Compound 1 was purified as a colorless crystal. According to the HRESIMS analysis showing a sodium adduct ion at m/z 303.1567 (calcd for 303.1567), its molecular formula was determined as C16H24O4, implying 5 degrees of unsaturation. The 1H NMR data (Table 1) of 1 showed obvious signals as three methyl protons (δH 1.08, 1.04, and 0.90), four methylene protons (δH 4.97/4.70, 1.62/2.08, 1.42/1.33, and 1.58), three methine protons (δH 6.05, 3.97, and 1.98) and one methoxy proton (δH 3.31). With the help of DEPT and HSQC spectroscopic analyses, the 13C NMR data (Table 1) of 1 demonstrated the existence of 16 carbon signals that were attributable to three methyls (δC 35.0, 23.1 and 18.5), one methoxy (δC 54.4), four methylenes (δC 69.0, 43.1, 30.5, and 18.4), three methines (δC 125.2, 76.3, and 46.1), four quaternary carbons (δC 135.6, 74.6, 41.7 and 33.8) and one ester carboxyl carbon (δC 175.2). The 1D and 2D NMR data of 1 were highly similar to those of the known 9α-hydroxy-5α-drim-7-ene-6-one-11,12-olide (6) , uncovering 1 and 6 to possess the same drimane sesquiterpenoid core skeleton. The significant difference of 1 and 6 was the existence of one methoxy group linked at C-6 in 1 instead of a conjugated ketone carbonyl (C-6) in 6, as further supported based on the key HMBC correlations (Fig. 2) of 6-OMe (δH 3.31) with C-6 (δC 76.3) and of H-6 (δH 3.97) with C-5 (δC 46.1), C-7 (δC 125.2), and C-8 (δC 135.6). In the NOESY experiment (Fig. 3), the key NOE correlations of H-6 with Me-14 (δH 1.04)/Me-13 (δH 0.90) and of H-5 (δH 1.98) with Me-15 (δH 1.08) suggested that H-6, Me-13 and Me-14 should all be β-oriented, while H-5 and Me-15 were all α-oriented. However, no useful NOE signals could be applied to verify the configuration of C-9. Fortunately, a suitable crystal of 1 was acquired by recrystallization and then furnished for X-ray crystallographic experiment (Fig. 4). According to a Flack parameter of 0.01(3), the absolute configuration of 1 was unequivocally confirmed as 5S, 6S, 9S, and 10S. Accordingly, the absolute structure of 1, named as ustusol F, was defined.
Compound 2, obtained as a white powder, was determined to possess a molecular formula of C15H24O3, as evidenced by its positive HRESIMS data at m/z 275.1618 (calcd for C15H24O3Na+, 275.1618). By comparing its 1H, 13C, and DEPT NMR data (Table 1) with those of the known 6-epi-pereniporin A (4) , we could speculate that both compounds were structural analogues, with the only distinction being that one hydroxy group linked at C-11 was absent in 2, as fully supported by the HMBC correlations (Fig. 2) of H2-11 (δH 3.45) with C-8 (δC 138.2), C-9 (δC 74.6), and C-12 (δC 61.5). Similar NOESY data (Fig. 3) and ECD curves (Fig. 5) between 1 and 2 proved that these two compounds possessed the identical absolute configuration. Accordingly, the absolute structure of 2, named as ustusol G, was defined.
Compound 3 was deduced to have a molecular formula of C15H24O3, as evidenced via its HRESIMS data. By comparing the 1D NMR data (Table 1) of 3 to those of the known ustusol D (11)  (Fig. 1) whose absolute structure was verified via crystallography experiment (Fig. 4), it revealed that both compounds possessed the identical drimane sesquiterpenoid core skeleton, with the only exception that a hydroxy group was linked at C-2 in 11 by C-3 in 3. This conclusion was further corroborated via the key 1H–1H COSY cross-peaks of H2-1/H2-2/H-3, as well as the HMBC correlations of both Me-14 and Me-15 with C-3, C-4, and C-5 (Fig. 2). The NOE cross-peaks (Fig. 3) of Me-15α/H-3/H-5 and H2-11/Me-13β demonstrated that OH-3 was β-oriented in 3. To validate this speculation, the 13C NMR chemical shifts of 3 were predicted at the B972/pcSseg-2 level showing the correlation coefficient (R2) value of 0.9985 (Fig. 6), which completely supported our proposed relative structure. Lastly, the quantum chemical electronic circular dichroism (ECD) calculation was employed for 3. To our expectation, the calculated ECD plot was closely similar to the experimental one (Fig. 7), proclaiming its absolute configuration as 3S, 5S, 9R, and 10S, and this compound was named as ustusol H.
Apart from new compounds 1–3, eleven known congeners were also isolated from A. calidoustus TJ403-EL05 and identified as 6-epi-pereniporin A (4) , 6-epi-O-methyl-pereniporin A (5) , 9a-hydroxy-5a-drim-7-ene-6-one-11,12-olide (6) , 6-dehydroxy-6-oxopereniporin A (7) , strobilactone A (8) , pereniporin B (9) , dendocarbin C (10) , ustusol D (11) , 9a,11,12-trihydroxydrim-7-en-6-one (12) , 12-hydroxyalbrassitriol (13)  and drim-8-en-6β,7a,11-triol (14) , by comparison of their HRESIMS and NMR data with those reported in the literature.
In the bioactivity assay, due to the limited amounts of 3, other compounds (1–2 and 4–14) were tested for anti-inflammatory activity by using LPS-induced murine macrophages RAW264.7 cells. As a result, only compound 14 was found to show an inhibitory effect against the NO release (IC50 = 25.6 μM), and the remaining compounds did not exhibit significant activity with IC50 values of > 40 µM (positive control MG132: IC50 = 0.32 µM).