In Vivo Angiogenesis Screening and Mechanism of Action of Novel Tanshinone Derivatives Produced by One-Pot Combinatorial Modification of Natural Tanshinone Mixture from Salvia Miltiorrhiza

Background Natural products present in low quantity in herb medicines constitute an important source of chemical diversity. However, the isolation of sufficient amounts of these low abundant constituents for structural modification has been a challenge for several decades and subsequently halts research on the utilization of this important source of chemical entities for drug discovery and development. And, pro-angiogenic therapies are being explored as options to treat cardio-cerebral vascular diseases and wound healing recently. The present study investigates the pro-angiogenic potential of tanshinone derivatives produced by one-pot synthesis using zebrafish model. Methodology/Principal Findings In order to address the difficulty of chemical modification of low abundant constituents in herb medicines, a novel one-pot combinatorial modification was used to diversify a partially purified tanshinone mixture from Salvia miltiorrhiza. This led to the isolation of ten new imidazole-tanshinones (Compounds 1–10) and one oxazole-tanshinone (Compound 11), the structures of which were characterized by spectroscopic methods in combination with single-crystal X-ray crystallographic analysis. The angiogenesis activities of the new tanshinone derivatives were determined in an experimental model of chemical-induced blood vessels damage in zebrafish. Of all the tested new derivatives, compound 10 exhibited the most potent vascular protective and restorative activity with an EC50 value of 0.026 µM. Moreover, the mechanism underlying the pro-angiogenesis effect of 10 probably involved the VEGF/FGF-Src-MAPK and PI3K-P38 signalling pathways by gene expression analysis and a blocking assay with pathways-specific kinase inhibitors. Conclusions/Significance Taken together, our study demonstrated the more distinctive pro-angiogenic properties of 10 than other tanshinones and revealed 10 has potential for development as a pro-angiogenic agent for diseases associated with insufficient angiogenesis. Our results highlighted the great potential of adopting a newly modified one-pot approach to enhance the chemical diversity and biological activities of constituents from natural products regardless of their abundances.


Introduction
Over 50% of commercial available drugs originated from the optimization of lead compounds isolated from natural products [1]. However, most reported chemical modification studies have mainly focused on the abundant compounds present in natural products as current organic synthesis technology does not efficiently modify small amount of compounds as starting materials. Thus, in the past, these minor compounds which have a higher structural diversity than the major compounds in natural products have been overlooked [2]. The development of an alternative novel approach is needed to address the bottleneck in the structural modification of these minor natural constituents.
Recently, there has been increasing interest in the application of the one-pot synthesis approach to create chemical structural diversity, in which chemical reactions were carried out by mixing starting materials including parental compounds and reactants into a single container to undergo multiple chains of modifying reactions without isolating intermediates. One-pot synthesis is also regarded as ''green'' mild approach as it improves the efficiency of chemical reactions and minimizes chemical waste generation by eliminating complicated separation and purification processes of the intermediate chemical compounds. For instance, even compounds with complex structure and multiple stereocentres, such as Oseltamivir, an anti-viral drug which is regarded as the last resort for controlling H5N1 infuenza pandemic, is successfully produced by sequential one-pot syntheses [3]. Interestingly, onepot chemical diversification has been performed on an inactive natural product mixtures; however, only one semi-synthetic pyrazole was isolated [4]. A proof of concept study is necessary to fully utilize the chemical diversity of natural products in herbal medicine and generate an array of drug-like derivatives for subsequent screening using a partial fractionated chemical mixture instead of a pure compound as a starting material. Thus, both the major and minor constituent would be diversified simultaneously, and the conventional intensive labour and inefficient chemical isolation processes could be bypassed.
Salvia miltiorrhiza (Danshen), is widely used botanical material in traditional Chinese medicine for treating cerebral and cardiovascular diseases such as stroke [5], coronary heart disease [6] and hyperlipidemia [7]. Tanshinones are naturally occurring abietanetype diterpenoid naphthoquinone, which exist widely in Salvia species and could be considered as a natural combinatorial library [8,9]. More than forty tanshinones have been isolated and identified. For instances, tanshinone IIA (Tan IIA, 0.29%), cryptotanshinone (CPT, 0.23%), tanshinone I (Tan I, 0.11%) are regarded as the major constitutes [10,11,12] while tanshindiol C, tanshinone VI, trijuganone A are examples of minor tanshinones [13]. However, structural modifications were exclusively carried out on the major tanshinones to improve their potency, solubility and bioavailability through a variety of modifications such as sulfation [14], acylation [15] and dimerization [16]. There has been no reported research on the modification of minor tanshinones. In particular, low abundant tanshinones in Salvia miltiorrhiza remain an unexplored natural source of starting materials for the discovery and development of new therapeutics.
Angiogenesis is a complex and critical step of various physiological and pathological processes. Many diseases such as cancers, psoriasis associated with excessive angiogenesis [17,18]. Nevertheless, insufficient angiogensis also occurs in numerous human disorders (e.g. stroke and myocardial infarction) due to dysfunctional blood vessel formation [19,20,21]. Angiogenesis therapies focused on the unique medical approachs to help people lead healthier lives by restoring balance to blood vessels growth. Thereinto, pro-angiogenic therapies are being explored as options to treat cardio-cerebral vascular diseases and wound healing. Therefore, improvement in endothelial cells function and enhancement of the de novo blood vessels formation are important steps in the recovery process following severe ischemic and traumatic situations [22].
One of major limitations of commonly used in vitro assays which can precisely reflect the pharmacodynamic change of a specific mode of action to a specific cell type and/or molecular target, is that these assays do not facilitate the assessment of absorption, distribution, metabolism, and excretion (ADME) [23]. Recently, zebrafish (Danio rerio) has become an ideal environmentally friendly and economical in vivo drug screening model for improving various drug-like properties such as pharmacodynamic, absorption and metabolism simultaneously [24]. In addition, our previous study suggested that phase I and II drug metabolism systems in zebrafish are highly conserved compared with those in rodents and mammals [25]. Furthermore, zebrafish is highly similar to human at the anatomical and physiological levels; therefore many human diseases including cardiovascular diseases [26], cancer [27], leukemia [28], have been successfully reproduced in zebrafish. More importantly, the studies on pharmaceutical and biomedical discovery and developement in zebrafish can provide safer commerial products by bridging the gap in the translation of a scientific hypothesis to a human compatiable prototype [29]. The Tg(fli-1a:EGFP)y1 zebrafish embryo provides a powerful tool for the discovery of novel angiogenesis modulating agents allowing observation of the responses of live embryo to drugs in real time. Moreover, our previous study established a novel chemicalinduced blood vessels loss model which can mimic certain pathological conditions relating to angiogenesis deficiencies. Thus, a more physiologically relevant assay is required for identifying pro-angiogenesis agents against diseases asscoiated with angiogenesis deficiencies [30]. The aim of the present study was to identify new angiogenic tanshinone derivatives from the chemically modified total tanshinone extract of Salvia miltiorrhiza using the zebrafish angiogenesis assays.
In present study, new tanshinone derivatives were synthesized by mixing total natural tanshinones with reactants using a one-pot synthesis approach. A series of new tanshinone derivatives were isolated and subjected to both structural elucidation and zebrafish angiogenesis assays. Furthermore, the involvement mechanism underlying the angiogenesis effect of the new tanshinone derivatives was investigated using real-time qPCR analysis and testing for blocking effects with pathway-specific inhibitors.

General Remarks in Chemistry Assays
Ultraviolet (UV) spectra were determined in CH 3 Cl 3 on a Jasco V-550 UV/vis spectrophotometer. ESI-MS spectra were carried out on a Finnigan LCQ Advantage Max ion trap mass spectrometer. HR-ESI-MS data were obtained on an Agilent 6210 ESI/TOF mass spectrometer. Optical rotation was recorded in CH 3 Cl 3 on Jasco P-1020 polarimeter at room temperature. Infrared (IR) spectra were measured on a Jasco FT/IR-480 plus Fourier Transform infrared spectrometer using KBr pellet. Nuclear magnetic resonance (NMR) spectra were measured on Bruker AV-300 or AV-400 spectrometers. Thin-layer chromatography (TLC) analyses were carried out using pre-coated silica gel GF 254 plates (Qingdao Marine Chemical Plant, P.R. China). Column chromatographies were performed on silica gel (200-400 mesh, Qingdao Marine Chemical Plant, P.R. China), reversephase C 18 silica gel (Merck, Germany) and Sephadex LH-20 (Pharmacia Biotec AB, Sweden). The solvents used in column chromatography and high-performance liquid chromatography (HPLC) were of analytical (Shanghai, P.R. China) grade and chromatographic grade (Fisher Scientific, USA), respectively.

Plant Material
The dried roots of Salvia miltiorrhiza were purchased from Qingping Herbal Medicine Market (Guangzhou, P.R. China) and identified by Prof. Guang-xiong Zhou in Jinan University. A voucher specimen (No. 20090901) was deposited in the Institute of Traditional Chinese Medicine and Natural Products, Jinan University, China.

Extraction and Enrichment of Total Tanshinones
The dried roots of Salvia miltiorrhiza (10 kg) was crushed and extracted using 95% ethanol. The crude extract was suspended in distilled water and partitioned with dichloromethane to afford the dichloromethane fraction 178.5 g, which was revealed to be the total tanshinones (95%, estimated based on the online UV spectrum and the total peak areas of individual peaks) by HPLC analysis.

Maintenance of Zebrafish
The zebrafish strains, including Tg(fli-1a:EGFP)y1 and Widetype, were maintained as previously described in our previous paper [31].

Embryo Collection and Drug Treatment
Zebrafish embryos were generated by natural pair-wise mating (3-12 months old) and were raised at 28uC in embryo medium. Healthy, hatched zebrafish were selected at 21 hour  (PCV) to dorsal longitudinal anastomotic vessels (DLAVs) were defined as intact. And in the VRI-only treatment group or some drug treatment groups, some of the ISVs observed sprouting from DA or PCV, but did not form complete ISVs, were defined as defective. The number of intact and defective ISVs in each zebrafish embryo was counted.
Total RNA Extraction, Reverse Transcription and Realtime qPCR The wild-type zebrafish embryos were treated with the protocols described above. At 48 hpf, total RNA was extracted from 30 embryos in each treatment group, using the RNeasy Mini Kit (Qiagen, USA) in accordance with the manufacturer's instructions. RNA was reverse transcribed into single-strand cDNA using III First-Strand Synthesis System for RT-PCR (Invitrogen, USA), followed by real-time PCR using the Universal PCR Master Mix, and 250 nM custom primers for zebrafish kdrl
The cancer cells were plated into a 96-well plate at a concentration of 0.8-1610 5 cells/well overnight to allow for cell attachment. The working solutions of compound 10 and tanshinones were dispensed appropriately into the partitioned 96-well plates then incubated for another 48 h. The cells were also incubated with DMSO (0.1%) which served as the vehicle control. Then the medium was discarded and cells were incubated for 4 h at 37uC in MTT solution (final concentration 0.5 mg/ml). The solution was then replaced by 100 ml DMSO to dissolve the violet formazan crystals in intact cells. The absorbance was measured at a wavelength of 570 nm by a SpectraMaxR M5 Multi-Mode Microplate Readers (Molecular Devices, USA). Cell viability was expressed as a percentage of the vehicle control.

Statistical analysis
Each experiment was done in triplicate at least. The mean and standard deviations were compared using Students' t-test with the following statistical criteria: p,0.05, significant; p,0.01, highly significant; p,0.001 extremely significant. All relevant data are presented as mean6SEM.

Synthesis of New Tanshinone Derivatives
In our study, total tanshinones (95%) fraction was isolated and enriched from an ethanol extract of Salvia miltiorrhiza and identified by HPLC analysis. One-pot synthesis was then carried out by reacting the total tanshinones fraction with benzaldehyde and ammonium acetate under 35uC to obtain a semi-synthetic mixture (Fig. 1). A comparison of the HPLC chemical profiles of the semisynthetic mixture with that of the unreacted total tanshinones revealed a marked change in the chemical composition (Fig. 2). The semi-synthetic mixture showed a group of peaks with retention times at 42-50 min, whereas the most prominent peaks of the parental total tanshinones were concentrated at 30-42 min. Further subsequent isolation of the semi-synthetic mixture resulted in eleven new derivatives named as compounds 1--11 (Fig. 3), which could be classified into three groups according to their structural scaffolds.
Compounds 1--5 were classified into a group which was composed of the 2-phenylimidazole and C18 tanshinone moieties. Compound 1 possessed a fully conjugated ring system with methyl groups located at C-4 and C-15. Compounds 2 and 3 were similar to those of 1 with the absence of D 1,2 and D 15,16 double bonds, respectively. Compound 4 was also an analog of 1, however, the three aromatic methines in ring A were replaced by three alicyclic methylenes and the methyl group at C-4 was replaced by an exomethylene. The structure of 5 was similar to 4 except that the exomethylene moiety in 4 was replaced by a hydroxyl and a hydroxymethylene groups.
Compounds 6--9 were classified into another group which was composed of the 2-phenylimidazole and C-19 tanshinone moieties. Compound 6 was a typical example of this group. The skeleton of 6 was the same as that of 5, but the substituents at C-4 were changed to two methyl groups. Compounds 7 and 8 were analogs of 6 with one of the methyl group at C-4 changed to a hydroxymethylene and a methyl carboxylate group, respectively. Compound 9 was also similar to 6, however, the double bond D 15,16 in the furan ring was absent.
The last miscellaneous group included compounds 10 and 11. In 10, no signals for the furan ring were observed in the 1 H and 13 C NMR data. In contrast, two conjugated carbonyl groups were found in C-13 and C-14. In 11, there was only one nitrogen atom in the molecule as revealed by the HRESI-MS (m/z 384.1588 [M+ H] + ), suggesting that the phenylimidazole might be replaced by a phenyloxazole ring, which was further confirmed by the absence of a proton signal for the NH group.
All those compounds were identified by extensive spectroscopic methods (For detailed structural elucidation, see Section 2 of File S1; for the high resolution ESIMS data, 1 H, 13 C and DEPT-NMR spectra, see Section 3 of File S1), and the structures of 1, 2, 3, 6 and 10 were further confirmed by single crystal X-ray analysis (Fig. 4) (Crystal data and CCDC numbers for 1, 2, 3, 6 and 10, see Table S1 in File S1).
It should be noted that there is a chiral atom at C-16 for 3, 9 and 11, and at C-4 for 5, 7 and 8. The absolute configurations of these natural tanshinones were well established [33][34][35] (For the known configurations, see Section 6 of File S1), and the chemical reaction took place at the diketone moiety. Thus the tanshinone derivatives would be expected to share the same absolute configurations as the corresponding parental tanshinones (16R for 3a, 9a and 11a; 4S for 7a and 8a; 4R for 5a). Furthermore, final refinement of the CuKa X-ray diffraction data of 3 resulted in a small Flack parameter 0.1, confirming the absolute configuration of this compound.

Determination of Angiogenesis Activity in Zebrafish
The angiogenesis effects of compounds 1--11 were determined in a chemical-induced blood vessels loss zebrafish model [30]. The major natural tanshinones including Tan IIA, CPT and Tan I were used as reference controls in the angiogenesis assay. VRI, a pyridinylanthranilamide compound that displays anti-angiogenic properties, strongly inhibits the kinase activities of both vascular endothelial growth factor receptor 1 and 2 (VEGFR1 and VEGFR2) [30]. The 21 hpf zebrafish embryos pre-treated with 500 ng/ml of VRI for 3 h were then washed out and placed in embryo medium and incubated for another 48 h. Treatment of the zebrafish with VRI induced obviously blood vessels loss in ISVs, DLAVs and the subintestinal vessels (SIVs) (Fig. 5 B-B'-b). After incubating the VRI pre-treated embryos with 1--11 and three major tanshinone controls for another 24 h and 48 h, the VRI-induced blood vessels loss in ISVs, DLAVs and SIVs regions were restored to different degrees. Of 1--11, compound 10 (EC 50 = 0.026 mM) and 3 (EC 50 = 4.03 mM) exhibited significantly potency and efficacy in the recovery of blood vessels loss in VRI-treated zebrafish (see Table S2 in File S1). However, there was no significantly observable change in blood vessels following treatment with other nine derivatives at their maximum non-toxic concentrations (Table 1). Furthermore, 10 was shown to be the most potent and effective derivative and demonstrated a stronger vascular effect than Tan IIA, CPT and Tan I on the recovery of VRI-induced blood vessels loss in zebrafish (Fig. 5). Quantitative analysis of the average number of ISVs both intact and defective ISVs, confirmed the significant dose-dependent restorative effect of 10 (Fig. 5II).

Determination of mRNA Expression by Real-time qPCR
In order to confirm the vascular phenotypic effects observed in Fig. 5, the significant pro-angiogenesis activity of compound 10 in VRI-treated zebrafish was validated and confirmed by real-time qPCR analysis of selected key angiogenesis gene markers. Embryos were pre-treated with VRI (500 ng/ml) for 3 h, and then treated with 10 (0.01, 0.03, 0.1, 0.3 and 1 mM) for 24 h. Total RNA from different treatment groups of embryos was isolated and reverse transcribed to cDNA. The relative mRNA expressions of flt-1 encoding the VEGFR1 as well as kdrl and kdr encoding VEGFR2 were determined using real-time qPCR. The gene expression results clearly showed that increasing dosages of 10 exhibited a significant concentration-dependent recovery of down-regulation of kdrl, kdr and flt-1 mRNA expression in VRItreated zebrfish. Compared to the VRI-only treatment group, the relative change in gene expression level in 1 mM 10 treatment group was 2.01-fold for kdrl (p,0.05), 1.80-fold for kdr (p,0.01) and 2.98-fold for flt1 (p,0.01) (Fig. 6). Thus, these results suggested that 10 reversed the VRI-induced down-regulation of the expression of three key VEGF receptors in zebrafish.

Effect of Inhibitors of FGF and EGF Signaling Pathways on Angiogenesis Effect of Compound 10 in VRI-treated Zebrafish
In order to confirm whether the angiogenesis response of compound 10 also involved epidermal growth factor (EGF) and fibroblast growth factor (FGF) signaling pathways, the effects of specific small molecule inhibitors on 10-induced angiogenesis were studied. SU5402, a cell-permeable small molecule compound, is an ATP-competitive inhibitor of the FGFR1 tyrosine kinase [36], which has been shown to abolish FGF2-induced angiogenic effect on zebrafish embryonic development [37]. Both observable blood vessel phenotype change and quantitative analysis suggested that co-treatment of SU5402 (20 mM) with 10 for 24 h, almost entirely inhibited the restorative effect of 10 on blood vessels loss in VRItreated zebrafish (Fig. 7E) while treatment with 20 mM SU5402 alone showed no significant effect on normal blood vessels development of zebrafish (Fig. 7D). In addition, PD 153035, a highly potent ATP-competitive and specific inhibitor of the tyrosine kinase activity of EGF receptor, failed to significantly block the vascular action of 10 in VRI-treated zebrafish (Fig. 7G). Therefore, our results suggested that activation of FGF, rather than EGF, might be essential for the angiogenesis effect of 10 in VRI-treated zebrafish (Fig. 7II).

Effect of Inhibitors Downstream Signaling Targets of FGFR and VEGFR on Angiogenesis Effect of Compound 10 in VRI-treated Zebrafish
Previous studies suggested that PI3K, P38 and Raf/MEK/ERK are crucial downstream signaling targets that are involved in VEGF and FGF-induced angiogenesis [38,39,40]. Moreover, Src is required for the angiogenesis actions of VEGF/FGF, and VEGF/FGF-induced angiogenesis are inhibited by Src inhibitors [41,42]. Therefore, we further investigated the roles of Src, PI3K, P38, Raf, MEK1/2, ERK1/2, Akt kinases, eNOS and HIF-1 in the angiogenesis effect of compound 10 by observing the effect of their selective pharmacological inhibitors in the zebrafish angiogenesis assay. PP2, a Src kinase inhibitor, acts as a potent and selective inhibitor of the Src family of protein tyrosine kinases [43]. Besides, LY 294002 is commonly used to inhibit PI3K activity which acts on the ATP binding site of PI3K [44]. Our results showed that, eleven inhibitors (Table 2) including PP2, LY294002, P38i, Rafi, ERK1/2i, MEK1/2i could reverse vascular restorative effect of 10 in zebrafish with VRI-treated blood vessels loss (Fig. 8I). Under the same treatment condition, HIFi, L-NAME and three Akti, all failed to abolish the vascular restorative effect of 10. Meanwhile, in the blank control, treatment at 24 hpf with 10 mM PP2, 10 mM LY 294002, 5 mM P38i, 50 mM Rafi, 25 mM ERK1/2i and 20 mM MEK1/2i for 24 h did not significantly alter normal blood vessels development in zebrafish (Fig. 8II). Hence, our results suggest that Src, Raf, ERK1/2, MEK1/2, P38 and PI3K, but not Akt, eNOS-NO and HIF-1 pathways, were probably involved in the angiogenesis effect of 10 ( Fig. 9).

Anti-proliferation Effects of Compound 10 Against MCF-7, MDA-231 and A549 Cell Lines
Although angiogenesis therapy is widely recognized as an effective mean of restoring various ischemic pathological conditions, it also raises a safety concern regarding the promotion cancer cell growth. Thus, in present study, we intended to identify compound 10 whether only elicited an angiogenesis effect under pathological conditions of angiogenesis deficiencies such as ischemic cerebral and heart diseases or not. In addition, the anti-cancer activities of numerous natural tanshinones had been well characterized in several pre-clinical studies [45,46,47]. Therefore, we performed a pilot study testing the anti-proliferation effect of 10 compared to Tan IIA, CPT and Tan I in three cancer Table 2. The effects of inhibitors of EGF, VEGF and FGF signaling pathways on compound 10-induced angiogenesis in VRI-treated zebrafish.

Discussion
In our study, a series of new tanshinone derivatives were successfully synthesized from a mixture of natural tanshinones through one-pot synthesis approach. The present pilot study, for the first time, proved the concept of feasibility using a modified one-pot synthesis to modify minor tanshinones, which are hardly to be isolated with sufficient quantity for chemical modification by conventional organic synthesis approach. We found that some of the new tanshinone derivatives exhibited strong protective and restorative effects against blood vessels loss in zebraifsh. And the newly identified derivatives could serve as new lead candidates for further development as angiogenesis therapeutic agents.
Tanshinones are a group of diterpenoid quinones with unique structural scaffolds and promising biological activities. However, a substantial number of tanshinones which are present in very low quantities in Salvia miltiorrhiza [48] constitute as a hurdle in researching on their conventional organic modification because purification of those minor tanshinones are difficult. In previous study, one-pot combinatorial modification was used to improve the drug-like properties of tanshinones by incorporating a Ncontaining heterocyclic ring which is an important moiety in many drug molecules [49]. 6-mercaptopurine and thioguanine [50] are two examples of anti-cancer drugs which have an N-containing heterocyclic ring. Interestingly, our rapid one-pot synthesis of total tanshinone fraction leaded to isolate and identify eleven new derivatives which exhibited different degrees of angiogenesis activities. In this study, we also proved that the use of a modified one-pot synthesis to enhance the chemical diversity and biological activities of natural products was feasible.
Angiogenesis deficiencies are associated with various human diseases (e.g. cardio-cerebrovascular problems and ulcers). Our zebrafish angiogenesis assay which was produced by inactivation of the VEGF signaling pathway was more physiological relevant assay for identifying pro-angiogenesis agents against diseases associated with angiogenesis deficiencies [30]. In present study, screening of all chemically modified tanshinone derivatives in VRI-treated zebrafish, led to the identification of compound 10 as the most potent and effective angiogenic compound for restoring blood vessels damage which may have been the result of improvement in either one or a combination of various drug-like properties. Further studies to investigate the change of pharmacodynamic and pharmacokinetic properties of 10 were warranted. In addition, the new tanshinone derivatives exhibited limited close relatedness in their structures as they evolved from the random modification of different natural tanshinones. Therefore, no conclusive relationship between their structures and angiogenesis activity in VRI-treated zebrafish could be drawn.
It has been reported that the formation of new blood vessels is orchestrated by different proteins, including cell adhesion molecules, extracellular matric components and vascular endothelial growth factor receptors (VEGFRs) [51]. The VEGFRs play crucial roles in transducing regulatory signal from VEGF which was originally described as a potent angiogenic factor as well as an essential growth factor for vascular endothelial cells [52]. Although our gene expression data provided clues on the possible involvement of the VEGF-KDR/FLK pathway in restoring blood  vessels after 10-treatment, these results alone did not entirely reflect the activation of the downstream signaling targets (Src, PI3K, P38 and MAPK signaling pathways) involved in VEGF/ FGF-induced angiogenesis because their activation usually occurs via phosphorylation in post-translation regulation [53]. In order to address this aspect, we determined the effects of specific small molecule inhibitors against key downstream targets of EGF, FGF and VEGF signalling pathways on the pro-angiogenesis effect of 10. Our results demonstrated that 10 probably promoted angiogenesis through VEGF and/or FGF by activating Src, PI3K, P38, Raf, MEK1/2 and ERK1/2 kinases but not HIF-1, eNOS and Akt kinases. Taken together, all these evidences showed that the angiogenesis activity of 10 probably involved the interaction of VEGF/FGF-Src-MAPK and PI3K-P38 signaling pathways in VRI-treated zebrafish (Fig. 9). In sum, the above data indicated that a novel one-pot combinatorial modification was used to simultaneously diversify both the major and minor natural tanshinones which led to the isolation and identification of eleven new derivatives. In the present study, our results suggested that 10, the most potent new angiogenic tanshinone derivative identified without any promoting proliferation effects in three cancer cells, had potential for further development as a therapeutic agent for diseases associated with imbalanced angiogenesis such as ischemic heart disease and traumatic injuries. In addition, both the mRNA expression assay and blocking assay of specific kinases inhibitors provided insights into the mechanism underlying the angiogenesis effect of 10. It is worthy of note that the zebrafish screening model used in this study is based on embryonic development, but not a real pathology model. Further investigation of the pharmacodynamic data of 10 based on the vascular injury pathological model was warranted. In addition, physiological angiogenesis is a vital process in growth and development of organism. Since the process is complicated, and typically consists of cell proliferation, migration, invasion and tube formation [54], we further investigated if 10 could directly act on either one of the biological steps in angiogenesis on human umbilical vein endothelial cells (HU-VECs). In our results, 10 could significantly promote the migration and invasion of HUVECs in a concentration-dependent manner, as effective as the effects of VEGF. However, the HUVECs proliferation stimulating effect of 10 was mild when compared with that of VEGF (Section 8 in File S1). These in vitro findings further confirm our observation of angiogenic effect of 10 in zebrafish and provide further insight into pharmacological action of 10 indeed involving direct cellular action, at least on the two crucial steps of migration ( Figure S15 in File S1) and invasion ( Figure S16 in File S1) underlying angiogenesis.
In conclusion, one-pot modification of the total tanshinone fraction from Salvia miltiorrhiza could successfully generate new chemical structures with higher biological activity, of which 10, a novel angiogenic compound, was a good example. More importantly, our study should have a high impact in this research field as it provided valuable information on an alternative, relatively fast and green approach for the discovery of novel diversified compounds, particularly those low abundant components present in natural products.