Binpu-3 and Handan (a wild dandelion strain collected in the non-saline area of Handan, Hebei Province, China) seeds were initially sown in a greenhouse and subsequently transplanted to the field once they developed 3–4 leaves. The seedlings were transplanted at a depth of 2–3 cm, with row spacing of 50 cm and ditch spacing of 30 cm. Throughout the cultivation period, an automatic negative-pressure drip irrigation system was used to maintain a soil matrix potential of −15 kPa. Urea was applied as topdressing at a rate of 15–22.5 g/m² when the longest leaves reached 10–15 cm in length. A coastal, mildly saline-alkali soil with a salt content of 0.2%–0.3%, a soil bulk density of 1.4 g/cm³, a soil pH in the range of 7.8–8.5, and an organic matter content of 1.426% was used. We conducted a salt stress test in a salt tolerance identification pool with a soil salt content of 0.6% (Zhang et al., 2021).

Whole Binpu-3 (or Handan) plants were collected and washed thoroughly. The aboveground parts (primarily leaves with a few flowers) and the underground parts (roots) were separated and dried at 60°C until the moisture content was <13%. The dried materials were ground and passed through a number 4 sieve to produce dry powders, which were stored as 100 g samples in gauze bags. The gauze bags were immersed in 1,000 mL of deionized water and boiled for 1 h. Subsequently, the residue was decocted again as described above. Lastly, the twice extracts were combined, centrifuged, filtered through a 0.45 μm membrane, and diluted to a final volume of 200 mL. The Binpu-3 leaf extract (Binpu-3LE) and Binpu-3RE stock solutions had final concentrations of 500 mg/mL.

The contents of the extracts were detected using 0.75 g samples of dry powder incubated with 15 mL of 0.1% acid cellulase aqueous solution at 60°C for 30 min. Then, 15 mL of methanol was added and ultrasonicated over 30 min at 66°C. Lastly, the extract was centrifuged at 4,000 rpm for 10 min and filtered through a 0.45 µm membrane for high-performance liquid chromatography (HPLC), which was performed using a Shimadzu LC-20A system equipped with an Agilent Eclipse XDB-C18 column (4.6 × 250 mm, 5 μm; Agilent Technologies, Inc., city, CA, USA). The mobile phase flow rate was set at 1.0 mL/min. The mobile phase for phenolic acid (caftaric acid, chlorogenic acid, caffeic acid, and cichoric acid) detection was methanol (A) and 0.2% phosphoric acid (B) at a ratio of 35:65 (A:B). A wavelength of 327 nm was used to measure the different contents. The mobile phase used for the detection of flavonoids (quercetin, luteolin, and apigenin) was acetonitrile (C) and 0.5% phosphoric acid (D) at a ratio of 30:70 (C:D) and a measurement wavelength of 360 nm. The detection of triterpenoids (taraxasterol and taraxerol) was conducted using 0.01 mmol/L potassium dihydrogen phosphate as the mobile phase and measured at a wavelength of 210 nm.

A sample of the prepared Binpu-3RE solution was analyzed using a Vanquish ultra performance liquid chromatography (UHPLC) system (Thermo Fisher Scientific, Bremen, Germany) equipped with a ACQUITY UPLC HSS T3 column (2.1 mm × 100 mm, 1.8 µm) maintained at 35°C. The mobile phase consisted of 0.1% formic acid aqueous solution (A) and 0.1% formic acid acetonitrile solution (B); it was pumped at a flow rate of 0.3 mL/min. Gradient elution was performed according to Supplementary Table 2. Next, the Q-Exactive HFX mass spectrometer (Thermo Fisher Scientific, Germany) was coupled with the UHPLC system to collect the primary and secondary spectrograms of the sample in both positive and negative electrospray ionization (ESI) modes. The spray voltage was 3800V (ESI+)/3500V (ESI-), the sheath gas pressure was 45 arb, the auxiliary gas pressure was 20 arb, the ion transfer tube temperature was 320°C, and the atomization temperature was 350°C. The MS/MS spectrum was obtained using the top 10 MS1 ion, and the collision energies (CEs) were normalized by steps with energy levels of 20, 40, and 60. The scanning range of the first-order mass-to-charge ratio was 90–1300. Lastly, the format of the original data was converted, peak alignment, retention time correction, and peak extraction were carried out using XCMS software, and the chemical substances in the sample were identified by searching the local database of high-resolution mass spectrometry of traditional Chinese medicine.

Fly strains were stored at 25°C under 60% humidity in a 12 h light/dark cycle with a standard culture medium of 25 g sucrose, 30 g cornmeal, 5 g agar, 15 g yeast, 3 mL propionic acid, and 500 mL ultrapure water. The following Drosophila stocks were used: w ¹¹¹⁸ (Bloomington Drosophila Stock Center, BL#3605), UAS-mGFP (BL#32197), NRE-GFP (BL#30728), and UAS-scrib-IR (Vienna Drosophila RNAi Center, VDRC#27424). The ey-GAL4, ptc-GAL4 UAS-GFP (ptc>GFP), and eyeful strains were gifts from Professor Lei Xue of Tongji University (Ferres-Marco et al., 2006; Wu et al., 2015; Wang et al., 2020). The Delta-LacZ and Su(H)-LacZ strains were gifts from Professor Xianjue Ma of Westlake University. The construction of the ptc>scrib-IR cell migration and eyeful tumor invasion models has been previously described (Wang et al., 2020). For all fly crossing experiments, healthy, unmated male and female parents were randomly assigned to different groups. We directly added the prepared Binpu-3LE or Binpu-3RE to the standard food to achieve final concentrations of 6.25, 12.50, 25.00, and 50.00 mg/mL. The corresponding offspring with the genotype ptc>GFP/+ or ey-GAL4/+ were kept in the control medium without any vehicle treatment as the control group. Offspring with ptc>GFP/UAS-scrib-IR or eyeful/+ were kept in the control or extract-supplemented medium as the model or drug-treated groups.

Early w ¹¹¹⁸ 3^rd instar larvae were washed with phosphate-buffered saline (PBS), starved for 2 h under adverse food conditions (PBS+0.8% agar), and transferred to fresh medium containing 0.5% Brilliant Blue FCF (Sangon Biotech, #A606319, China) for 20 min. Then, the larvae were washed with PBS and homogenized in 100 μL lysis buffer (PBS+0.1% Triton X-100) using a micro high-speed tissue homogenizer for 3 min. After centrifugation, 2 μL of the supernatant was measured with a spectrophotometer at a wavelength of 630 nm. A standard curve was generated using 2 mg of dye-containing food dissolved in 480 μL lysis buffer (Matsuda et al., 2015; Cao et al., 2022).

TRIzol (Sigma, #M3148) and a PureLink™ RNA Mini Kit (Life Technologies, #12183018A) were used to isolate the total RNA from the heads of 50 adult flies of each group. Real-time quantitative polymerase chain reaction (RT-qPCR) was performed as previously described (Wang et al., 2003). The relative amounts of transcripts were calculated using the comparative C _T method, and ribosomal protein 49 (rp49) was used as an internal control gene. The primer sequences that were used in this study are listed in Supplementary Table 1.

The dissected larval tissues were fixed in Buffer A (50 mL 0.1% PBS containing Tween-20 (PBST) + 50 μL 1 mM MgCl₂ + 1.5 mL 5M NaCl) containing 1% glutaraldehyde for 15 min at 25°C. Then, the tissues were rinsed once in Buffer A containing 3.3 mM K₃Fe(CN)₆ + 3.3 mM K₄Fe(CN)₆.3H₂O, and incubated in Buffer A containing 3.3 mM K₃Fe(CN)₆ + 3.3 mM K₄Fe(CN)₆.3H₂O + 0.2% 5-bromo-4-chloro-3 indolyl-β-D-galactopyranoside (X-gal, TIANGEN, #RT119) at 20°C–25°C (room temperature). The dyeing time of Delta-LacZ was 1 h and that of Su(H)-LacZ was 15 min. The tissues were then washed with Buffer A mounted with 100% glycerol.

Dissected larval tissues were fixed in 4% formaldehyde for 20 min at room temperature. After three washes with 0.3% (v/v) PBST, the tissues were stained with primary antibodies overnight at 4°C followed by staining with secondary antibodies for 4 h at room temperature. The following antibodies were used: mouse anti-MMP1 (1:200, Developmental Studies Hybridoma Bank (DSHB), #3A6B4, #14A3D2, #5H7B11), mouse anti-β-integrin (1:100, DSHB, CF.6G11), and goat anti-mouse-Cyanine3 (1:1000, Life Technologies, #A10521). Vectashield medium with 4′,6-diamidino-2-phenylindole (Vector Laboratories, #H-1500) was used for mounting. The fluorophores were excited and visualized using an inverted fluorescence system (Olympus, IX51).

Thirty-two specific pathogen free (SPF) male SD rats (8 weeks, 180 ± 20 g) were obtained from the Experimental Animal Center of North China University of Science and Technology (Laboratory Animal Production license: SYXK (Ji) 2023–018). The animals were raised in an SPF room, on a light–dark cycle of 12 h each session, with an appropriate temperature (23 ± 2°C) and humidity (60 ± 5%). They were allowed free access to SPF-purified water. After 7 days of adaptive feeding, the rats were randomly divided into the blank control group, Binpu-3RE-low (Binpu-3RE-L, 2.73 g/kg), Binpu-3RE-medium (Binpu-3RE-M, 5.46 g/kg), and Binpu-3RE-high (Binpu-3RE-H, 10.92 g/kg) groups. Animals in all groups were given the corresponding dose of Binpu-3RE or the same volume of normal saline once a day. On the eighth day, they were anesthetized and blood was collected from the aorta, centrifuged at 3,000 rpm for 10 min to separate serum, inactivated in a water bath at 56°C for 30 min, and then filtered and sterilized using a needle filter. Lastly, the obtained rat control serum and Binpu-3RE-containing sera were stored at −20°C for future use.

Human breast cancer MDA-MB-231 cells obtained from FuHeng Biology (Shanghai, China, # FH0238) were incubated at 37°C in a water jacket incubator (Thermo Fisher Scientific, 311) under a 5% CO₂ atmosphere with Dulbecco’s modified Eagle’s medium (DMEM) (EallBio, #C008) supplemented with 10 U/mL penicillin-streptomycin (NCM Biotech, China, #C100C5) and 10% fetal bovine serum (FBS) (Procell, China, #164210–50), 10% rat control serum, or rat Binpu-3RE-containing sera. Cells were passaged when they reached 80% confluence.

MDA-MB-231 cells were passaged into 96-well plates at a density of 5 × 10³ cells/well. After 24 h, the cells were treated with DMEM supplemented with rat control serum (without any vehicle treatment) or Binpu-3RE-containing serum (Binpu-3RE-L, Binpu-3RE-M, Binpu-3RE-H) for 24 h, 48 h, and 72 h at 37°C under a 5% CO₂ atmosphere. Subsequently, the cells were treated with a mixture of DMEM (90 μL) and cell counting kit (CCK)-8 solution (10 μL, Mei5 Biotech, China, #MF128-01), then incubated for 2 h. The absorbance was then measured at 450 nm using a plate reader (Tecan, M200PRD).

MDA-MB-231 cells were cultured in 6-well plates at a density of 5 × 10⁴ cells/well at 37°C under 5% CO₂ saturation humidity until the degree of integration exceeded 90%. Next, a sterile 200 µL pipette tip was used to create a straight clean scratch wound across the center of each well, and marked “0 h”. Then, the cells were maintained in DMEM supplemented with rat control serum or Binpu-3RE-containing serum (Binpu-3RE-L, Binpu-3RE-M, Binpu-3RE-H) for 24 h, 48 h, and 72 h. Each wound was photographed at predetermined time points in the same position, and the migration was determined using ImageJ software. The following equation was used to determine the rate of cell migration: Cell migration rate (%) = (initial scratch area-final scratch area)/initial scratch area × 100%.

The upper chamber of the transwell instrument was coated with diluted matrigel (1:8), and MDA-MB-231 human breast cancer cells were suspended in serum-free DMEM and seeded in the upper chamber at a density of 5.0 × 10⁴ cells/200 μL of cell suspension. To promote cell invasion, the lower chamber was filled with medium supplemented with 10% rat control serum or 10% Binpu-3RE-containing serum (Binpu-3RE-L, Binpu-3RE-M, Binpu-3RE-H) as a chemoattractant. Following this, the chambers were incubated at 37°C in a humid atmosphere containing 5% CO₂ for 72 h. Next, non-invading cells on the upper surface were wiped off with cotton swabs, and the invading cells in the lower chamber were fixed with 1% crystal violet for 20 min, washed with PBS, and then stained and visualized using an inverted microscope. Lastly, the number of invading cells in each group was quantified using ImageJ software.

MDA-MB-231 cells were collected and extracted using radioimmunoprecipitation assay (RIPA) lysis buffer containing phosphatase inhibitors and 1% phenylmethanesulfonyl fluoride (PMSF). The total protein was quantified using the BCA quantification assay. Equal amounts of protein (10–30 µg) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), transferred to a polyvinylidene fluoride (PVDF) membrane, and subjected to the standard Western blot protocol, as previously described (Roth et al., 2018). The antibodies used were rabbit anti-Notch1 (1:1000, HUABIO, #SJ205), rabbit anti-Jagged1 (1:1000, HUABIO, #JF96-06), rabbit anti-Hes1 (1:1000, HUABIO, #SC06-21), rabbit anti-α-Tubulin (1:1000, Cell Signaling Technology, #2125S), and goat anti-rabbit IgG (H + L, HRP) (1:1000, Report, #S1002).

GraphPad Prism (version 8.0) was used to create bar graphs of the study results. Comparison of data was performed using one-way analysis of variance (ANOVA) with Bonferroni’s multiple comparison test, unpaired t-test, chi-square test, Kruskal–Wallis one-way ANOVA test, or two-way ANOVA with Dunnett’s multiple comparison test. Statistical significance was set at P < 0.05, and the center values were considered the mean.

To screen high-quality dandelion suitable for growth in saline-alkali areas, 22 dandelion strains were obtained via field collection or combined salt stress mutation breeding (Zhang et al., 2021). The Binpu-3 dandelion strain (initially termed Tanghai-6, according to the area and number at the time of collection) exhibited obvious advantages in terms of growth, fingerprint quality, in vivo antioxidant capacity, and anti-bacterial ability. Therefore, this strain was used to create a new variety (Hebei Province Drug Recognition, China, #2023–015), and was selected for use in the current study.

Binpu-3 and Handan seeds were sown in a greenhouse and then transplanted into slight saline-alkali soil with a salt content of 0.2%–0.3%. Compared with the control Handan strain, Binpu-3 plants had a longer leaf length, longer leaf width, and greater height (Figure 1A). In addition, the growth mode of Binpu-3 was nearly vertical, with a narrow leaf shape, dark and hairless leaves, teeth at the crack edge, deep leaf cracks, and bright flowers under mild saline-alkali conditions (Figures 1C–G). Binpu-3 grew well in an environment of 0.6% salt stress (Figure 1B), indicating its strong salt resistance.

When the plant grew to maturity, whole plant samples of Binpu-3 and Handan were collected and separated into aboveground (mainly leaves, with a few flowers, hereafter referred to as leaves) and underground (roots) parts to make dry powders. The plant parts were cleaned, dried, crushed, and prepared into extracts for subsequent experiments (Figures 1H–J). Four effective components (phenolic acids) were detected using HPLC (Figures 1K–N; Supplementary Figure 1A). The caftaric, chlorogenic, caffeic, and cichoric acid levels in the leaves and roots of Binpu-3 were significantly higher than those in the leaves and roots of Handan (Figures 1O, P). Therefore, the Binpu-3 strain was considered of good quality and suitable for popularization and application in saline-alkali soil below moderate levels and in most regions of the world.

Drosophila melanogaster was used to construct an ocular tumor metastasis model, termed eyeful (Ferres-Marco et al., 2006; Wang et al., 2022). Compared with the control (Figures 2A, E), the eye tissues of the eyeful tumor model exhibited different degrees of tumor in situ proliferation (Figures 2B–D). Red-eye tumor cells migrated to other parts of the body including the head, thorax, and abdomen, at a migration rate of 28.99% (Figures 2F–I). Binpu-3LE and Binpu-3RE were added to the Drosophila standard culture medium at final concentrations of 6.25, 12.50, 25.00, and 50.00 mg/mL for a total of approximately 12 days from the embryonic stage until 2 or 3 days after emergence into adults. In detail, compared with the eyeful model, there was no significant difference in the composition ratio of primary growth and migration location of Drosophila eye tumors in the Binpu-3LE or Binpu-3RE treated groups (Figures 2K, L). Binpu-3RE treatment caused a reduction in the migration rate of eye tumor cells from 28.99% to 17.47% when applied at a concentration of 25.00 mg/mL (Figure 2J). However, Binpu-3LE had no obvious effect on tumor cell metastasis at any concentration (Figure 2I). Together, these data suggest that Binpu-3RE inhibits tumor cell metastasis at a suitable concentration in the D. eyeful model.

Why was Binpu-3RE capable of inhibiting eyeful tumor metastasis, whereas Binpu-3LE did not have an obvious inhibitory effect? To explore the potential reasons for this, we analyzed the contents of several active ingredients in Binpu-3 leaves and roots dry powder samples with HPLC (Supplementary Figure 1). The levels of four phenolic acids (caftaric acid, chlorogenic acid, caffeic acid, and cichoric acid) (Figures 3A, B), three flavonoids (quercetin, luteolin, and apigenin) (Figures 3C, D), and two triterpenes (taraxerol and taraxasterol) (Figures 3E, F) were assessed in Binpu-3 roots and Binpu-3 leaves. The caftaric acid, chlorogenic acid, cichoric acid, luteolin, and taraxerol levels were lower in leaves than in roots. However, the caffeic acid, quercetin, apigenin, and taraxasterol levels were significantly higher in Binpu-3 roots than those in Binpu-3 leaves (Figures 3G–O). These differences may account for the different effects of the two extracts.

To comprehensively analyze the components in Binpu-3RE, we employed ultra-performance liquid chromatography combined high resolution mass spectrometry (UPLC-HRMS). The analysis data showed that 1349 components were detected in positive ion mode and 573 components were detected in negative ion mode, with a total of 1826 components being detected. Subsequently, we confirmed the peak shape of the chromatographic peaks with high abundance, checked the secondary chromatograms, and marked the chromatographic peak numbers of the positive and negative ion base peak chromatograms (BPCs) in numerical order (Figures 4A, B; red numbers). We discovered 33 major components with relatively high content, including pantothenate (0.1%), butein (0.53%), chlorogenate (0.78%), chicoric acid (1.96%), azelaic acid (0.23%), and [6]-gingerol (0.13%) (Figure 4; Supplementary Table 4).

A well-recognized cell migration model of third-instar larval wing imaginal discs in D. melanogaster was used in this study (Zhang et al., 2019). Compared to the control (Figures 5A, G), RNAi-mediated downregulation of the cell polarity gene scrib (UAS-scrib-IR) driven by ptc-GAL4 along the anterior/posterior axis boundary area marked by GFP (UAS-GFP) (overall genotype abbreviated as ptc>scrib-IR) triggered a large number of cells to move to the rear within the wing pouch region, generating the cell migration phenotype (Figures 5B, H, M, N). The number of migrating cells was significantly reduced after treatment with 6.25, 12.50, 25.00, and 50.00 mg/mL Binpu-3RE (Figures 5C–F, I–M), with inhibition of the maximum migration distance after treatment with 25.00 mg/mL Binpu-3RE (Figure 5N). Therefore, this concentration was used for mechanistic research. Based on the above results, we concluded that Binpu-3RE could inhibit the ptc>scrib-IR triggered cell migration in Drosophila at an appropriate concentration, and 25.00 mg/mL was selected as the best dosage to carry out the follow-up mechanistic research.

Additionally, the experimental study of the corresponding Drosophila food intake was performed to verify whether the administration medium at each concentration affected feeding rate. The Drosophila fed with Binpu-3RE medium containing concentrations of 6.25, 12.50, 25.00, and 50.00 mg/mL, after starvation for 2 h, were adequately fed for 20 min in contrast with those fed with ordinary culture medium. No significant difference (ns, P > 0.05) was observed, excluding the possibility that changes in feeding rates may interfere with the tumor invasion and cell migration phenotype, and could be used for subsequent experiments (Supplementary Figure 2).

Due to the ectopic expression of Delta, a ligand for the Notch pathway, in the eyeful tumor model and the key role of Notch signal in tumor metastasis, the mRNA expression levels of key factors in the Notch signal pathway were detected using RT-qPCR including Delta (FlyBase ID: FBgn0000463), Serrate (FBgn0004197), Notch (FBgn0004647), Suppressor of Hairless (Su(H), FBgn0004837), and Enhancer of split (En(spl), FBgn0000591) (Figure 6A). The mRNA levels of these genes were significantly enhanced in the D. eyeful tumor metastasis model compared to those in the control model. After treatment with 25.00 mg/mL Binpu-3RE, the enhanced mRNA expressions of these genes’ mRNA were impeded (Figures 6B–F), indicating that Binpu-3RE could downregulate the transcription levels of Delta, Serrate, Notch, Su(H), and En(spl) in the Notch pathway.

The reporter markers Delta-LacZ and Su(H)-LacZ were used to detect the expression of Delta and Su(H) in vivo using an X-gal staining assay. The eyeless (ey) protein is expressed in all cells of the eye discs at the second-instar stage and is expressed anterior to the morphogenetic furrow (MF) but not in the anterior-most part of the eye discs (where it contacts the antennal disc) at the third-instar larval stage (Bessa et al., 2002). The enhanced membrane GFP (mGFP)-labeled expression region of the ey-GAL4 driver was mainly located near the MF in the third instar larval eye-antennal imaginal discs (Figures 6G, G’). Compared with the ey-GAL4 control (Figures 6H, K), the expression of Delta-LacZ or Su(H)-LacZ was significantly upregulated both in the antennal and eye discs of the eyeful tumor model at the third instar larval stage, and an increased volume was observed (Figures 6I, L). Therefore, the transcription of Delta or Su(H) was enhanced in the eyeful tumor in both a cell-autonomous and non-autonomous manner. After treatment with 25.00 mg/mL Bingpu-3RE, the increased Delta-LacZ and Su(H)-LacZ levels were inhibited (Figures 6J, M), implying that Binpu-3RE restrains the mRNA levels of Delta and Su(H) in vivo.

A transgenic fly line for the NRE-GFP construct, in which enhanced GFP expression marked cells with active Su(H)-dependent Notch signaling, was used to visualize canonical Notch signaling activity in D. eyeful tumors in vivo (Saj et al., 2010; Beira et al., 2018). The expression of NRE-GFP was low in the control group (Figures 6N, N’) and strongly promoted in the third instar larval eye-antennal imaginal discs (Figures 6O, O’). However, this increased expression and enlargement of the eye-antennal discs was inhibited after treatment with 25.00 mg/mL Binpu-3RE (Figures 6P, P’) as well as the enlarged size of the eye-antennal discs. Together, these data suggest that Binpu-3RE can effectively inhibit the activity of the Notch signaling pathway in vivo.

Based on the close relationship between Notch signal and EMT and their central roles in tumor cell migration and metastasis (Espinoza et al., 2013; Gonzalez and Medici, 2014), the expressions of β-integrin and MMP1 were measured using immunohistochemistry staining to identify potential mechanisms. Compared to the control (Figures 7A, A’, D, D’), protein expression levels of β-integrin and MMP1 were upregulated in the eyeful tumor model, though the upregulation was inhibited after treatment with 25.00 mg/mL Binpu-3RE (Figures 7B, B’, C, C’, E, E’, F, F’). Hence, we speculated that Binpu-3RE may interfere with the EMT process by inhibiting the expression of β-integrin and MMP1, thereby regulating tumor metastasis.

To evaluate the effect of Binpu-3RE on human breast cancer cells, we prepared rat control serum and Binpu-3RE-containing serum (Binpu-3RE-L, Binpu-3RE-M, Binpu-3RE-H) for use in treating MDA-MB-231 TNBC cells. Compared with the control (treated rat control serum), the cell viability of MDA-MB-231 cells was obviously impeded by Binpu-3RE-H at 24 h, by Binpu-3RE-M and Binpu-3RE-H at 48 h, and by all Binpu-3RE-containing sera at 72 h (Figure 8A), as assessed using the CCK-8 assay. Next, we employed the in vitro wound healing and transwell assays to assess the cell migration and cell invasion capabilities of MDA-MB-231 cells, respectively. Following exposure of Binpu-3RE at different doses, the cell migration and invasion rates were both largely suppressed in the MDA-MB-231 cell line (Figures 8B–E).

As the Notch signaling pathway is highly conserved from fly to human, we further detected the level of Notch signaling in MDA-MB-231 breast cancer cells. As hypothesized, mRNA and protein expression levels of the key factors in the Notch signaling pathway in mammals, viz. Notch1, Jagged1, and HES1 were distinctively reduced by Binpu-3RE treatment (Figure 9). Together, these data imply that the positive effect of Binpu-3RE on tumor cell migration and invasion is through inhibiting Notch signaling with evolutionary conservatism.