Crude extracts from S. xiamenensis were assayed for their inhibitory effect on the adhesion of THP-1 cells on a monolayer of WI26 cells. After incubating the crude extract (50 μg/mL) for 3 h, the adhesion of the THP-1 cells was about 15% lower than that with the solvent control (0.01% DMSO). Thus, the extract showed inhibitory adhesion activity and S. xiamenensis was considered as a promising bioactive resource for further chemical analyses. Guided by this bioactivity, a series of alkaloids were detected, using TLC and UPLC-MS, as the main secondary metabolites from the mixture extracted in a solution of ethyl acetate: methanol: formic acid = 80:15:5. The presence of pseudo-molecular ions [M + H]⁺ at m/z 392.2069 and [M − H]⁻ at m/z 390.1886 in positive and negative ion modes, respectively, in the mass spectra, led to the purification of the main component, compound 1.

The molecular formula of 1 is C₂₁H₂₉NO₆ based on HRESIMS, which is in accordance with the previous literature [14]. In the ¹H NMR spectrum of 1, three aromatic protons with signals at δ_H 7.63 (1H, d, J = 8.4 Hz), 6.81 (1H, d, J = 8.4 Hz), 7.76 (1H, s), and 7.78, (1H, d, J = 7.8 Hz), as well as a high-field double-bond proton at δ_H 5.12 (1H, dd, J = 7.1, 1.3 Hz) were detected (Table 1). Carbon signals in the ¹³C NMR spectrum consisted of six aromatic carbons, plus two olefinic carbons, two carbonyl groups, two hydroxylated carbon signals at δ_C 79.8, 66.3, and a methylene at δ_C 31.2 (Table 1), which is in accordance with the structure of a benzopyran moeity, attached to two side chains, threonine and isoprenylmethyl group. The HMBC correlations (Table 1) further confirmed the planar structure of 1, named N-[[3,4-dihydro-3-hydroxy-2-methyl-2-(4′-methyl-3′-pentenyl)-2H-1-benzopyran-6-yl]carbonyl]-threonine [14].

The absolute configurations of 1, with four stereogenic centers, were identified by Mosher’s method, Marfey’s reagent, NOESY experiment, and QM calculations. Compound 1 was treated with (R)- and (S)-α-methoxy-α-phenylacetic acid (MPA) to form (R)- and (S)-MPA diesters. Diagnostic Δδ^RS sigals of the (R)- and (S)-MPA diesters resulted in negative Δδ^RS values for 4-H₂ (Table 1), which were in contrast to the positive Δδ^RS values for 9-H₂, 10-H₂, 11-H, 13-H₃, 14-H₃ and 15-H₃, which are all related to the dihydropyran moiety. These data are in accordance with a 3S configuration of 1. The absolute configuration of the neighboring position 2 was then determined by elucidating the relative configuration related to 3S. The NOE correlation between 2-H₃ and 4-H_a demonstrated that 2-H₃ was in an axial orientation. The other NOE interaction between 2-H₃ and 3-H demonstrated that 3-H was equatorially oriented and 3-OH was in a trans configuration to 2-H₃ (Figure 2). The relative configuration of 2-H₃, 3-H and 4-H_a/e were further confirmed from the values of the coupling constants, which were 4-H_a (2.71, dd, 17.3, 7.4) and 4-H_e (2.98, dd, 17.3, 5.2).

QM calculations on a model molecule of 2,2-dimethyl-3,4-dihydro-2H-pyran-3-ol indicate that, compared to that at the equatorial position, the 3-OH group prefers the axial position by about 1.2 kcal/mol both in the gas phase and in DMSO solvent due to the gauche effect [15]. In the gauche conformation, the H-H distances of 2-H₃ are as close as 2.55 Å to 3-H, and 2.30 Å to 4-H_a. The results are consistent with the key NOE correlations (Figure 2). Based on the absolute configuration of C-3 and the relative configuration of C-2, we therefore identified the absolute configuration of position 2 to be 2S.

As far as the other two stereocenters in the threonine moiety, diagnostic Δδ^RS sigals of the (R)- and (S)-MPA diesters resulted in negative Δδ^RS values for 3'-H and 2'-NH (Table 1), which were in contrast to the positive Δδ^RS values for 5'-H. The configuration of the position 4' was thus assigned to be R. In order to determine the configuration at the position 3', 1 was hydrolyzed in hydrochloric acid and then derivatized with 1-fluoro-2,4-dinitrophenyl-5-L-alanine (FDAA, Marfey’s reagent) [16]. Based on the results of hydrolyzation of 1 and derivatization with Marfey’s reagent afterwards, the threonine moiety was assigned as the naturally occurring structure of L-threonine [(2S,3R)-2-amino-3-hydroxy butanoic acid]. The absolute configuration of 1 was therefore determined to be N-[[3,4-dihydro-3S-hydroxy-2S-methyl-2-(4′R-methyl-3′S-pentenyl)-2H-1-benzopyran-6-yl]carbonyl]-threonine (1).

Fibrosis is characterized by the overgrowth, hardening, and/or scarring of various tissues and is attributed to an excess deposition of ECM components including collagen [1]. To combat this, an ability to slow down rapid fibrotic response during typical wound repair and tissue regeneration processes would seem to be a reasonable intervention, with the goal to avoid an accumulation of non-functional fibrotic tissue and thereby enable transformation to more regenerative conditions [3]. We thus investigated the anti-proliferative activity of 1 to this end. Lung fibroblast WI26 cells were exposed to 1, at a concentration of 30 µg/mL, for 0, 2, 3, 4 and 6 days, and the optical density (OD) λ = 450 nm was monitored as a function of time. As illustrated in Figure 3, compound 1 significantly inhibited the proliferation of WI26 cells. The extent of this inhibition increased with time (Figure 3), from a decrease of 7% at day 2 to 27% at day 6, compared with solvent control. Compound 1 therefore indeed exhibits anti-proliferative effects on lung fibroblast cells.

Monocytes and macrophages have been shown to play pivotal roles in inflammation and fibrosis by modulating the secretion of cytokines as well as recruiting and activating fibroblasts and other inflammatory cells [17]. Direct contact between monocytes/macrophages and fibroblasts is important for the initiation, perpetuation, and resolution of fibrosis [18]. We therefore also evaluated the anti-adhesion bioactivity of 1. WI26 cells were first grown as a cell monolayer on a rigid tissue culture support, and then monocytic THP-1 cells were added to this monolayer. The number of adhered cells was then measured by optical microscopy after a 3 h incubation with either 1 (30 μg/mL) or 0.1% DMSO. As shown in Figure 4, there were obvious differences between the experiments with 1 present compared with those of the solvent control (Figure 4). Namely, the adhesion of THP-1 was about 41% lower in the presence of 1 than that in the solvent control. Thus, compound 1 exhibits significant anti-adhesion bioactivity for monocytes on fibroblasts.

It is widely known that mechanical force plays a contributing role in inflammation and fibrotic processes [19,20]. Not surprisingly, the contractible force generated by fibroblasts is a focus of much research [21]. Activated fibroblasts express many stress fibers which can facilitate wound repair [3] and fibrogenesis [22], while at the same time mechanical stress can promote inflammation and fibrosis by activating fibroblasts and inflammation cells [19,23]. In order to evaluate the anti-contractility activity of 1, three-dimensional collagen lattices were used as a richer, perhaps more physiologically relevant physical environment, that can enable direct investigation of the interactive process of fibroblast-collagen matrix remodeling [24]. WI26 cells were seeded in free-floating collagen gel lattices, followed by the application of 1 or solvent control. As shown in Figure 5A, the presence of compound 1 significantly delayed contraction of the collagen gel. After 1, 2, 3 and 6 h, the contraction was markedly decreased compared to solvent control by 21%, 22%, 18% and 8%, respectively (Figure 5B).

Today, fibrotic diseases are becoming more threathening than ever, due to enviromental pollution. It has been reported that high levels of air pollutant paticles can induce chronic fibrotic responses in airways [25]. Identification of bioactive molecules that affect fibrosis is thus becoming a promising theraputic area and also an urgent task. For example, the synthetic drug TAS-301, a constrictive remodeling regulator, was shown to have high inhibitory potency on coronary artery stenosis, mainly due to inhibition of adventitial fibroblast proliferation and the contractile ability of myofibroblasts [26]. During excessive fibrogenesis, it is well known that self-perpetuating circuits of inflammation and ECM accumulation, and constriction by inflammation and mechanical force, directly influence the development of fibrotic diseases [4]. Hence, the development of leading compounds aimed at these circuits may target the effects on these diseases, while avoiding severely disturbing physiological processes induced by steroidal anti-inflammatory drugs.

Fibroblasts play pivotal roles in establishing and maintaining these circuits and are in fact the main ECM generating cells in fibrotic diseases [27]. During these diseases, resident fibroblasts proliferate and are activated as a result of stimulation by injury or inflammation. The activated fibroblasts contribute to ECM production and inflammatory modulation [28]. Radiotherapy and chemotherapeutics are useful for treating hypertrophic scar and keloid development by inducing apoptosis of fibroblasts [29]. Compound 1 showed inhibitory effects on fibroblast proliferation, and thus may contribute to the alleviation of excessive fibrotic accumulation.

Activated fibroblasts may also convert into myofibroblasts, which express many more stress fibers and generate much more mechanical tension [1]. The mechanical force generated by endogenous fibroblasts has been found to play important roles in the activation and secretion of inflammatory cells [19,23,30]. It has been demonstrated that reduction of mechanical stress can alleviate inflammation and fibrosis in vivo [12]. Therefore, mechanical force and local inflammation together comprise one of the most important self-perpetuating inflammatory circuits that, underlies many fibrotic diseases. Compound 1 can reduce the contractile capability of fibroblasts, thus indicating a potential to inhibit this key aspect of fibrogenesis.

The interaction of monocytes and fibroblasts is known to play an important role in inflammation [17,31]. Direct contact between monocytes and fibroblasts can faciliate the activation of both cells [32]. Compound 1 blocked the adhesion of fibroblasts and monocytes, which may ameliorate excessive inflammatory processes in fibrotic diseases. In previous investigation, 1 was reported as an antagonist of the LFA-1/ICAM-1 interaction [14]. Since LFA-1/ICAM-1 can mediate direct contact of inflammatory cells and fibroblasts [33], the LFA-1/ICAM-1 binding inhibitory effect may explain the results of the adhesion experiment.

In conclusion, 1 has multiple biological effects on lung fibroblast properties, including proliferation, contractile capacity and interaction with monocytes, which are likely to be involved in the establishment of positive feedback circuits that lead to excessive fibrosis.

Although pathological fibrosis of different organs is common in many diseases [1], there are presently very few therapeutic options [34]. Anti-inflammatory drugs such as corticosteroids can prevent pathological fibrosis [29,35], but long-term use and high dosage are unsuitable owing to their extensive adverse effects [36,37]. Superficial compression is useful for preventing the development of skin hypertrophic scars [29], but this procedure cannot be used in visceral organs. The “vicious cycle” of fibroproliferation is one of the most important characteristics of pathological fibrosis [19,23]. Hence, small molecules that target the positive feedback circuits may prove tremendously effective in exhibiting antifibrotic activities without severe side effects. The results presented herein illustrate that compound 1 may serve as a useful antifibrotic leading compound for treating pulmonary fibrosis and other pathological fibrotic diseases.

¹H and ¹³C NMR spectra were recorded with Bruker DRX-500 and Avance III-600 NMR spectrometers with the solvent as an internal standard (DMSO-d₆ δ = 2.51 and 40.0 ppm, respectively). Chemical shifts (δ) expressed in parts per million (ppm) and coupling constants (J) are reported in Hertz (Hz). LC-HRMS was performed on a Waters ACQUITY UPLC system equipped with a binary solvent delivery manager and a sample manager, coupled with a Waters Micromass Q-TOF Premier Mass Spectrometer equipped with an electrospray interface (Waters Corporation, Milford, MA). Column chromatography was performed with silica gel (200–300 mesh, Qingdao Marine Chemical, Inc., Qingdao, China), silica gel H (10–40 μm, Qingdao), Sephadex LH-20 (40–70 μm, Amersham Pharmacia Biotech AB, Uppsala, Sweden) and Lichroprep reversed-phase RP-18 silica gel (40–63 μm, Merck, Darmstadt, Germany). The chemical reagents used for chromatography were purchased from Shanghai Chemical Works Co. Ltd. (Shanghai, China). Optical rotations were recorded on a JASCO P-2000 polarimeter. CD spectra were taken on a J-815 spectropolarimeter (JASCO, Gross-Umstadt, Germany) at room temperature. An Agilent 1200 series system was used for the analytical scale method. Analytical HPLC was carried out on an Welch XB-C18 column (4.6 × 150 mm, 5 μm, Welch Materials, Inc., USA), flow 1 mL/min. Organic solvents for HPLC were analytical grade and were purchased from Shanghai ANPEL Scientific Instrument Co. Ltd. (Shanghai, China).

Streptomyces xiamenensis strain 318 was isolated from a mangrove sediment sample collected in the national mangrove reserve in Fujian province of China [13]. Spores of strain 318 were inoculated on GYM (Glucose 4 g, Yeast extract 4 g, Malt extract 10 g, CaCO₃ 2 g, Agar 12 g, Distilled water 1 L, adjusted to pH 7.2 with KOH before adding agar.) plates were then incubated at 28 ± 1 °C for 12 days before harvesting.

The 30 L plate culture of S. xiamenisis was extracted with a solvent mixture of ethyl acetate/methanol/acetic acid (80:15:5, v:v:v) at room temperature overnight. The supernatant was filtered and the residue was then extracted twice as described above. The total supernatants were combined and concentrated under vacuum at 37 °C in order to remove the organic part to afford a crude extract (12 g). Half of the crude extract was subjected to column chromatography on silica gel, eluted by CH₂Cl₂:MeOH (gradient from methanol, 100:1, 50:1, 20:1, 10:1, 5:1 to MeOH). The fraction [eluted with CH₂Cl₂:MeOH (10:1, v:v)] was collected and purified by repeated column chromatography on Sephadex LH-20, which was washed by MeOH. The fractions were collected and guided by HPLC fingerprints to find peaks containing the target UV profile (λ_max 206, 260 nm). Finally, the combination was separated by ODS column (eluted with MeOH:H₂O = 75:25) to obtain compound 1 (15 mg).

N-[[3,4-dihydro-3S-hydroxy-2S-methyl-2-(4′R-methyl-3S-pentenyl)-2H-1-benzopyran-6-yl]carbonyl]-threonine (1): Yellow amorphous powder (MeOH); [α]²²_D +39.5° (c 0.044, MeOH); UV λ_max (MeOH) 206, 260 nm; CD (MeOH) Δε₂₀₁ +0.15, Δε₂₀₂ +0.55, Δε₂₀₅ +0.3, Δε_207.5 +0.5, Δε₂₂₉ −0.05, Δε₂₅₉ +0.28; ¹H and ¹³C NMR data, see Table 1; HRESIMS m/z 392.2069 [M + H]⁺, (calcd for C₂₁H₃₀NO₆, m/z 392.2073), 390.1886 [M − H]⁻, (calcd for C₂₁H₂₉NO₆, m/z 390.1917).

(R)-MPA (4.0 mg), dicyclohexylcarbodiimide (DCC), and catalytic N,N-dimethylpyridin-4-amine (DMAP) were added to a CHCl₃ solution of compound 1. The mixture was stirred for 2 h to yield (R)-MPA ester of 1. The (S)-MPA ester (1.1 mg) was obtained by the same protocol as for (R)-MPA ester (1.2 mg).

Compound 1 (1.0 mg) was hydrolyzed in 1:1 hydrochloric acid-acetic acid at 100 °C for 18 h, 500 μL H₂O was added and the mixture was lyophilized overnight [16]. Then 200 μL of 1 M NaHCO₃ and 400 μL of 38.7 μM FDAA (Sigma-Aldrich, St. Louise, MO, USA) in acetone were added to the mixture for derivatization. The solution was vortexed and incubated at 40 °C for 60 min. Reactions were quenched by addition of 100 μL of 2M HCl. Samples were diluted 1:2 with HPLC initial mobile phase. The derivatization of L-Threonine (Sangon, Shanghai, China) and D-Threonine (Sangon, Shanghai, China) were prepared in the same way. Then the samples were analyzed by HPLC with UV detection at 340 nm and further confirmed by UPLC-MS.

The WI26-SV40 transformed human lung fibroblast [American Type Culture Collection (ATCC) CCL 95.1] cells were maintained in Dulbecco's Modified Eagle Medium (DMEM, Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS, GIBCO-BRL, Gaithersburg, MD, USA), 2 mM L-glutamine, 100 U/mL penicillin and 100 μg/mL streptomycin (Solarbio, Beijing, China). THP-1 (human acute monocytic leukemia cell line) cells were maintained in RPMI 1640 meted with 10% fetal bovine serum (FBS, GIBCO-BRL, Gaithersburg, MD, USA), 2 mM L-glutamine, 100 U/mL penicillin and 100 μg/mL streptomycin (Solarbio, Beijing, China). THP-1 (human acute monocytic leukemia cell line) cells were maintained in RPMI 1640 medium (Gibco, Grand Island, NY, USA), supplemented with 10% FBS, 2 mM L-glutamine, 100 U/mL penicillin, 100 μg/mL streptomycin, and 0.5 mM/L β-mercaptoethanol (Gibco, Grand Island, NY, USA). Both kinds of cells were incubated at 37 °C humidified atmosphere of 5% CO₂ in air.

The effect of compound 1 on cell proliferation was determined using a standard Cell Counting Kit-8 (CCK-8, Dojindo, Kumamoto, Japan) assay according to the manufacturer’s instruction. WI26 cells at 70%–80% confluent were trysinized by 0.25% trypsin-0.02% EDTA solution, centrifuged and re-suspended in DMEM supplemented with 10% FBS and antibiotics and seeded in 96-well plates (100 μL/well) at an initial density of 2.5 × 10⁴ cells/mL. The medium was replaced 24 h later by fresh DMEM with 10% FBS and antibiotics containing 30 μg/mL compound 1 or 0.1% DMSO (AppliChem, Darmstadt, Germany), then the medium was refreshed and the cells viability was measured by using CCK-8 solution at day 0, 2, 3, 4 and 6, respectively. Proliferation measurement was applied by adding 10 μL CCK-8 solutions to each well and incubating at 37 °C for 1 h. The OD values of each well were measured at the primary wavelength λ = 450 nm by using a Microplate Spectrophotometer (PowerWaveXS, BioTek, Seattle, WA, USA). Data are shown as means ± standard deviation (SD) of three independent experiments, each performed in triplicate.

Two days before adhesion assay, WI26 cells were seeded in 24-well plates (500 μL/well) at a density of 3 × 10⁵ cells/mL and allowed to grow 24 h to confluence. The medium of THP-1 and WI26 cells was replaced by DMEM supplemented with 0.5% FBS and antibiotics 24 h before the adhesion experiment. The culture of THP-1 cells was centrifuged at 1000 rpm for 5 min. The supernatant was discarded. The THP-1 cells were re-suspended in DMEM with 10% FBS and antibiotics, then diluted to 5 × 10⁵ cells/mL. The medium of WI26 cells in 24-well plates was discarded to get a monolayer of WI26 cells in each well, then 500 μL THP-1 suspension was added into each well as previously published [38,39]. For the biotest groups, 30 μg/mL compound 1 or 50 μg/mL crude extract of S. xiamenensis culture was added in each well, respectively. For the control groups, 0.1% DMSO was used as solvent control for 1 and 0.01% DMSO as solvent control of crude extract, respectively, according to the different final DMSO concentration of 1 and crude extract. Two kinds of cells were co-cultured for 3 h, then the medium and the non-adherent THP-1 cells were removed by washing with PBS three times. Adherent THP-1 cells were quantified microscopically at 400-fold magnification in 4 random visual fields of each well and photographed with a digital camera (LY-WN-HPCCD, China) mounted on a microscope (Olympus, CKX31, Japan). Data are shown as means ± standard deviation (SD) of three independent experiments, each performed in triplicate.

WI26 cells at 70%–80% confluent were trysinized by 0.25% trypsin-0.02% EDTA solution and seeded at a density of 4 × 10⁵ cells/mL into 32 mm bacteriological plates (2 mL/dish) in DMEM supplemented with 10% FBS, antibiotics, sodium ascorbate (50 μg/mL), and containing 0.3 mg/mL of acid-extracted collagen I from newborn calf skin (IBFB, Leipzig, Germany) as previously described [40]. For the biotest groups, 30 μg/mL pure extract of compound 1 was added in each well; for the control groups, 0.1% DMSO was used as solvent control. The cultures were placed at 37 °C for 60 min to allow collagen polymerization, and then the gels from plates were released by tilting the plates slightly. Gradual lattice contraction was monitored by measuring the gel area of triplicate setups at successive time points up to 6 h. Data are shown as means ± standard deviation (SD) of three independent experiments, each performed in triplicate.

Statistic differences were calculated using student’s paired t-test at a significance level of p < 0.05 to 0.001.

The QM calculations were carried out with the M06-2X hybrid meta DFT functional with the 6 − 31 + G(d,p) basis set using Gaussian 09. Conformations were fully optimized and characterized by harmonic vibrational frequency analysis. The solvation energies in DMSO were calculated with Truhlar’s SMD solvation model [41,42,43].