Antiproliferative Sorbicillinoids From the Deep-Sea-Derived Penicillium allii-sativi

Two new (1–2) and three known (3–5) sorbicillinoids were isolated from the deep-sea-derived fungus Penicillium allii-sativi MCCC 3A00580. Compounds 1 and 2, named sorbicatechols C and D, were two new hybrid dihydrosorbillinoids. Their structures were established mainly by spectroscopic analyses and electronic circular dichroism (ECD) calculations. All five isolates were tested for antiproliferative activities against four tumor cell lines of MCF-7, HT-29, HuH-7, and LNCap. Compounds 2 and 5 inhibited HT-29 cells in a good dose-dependent manner. Mechanism investigation uncovered that they could significantly induce cell cycle G2-M phase arresting by increasing the protein levels of p-H3 and cyclin B1.


INTRODUCTION
Sorbicillinoids are hexaketide metabolites that possess complex and highly oxygenated frameworks. Structurally, they can be divided into four groups: monomeric, dimeric, trimeric, and hybrid sorbicillinoids (Harned and Volp, 2011;Meng et al., 2016). The unique structural features of the sorbicillinoids make them attractive candidates for developing new pharmaceutical and agrochemical agents (Abe et al., 2000(Abe et al., , 2001Fahad et al., 2013;Meng et al., 2019;Kahlert et al., 2020). Up to now, around 100 sorbicillinoids have been reported from fungi, especially marine Penicillium (Meng et al., 2016). During our ongoing search for structurally novel and biologically interesting secondary metabolites from deep-sea-derived microorganisms (Yang et al., 2013;Niu et al., 2017Niu et al., , 2020Xie et al., 2019a), the rice static fermentation extract of Penicillium allii-sativi MCCC 3A00580 exhibited potent in vitro antitumor activity. Previously, three meroterpenoids were obtained and andrastones A showed significant inhibitory effect against HepG2 tumor cells by activating caspase-3 and regulating the transcriptional activation function of RXRα (Xie et al., 2019b). Further investigation on this strain led to the discovery of two new and three known sorbicillinoid derivates (Figure 1). Structurally, compounds 1 and 2 are two novel hybrid dihydrosorbicillinoids. All isolates were tested for antiproliferative bioactivities, compounds 2 and 5 could inhibit HT-29 tumor cells in a good dose-dependent manner. Herein, we report the isolation, structures, and bioactivities of these compounds.

ECD Calculation
As reported previously (Niu et al., 2020), conformational analyses were carried out via random searching in the Sybyl-X 2.0 using the MMFF94S force field with an energy cutoff of 2.0 kcal/ mol. Subsequently, the lowest energy conformer was re-optimized using DFT at b3lyp/6-31+g(d,p) level in MeOH by the GAUSSIAN 09 program. The energies, oscillator strengths, and rotational strengths (velocity) of the first 30 electronic excitations were calculated using the TDDFT methodology at the cam-b3lyp/ TZVP level using the polarizable continuum model in MeOH. The ECD spectrum were simulated by the overlapping Gaussian function (half the bandwidth at 1/e peak height, σ = 0.35, UV correction = 23 nm).

Western Blotting
Cell lysates were boiled in sodium dodecyl sulfate (SDS) sample loading buffer, resolved by 10% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose. The membranes were blocked in 5% milk in Tris-buffered saline and Tween 20 [TBST; 10 mM Tris-HCl (pH 8.0), 150 mM NaCl, and 0.05% Tween 20] for 1 h at room temperature. After washing twice with TBST, the membranes were incubated with appropriate primary antibodies in TBST for 1 h and then washed twice, probed with horseradish peroxide-linked anti-immunoglobulin (1:5000 dilution) for 1 h at room temperature. After three washes with TBST, immunoreactive products were visualized using enhanced chemiluminescence reagents and autoradiography.

RESULTS AND DISCUSSION
Compound 1 was isolated as a yellow oil.  Figure 2). These two segments along with the ABX aromatic moiety could be connected by the HMBC correlations of Me-21 to C-1 (δ C 64.1 s)/C-2 (δ C 196.0 s)/C-6 (δ C 209.6 s)/C-7 (δ C 45.3 d), Me-22 to C-4 (δ C 41.1 d)/C-5 (δ C 72.9 s)/C-6, H-4 (δ H 3.15, t, J = 2.8 Hz) to C-2/C-3 (δ C 112.5 s)/C-9 (δ C 178.7 s), H-7 (δ H 3.11, dd, J = 10.6, 5. deduced the Z-orientation of the olefinic bonds of C-3/C-9. Since the chemical shifts of H-12 and H-13 were overlapped, it was impossible to assign the configuration by their coupling constants. However, the NOESY correlations of Me-14 to H-12(H-13) but not H 2 -11 could suggest the E-orientation of C-12/C-13. On the basis of the above evidences, the relative configuration of compound 1 was, therefore, determined undoubtedly. To further assign its absolute stereochemistry, the theoretical calculation of the electronic circular dichroism (ECD) was conducted. As shown in Figure 3, the experimental ECD spectrum of compound 1 showed the same Cotton effects as those of 1R,4S,5S,7R-1 (1a). Therefore, the absolute stereochemistry of compound 1 was assigned as 10,11-dihydrosorbicatechol A, and named sorbicatechol C.
Compound 2 was also isolated as a yellow oil. The molecular formula was assigned as C 23 H 28 O 6 on the basis of the HRESIMS at m/z 399.1790 [M−H] − (calcd for C 23 H 27 O 6 , 399.1808). The 1 H and 13 C NMR spectroscopic data ( Table 1) indicated 23 carbon resonances including two methyl singlets, one methyl doublet, three methylenes, seven methines (two olefinic and three aromatic), nine quaternary carbons (two ketones at δ C 196.0 and 209.6, two olefinics at δ C 112.5 and 178.6, and three aromatics at δ C 132.8, 145.5, and 147.2), and one methoxyl. These signals were very similar to those of compound 1, except that the methoxyl group at the C-18 position in compound 1 was changed to a hydroxy moiety in compound 2. This was evidenced by an additional hydroxy unit and the absence of a methoxyl group in compound 2. Further confirmation could FIGURE 4 | Antiproliferative effects of compounds 1-5 against four cancer cells. Cells were treated with five compounds in a dose-course (12.5, 25, and 50 μM) for 48 h, which were analyzed by MTT assays to compare the cell proliferation levels. All the bar graphs represent mean ± SEM of three independent experiments. **p < 0.01, ***p < 0.001, ****p < 0.0001 vs. the DMSO group.
be found by the HMBC correlations of the hydroxy at δ H 8.92 s to C-19 (δ C 115.2 d). By detailed analysis of its heteronuclear single quantum correlations (HSQC), correlation spectroscopy (COSY), heteronuclear multiple-bond correlation (HMBC), and nuclear overhauser effect (NOESY) spectroscopic data (Supplementary Table S1 of the Supporting Information), the structure and relative configuration of the compound 2 were determined. The absolute configuration of compound 2 was assigned to be identical to that of compound 1 based on their similar specific rotations and ECD data. Compound 2 was Hybrid sorbicillinoids are derived from either a Diels-Alder or a Michael reaction of a monomeric sorbicillinoid diene and a second non-sorbicillinoid dienophile (Meng et al., 2016). Noteworthily, compounds 1 and 2 are two novel hybrid sorbicillinoids derived from dihydrosorbicillinols which were very rarely in nature. Their biogenetic origin might be involved ferulic acid and 2',3'-dihydrosorbicillinol via an intermolecular Diels-Alder condensation followed by a decarboxylation (Peng et al., 2014).
All five isolates were subjected to the preliminary screening tests for antiproliferative activity against MCF-7, HT-29, HuH-7, and LNCap tumor cells. Compounds 2 and 5 could inhibit the proliferation of HT-29 cells in a dose-dependent manner (Figure 4).
Interestingly, after treatment by compounds 2 and 5, many HT-29 cells became more rounded and less adherent under the microscope, which suggested cell cycle was disrupted and resulted in M-phase arresting. Therefore, further investigation of the cell cycle was conducted by flow cytometry. As expected, they significantly blocked 40.96 and 41.69% tumor cells in G2-M phase, respectively ( Figure 5). Therefore, compounds 2 and 5 could blocked HT-29 cells in G2-M phase.
Phosphorylation of histone H3 (p-H3) is one of the methods of histone modification. It occurs at specific periods and chromosomal sites during mitosis and meiosis. Cyclin B1 also plays an important role in cell cycle regulation. The overexpression of cyclinB1 can promote G2/M phase conversion and even lead to uncontrolled cell proliferation and malignant transformation (Hartwell and Kastan, 1994;Hwang et al., 1995). Therefore, the expression of M phase markers p-H3 and cyclin B1 was further detected. As shown in Figure 6, they could significantly increase the protein levels of p-H3 and cyclin B1 in a dose-dependent effect, confirming compounds 2 and 5 indeed induced M phase arresting. The effective concentrations of compounds 2 and 5 were 30 and 5 μM, respectively.
Although several examples were found for monomeric, dimeric, and trimeric sorbicillinoids in which C-2'/C-3' double bonds were reduced, compounds 1 and 2 are the first two examples for hybrid sorbicillinoids. Therefore, the discovery of compounds 1 and 2 has expanded the diversity and complexity of sorbicillinoids. Compounds 2 and 5 could inhibit HT-29 tumor cells in a good dose-dependent manner. They significantly induced cell cycle G2-M phase arresting by increasing the protein levels of p-H3 and cyclin B1.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

AUTHOR CONTRIBUTIONS
C-LX performed chemical investigations. DZ conducted biological experiments. TL, Z-HH, and Q-XY assisted C-LX's chemical FIGURE 5 | Effects of compounds 2 and 5 on cell cycle by flow cytometry. HT-29 cells were treated with compound 2 (50 μM, 48 h) or 5 (30 μM, 48 h), which were stained with DAPI for the following cell cycle FCM assay. The blue part indicated the cells in G0-G1 phase, the green part showed the G2-M phase cells, and the green numbers showed the G2-M phase proportion.