized from one activated acetyl-CoA starter, seven malonyl-CoA extender units and phenylalanine. An intramolecular aldol condensation generates pyrrolinone, which reacts via a [4 + 2]-cycloaddition, hydroxylation, and dehydrogenation to generate the same biosynthetic precursor **7**. Subsequent acetylation and oxidation of 7 formed **6** and **8**, respectively. Compound **5**, which contains a 5/6/6/7/5-fused pentacyclic ring system, might originate by dehydration, epoxidation, intramolecular nucleophilic addition, hydroxylation, and dehydration reactions. In another pathway, **7** undergoes dehydration, intramolecular rearrangement, and hydroxylation, to produce the 5/6/5/8-fused tetracyclic intermediate **4**. The subsequent intramolecular dehydration and acetylation led to the formation of **1**–**3** (Figure 5). Of these, **1** possesses an unprecedented 5/6/5/8/5-fused pentacyclic ring system.

**Figure 5.** Plausible biogenetic relationship of isolated compounds.

The isolated compounds (**1**–**8**) were evaluated for their immunosuppressive activities against the proliferation of ConA-induced T and LPS-induced B murine splenic lymphocytes, according to previously described protocols [33–35]. The results showed that Compounds **2** and **4**–**6** remarkably inhibited the proliferation against splenic lymphocyte growth, with IC50 values ranging from 11.2 ± 0.3 μM to 154.4 ± 0.4 μM, of which **2** and **6** displayed the most promising inhibitory effects (Table 3). The cytotoxicity of immunosuppressive Compounds **2** and **4**–**6** was tested in murine splenocyte cultures for 72 h using the tetrazolium salt-based CCK-8 assay. The results (Table 4) showed that even **6** exhibited better suppression of the overproduction of the cell stimulated by ConA compared with that of **2**, but **2** exhibited relatively lower toxicity for the survival of normal splenocytes (IC50 = 111.7 ± 1.1 μM) than that of **6** (IC50 = 42.2 ± 1.7 μM) and it was approximately 11-fold lower in comparison with that of CsA (IC50 = 10.9 ± 0.8 μM) and cytochalasin D (IC50 = 1.0 ± 0.0 μM), indicating that the compound has a relatively low toxicity toward the survival of normal splenic cells. Thus, we selected this particular cytochalasin for the mechanism of action studies.


**Table 3.** Immunosuppressive activities of isolated compounds a.

a Compound **1**, **3**, **7**–**8** were inactive (IC50 > 200 μM). b Data are presented as mean ± SD from three separate experiments. c Positive control.


**Table 4.** Cytotoxicity data of immunosuppressive compounds a.

a ResultsareexpressedasIC50 valuesofmean± SD(*n*=7)inμM.

To directly examine whether **2** specifically inhibits the CaN/NFAT pathway, we first investigated the CaN inhibition rate of **2**. As expected, **2** was found to be significantly active and inhibited CN in a dose-dependent manner with an IC50 value of 17.89 ± 0.40 μM, which has a greater potency than that of the clinically used immunosuppressant cyclosporine A (CsA, IC50 of 31.7 ± 0.7 μM) (Figure 6A). The effect of **2** on ConA-stimulated NFAT1 and NFAT-P expression was determined by Western blotting. In unstimulated cells, NFAT was found exclusively in the phosphorylated form, reflecting that calcineurin is inactive under resting conditions. In contrast, in the ConA-stimulated cells in the absence of **2,** the dephosphorylated form of NFAT1 was detected, as well as the phosphorylated form. The presence of 50 μM **2** strongly inhibited the dephosphorylation of NFAT (Figure 6B). Correspondingly, immunofluorescence analysis demonstrated that NFAT1 protein was diffusely distributed in the cytoplasm and was absent from the nucleus. After 48 h of stimulation with 5 μg/mL Con A, the fluorescent NFAT1 translocated to the nucleus. In contrast, the presence of **2** blocked the Con A-stimulated translocation of NFAT1 from the cytoplasm to the nucleus in a dose-dependent manner (Figure 6C). With respect to the effect of **2** on the expression levels of IL-2 mRNA, as determined by real-time quantitative PCR (q-PCR), the transcription level of Con A-stimulated IL-2 mRNA decreased with increasing concentrations of **2** (Figure 6D). ELISA experiments further confirmed the effects of **2** at the IL-2 protein level (Figure 6E). Molecular docking was then performed to further understand the possible binding modes and binding affinities of highly active **2** with the active sites of CN using AutoDock 4.2. As shown in Figure 6F, **2** formed four key hydrogen bonds with residues GLN-130 and HIS-339 with a binding energy of -5.92 kcal·mol−1. In addition, **2** formed

hydrophobic interactions with residues ILE-436, ILE-474, ARG-477, VAL-490, and VAL-498 and intermolecular interactions with residues THR-126, GLN-127, ASP-313, and PRO-340. Therefore, the immunosuppressive activity of **2** is possibly, at least in part, mediated via CaN/NFAT signaling pathway-regulated Con A-stimulated activation of splenocytes (Figure 7), highlighting its potential for use as an effective noncytotoxic natural immunosuppressant.

**Figure 6.** The effect of **2** on calcineurin activity (**A**). Effects of **2** on the expression of ConA-induced NFAT1 protein and analyzed by Western blot (**B**). Effects of **2** on the expression of ConA-induced NFAT protein and analyzed by Western blot (**C**). Effect of **2** in ConA-induced mouse T lymphocytes on IL-2 mRNA expression by q-PCR (**D**) and IL-2 secretion by ELISA (**E**). Molecular docking analysis of the binding of **2** to calcineurin (**F**). \*\* *p* < 0.01 compared to the stimulated group. ## *p* < 0.01 compared to the control group.

**Figure 7.** Schematic view of **2** acting on the CN/NFAT signaling pathway.

#### **3. Materials and Methods**

## *3.1. General Experimental Procedures*

A WYA-2S digital Abbe refractometer (Shanghai Physico-optical Instrument Factory) was used to measure optical rotations. Circular dichroism (CD) spectra were measured on a JASCO J-715 spectra polarimeter. An LTQ Orbitrap XL instrument (Thermo Fisher Scientific, Bremen, Germany) was used to record HRESIMS data. 1D and 2D NMR spectra were recorded on a Bruker AV-400 spectrometer for 1H nuclei and 100 MHz for 13C nuclei in CD3OD. An Agilent 1100 instrument was applied for HPLC analysis and semipreparative HPLC separation. Sephadex LH-20 (18−110 μm, Merck, Darmstadt, Germany), RP-18 gel (25–40 μm, Daiso Inc., Osaka, Japan), or silica gel (200–300 mesh, Qingdao Marine Chemical Inc., Qingdao, China) were employed for column chromatography. Thin-layer chromatography was performed over F254 glass plates (200–400 mesh, Qingdao Marine Chemical Inc., Qingdao, China) and analyzed under UV light (254 and 366 nm). The purity of the isolated compounds was determined by high-performance liquid chromatography (HPLC), which was performed on an Agilent 1200 instrument and a reverse-phase column (4.6 × 150 mm, 5 μm). The UV wavelength for detection was 210 nm. All compounds were eluted with a flow rate of 0.7 mL·min−<sup>1</sup> over a 15-min gradient, as follows: T = 0, 95% B; T = 15, 100% B (A, H2O; B, MeOH) and the purity of tested compounds were proven to exceed 95% (Figure S43).
