Background
Microvessels deliver oxygen and nutrients and remove waste from the tumor microenvironment [
1,
2]. Although solid tumors have numerous blood vessels, the function and structure of these vessels are usually abnormal, characterized by high permeability, the absence of pericytes and non-specific extravasation of blood components. Abnormal microvessels can result in a hypoxic and an acidic microenvironment and increase the interstitial fluid pressure (IFP), which can induce epithelial-mesenchymal transition (EMT) and resistance to radiation therapy and chemotherapy, and these changes can even increase metastasis [
3,
4].
Microvascular structural entropy (MSE), a measure of the degree of disorder, is used to assess the structural organization of microvessels. MSE combines measurements of the size and distance between vessels. Furthermore, we used fractal dimension analysis to quantitatively assess vascular complexity and tortuosity [
5‐
7]. Until now, there has been a lack of studies on the role of MSE in tumor pathology. A regular or ordered microvascular distribution improves the tumor microenvironment and decreases malignancy by promoting microvascular normalization [
8‐
11].
Ginger has been used for thousands of years in China as an “interior-warming drug”. It is used in the “Sini Decoction” and “Decoction for Restoring Yang from Collapse” as an adjunctive drug. Ginger can alleviate gastrointestinal paralysis and spasms and improve dysmenorrhea and cold stress [
12‐
14]. Despite being used for more than a millennium, the mechanism of action of ginger is still unclear. Being the major bioactive component in ginger, 6G has similar effects to those of ginger [
15‐
17].
Anti-angiogenic therapies attempt to “block” vessels to starve tumor cells (TCs) by inhibiting angiogenesis and/or destroying existing vessels [
18‐
21]. However, anti-angiogenic therapies that inhibit angiogenesis often have side effects, such as the deterioration of the tumor microenvironment, leading to tumor progression, increased EMT and drug resistance [
22]. In this study, we report that 6G can regulate the normalization and order of tumor microvessels and further suppress tumor progression. In addition, 6G can also increase the integrity of the endothelial cell (EC) barrier and tight junctions between and around the periphery of ECs. 6G alters the vascular morphology from multiple branches with short intervals to minor branches with long intervals, resulting in the reconstruction of the tumor microvasculature network and reduction of MSE and fractal dimension. The results of proteomics analysis indicated that 6G directly targets and activates VEGFR2. 6G enhances the binding of VEGFa to VEGFR2 to promote VEGFR2 phosphorylation. Phosphorylated VEGFR2 (p-VEGFR2) recruits VE-cadherin (VE-cad)/β-catenin/actin to regulate F-actin cytoskeletal remodeling in ECs, altering vascular remodeling, thereby promoting the normalization of tumor microvessel networks and improving the tumor microenvironment. These changes can increase the cytotoxic effects of chemotherapy drugs and prevent cancer progression.
Materials and methods
Cell culture
Alexander hepatoma cell line PLC/PRF/5 and breast cancer cell line MCF-7 were maintained in RPMI 1640 supplemented with 10% FBS and 1% PS (penicillin and streptomycin) and under 37 °C and 5% CO2. The cells were purchased from KeyGen Biotech (Nanjing, China) and the company provides complete cell identification. The cells periodically authenticated by biomarkers detection of hepatocellular carcinoma, morphologic inspection, growth curve analysis, and mycoplasma testing, and not used beyond 10 passages.
Microvascular structural entropy (MSE)
HE-stained sections and CD34-stained tissues were quantitatively assessed using a Nikon NIS-Elements analysis system (Nikon, Japan). A 40× apochromatic objective was used to capture the images. To enhance the contrast between microvessels and their lumens, each image was digitally processed. In each immunostained preparation, 5 random fields were selected, and MSE was calculated. The following structural entropy equation was used to calculate MSE:
Here, MSE represents microvascular structural entropy; the microvessel area is represented by a, and d is the distance between neighboring microvessels. In the calculation of MSE using the minimum spanning tree method as the basic procedure, each microvessel should be related to only one neighboring microvessel [
23]. Δa represents the mean area of all microvessels in one calculation field; Δd represents the mean distance between all neighboring microvessels. The detailed parameters are explained in the study by Kovacs et al. [
24] When both the area and distance between the nearest-neighboring microvessels were constant for all microvessels, the value of MSE was zero. A greater difference in the areas and distances of the microvessels results in a larger MSE.
Immunofluorescence staining
Adherent cells at 80% confluence were treated with 0, 15 and 30 μM 6G or DMSO for 12 h. Then, the medium was removed, and the cells were washed three times with PBS. The cells were fixed with 3.7% paraformaldehyde for 20 min, washed three times with PBS, treated with 0.1% Triton X-100 and 5% BSA for 30 min and washed again with PBS. Next, phalloidin-tetramethylrhodamine (TRITC) (1:50) was added, and the cells were incubated for 30 min. Finally, the cells were stained with 4′,6-diamidino-2-phenylindole (DAPI) (Sigma, USA), mounted and viewed using a laser-scanning confocal A1 microscope (Nikon, Japan).
Western blotting
The cells were lysed in RIPA buffer with protease inhibitor cocktail (Sigma). Proteins were quantified and separated on a 10% SDS-PAGE gel, transferred to PVDF membranes, and incubated with primary antibodies. Pierce detection reagents were used to visualize the horseradish peroxidase-labeled secondary antibodies. Western blot results were collected by automatic analysis system of electrophoresis gel imaging (GenoSens, Shanghai). The relative band intensity was quantified using ImageJ v1.37 software (U.S. National Institutes of Health, Bethesda, MD, USA).
Co-immunoprecipitation (IP)
Lysates were prepared by incubating the cells on ice with cold TBS-Nonidet P-40 lysis buffer (20 mM Tris, pH 7.4, 150 mM NaCl, and 1% Nonidet P-40) in the presence of protease and phosphatase inhibitor cocktails (Roche) for 30 min at 4 °C. Then, the cell lysate was centrifuged at 12,000×g for 10 min at 4 °C. For immunoprecipitation, proteins (approximately 0.5 mg) were incubated with control or specific antibodies (3–5 μg) for 12 h at 4 °C under constant rotation; 30 μl of protein G magnetic beads (Invitrogen) was then added, and the mixture was further incubated for 2 h. The beads were washed with lysis buffer and then collected using magnetic stands at 4 °C. The proteins were eluted by boiling for 10 min with 2× SDS-PAGE loading buffer. Immune complexes were analyzed by SDS-PAGE using appropriate antibodies.
Murine xenograft model
All mice were purchased from Charles River Laboratories (Beijing, China). Four-week-old female nude mice were maintained at a specific pathogen-free animal facility at Tianjin International Joint Academy of Biomedicine. The mice were randomly divided into 3 groups (n = 10 mice/group), and 1 × 107 MCF-7 cells were subcutaneously injected into the axilla. When the volume of the tumor reached approximately 10 mm3, the mice were treated with 100 or 200 mg/kg of 6G or saline (as a control) for four weeks. All mice were sacrificed through carbon dioxide inhalation. Subsequently, the tumor, lung and liver tissues were fixed for analysis. All animal experiments were performed under approved protocols of the institutional animal use and care committee of Tianjin International Joint Academy of Biomedicine.
For the CDX model with PLC/PRF/5 cells, mice were randomly divided into six groups (eight mice each) and treated with saline, cisplatin, sorafenib, sorafenib/cisplatin, 6G or 6G/cisplatin for four weeks.
Immunohistochemical (IHC) analysis
Tumor tissues were embedded in paraffin and cut into 4-μm thick sections. Following uplink dehydration, microwave antigen retrieval and blocking, the samples were incubated with the following primary antibodies overnight at 4 °C: rabbit polyclonal anti-HIFα (Affinity, 1:100 dilution), rabbit polyclonal anti-MMP2 (Zhongshan, 1:100 dilution) and rabbit polyclonal anti-MMP9 (Abcam, 1:100 dilution). Diaminobenzidine was applied for color development, and hematoxylin was used for counterstaining.
Hypoxia-induced zebrafish retinal angiogenesis
We selected 12- to 24-month-old zebrafish with EGFP-positive vessels in the retina for the experiments. Zebrafish in the control group were bred and reared normally, and zebrafish in the hypoxia-induced group were exposed to hypoxic water (10% air saturation) for 12 d. The 6G treatment groups were exposed to hypoxic water (10% air saturation for 12 d) with 15 or 30 μM of 6G. Many zebrafish were kept in the hypoxic water. After 12 d, the zebrafish were fixed at 4 °Covernight in 3.7% paraformaldehyde (solution volume/sample volume = 4:1). The zebrafish retina (attached to the lens) were removed with precision ophthalmic devices and then fixed on a glass slide. The detailed parameters and the dissecting and flatting methods applied to the adult zebrafish retinas have been previously described [
25]. Finally, the samples were illuminated under low magnification at 488 nm using a Nikon confocal microscope (Nikon, Japan).
Calculation of the fractal dimension
The fractal dimension of the retinal vasculature of adult zebrafish was calculated in MATLAB (MATLAB R2014b). Briefly, the image of retinal vasculature was loaded in MATLAB, and the green color was extracted. Then, the extracted binary data was further calculated using Box dimension in the Fractal module of MATLAB.
Calculation of the coefficient of variation
The coefficient of variation is calculated by dividing the standard deviation by the mean. Use formula is expressed as: C.V. = σ/|μ|, σ = √∑(xi-u)^2/(n-1), u = (∑xi)/n.
Microscale thermophoresis (MST)
The binding between 6G and VEGFR2 was analyzed using the MicroScale Thermophoresis Instrument Monolith NT.115 (NanoTemper Technologies, Munich, Germany). Briefly, the VEGFR2 protein was labeled with MST dye, and 6G was prepared using 16 geometric proportion serial dilutions (0.0046–150 μM). The labeled VEGFR2 and 6G working solutions were mixed and measured using an MST device. In the present study, the LED channel was set as “red” for the Cy5 dye, the temperature was set at 25 °C, and the MST was set at 40% power.
Biacore assay
Surface plasmon resonance (SPR) experiments were performed using a Biacore 3000 instrument (GE Healthcare, Piscataway, NJ, USA). VEGFR2 was immobilized on CM5 sensor chips using the Biacore Amini Coupling Kit according to the manufacturer’s instructions. Different amounts of 6G were diluted and then injected into VEGFR2-immobilized CM5 sensor chips at concentrations of 6 μM, 12.5 μM, 25 μM and 50 μM. The corresponding concentration of 6G was set as control, and the background was subtracted. The surface of the control chip was prepared in the same manner and used for data correction. Data analysis was performed using BIA evaluation software.
Scanning electron microscopy (SEM)
HUVECs grown on coverslips (KeyGen Biotech, Nanjing, China) were treated with solvent control, VEGFa or VEGFa/6G. After 12 h of incubation, the cells were fixed using ethyl alcohol, dried with a gradient concentration of tertiary butyl alcohol and finally coated with gold. Images of the cells were obtained using a scanning electron microscope (SEM, LEO1530VP, Germany).
N-STORM microscopy of VEGFR2, VE-cad and β-catenin
Samples were prepared for single-molecule imaging to determine co-localization using super-resolution microscopy. HUVECs were seeded in 35-mm N-STORM super-resolution microscope dishes and grown to 60% confluence. The cells were cultured in RPMI-1640 medium (HyClone, USA) supplemented with 10% fetal bovine serum (HyClone) and antibiotics (50 units/mL penicillin and 50 μg/mL streptomycin). Next, the cells were cultured in the medium described above, medium containing 10 ng/mL VEGFa, or medium containing 10 ng/mL VEGFa and 15 μM 6G. The cells were incubated at 37 °C for 12 h in the presence of 5% CO2. The cells were fixed with 4% formalin for 30 min. Non-specific binding was blocked by incubation with 3% BSA for 30 min, and then the cells were incubated with rabbit polyclonal anti-VE-cad (Cell Signaling Technology (CST), USA), mouse polyclonal anti-β-catenin (CST) and goat polyclonal anti-VEGFR2 (Abcam, USA) antibodies in a humidified chamber at 4 °C overnight. Then, the cells were incubated with Alexa Fluor647-conjugated donkey anti-goat secondary antibody (1:200; Millipore) for 1 h at room temperature. Next, non-specific binding was blocked again by incubation with goat serum for 30 min. Then, the cells were incubated with Alexa 488-conjugated goat anti-rabbit secondary antibody and Alexa Fluor 568-conjugated goat anti-mouse secondary antibody for 1 h at room temperature. The N-STORM super-resolution microscope (Nikon, Japan) was used to assess the co-localization of VEGFR2, VE-cad and β-catenin at the nanoscale level.
Proteomics analysis
HUVECs were seeded into a 60-mm dish. After reaching 70–80% confluence, the cells were treated with 6G (15 μM) for 12 h. Mass spectrometry was performed using a Q-Exactive Mass Spectrometer (Thermo Fisher Scientific). GO analyses for molecular functions, biological processes and cellular components were performed using ClueGo plug-in in CytoScape. Pathway analysis was performed using KEGG (Kyoto Encyclopedia of Genes and Genomes) databases.
Molecular docking
An X-ray structure of the VEGF/VEGFR2 complex is available in the PDB database (PDBID: 3V2A). In the preparation steps, water molecules were removed, and the VEGFA/VEGFR2 crystal structure was treated as follows: hydrogen molecules were added, partial changes were assigned to relax amino residue side chains, and energy minimization was performed to relax the entire structure. The pocket of the VEGFA/VEGFR2 structure was selected to determine the active site of the docking grid box, and Glide XP (extra precision) was used for the 6G docking calculations.
Patient-derived tumor xenograft (PDTX) models
Fresh tumor tissues were collected immediately after surgery from S. G. Hospital (Shandong, China), Tianjin Medical University General Hospital, and the Hospital of Shunyi District, Beijing. Written informed consent was obtained from each patient. The tumors were cut into 1–2 mm
3 pieces in antibiotics-containing RPMI medium. Tumor fragments were implanted in subcutaneous pockets, which were made on each side of the lower back of BALB/c nude mice (4–6 weeks old). Samples with tumor sizes of 100–200 mm
3 were called F0. Subsequently, the samples were divided for passaging in vivo to obtain F1 and then F2 xenograft tumors as described above [
26,
27]. When the F2 tumor size reached 100–200 mm
3, the mice were randomly divided into six groups (eight mice each) and treated with saline, cisplatin, sorafenib, sorafenib/cisplatin, 6G or 6G/cisplatin to serve as the xenograft tumor model.
Statistical analysis
All data are expressed as the means ± standard deviations (SD). After testing for normality and equal variance across the groups, intergroup differences were assessed using Student’s t-tests, ANOVA, kaplan-meier survival analysis, GO and KEGG enrichment analysis. All experiments were repeated three times. *p < 0.05 or **p < 0.01 was considered statistically significant.
Discussion
The growth and metastasis of malignant tumors depends on an adequate blood supply, which can be achieved by the recruitment of vessels to establish microcirculation in tumors. Some factors, such as hypoxia, IFP and pH, can affect tumor microcirculation [
32]. A hypoxic microenvironment and excess IFP within the tumor can affect microcirculation and promote an aggressive behavior along with the acquisition of TC stemness. Several critical signaling pathways have been identified to have important roles in tumor angiogenesis, including the PI3k/Akt/mTOR, NF-휅B, Notch, Wnt/훽-catenin and Hedgehog signaling pathways.
Previous data provide evidence that MVD does not fully mirror the interaction between blood supply and TC proliferation [
24]. To evaluate tumor vasculature normalization, an independent parameter, MSE, was calculated using histological features to measure microvascular heterogeneity. Histopathological evaluation of liver cancer demonstrated that MSE was associated with clinical stage and clinical prognosis. Previous studies have focused more on the number and area of vessels, but not the MSE. Tumor metastasis is related to MSE but not to MVD. The MSE may be an important parameter of vascular normalization.
6G is a polyphenolic compound derived from ginger that demonstrates anti-tumor effects in multiple tumors [
33]. In the present study, 6G exerted anti-tumor effects and decreased MSE by promoting vascular normalization. VEGFR1 regulates the growth of ECs and the extension and repair of vessels, whereas VEGFR2 regulates the differentiation of ECs and the remodeling of vessels [
34,
35]. 6G increased the binding of VEGFa with VEGFR2 and maintained VEGFR2 phosphorylation. However, it did not affect the binding of VEGFa to VEGFR1. While, the effect of 6G on the remodeling of ECs was not dependent on the PI3K/Akt pathway, but depend on p-VEGFR2/VE-cad/β-catenin/actin protein complex. 6G binds to the VEGFa/VEGFR2 complex and induces conformational changes in VEGFR2, resulting in its dimerization and autophosphorylation to p-VEGFR2. Thus, the interaction among the members of the p-VEGFR2/VE-cad/β-catenin/actin protein complex was modulated by tyrosine phosphorylation and the subsequent recruitment of β-catenin by p-VEGFR2, leading to actin polymerization and stress fiber formation, which eventually regulated the migration of ECs. Indeed, 6G induced the p-VEGFR2/VE-cad/β-catenin/actin complex formation and the actin cytoskeletal remodeling to increase the supply of oxygen and nutrients to TCs by normalizing the tumor vasculature, resulting in alleviated microenvironmental stress and prevented tumor progression. 6G also shown to enhance the delivery of cisplatin into the tumor core via increased permeability and normalized tumor vasculature [
36]. Both CDX and PDX models showed that compared with the untreated group, the 6G and cisplatin co-treatment group displayed inhibited tumor growth. 6G increased the distribution of oxygen and chemotherapeutic drugs in tumors, leading to profound anti-tumor effects by increasing ROS to directly or indirectly kill the TCs. In addition, published data have demonstrated the benefit of ginger over placebo in Functional Living Index Emesis nausea score (day 6-day 1) for female patients (
P = 0.048) and head and neck cancer (HNC) patients (
P = 0.038). The results showed that the daily use of ginger is safe in patients with high dose cisplatin therapy. Similarly, ginger shows good results in specific nausea risk subgroups, such as in female patients and HNC patients [
37].
The “feeding” and “starvation” of TCs are the two sides of a coin that can influence the choice of tumor therapy. Because of disrupted structure and function of vessels and high IFP, tumors are poorly perfused with blood, leading to extreme hypoxia and drug resistance and increased microenvironmental stress. These alterations in turn induce the progression, invasion and metastasis of TCs. Anti-angiogenesis drugs are designed to inhibit blood supply to tumors. However, many of these agents often promote tumor progression despite decreasing tumor volume, thus leading to increased metastasis [
19,
38]. Many investigators have shown that the anti-angiogenesis drugs, such as sorafenib can cause tumor recurrence and metastasis. This phenomenon has become an important bottleneck in anti-vascular treatment. Interestingly, we also reported for the first time that anti-angiogenesis drugs decrease MVD but increase MSE and thus may be associated with tumor metastasis. Therefore, MSE might have potential as a biomarker to assess the proliferation and metastasis of TCs. Appropriate “feeding” of TC scan reduce stress on the tumor microenvironment. Although this view contradicts the idea of anti-vascular therapy and has not yet been fully recognized, some studies on the tumor microenvironment, immunity, and metabolism have shown that “feeding” TCs may not be wrong. In this study, 6G increased the remodeling of tumor vasculature and prevented distant metastasis. Thus, the strategy of increasing the remodeling of ECs in tumors to enhance the delivery of chemotherapeutic drugs and oxygen into the tumor core and to generate a favorable tumor microenvironment may be a promising therapeutic approach in the clinic. Normalization of the tumor vasculature can affect HIF1α/VEGFa feedback and reduce TC recruitment or vasculogenic mimicry (VM) formation, leading to better clinical responses in cancer patients [
39,
40].
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.