Using of animal models of angiogenesis to confirm a Bidens pilosa-sourced polyacetylenic glucoside inhibits angiogenesis targeting hypoxia through VEGF and PDIA4 suppression

Anti-angiogenesis is a pivotal combination treatment approach in cancer therapy but rare using on companion animals. This study aimed at evaluating the anti-angiogenic effect of a B. pilosa sourced polyacetylenic glucoside, cytopiloyne, on various in vitro assays and in vivo models. We provide evidences showing that CP has anti-angiogenic activities. Firstly, CP inhibited sponge/ Matrigel plug angiogenesis from tumor cells and decreased the survival of tumor cells on hypoxic conditions. Besides, CP declined PKCα protein expression which a protein leads to the growth and spread of tumors under hypoxia. Secondly, inhibitory effects of CP on endothelial angiogenesis were confirmed by chick chorioallantoic membrane assay, tube formation of SVEC4-10 cells and Matrigel plug assay. A dose-dependent CP treatment inhibited 4T1 cells proliferation under hypoxia and migration. It also suppresses VEGF transcription under hypoxia. Finally, we found that CP decreased Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 4 November 2020 doi:10.20944/preprints202011.0198.v1 © 2020 by the author(s). Distributed under a Creative Commons CC BY license. PDIA4, a novel regulator of cancer growth, expression in endothelial cells. This effect was confirmed by PDIA4 knockout mice with reduced angiogenesis in Matrigel plug assay. Taken together, these results suggest that CP might act as a promising anti-angiogenic herbal agent candidate to be used in animal hypervascularized cancer of veterinary medicine or in combination to control human cancer as adjuvant therapy.


Introduction
Translation of cancer therapeutic strategy between pets and humans became an important target on comparative oncological investigations [1]. A  cancer. Cancer is often more common in dogs than cats. 50% of dogs over ten years of age develop cancer. Cancer is the leading cause of death in dogs and cats, mortality of cancer dogs and cats were about 47% and 32%, respectively. As for treatments, common treatment methods include surgery, chemotherapy, and radiation therapy.
Anti-angiogenesis is a pivotal combination treatment approaches in cancer therapy but rare using on companion animals. Current methods using anti-vascular endothelial growth factor (VEGF) antibodies or inhibitors targeting VEGF receptors after surgeon on cancer patients [2]. Due to their low blocking VEGF efficiencies on signaling transfer, toxicity and high risk of adverse effects in clinic [3], a mild and auxiliary method co effectively block cancer angiogenesis during treatment is necessary in veterinary clinic. Recent studies in pet dogs with cancer were undertaken to assist in the evaluation of anti-angiogenic peptide mimetics of thrombospondin 1 (TSP1) [1].
Follow-up study for cooperative activity between cytotoxic chemotherapy and TSP1 anti-angiogenic treatment in dogs with lymphoma has now supported these potential combinational therapies [4].
However, the anti-angiogenic mechanism with hypoxia underlying CP is not clear.
Further, over 200 compounds were identified in B. pilosa [12], the using angiogenic models to confirm its anti-angiogenic compounds are deficient, which limits the clinical use in the veterinary medicine.
In this study, we first evaluate the anti-angiogenic effect of CP, on various in vitro systems and in vivo models. Next, using hypoxia system to confirm CP owned anti-angiogenic mechanism through protein kinase C (PKC) and VEGF modulation.

The Sponge/Matrigel Angiogenesis Assay
We used a sponge angiogenesis assay modified from the method described by previous studies [27,28]. Matrigel (500 μL) was injected subcutaneously in the midventral abdominal region of C57BL/6 mice (10-12 weeks of age) and permitted to solidify. Subsequently (after 20-30 min), mice were anesthesized with ketamine 100 mg/kg and xylazine 7.0 mg/kg. The skin overlying the Matrigel plug was gently shaved, after which a small (0.5 cm) nick was made in the skin using a #15 surgical blade. Using the same blade, a smaller nick was made in the Matrigel plug. A sterilized polyvinyl sponge (appx. 3×2×1.5 mm) containing with/without 0.110 6 4T1 cells to induce angiogenesis was introduced through the nick in the Matrigel and advanced to the center of the plug with the help of tweezers. The wound was then closed with a suture. Mice were observed after 24 h to monitor condition of the wound. Plugs could be recovered for several weeks but typically tumor-induced angiogenesis was measured after 7 and 14 days. Mice were killed after 3-5 min; the Matrigel plug with sponge was removed, separated from the abdominal muscle, fixed in 10% formalin and stained following sections with H&E. Light microscopy was used to visualize these sections for blood vessels formation and metastasis potential.

Mouse Matrigel Plug Assay
Flanks of C57BL/6 mice (8 weeks of age) were injected subcutaneously 500 μL of Matrigel (BD Biosciences, MA, USA) with an ice-cold syringe containing bFGF (100 ng/ mL) and heparin (50 U/ 500 μL) with or without CP (2.5 or 5 μg). After seven days, the skin of the mouse was pulled back to expose the Matrigel plug, which remained intact. The Matrigel plugs were weighted and photographed. To quantitate the formation of functional blood vessels, the amount of hemoglobin (Hb) was measured using the Drabkin hemoglobin assay with Drabkin reagent kit 525 (Sigma, MO, USA) as described previously [30]. Plugs were fixed in 10% formalin and stained following sections with H&E. Moreover, same experiment and operation was also completed on PDIA4 knockout mice. Another angiogenic factor VEGF (500 ng/mL) was used in Matrigel (500 μL).

SVEC4-10 endothelial cells (2.5×10 4 cells/ well) were added with serum-free DMEM
with/without CP (5 μg/mL) for 24-48 hr pretreatment. Tubes will begin to form within 2-4 hr. The final tube phenomenon will be examined under light microscopy after 6 hr.
2.9. In Vitro Migration Assay [32] 4T1 cells in medium containing 10% FBS were seeded into wells of 24-well plates.
After the cells grew to confluence, wounds were made by sterile pipette tips. Cells were washed with PBS and refreshed with medium with/without 10% FBS and various concentrations of CP (1, 2.5 and 5 g/mL). After overnight incubation at 37°C, the cells were fixed and photographed.

Statistical Analysis
Data from three independent experiments or more are presented as mean  SEM. Two-tailed Student's t test, Kruskal-Wallis test and ANOVA test were used for statistical analysis of differences between Groups according to the data type, and a P value of less than 0.05 was considered to be statistically significant.

In vivo Effects of CP on Tumor-induced Angiogenesis in Sponge/Matrigel Assay
There are many murine in vivo tumor models for investigating angiogenesis, especially work on oxygen supply and metastasis [35,36]. These tumor dissemination methods include tumor cells injection (subcutaneous, intravenous, intraperitoneal and intrahepatic etc.) and graft transplantation but without direct interaction data between angiogenesis and metastasis. For this reason, we use sponge/Matrigel angiogenesis assay [27] was used to check anti-angiogenic effect of CP on hypoxia condition. We  Black, blue and green arrows indicated the location of methylcellulose disk, new formation of blood vessels and inhibition of angiogenesis, respectively.

Influence of CP on Angiogenesis in Chick Chorioallantoic Membrane Assay
The CAM has served as an in vivo platform for advantages to research and manipulate vascular functions near 60 years. This system has been used for the study of vascular development and angiogenesis, especially on tumor growth and metastasis [37]. It can also be advanced studies on respiratory properties, ion transport [34,35], selective vascular occlusion therapies, biocompatibility of engineered materials, drug distribution and toxicology [38][39][40]. Control CAMs implanted on the empty methylcellulose disks without angiogenic inhibitor did not develop avascular zones as determined by visual examination (Fig. 1C; Mock group). We found that an inhibitor of urokinase plasminogen activator, amiloride, cause a significant reduction of angiogenesis ( Fig. 1C; Amiloride group). A larger avascular zone outside of area covered by disk containing this inhibitor was observed. As the positive controls, we used VEGF and bFGF disks, dense areas or increased newly formed vessels were developed ( Fig. 1C; VEGF and bFGF groups). In contrast, CP-contained disks apparently showed inhibition of angiogenesis ( Fig. 1C; CP2.5 and CP10 groups) under concentrations (2.5 and 10 µg/ disk) lower than amiloride group (30 µg/ disk).

CP Inhibits Angiogenesis in in vivo Murine Matrigel Plug Assay
We used matrix gel plug assay, an efficient method, to assess angiogenic and antiangiogenic compounds. First, bFGF was used as positive control. Removed Matrigel plugs were photographed. An increased redness level indicated that angiogenesis was induced by angioenic factor, bFGF ( Fig. 2A). More than 30 folds of the content of hemoglobin of the plugs provided a synchronous data on bFGF-induced angiogenesis (Fig. 2B). This significant data was confirmed by higher erythrocytes distribution in Matrigel section with H&E stain (Fig. 2C). In contrast, CP significantly inhibited angiogenic effect of bFGF in Matrigel plug. The redness level of plugs was decreased, especially on 5 µg CP/mL plug ( Fig. 2A) and confirmed by significant lower content of hemoglobin with almost same as control ones (Fig. 2B). Almost same erythrocytes distributions in plug sections between Mock group and CP groups were observed (Fig. 2C). with or without CP (2.5 or 5 μg) were subcutaneously injected into C57BL/6 mice (8 weeks of age). After 7 days, the Matrigel plugs were removed for weight and photograph (A). The concentration of hemoglobin (Hb) was measured using the Drabkin hemoglobin assay (B). Sampling part of plugs were fixed and stained with 10% formalin and H&E, respectively. These sections were photographed by light microscope observations (C).

CP Inhibits the Tube Formation of SVEC4-10 Endothelial Cells
To estimate the effect of CP on the differentiation of SVEC4-10 cells, we carried out tube formation assay [31]. SVEC4-10 cells were placed on a growth factor-reduced Matrigel-coated plate and were incubated for 24-48 h. As shown in Figure 3A, SVEC4-10 cells on Matrigel formed blood vessel network in the absence of CP, whereas the treatment of CP strongly inhibited the formation of tube-like structure. Moreover, CP did not show toxic effect on SVEC4-10 cells for 24-48 hr (data not shown).

4T1 Cell Viability after CP Treatments under Normoxia and Hypoxia
As other investigation description, tolerated cancer cells in the hypoxia region will show faster proliferation. But as tumor size increased, the part located away from blood vessels will enhance deterioration on oxygen deficiency and form necrotic zone [41]. This vicious effect on hypoxia will induce advanced angiogenesis [42].
Therefore, screening toxic effect of compounds on cancer cells with angiogenesis under different oxygen levels, normoxia and hypoxia, is a pivotal step. We found that CP (1 and 5 g/ mL) did not induce any toxic response on 4T1 cells under normoxia by microscopic examination and MTT test (Fig. 3B and 3C). A significant concentration-dependent of increment of dead 4T1 cells after CP treatment under hypoxia was notice ( Fig. 3B and 3C).

CP Decreased PKC Expression VEGF Transcription of 4T1 cells under Hypoxia
In the early 1980s, PKC was identified as the target of phorbol esters of natural products [43]. Therefore, many dietary phytochemicals were screened for cancer chemoprevention based on PKC pathway related with tumor-promoting activity [44].
PKC, one of PKC isoforms, which can accelerate tumor cells proliferation/ metastasis and inhibit cancer cells apoptosis [45]. Hypoxia will also induce PKC expression to promote tumor VEGF production and malignancy [46][47][48]. We found that CP did not modify 4T1 cells PKC expression on normoxia. But a decreased PKC level in 4T1 cells with CP treatment was noticed in hypoxia (Fig. 3D). Hypoxia can significantly increase VEGF transcription and production [49]. We used RT-PCR confirmed that 4T1 cells under dose-dependent CP treatment showed lower VEGF transcription under hypoxia, but not normoxia (Fig. 4F).

Inhibition of 4T1 Cells Migration by CP
Apart from oxygen deficiency, fast growing tumor presented nutrition demands.

CP Decreased PDIA4 Expression of SVEC4-10 Endothelial cells
Abnormal PDIA4 expression combined endoplasmic reticulum stress is related with a self-protection to various diseases, including angiogenesis related to survival and progression of different cancer types [50][51][52]. A dose-and time-dependent CP treatment on SVEC4-10 endothelial cells will apparently decrease 30-70 % PDIA4 expression by western blot detection (Fig. 4A).

Angiogenesis in in vivo Murine Matrigel Plug Assay on PDIA4 Knockout Mice
Previous study indicated that PDIA4 is not an essential protein because mice with PDIA4 knockout survive without obvious phenotypes [51]. But no investigation worked on the relationship among angiogenesis, angiogenic factor and PDIA4. VEGF can also induce angiogenesis ( Fig. 4B and 4C) which is same as our data in this study of bFGF-induced angiogenesis (Fig. 2) in in Matrigel plug in C57BL/6 mice.
An increased redness level in Matrigel plug indicated that angiogenesis was induced by angioenic factor, VEGF (Fig. 4B). About 5 folds of the content of hemoglobin of the plugs provided a synchronous data on VEGF-induced angiogenesis (Fig. 4C). The VEGF-induced angiogenesis in Matrigel plug on PDIA4 KO mice were significantly decreased ( Fig. 4B and 4C). PDIA4 expression from these cell lysates were analyzed by western blot (A). Matrigel (500 μL) containing VEGF (500 ng/mL) and heparin (50 U/ 500 μL) were subcutaneously injected into PDIA4 knockout mice (8 weeks of age). After 7 days, the Matrigel plugs were removed for weight and photograph (B). The concentration of hemoglobin (Hb) was measured using the Drabkin hemoglobin assay (C).

Discussion
Arthur T. Hertig first name the term angiogenesis in 1935 [53]. Judah Folkman found a revolutionary novel way to consider about cancer and advanced therapy on tumor angiogenesis more than 40 years [54][55][56][57]. Based on anti-angiogenic mechanism worked on blocking nutrition and oxygen supply to tumors, more than 10 angiogenesis inhibitors were discover from 1980 to 2005 [57]. After that phytochemicals targeting angiogenic factor and other related inflammatory pathway (eg. upstream PKC activate a distinct set of transcription factors, including NF-κB) were screened for cancer prevention and adjuvant therapy [7,44,58]. There are many consistencies on tumor types and progression between bet animals and human.
Therefore, translational drug development studies became an optimal choice on integrated approach to link the pet dog with cancer and conventional preclinical models (mouse, research-bred dog and non-human primate) and the human clinical trial [1]. Tolerated cancer cells in the hypoxia region will show faster proliferation.
But as tumor size increased, the part located away from blood vessels will enhance deterioration on oxygen deficiency and form necrotic zone [41]. This vicious effect on hypoxia will induce advanced angiogenesis [42]. Here, we report a study on a polyacetylene glucoside, cytopiloyne (CP), purified from an edible herbal medicine B.
pilosa in inhibiting angiogenesis targeting hypoxia through VEGF and PDIA4 suppression through various in vitro and in vivo angiogenic models. Amazingly, our data revealed that CP can inhibit angiogenesis in different animal models researched on angiogenesis ( Fig. 1 and 2 Normal tissues with acute and chronic diseases will present hypoxia under inflammation and toxic response. Cancer cells can tolerate and proliferate in hypoxia but as tumor cells located > 180 μm from the blood vessels were observed to become necrotic [59]. Therefore, metastatic tumor cells or new blood vessels formation will prevent the progressive damage on cancer mass. The relationship between hypoxia and angiogenesis on tumor tissue organization were investigated by previous studies. Results of these investigation confirmed that hypoxia-induced angiogenesis will be occurred once the tumor size exceeds 1000 m diameter and/or the distance to the nearest blood vessel exceeds 180 m [60]. Direct cancer cells injection or tumor graft transplantation cannot fully provide a mimic animal model link hypoxia, angiogenesis and metastasis. So, we used a cancer cells-loaded sponge/Matrigel assay to confirm that CP can inhibit 4T1 cells migrate from sponge into Matrigel and form new blood vessel from implanted mice ( Fig. 1A and 1B) to overcome oxygen and nutrient deficiency. CAM assay and bovine aortic endothelial cells (BAECs) model are easy and quick ways to screen out drugs or phytochemicals influence angiogenesis. Based on these assays, there are 3 herbs enhanced and 7 herbs significantly inhibited angiogenesis which were screened from 24 traditional Chinese medicines used as curing ischemic heart disease in clinic [61]. Comparing positive angiogenic factors, VEGF and bFGF, CP can direct suppress chick embryo blood vessels development as amiloride, an inhibitor of urokinase plasminogen activator, inhibiting angiogenesis ( Fig. 1C). Besides the direct inhibitory effect of angiogenesis of CP, we also check whether or not CP can lower the angiogenic effect on bFGF stimulation. A dose-dependent CP treatment significantly curbs bFGF-induced angiogenesis used in vivo murine Matrigel plug assay (Fig. 2). At least 10-30 folds lower concentrations of hemoglobin extract from erythrocytes in plugs were analyzed in CP groups (Fig. 2B).
Moreover, the matrix gel plug assay has proven to be a convenient and powerful method to evaluate gene regulation in angiogenesis, angiogenic and antiangiogenic compounds in vivo, and to supplement in vitro tests [62]. As above mention, PDIA4 is gene related with regulating cancer growth [50][51][52]. Similar as Fig. 2, another angiogenic factor VEGF can induce angiogenesis in Matrigel plug in implanted mice ( Fig. 4). Significantly, PDIA4 KO mice did not show VEGF-induced angiogenesis in Matrigel plug ( Fig. 4B and 4C). The hemoglobin levels decreased about 5 folds in PDIA4 mice implanted with VEGF plug (Fig. 4C). It meant that PDIA4 is related with angiogenesis on tumor progression.
As other investigation description, polyacetylenes isolated from B. pilosa possess significant anti-angiogenic effects and regulate the expression of cell cycle mediators, p27(Kip1), p21(Cip1), or cyclin E, on human umbilical vein endothelium cells (HUVEC) [25]. Advanced data following this study indicated that a novel polyacetylene, 1,2-dihydroxy-5(E)-tridecene-7,9,11-triyne, structure similar with CP can promote apoptosis of HUVEC through overexpress death ligand FasL, activate caspase-7 and CDK inhibitors to inhibit angiogenesis [26]. These pilot studies worked on anti-angiogenesis of polyacetylenes from B. pilosa did not provide direct data on angiogenesis animal models and responses of cancer cells under hypoxia. But they suggest that phytocompounds such as polyacetylenes owned potential as candidates for anti-angiogenic therapeutics [26]. In our study, we provided different in vivo data to confirm that CP can inhibit angiogenesis and probably suppress metastasis (Fig. 1, and 2). Correlated in vitro data also prove that CP can slow 4T1 cells migration and endothelial cell tube formation ( Fig. 3A and 3E). More important, the toxic effect of CP on 4T1 cells only presented in hypoxia ( Fig. 3B and 3C). It meant that polyacetylenes with anti-angiogenic effects, especially CP, showed safe characteristic and owned potential in combinational therapy with other cancer treatment. Apart from these in vitro and in vivo data of CP used to evaluate anti-angiogenic effects, this is the first report to demonstrate that polyacetylene worked on angiogenic factors under hypoxia. First, we checked PKC expression in 4T1 cell under hypoxia. CP can suppress PKC expression in 4T1 cells under hypoxia but not nomoxia (Fig. 3D).
Because elevated PKC expression will increase VEGF levels in tumors under hypoxia and enhance malignancy progression [46][47][48]. Next, we found that higher VEGF transcription in 4T1 cells under hypoxia was decreased by CP co-incubation ( Fig. 3F).
Of note, the anti-angiogenic mode of action of CP needs to be further addressed and is under way. At the outset, the data in this study indicated that CP will be a safe phytocompound to be used in cancer patient and pet animals, especially on early stage.
CP did not enhance angiogenesis and toxic effect under normoxia (Fig. 3B-D and 3F).
Second, the data in previous publications [16,21,63] and this work suggest that CP owned immune modulatory effects, especially on anti-inflammation. Approximately 15% of all cancers are linked to inflammation which contributes to the development of neoplasms. Related intracellular signaling mechanisms involved in NF-κB activation and the induction of iNOS response [64]. Our other investigation found that CP inhibited inflammatory cytokines production by lymphocyte [16,21] and suppressed NF-κB pathway and iNOS activity (data not shown). Usually, anti-inflammatory drugs will be used in cancer patients or pets, but non-steroidal anti-inflammatory drugs (NSAIDS) will raise the risk of cardiovascular incidence and gastrointestinal bleeding, make herbs and spices potentially appealing alternatives [65]. Extract from B. pilosa can improve gastric ulcer [66,67] and cardiovascular function [68]. Taken together with our earlier publications [16,21,63], our data illustrate the applicability of CP for prophylactic use or combination therapy to control abnormal angiogenesis in progressive tumor development.

Conclusions
Finding anti-angiogenic drugs and/or alternative herbal drugs is a key and difficult works to overcome. For these reasons, we tried to evaluate the anti-angiogenic effect and related mechanism of CP purified from B. pilosa. We found this polyacetylenic glucoside from a safe and edible herbal medicine can significantly inhibit angiogenesis in various animal models (sponge/Matrigel angiogenesis assay, CAM and Matrigel plug assay) and in vitro angiogenic systems (tube formation assay and migration assay). The anti-angiogenic effect of CP is especially significantly presented under hypoxia state through inhibiting PKC expression and VEGF expression. We also confirmed that a CP docking gene PDIA4 on a knockout mouse