Effects of Xuefu Zhuyu Decoction on Cell Migration and Ocular Tumor Invasion in Drosophila

Tangshan Key Laboratory of Traditional Chinese Medicine Pharmacology, College of Traditional Chinese Medicine, North China University of Science and Technology, 21 Bohai Road, Tangshan 063210, China School of Life Science and Technology, ShanghaiTech University, 393 Middle Huaxia Road, Shanghai 201210, China Department of Neuroscience, Scripps Research Institute, 130 Scripps Way Jupiter, Florida 33458, USA College of Integrative Medicine, Fujian University of Traditional Chinese Medicine, Fuzhou 350122, China


Introduction
With the increased incidence, cancer has become a worldwide disease and represents one of serious health-care issues for human race [1,2]. More than 90% fatalities of human cancers are due to the ability of tumor metastasis at late stage, or form discontinuous secondary tumor foci at distant locations, instead of the localized primary tumor growth [3,4]. Tumor invasion development is an extensive complex progress with multiple steps. Firstly, with the metastatic feature of decrease in adhesive force and increase in superficial charges, cancer cells could separate from the initial tumor. Through releasing numerous proteolytic enzymes and other factors, they possess the abilities of penetrating basement membrane (BM) and destroying cellular matrix, thus infiltrating into peripheral tissue [5]. After cancer enters the vascular and lymphatic vessels, these tumor cells survive and spread across the body through the circulatory system. Finally, they pierce through the vessel wall and settle at a secondary location to expand the metastatic focus [6][7][8]. A large number of genes and signaling pathways participate in this biological behavior, providing potential drug targets and valuable therapeutic strategies for clinical cancer treatments [9][10][11][12]. Due to the adverse effects and high cost, synthetic chemotherapeutic anticancer drugs have not been as successful as expected in preventing and fighting against cancers in the clinic. Hence, the need to develop effective, safe, and affordable anticancer drugs from alternative and complementary herbal resources is becoming more and more urgent.
Traditional Chinese Medicine (TCM), with ancient application history in China or other regions of Asia, has gained considerable importance in defending against cancer during the last two decades [13][14][15]. Blood stasis, characterized by blood flow retardation and blood stagnation [16][17][18], is an essential factor for tumor metastasis during tumorigenesis in the TCM basic theory system [19,20]. Usually, researchers utilize the hemorheology indexes to evaluate the patients with blood stasis syndrome or the established blood stasis animal models rather than pathological methods [21,22]. Gao et al. checked the nail fold microcirculation in tumor patients and found that microcirculatory disturbance makes blood flow slow down, whose erythrocyte aggregation, exudation, and hemorrhage in turn promote the generation, development, and metastasis of tumor [23]. And, the association between blood stasis and cancer has received increasing attention in the studies related to TCM and integrative Chinese and Western medicine [24][25][26]. Thus, promoting blood circulation to remove blood stasis has become one of the major treatment principles for cancer in TCM. While the blood-activating stasis-resolving drugs (BAHRDs) have been well documented in antitumor therapy, their roles in cell migration and tumor metastasis remain controversial. Several reports have indicated that BAHRDs could inhibit tumor cell proliferation, invasion, and metastasis and reduce adverse reactions of radiotherapy and chemotherapy [27][28][29]. Conversely, other studies suggested that BAHRDs can improve local microcirculation and provide a richer blood supply for tumor growth, thereby promoting tumor metastasis [30]. A possible explanation for the discrepancy is that this category of drugs has multiple active ingredients, which are mainly categorized into the following: flavonoids, terpenoids, alkaloid, and fatty oil [31]. Different compounds and various combinations may trigger different gene expression profiles and signaling pathways, thus producing diverse even opposite effects. Additionally, most work was done in vitro or in cultured cancer cell lines that could not adequately recreate the in vivo condition.
Xuefu Zhuyu Decoction (XFZYD), a traditional Chinese herbal formula, has been widely investigated and applied in the clinic mainly for treating blood stasis syndrome. It contains 11 herbs (Supplementary Table 1 showing the drug composition and medicinal part, as well as the amount of each herb in one dose XFZYD) [32] and is originated from the well-known medical book Yi Lin Gai Cuo (Corrections of Errors Among Physicians) which was written by Wang Qingren in Qing Dynasty [33]. As a representative of blood-activating stasis-resolving recipe, XFZYD is mainly used to treat metabolic syndrome and cardiac-cerebral vascular diseases, including hyperlipidemia, hypertension, coronary heart disease, traumatic brain injury, and ischemic stroke [34][35][36][37][38][39][40]. It has been reported that XFZYD displays a significant effect on the pretreatment of the myocardium in sepsis rats by inhibiting myocardial cell apoptosis and antioxidation [41]. In addition, researchers found that XFZYD could protect against retinal ischemic by downregulating HIF-1α and VEGF via a synergistic inhibition of RBP2 and PKM2 [42]. According to the Chinese medical theory, blood stasis is one of main leading causes for malignant tumor formation. With the efficacy in promoting circulation to remove blood stasis, we considered that XFZYD could be potentially used to develop an effective treatment for malignant tumors. However, the effects of XFZYD on tumor metastasis and the underlying mechanisms remain poorly understood.
With less genome redundancy and more genetic tools, Drosophila melanogaster serves as an elegant in vivo model for the study of tumor metastasis [43][44][45]. Many genes and signal transduction pathways pivotal for cancer progression were first identified in flies and subsequently verified in mammals. Moreover, several powerful tumor invasion models have been well established in Drosophila melanogaster. For instance, the loss of C-terminal SRC kinase (csk) or cell polarity genes, like scribbled (scrib), disc large (dlg), or lethal giant larvae (lgl), in the anterior/posterior (A/P) compartment boundary region triggers invasive cell migration phenotype [10,46,47]. Drosophila is also becoming a potential model for studying ocular tumors [48]. Ferres-Marco et al. found that coupled with overexpression of Delta (the ligand for Notch signal), downregulation of the expression of two epigenetic genes pipsqueak (psp) and longitudinal lacking (lola) induces the formation of Drosophila eye metastatic tumors during development, which is termed eyeful [49]. By using these important in vivo models, the purpose of our study was to analyze the effects of XFZYD extract on cell migration and ocular tumor invasion in Drosophila and to provide the potential molecular mechanisms for XFZYD in clinical treatment on malignant tumors.    Table 1 showing the amount of each herb in one dose) was weighed and soaked overnight [32]. After being soaked for 12 h, the herbs were extracted in boiled double distilled water (ddH 2 O) (100 g herbs to 1000 mL ddH 2 O) at 60°C for 3 hours. The extract was centrifuged at 4200 × g for 30 min at 25°C, followed by 2

Materials and Methods
BioMed Research International vacuum filtration with a 0:45 μm filter. The XFZYD extract was stored at -80°C and then freeze dried with a lyophilizer and then reconstituted in ddH 2 O in order to obtain a final stock concentration of 0.78 g/mL (Supplementary Figure 1l). XFZYD extract was added directly to regular food from a 0.7800 g/ml aqueous stock to a final concentration of 0.0100, 0.0125, 0.0200, 0.0500, and 0.1000 g/ml. For control food, only water was used.

Quality Control of Xuefu Zhuyu Decoction.
To control the quality of XFZYD, a high-performance liquid chromatography (HPLC) method was performed to establish the fingerprint spectrum. The analyses were performed with Shimadzu LC-20A. The chromatographic column was Agilent Eclipse XDB-C18 (4:6 × 250 mm, 5μm). The mobile phase flow rate was 1 ml/min, and column temperature was maintained at 25 ± 1°C. The paeoniflorin, ammonium glycyrrhizinate, and naringin were dissolved in methanol as control detected by the HPLC method. Paeoniflorin was eluted with

Drosophila Strains and
Genetics. Flies were kept on a cornmeal and agar medium at 25°C with a 12 h light-dark cycle incubator according to standard protocols unless indicated. Drosophila stocks used include the following: w 1118 (#3605) and Oregon RC (#0005) were obtained from Bloomington Drosophila stock center (BDSC); UAS-scrib-IR (#27424) was obtained from Vienna Drosophila RNAi center (VDRC); ptc-GAL4 UAS-GFP (ptc>GFP) was previously described [51,52]. eyeful/Cyo fly is a kind gift of Professor Lei Xue from Tongji University, which has been established by the homologous chromosome recombination (HCR) of ey-GAL4 (#8220, BDSC), UAS-Delta (#5614, BDSC), and GS88A8 (previously described [49]) fly strains via the balancer fly stock Sco/Cyo (#2555, BDSC). The crossing scheme for establishing the eyeful model is shown in Supplementary  Figure 2. For all fly crossing experiments, healthy unmated male and female parents were randomly assigned to different groups.

Drosophila Cell Migration Model.
To set up the cell migration model, female flies with genotype ptc>GFP were crossed to male flies with genotype UAS-scrib-IR. These fruit flies mated and laid eggs in one tube, which contains normal food. When the offspring reach the 3 rd instar larval stage, the larvae with genotype ptc>GFP/UAS-scrib-IR were collected and dissected (Supplementary Figure 1n). After fixation, the migrated cells were observed and recorded, which are marked by green fluorescent protein (GFP) in the wing pouch region through a fluorescence microscope. Simultaneously, the migrating cell number was counted and the migrating distance was measured (Supplementary Figure 3). For the XFZYD-treated groups, the offspring were kept on the XFZYD-added medium. Meanwhile, female flies with genotype ptc>GFP were crossed to male files of the strain w 1118 on the regular food, and the progeny larvae with genotype ptc>GFP/+ were collected as the control group. For cell migration experiments, all crosses were raised at 25°C for 2 days, then shifted to 29°C for additional 3 days; the 3 rd instar larvae were dissected [10].
2.6. Drosophila Ocular Tumor Invasion Model. For ocular tumor invasion assays, female w 1118 flies and male files with genotype eyeful/Cyo were crossed. The parental flies mated and laid eggs in one tube, which contains normal food. When the offspring reach the adult stage, the flies with genotype eyeful/+ were collected as the model tumor invasion group. Their ocular cells were observed under a stereo microscope, their primary growth and invasion location were recorded (Supplementary Table 3), and the percentage was calculated. For the XFZYD-treated groups, the offspring were kept on the XFZYD-containing medium. Meanwhile, Oregon RC flies which were raised on the regular fly food were used as the control group. 2.8. qRT-PCR. TRIzol (Invitrogen) was used to isolate total RNA from ten wing imaginal discs dissected from the third instar larvae or thirty adult heads collected from freshly eclosed flies of indicated genotypes, and qRT-PCR was performed as previously described [53] using the following primers: (1) For rp49: sense-5 ′ -TACAGGCCCAAGATCG TGAA-3 ′ ; antisense-5 ′ -TCTCCTTGCGCTTCTT GGA-3′ (2) For Su(H): sense-5 ′ -CTTGCTGCCGGGTCCT TAC-3 ′ ; antisense-5 ′ -CTCGCGCATGTACTTC TCCA-3 ′ 3 BioMed Research International 2.9. Data and Statistics. All data were verified in at least three independent experiments. Results were presented as bar graphs or scatter plots created with GraphPad Prism 8.0. For statistical significance, one-way ANOVA with Bonferroni's multiple comparison test or chi-squared test was applied. P value less than 0.05 was considered significant and center values as the mean. Error bars indicated standard deviation. P value less than 0.05 was considered significant (ns was not significant; P ≥ 0:05, * P < 0:05, * * P < 0:01, * * * P < 0:001).

Xuefu Zhuyu Decoction Preparation and HPLC Analysis.
To examine the quality of the Xuefu Zhuyu Decoction (XFZYD), three major compounds were used as quality control standards. When XFZYD high-performance liquid chromatography (HPLC) chromatogram was compared with the standard controls ( Supplementary Figure 4a-c), we observed the distinguished peaks for paeoniflorin, ammonium glycyrrhizinate, and naringin in the chemical fingerprints of XFZYD (Figures 1(a)-1(c)). In addition, the content of these three compounds in XFZYD determined by the HPLC method is 0.140%, 0.039%, and 0.220%, respectively. Taken together, these results indicated that the prepared XFZYD consists of the necessary active ingredients and is suitable to carry out the study.

XFZYD Inhibits Cell Migration in Drosophila Wing Discs.
To investigate the XFZYD's suppressive effect on cell migration in vivo, we employed a well-established cell migration model in Drosophila 3 rd instar larval wing imaginal discs [10,47]. Compared with the control (Figures 2(a), (a')), depletion of the cell polarity gene scribble (scrib) driven by patched-GAL4 (ptc-GAL4), which was mediated by RNA interference (RNAi), triggered massive cell migration phenotype (Figures 2(b), (b')). In the ptc>GFP/UAS-scrib-IR-(ptc>scrib-IR-) induced cell migration model, the green fluorescent protein-(GFP-) labeled migrating cells were detached from the anterior/posterior (A/P) compartment boundary and moved toward the posterior part of the wing imaginal discs (Figure 2(b')). To further quantify the cell migration phenotype, we calculated the total number of migrating cells and measured the median or maximum (max) migrating distance in the selected wing pouch region (Supplementary Figure 3) and observed a significant increase in migrating cell number and migrating distance (Figures 2(h)-2(j)).
Next, to test whether there is the inhibitory effect, we prepared feeding media consisting of XFZYD aqueous extract (Supplementary Figure 1m) and raised the ptc>scrib-IR flies from the egg stage to the 3 rd instar larval stage. Results showed that the loss-of-scrib-induced increased migrating cell number was notably suppressed by XFZYD extract at the concentration of 0.0125 g/ml or 0.0200 g/ml, but not that of 0.0100 g/ml or 0.0500 g/ml (Figures 2(b)-2(f), (b')-(f'), (h)). For the max or median migrating distance, XFZYD at the concentration of 0.0125 g/ml showed a stronger inhibitory effect than that of 0.0200 g/ml (Figures 2(i) and 2(j)), while the concentrations of 0.0100 g/ml or 0.0500 g/ml did not perform an obvious inhibitory effect. Intriguingly, we found that XFZYD at the concentration of 0.1000 g/ml could strongly aggravate the ptc>scrib-IR-triggered cell migration in cell number, but not in distance (Figures 2(g)-2(j), (g')). Collectively, XFZYD displayed a dose-independent and biphasic effect on the cell migration in the Drosophila larval stage. And the concentration of 0.0125 g/ml for XFZYD water extracts was chosen, which significantly suppressed cell migration, to further explore the mechanisms.

XFZYD Alters MMP1, β-Integrin, and E-Cadherin
Expression. To further elucidate the underlying mechanisms of XFZYD in inhibiting cell migration, we analyzed several cell migration-related key factors. Matrix metalloproteinase-1 (MMP1), as a member of the MMPs family, has a direct role in enhancing tumor invasion, whose overexpression indicates a poor prognosis [54][55][56]. Clearly, inhibition of the well-characterized tumor suppressor E-cadherin's function or expression will lead to cell epithelial-mesenchymal transition, promoting cell migration and metastasis [57]. In addition, as a transmembrane receptor which is composed of αand β-subunits, integrin adheres to and migrates between cells and is phosphorylated after being subjected to intracellular and extracellular metastasis signals to produce tumor cell migration [58,59]. In agreement with these previous works, results showed that lossof-scrib under ptc-GAL4 control along the A/P boundary could trigger strong epithelial-mesenchymal transition (EMT) like phenotype in the wing pouch region, revealed by upregulated expression of MMP1 (Figures 3(a) and 3(b)) and the major β-subunit of integrin (βPS-integrin) (Figures 3(d) and 3(e)) [60,61] and downregulation of cell adhesion molecule E-cadherin (Figures 4(a) and 4(b)). And the fluorescence intensity analysis results showed that the altered expression of MMP1, βPS-integrin or Ecadherin is nearly located in the ptc>scrib-IR region which was visualized by GFP (Figures 3(b"'), 3(e"'), and Figure 4(b"')), indicating that depletion of scrib triggers the increase of MMP1and integrin, or the decrease of Ecadherin in a cell-autonomous manner. Moreover, the upregulation of MMP1 and β-integrin or the reduction of E-cadherin induced by ptc>scrib-IR was remarkably ameliorated by XFZYD at the concentration of 0.0125 g/ml (Figures 3(c) and 3(f), and Figure 4(c)). Together, these evidences suggest that the strong inhibitory effect of XFZYD on cell migration may be due to altering the expression of MMP1, β-integrin, and E-cadherin.

Xuefu Zhuyu Decoction Suppresses Ocular Tumor
Invasion. Considering cell migration is a central step for metastasis during tumorigenesis, we introduced a Drosophila ocular tumor invasion model, which is named eyeful [49], to further analyze the roles of XFZYD in metastasis. In line with a previous study, we found that exogenous ectopic expression of Delta, which functions as the ligand for the Notch signaling pathway, under the control of ey-GAL4 in Drosophila compound eye, and simultaneous depletion of epigenetic genes psp and lola leads to overproliferation in situ and metastatic tumor formation of Drosophila eye tissue cells   5(a)-5(h)). Again, 0.0200 g/ml or 0.0500 g/ml XFZYD significantly suppressed the migration rate of the eyeful model ( Figure 5(i)), whereas it slightly increased the migration level at the concentration of 0.1000 g/ml ( Figure 5(i)). Meanwhile, XFZYD had almost no suppressive effect on Drosophila ocular tumor invasion at the concentra-tion of 0.0100 g/ml or 0.0125 g/ml (Figures 5(i)-5(k)). Given that the Notch signal is upregulated in the eyeful tumor model, it was hypothesized that XFZYD may inhibit tumor invasion through modulating Notch signaling activity. To verify our assumption, a real-time quantitative PCR (qRT-PCR) assay was performed. In accordance with a previous Error bars indicate the standard deviation. One-way ANOVA with Bonferroni's multiple comparison test was used to compute P values: * * * P < 0:001, * * P < 0:01, and * P < 0:05 * ; ns: no significant difference. Scale bar: 50 μm (a-g and a'-g').
DNA GFP PS-integrin PS-integrin GFP Figure 3: XFZYD downregulates MMP1 and β-integrin expression. Merged fluorescence micrographs of the 3 rd instar larval wing discs stained with anti-MMP1 (a-c) or anti-βPS-integrin antibody (d-f) are shown. The individual channels detecting only GFP (green, a'-f'), only MMP1 (red, a"-c"), and only βPS-integrin signal (red, d"-f"). Nuclei (DNA) were labelled with DAPI (4',6-diamidino-2phenylindole, blue). (a"'-f"') Graphs of total intensity sum of each fluorophore with respect to the region of interest (indicated by the yellow solid line) as shown in (a-f), respectively. Fluorescence intensities were measured in pixels using Image-Pro Plus 6.0. Scale bar: 50 μm (a-f). 7 BioMed Research International study, we found that the mRNA level of the Suppressor of Hairless (Su(H)), which acts as the transcription factor in Notch pathway [62], was upregulated in the eyeful invasion tumor model ( Figure 5(l)). And XFZYD remarkably impeded the eyeful-induced elevated transcription level of Su(H) at the concentration of 0.0500 g/ml ( Figure 5(l)). Hence, we conclude that XFZYD inhibits tumor invasion through downregulating Notch signaling activity.

Discussion
Cancer has been a worldwide fatal disease that seriously threatens human health. As more than 90% of cancer patients died from tumor migration, suppression of tumor cell migration has emerged as the central target for cancer treatments [3,4]. Drosophila melanogaster is regarded as an excellent model system to study the genetic and molecular mechanisms of tumorigenesis and metastasis [63,64]. In present study, the ptc>scrib-IR-induced cell migration model and eyeful ocular tumor invasion model were established and used to investigate the protective effect of XFZYD.
As a well-known TCM herbal formula, XFZYD is categorized into the BAHRDs and has been extensively applied for treating metabolic syndrome and cardiac-cerebral vascular diseases [34][35][36][37][38][39][40]. Although blood stasis is also a main pathogenic mechanism for tumor metastasis in the TCM theory, the application of XFZYD in treating malignant tumors is largely unknown. Interestingly, our in vivo data showed that XFZYD decreased cell migration and invasion at the low dose (the concentration of 0.0125 g/ml for cell migration and that of 0.05 g/ml for ocular tumor invasion) but increased cell migration and invasion at a high dose (the concentration of 0.1000 g/ml for both). Collectively, the effects of XFZYD on cell migration and tumor invasion in Drosophila displayed a dose-independent and biphasic manner, whereas, in a septic shock rat model, XFZYD may reduce myocardial damage and a protective role for the heart structure and function in a dose-dependent manner [41]. Besides, we have noticed that the doses of XFZYD that affect cell migration and tumor invasion are different, which may be due to distinct genetic backgrounds of two fly models. Several previous studies showed that multiple cancer-related factors and signal 0.1000 0.0500 0.0200 9 BioMed Research International transduction pathways exhibit varied expression levels and activities in the ptc>scrib-IR cell migration and eyeful tumor models [10,[47][48][49]. And another reasonable explanation is that the best inhibitory effects of different doses on the two models may be caused by different active ingredients (or different combinations of active ingredients) in XFZYD.
Based on quantifications of daily Drosophila food intake [65], it is estimated that flies raised on medium containing 0.0125 g/ml, 0.05 g/ml, or 0.1000 g/ml XFZYD ingest about 0.4 mg/kg body weight of the drug per day [66], which is, respectively, comparable to the treatment dosage of 26.04 g, 104.2 g, or 208.3 g per day for human patients. In the clinic, the patients with coronary heart disease were treated with XFZYD one dose (78 g) a day [32]. Thus, according to the normal dose, taking XFZYD 26.04 g/day for human (equivalent to 0.0125 g/ml for fly) is far below the usual medication dosage (78 g/day). Above all, the data may provide a key clue for the application of XFZYD in cancer clinical treatment.
Notch signaling is an evolutionarily conserved pathway that regulates many cellular processes, including cell proliferation, survival, apoptosis, invasion, angiogenesis, and stem cell self-renewal [67][68][69]. Recently, Notch signal has been regarded as a central regulator in the induction of EMT, which is important for migration and metastasis of cancer cells [70][71][72]. In addition, several studies revealed the molecular mechanisms behind Notch-mediated EMT regulation during non-small-cell lung cancer and breast tumorigenesis [73,74]. Consistently, in this study, XFZYD showed its anticancer activities through inhibiting EMT-mediated cell migration (Figures 3 and 4) and downregulating Notch signal activity ( Figure 5) in Drosophila. Overexpression of Delta (the ligand in fly) is a very powerful strategy to increase Notch signal by releasing the Notch intracellular domain (NICD), which enters the nucleus and regulates the transcriptional activity of Su(H) [75]. Thus, we would like to examine the subcellular distribution of the NICD protein (tagged by GFP/YFP) and the expression of Notch/Su(H) target genes (Enhancer of split (E (spl)), Cut and Wingless) in vivo to directly monitor Notch pathway activity for further investigation [76][77][78][79].
In conclusion, our work proved that XFZYD has anticancer activity on cell migration and tumor invasion at the moderate concentrations. Moreover, XFZYD could alter the expression of MMP1, integrin, or E-cadherin to suppress cell migration. Finally, it is demonstrated that XFZYD impedes ocular tumor invasion presumably by inhibiting Notch signaling activity. In sum, a positive function of XFZYD in suppressing cell migration and tumor metastasis has been shown, and the results indicated the potential drug targets and provide key clues for cancer clinical treatment strategies.