2020 年 26 巻 6 号 p. 891-896
Angiogenesis is a fundamental step in the transition of tumors from a dormant state to a malignant one. Food factors capable of inhibiting angiogenesis, if found, would be useful tools to stop the progression of small cancers. In this study, we examined the antiangiogenic activities of four flavonols commonly found in foods: galangin, kaempferol, quercetin, and myricetin. The effects of these flavonols were tested in an in vivo model of angiogenesis, the chorioallantoic membrane (CAM) assay. These compounds showed strong antiangiogenic activities. Furthermore, we compared the results of the CAM assay with those of an in vitro model of angiogenesis, the tube formation assay. Flavonols with strong antiangiogenic activities in the CAM assay also showed strong inhibitory activities in the tube formation assay. We discuss the structure-activity relationship for the flavonols with regard to their antiangiogenic activities.
Flavonoids are polyphenolic compounds that are integral components of the human diet, and are a family of compounds with a C6-C3-C6 skeleton structure. They are ubiquitously present as constituents of flowering plants, particularly of food plants. Epidemiologic studies have recently revealed an association between higher flavonoid intake from fruits and vegetables and a decreased risk of cardiovascular diseases. The antioxidant activity of flavonoids could be considered to be, at least partly, responsible for such prevention. Furthermore, regular consumption of certain foods and beverages, such as apples, berries, wine, coffee, and tea, may significantly influence the quantity of antioxidants in a diet (Fernandes et al., 2017; González-Paramás et al., 2019; Maleki et al., 2019). Flavonoids have been reported as natural chemopreventive and anticancer compounds. In particular, many in vitro studies have been conducted on the direct and indirect actions of flavonoids on tumor cells, and a variety of anticancer effects, such as cell growth and kinase activity inhibition, have been reported (Chen et al., 2013; Lall et al., 2016; Li et al., 2018). In addition, studies on angiogenesis inhibition by flavonoids have been reported (Mirossay et al., 2018; George et al., 2018). Angiogenesis, or new blood vessel growth, is defined as a process in which a network of new blood vessels emerges from preexisting vessels (Folkman, 1971). In the 1970s, it was shown that exponential tumor growth over a few cubic millimeters in size was dependent on the induction and recruitment of the tumor's own new blood vessels, i.e., angiogenesis, for nutrition and oxygen supply (Naumov et al., 2006). Since then, many investigators have conducted studies to prevent or delay cancer growth, or even to completely eliminate cancer from a patient's body, by suppressing such neovascularization (Folkman, 1995). Food factors capable of inhibiting angiogenesis, if found, would be useful tools to stop the progression of small cancers (Tosetti et al., 2002).
In our previous study, we reported the antiangiogenic effects of one of the honeybee products, propolis, both in vitro and in vivo (Ahn et al., 2007). We have also evaluated the in vitro antiangiogenic effects of various compounds, including flavonoids from propolis (Ahn et al., 2009; Tsuchiya et al., 2013; Ishizu et al., 2019). Although we and others have reported on in vitro antiangiogenic studies of flavonoids, there have not been many reports on in vivo studies thus far. In the present study, we investigated the in vivo antiangiogenic activities of four flavonols, galangin, kaempferol, quercetin, and myricetin, which are commonly found in foods (Fig. 1). The effects of these flavonols were tested in an in vivo model of angiogenesis, the chorioallantoic membrane (CAM) assay. The CAM assay is an in vivo test that uses the CAM of fertilized hen eggs to determine the antiangiogenic effects of various compounds and extracts. It is a simpler method than other in vivo tests that use mice and rats (Demirci et al., 2004; Weng-Yew et al., 2009). We also compared the results of the CAM assay with those of an in vitro model of angiogenesis, the tube formation assay, which uses human umbilical vein endothelial cells (HUVECs) cultured in collagen gel and stimulated with angiogenic factors. The CAM assay shows whether flavonols are capable of inhibiting angiogenesis in the actual in vivo environment with not only endothelial cells, but also all the other cells involved in angiogenesis in place. Thus, the aim of this study is to demonstrate and compare the antiangiogenic effects of four flavonols in vivo, and also to compare them with the antiangiogenic effects in vitro.
Chemical structures of galangin (a), kaempferol (b), quercetin (c), and myricetin (d).
Materials Galangin, kaempferol, quercetin, and myricetin were purchased from Funakoshi (Tokyo, Japan). Medium 199 and all other chemicals were purchased from Sigma (St. Louis, MO, USA) unless noted otherwise. Medium MCDB-104 was a product of Nihon Pharmaceutical (Tokyo, Japan). Fetal bovine serum (FBS) was purchased from Moregate (Brisbane, Australia). Atelocollagen was obtained from Koken (Tokyo, Japan). Epidermal growth factor (EGF) was purchased from BD Biosciences (Bedford, MA). Endothelial cell growth factor (ECGF) was purified according to our previous report (Kondo et al., 2002). Human basic fibroblast growth factor (bFGF; recombinant) was purchased from Austral Biologicals (San Ramon, CA, USA). Fertilized chick eggs were purchased from Goto Farm (Gifu, Japan). Dimethyl sulfoxide (DMSO) and methylcellulose was purchased from Wako Pure Chemical Industries (Osaka, Japan). Intralipos® (Injection 20%) was purchased from Otsuka Pharmaceutical (Tokyo, Japan).
Chick embryo CAM assay Antiangiogenic activity was determined using the CAM assay as previously described (Nakayama et al., 2008) with slight modifications. In brief, the test compounds (galangin, kaempferol, quercetin, and myricetin) were first dissolved in vehicle (DMSO), then prepared to the final concentrations of 1 to 10 nmol/egg for the assay (sample (test compound) : 0.9% saline : 2% methylcellulose/saline = 1 : 9 : 10). The fertilized chicken eggs were kept in a humidified egg incubator at 37 °C. After 4 days of incubation, approximately 4 mL of albumen was aspirated from the eggs with an 18-gauge hypodermic needle through a small hole drilled in the narrow end of the eggs. The shell covering the air sac was punched out and removed with forceps, and the shell membrane was peeled away. Embryos with chorioallantois, 3 to 5 mm in diameter, were used for the assay for antiangiogenic activity. On day 6, each sample (10 µL) was applied to a silicone ring placed on the CAM surface. As a blank negative control, CAMs treated only with DMSO (DMSO : 0.9% saline : 2% methylcellulose/saline = 1 : 9 : 10) were also included. After covering the open ends of the embryo shells with stainless steel caps, the eggs were incubated for 2 more days. Intralipos® was injected into the 8-day-old embryo chorioallantois using a 27-gauge needle so that the vascular network of the CAM would stand out against the white background of the lipid. The antiangiogenic response was assessed by measuring the avascular zone of the CAM beneath and around the ring.
Antiangiogenic activities were estimated by the avascular zone to obtain angiogenic points: no inhibition (–) = 0 points; barely detectable inhibition of capillary vessels (±) = 1 point; apparent partial inhibition of capillary vessels (+) = 2 points; complete inhibition of capillary vessels (++) = 3 points; and inhibition of large vessels (+++) = 4 points. Then, the antiangiogenic rates (%) were calculated by the following formula:
 |
For every test compound, 3 to 10 eggs were used, and each experiment was performed at least three times.
Cell culture HUVECs were isolated from human umbilical cord, and were grown in HUVEC growth medium (MCDB-104 medium supplemented with 10 ng/mL EGF, 100 µg/mL heparin, 100 ng/mL ECGF, and 10% FBS) as previously reported (Kondo et al., 2002). Incubation was carried out at 37 °C under a humidified 95%–5% (v/v) mixture of air and CO2. The cells were seeded on plates coated with 0.1% gelatin, and allowed to grow to sub-confluence before experimental treatments. Cells of passages 5 to 8, equivalent to population doubling levels of 20.8 to 30.8, in the actively growing condition were used for the experiments.
Tube formation assay HUVECs were induced to form capillary tube-like structures in type I collagen gel (Atelocollagen) as previously described with slight modifications (Kondo et al., 2002; Ahn et al., 2007). Aliquots (200 µL) of collagen solution (0.21% in Medium 199) were poured into the wells of 24-multiwell culture plates, and the plates were incubated at 37 °C for 30 min to solidify the gels. HUVECs (6.0 × 104 cells/cm2) in MCDB-104 with 0.5% FBS were seeded onto the collagen gels and left at 37 °C for 1 h in a 5% CO2 incubator for attachment. After removing the medium, a 150-µL aliquot of the collagen solution was overlaid and subjected to gelation as described above. Subsequently, 650-µL aliquots of MCDB-104 with 0.5% FBS supplemented with 10 ng/mL bFGF, 8 nM/mL PMA, and 25 µg/mL ascorbic acid with various flavonols (10 and 50 µM) were added to the wells and incubated for 36 h. The final concentration of DMSO was 0.4% for all test compounds and the control. The resulting web-like capillary structure was viewed with a microscope under 100× magnification, and captured with an Olympus C 4040-ZOOM digital camera. The experimental procedures used in the present study were carried out in accordance with the guidelines of the Animal Usage Committee of the University of Shizuoka.
Four flavonols inhibited angiogenesis in the CAM assay, an in vivo model of angiogenesis Figure 1 shows the chemical structures of the four flavonols, galangin, kaempferol, quercetin, and myricetin. To examine the antiangiogenic activities of these four flavonols, an in vivo CAM assay was performed. As shown in Fig. 2, all four flavonols had strong antiangiogenic effects in this in vivo assay. During normal development in the control, three types of vessels were observed: thick large vessels, thin small vessels, and capillaries that were visible only as red areas between the large and small vessels. The four flavonols inhibited the development of capillaries, and changed the color of the areas between the large and small vessels from red to white. When the antiangiogenic activities of the four flavonols were quantified, galangin, kaempferol, quercetin, and myricetin all showed stronger inhibitory effects at the higher concentration of 10 nmol/egg than at the lower concentration of 1 nmol/egg (Fig. 3). At 10 nmol/egg, all four flavonols seemed to show almost the same degree of antiangiogenic effects. At 1 nmol/egg, the degree of angiogenesis inhibition was in the following order: quercetin = kaempferol > galangin = myricetin.
Photos of the inhibitory effects of four flavonols on chick embryo chorioallantoic membrane: galangin (a), kaempferol (b), quercetin (c), myricetin (d). Large vessels are indicated by arrowheads and some representative capillary vessels are indicated by circles.
Quantification of inhibitory effects of four flavonols on chick embryo chorioallantoic membrane.
Four flavonols inhibited HUVEC tube formation, an in vitro model of angiogenesis During normal tube formation, the endothelial cells migrated and gathered together, and then became elongated and adhered to each other to form a network of capillary-like tubes. We examined the inhibitory effect of four flavonols in this in vitro model of angiogenesis (Fig. 4). Treatment with galangin had only a small inhibitory effect on tube formation, and the endothelial cells were able to form capillary-like tubes almost normally. Kaempferol had a strong inhibitory effect on tube formation: it disturbed the tube morphology and caused fragmentation of the network at 10 µM, and it also strongly inhibited the elongation of the cells at 50 µM. Quercetin had an even stronger inhibitory effect on tube formation: it disturbed the tube morphology, caused fragmentation of the network, and inhibited the elongation of the cells at 10 µM; it even seemed to induce cell death in some cells at 50 µM Myricetin had a very small inhibitory effect, and the endothelial cells were able to form capillary-like tubes normally at 10 and 50 µM. The degree of angiogenesis inhibition was in the following order: quercetin ≥ kaempferol > galangin ≥ myricetin.
Inhibition of tube formation of HUVECs by four flavonols: galangin (a), kaempferol (b), quercetin (c), myricetin (d).
In this study, we evaluated the antiangiogenic activities of four flavonols, galangin, kaempferol, quercetin, and myricetin, both in vivo and in vitro. The comparison of these flavonols can be a good model case to investigate the structure-activity relationships for flavonoids and other polyphenolic compounds, since these flavonols differ only in the number of OH groups that are attached to the B-ring of the common structure. Our results revealed that quercetin with two OH groups, and not myricetin with three OH groups, had the strongest antiangiogenic activity among the four flavonols tested. Our study is also the first report to compare the antiangiogenic activities of these four flavonols in vivo with a CAM assay.
Although some investigators have reported that flavonoids, which include flavonols, can suppress tumor growth both in vitro and in vivo (Chen et al., 2013; Lall et al., 2016; Li et al., 2018), the actual mechanisms of these effects are not yet fully understood. As previously reported, one possible explanation for such tumor-suppressing effects of flavonoids is their antiangiogenic activities (Ahn et al., 2009; Mirossay et al., 2018). In order to elucidate the antiangiogenic mechanisms of flavonols in detail, we investigated their antiangiogenic activities both in vitro and in vivo, and looked into the structure-activity relationships among these four flavonols.
When the avascular areas in the CAM were observed, these four flavonols showed strong inhibitory effects on angiogenesis in vivo in the following order: quercetin = kaempferol > galangin = myricetin. It was shown that as the number of OH groups increased, the degree of antiangiogenic activity increased in the case of galangin (zero OH groups), kaempferol (one OH group) and quercetin (two OH groups). The only clear exception was myricetin (three OH groups). One other notable point was that the difference between quercetin and kaempferol was very small. When the areas of all tubes constructed in the in vitro tube formation model were observed, these four flavonols showed strong inhibitory effects on tube formation by HUVECs in the following order: quercetin ≥ kaempferol > galangin ≥ myricetin. Once again, the more OH groups there were in the flavonol, the stronger the antiangiogenic activity, with the major exception of myricetin and the minor exception of the ambiguous difference between quercetin and kaempferol. Thus, the results of the in vivo experiment and the in vitro experiment were mostly in agreement with only a minor difference.
We have previously shown that there was a good correlation between antiangiogenic and antioxidative activities (Ahn et al., 2009). The correlation between these two activities can be explained by the role of superoxide and hydrogen peroxide in angiogenic signal transduction. In response to angiogenic factors such as VEGF and PMA, NADPH oxidase has been reported to produce superoxide and, subsequently, its metabolite, hydrogen peroxide, which then activates multiple intracellular signaling pathways that lead to the proliferation, migration, and tube formation by endothelial cells (Griendling et al., 2000; Ushio-Fukai, 2006). Kim et al. has reported that the number of OH groups in the four flavonols (myricetin, quercetin, kaempferol, and galangin) is important for their antioxidant activity and, to some extent, for their effects on modulating endothelial cell angiogenesis (Kim et al., 2006). However, their results and our data in this study both suggest that antioxidant activity alone cannot fully explain the difference in antiangiogenic activity among the four flavonols.
Flavonoids have been suggested to have several potential health benefits due to their antioxidant activities, which are attributed to the presence and the number of phenolic hydroxyl (−OH) groups on the structure (Pietta, 2000; Amic et al., 2003). However, since myricetin was a clear exception to this rule, we had to devise a new principle to fully explain the structure-activity relationships of the flavonols with regard to their antiangiogenic activities. Kajiya et al. previously reported that lipophilicity also has an important impact on the biological activities of these flavonols (Kajiya et al., 2001). They showed that myricetin and quercetin have about 40% and 80% lipophilicity, respectively, compared to about 100% lipophilicity for kaempferol and galangin, and that their lipophilicity correlated well with their cytotoxicity. Our results could be explained well when we took the lipophilicities of flavonols into consideration. Although myricetin should have possessed the strongest antiangiogenic activity based on the number of OH groups, which correlates with antioxidant activity, it failed to exhibit the strongest antiangiogenic activity due to the low lipophilicity and, presumably, the subsequent low binding to and low incorporation into the cells. Thus, the combination of antioxidant activity and lipophilicity could be an effective way to predict the antiangiogenic activities of a group of structurally related compounds with minor variations, such as the number of OH groups.
In the present study, we found a very good correlation between the antiangiogenic activities of flavonols in vivo and in vitro. We also suggest that, in addition to the numbers of OH groups, taking lipophilicity into consideration for structure-activity relationships can explain well why these flavonols exhibit various degrees of antiangiogenic activity. We would like to further investigate the validity of this hypothesis by investigating the structure-activity relationships of other flavonoids and phytochemicals as well.
