2023 年 46 巻 4 号 p. 608-613
Angiogenesis is involved in the malignant transformation of cancers. Vascular endothelial growth factor (VEGF) is important in inducing angiogenesis. Cultured cells play an important role in analyzing the regulation of VEGF expression, and it is revealed that VEGF expression is induced under hypoxia. However, it has been shown that there are differences in the pathway for gene expression between two-dimensional (2D) cells and in vivo cells. Three-dimensional (3D) spheroids constructed in 3D culture with a gene expression pattern more similar to that of in vivo cells than 2D cells have been used to solve this problem. This study analyzed the VEGF gene expression pathway in 3D spheroids of human lung cancer cells, A549 and H1703. Hypoxia-inducible factor-1α (HIF-1α) and aryl hydrocarbon receptor nuclear translocator (ARNT) regulated VEGF gene expression in 3D spheroids. However, VEGF gene expression was not regulated by HIF-1α in 2D cells. To conclude, we found that the regulatory pathway of VEGF gene expression is different between 2D cells and 3D spheroids in human lung cancer cells. These results suggest the possibility of a new VEGF gene expression regulation pathway in vivo. In addition, they show useful knowledge for the analysis of angiogenesis induction mechanisms and also demonstrate the usefulness of 3D spheroids.
Cancer cells form spheroids with a three-dimensional (3D) structure through carcinogenesis and proliferation in vivo. Cancer cells in the spheroids are hypoxic and hypoglycemic stress during the early stages because of the lack of blood vessels supplying glucose and oxygen. Hypoxia—one of the stress conditions induces vascular endothelial growth factor (VEGF) expression.1) Hypoxia causes accumulation and activation of hypoxia-inducible factor-1α (HIF-1α) in the cells. Two-dimensional (2D) cells analysis has revealed that HIF-1α is involved in VEGF expression.2–4) HIF-1α is degraded, usually within 5 min,5) in a normoxic environment by ubiquitination and proteasomes via a pathway involving the von Hippel-Lindau protein (pVHL), a tumor suppressor protein, and one of the recognition components of E3 ubiquitin-protein ligase.6) In vitro analysis of cancer cells has been performed using 2D monolayer cells. However, the results in 2D cells are different from those in vivo models.7) It has been reported that the drug-metabolizing enzyme expression is lower in cultured cells than in tissues.8) In recent years, 3D cultures have been used to resolve the differences between 2D cells and in vivo models. The construction of 3D spheroids resembles human cancer tissue due to increased expression of drug-metabolizing enzymes9) and restoration of cell adhesion.10) 3D spheroids are obtained by 3D culture, they have a 3D structure that mimics the structure of tumors in vivo. In addition, the 3D spheroids gene expression pattern is similar to that of in vivo systems.11) We have previously reported that CYP gene expression pathways are different between 2D cells and 3D spheroids in human lung cancer cells and human liver cancer cells.12,13) CYP expression is induced by stress. Cells in the core of the 3D spheroids are stressed due to a lack of glucose and oxygen.14) However, 3D spheroids and 2D cells under stress have different CYP expression pathways.12,15) The altered stress response of 3D spheroids causes the differences in gene expression between 3D spheroids and 2D cells. VEGF expression is triggered by hypoxia and induced via HIF-1α. The accumulated HIF-1α forms a complex with the aryl hydrocarbon receptor nuclear translocator (ARNT) after translocating into the nucleus and binds to the hypoxia-response element (HRE) to induce VEGF.16) ARNT forms a complex with aryl hydrocarbon receptor (AhR) and is involved in the induction of CYP1A expression. CYP1A expression induced by AhR/ARNT differs between 2D cells and 3D spheroids in lung cancer cells.13) Therefore, we expected that VEGF gene expression induced by HIF-1α/ARNT, and is different between 2D cells and 3D spheroids. There are several subtypes of VEGF, each VEGF binds to a different vascular endothelial growth factor receptor (VEGFR).17) VEGFR-1 and VEGFR-2 are involved in angiogenesis, and VEGFR-2 is reported to have higher tyrosine kinase activity and bind VEGFA.18) We focused on the VEGFA gene in 2D cells and 3D spheroids of human lung cancer cells and analyzed the differences in molecules involved in its expression. As a result, VEGF gene expression is regulated by different molecules between 3D spheroids and 2D cells. In 2D cells, VEGF gene expression is not regulated by HIF-1α/ARNT, but by ARNT and an unknown molecule. These suggest the existence of an unknown VEGF gene expression pathway that is different from previous reports. These results are expected to contribute to our understanding of the mechanism of angiogenesis induction in lung cancer cells in vivo.
A549 cells (a human lung adenocarcinoma cell line, ATCC No. CCL-185) were cultured in Dulbecco’s modified Eagle’s medium (DMEM; high glucose: 4.5 g/L D-glucose; Gibco, Grand Island, NY, U.S.A.). H1703 (a human lung squamous carcinoma cell line, ATCC No. HTB182, Cal-5889) cells were cultured in Rowell Park Memorial Institute medium (RPMI-1640; Gibco) supplemented with 10% fetal bovine serum (FBS; Capricorn, lot#CP20-3423) and 1% antibiotic–antimycotic (Gibco). For 3D culture, 100 µL of the cell suspension (1.0 × 104 cells/mL) was applied to a Prime Surface 96V Plate (Sumitomo Bakelite, Tokyo, Japan). For the 2D culture, 3 mL of the cell suspension (1 × 105 cells/mL) was applied to a 60 mm culture dish and incubated at 37 °C in a CO2 incubator for 48 h as a preculture. After preculture, the cells were incubated under hypoxic conditions at 37 °C with 1% oxygen or normoxic conditions at 37 °C with 20% oxygen.
RNA Interference (RNAi) AssaySmall interfering RNA (siRNA) sequences were constructed to target human HIF-1α and ARNT mRNA (HIF-1α: HSS104775 and ARNT: HSS100700, Invitrogen, Carlsbad, CA, U.S.A.). The negative control siRNAs used were nonspecific. A549 and H1703 cells were individually transfected with siRNAs using Lipofectamine RNAiMAX reagent (Invitrogen) or ScreenFect siRNA (FUJIFILM, Tokyo, Japan) for 48 h according to the manufacturer’s instructions. siRNA transfection was performed by treating the cells with 50 nM HIF-1α siRNA and 20 nM ARNT siRNA. The negative control group was treated with the same concentration of negative control siRNA. All alternative control groups were treated with ScreenFect siRNA or Lipofectamine RNAiMAX reagent without siRNA.
RNA Extraction and Quantitative RT-PCRTotal RNA was extracted from A549 and H1703 cells using the silica base method with EconoSpin IIa (INA·OPTIKA, Osaka, Japan). Total RNA was used for the first-strand cDNA synthesis using the PrimeScript II 1st Strand cDNA Synthesis Kit (TaKaRa, Clontech, Shiga, Japan). We performed real-time PCR with the first-strand cDNA using a 7500 Real-Time PCR system (Applied Biosystems, Tokyo, Japan). The primer sequences are shown in Table 1. Data are represented as the mean ± standard deviation of three independently cultured samples. Statistical analysis was performed using Tukey’s multiple comparison test, and the statistical significance was set at p < 0.05.
| Genes | Primer sequence (5′-3′) |
|---|---|
| β-actin | |
| Forward | TGAAGTGTGACGTGGACATC |
| Reverse | GAGGAGCAATGATCTTGATC |
| VEGF | |
| Forward | AAGGAGGAGGGCAGAATCAT |
| Reverse | ATCTGCATGGTGATGTTGGA |
A549 and H1703 cells in the spheroids or monolayer cultures were lysed using NE-PER Nuclear and Cytoplasmic Extraction Reagents (Thermo Scientific, Rockford, IL, U.S.A.). Protein samples were subjected to 10% sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) (10 mg/well) and transferred to a nitrocellulose membrane. Blots were blocked with Blocking One (Nacalai, Kyoto, Japan) at room temperature for 1 h and rinsed with Tris-buffered saline containing 0.1% Tween-20 (TBS-T). Then, the blots were incubated for 1.5 h in a hybrid bag with primary antibodies (1 : 100, 1 : 250, or 1 : 500 dilutions): HIF-1α (1 : 250, GTX127309, GeneTex), ARNT (1 : 100, sc-17811, Santa Cruz Biotechnology), or β-actin (1 : 500, A1978, Sigma) at 25 °C in Can Get Signal Solution 1 (TOYOBO, Osaka, Japan), followed by washing with TBS-T. The cells were then incubated with the secondary antibody in a hybrid bag overnight with Mouse-immunoglobulin G (IgG) horseradish peroxidase (HRP) (1 : 500, 7076S, Sigma) or rabbit-IgG HRP (1 : 250, sc-2357, Santa Cruz Biotechnology) secondary antibody in Can Get Signal Solution 2 at 25 °C. The bands were visualized using a chemiluminescent reaction using Laminate Classico Western HRP substrate (Millipore, Billerica, MA, U.S.A.).
HIF-1α accumulates intracellularly under hypoxic conditions and is a mediator of the transcriptional response to hypoxia.5) 2D and 3D culture schedules are shown (Fig. 1A). The 3D spheroids were observed after 72 h of 3D culture initiation (Fig. 1B). We established the preculture under normoxia, attached the 2D cells to the bottom of the culture dish, and stabilized the 3D spheroids by constructing spheroids. After 48 h of preculture, both cell lines were cultured under normoxia or hypoxia for 24 h. In 2D cells and 3D spheroids, HIF-1α protein levels increased under hypoxic conditions (Fig. 1C). HIF-1α was detected in almost every cell of the 3D spheroids under hypoxic conditions (Fig. 1D). Previous studies have reported that VEGF gene expression increases in hypoxic 2D cells.2–4) In our results, VEGF gene expression increased under hypoxic conditions in both 2D cells and 3D spheroids (Fig. 2A). These results are consistent with the previous findings that 2D cells accumulate HIF-1α and VEGF expression increases under hypoxic conditions.
(A) Culture scheme with hypoxic stress. (B) A549 and H1703 spheroids under bright field microscopy. (C) Western blot analysis using antibodies against HIF-1α. β-actin was used as a control. 2D and 3D indicate 2D cells and 3D spheroids. (D) Immunofluorescence using an antibody against HIF-1α. Nuclear staining was performed using DAPI. Accumulation of HIF-1α due to hypoxia was seen throughout the spheroid.
(A) Gene expression levels of VEGF were normalized with the control gene, β-actin. Bars indicate standard deviations of three independent experiments; differences between 2D cells and 3D spheroids were analyzed to determine statistical significance (p < 0.05).
To determine whether HIF-1α was involved in hypoxia-induced VEGF gene expression, we knocked down HIF-1α using an RNAi assay according to the culture schedule shown in Fig. 3A. This suppressed the HIF-1α protein synthesis and intracellular accumulation under normoxia and hypoxia (Fig. 3B). VEGF gene expression under normoxic and hypoxic conditions was also reduced in HIF-1α knockdown 3D spheroids (Fig. 3C). However, VEGF gene expression remained unaffected by HIF-1α knockdown in 2D cells, implying different VEGF gene expression pathways between 2D cells and 3D spheroids (Fig. 3C).
(A) Culture scheme for HIF-1α knockdown. (B) The data, derived by Western blotting indicates the HIF-1α protein levels in 2D and 3D cultures after 72 h treatment with siRNA for HIF-1α knockdown. N: none; C: addition of control RNA; and i: addition of siRNA for HIF-1α. (C) Gene expression levels of VEGF under HIF-1α knockdown. Statistical analysis was performed using Tukey’s multiple comparison test and p < 0.05 is considered to indicate statistical significance.
HIF-1α forms a complex with ARNT to induce VEGF expression.2) If ARNT is involved in VEGF gene expression in A549 and H1703 cells, ARNT knockdown should prevent HIF-1α from forming a complex, affecting VEGF gene expression. We knocked down ARNT through RNAi using the same culture schedule as that for HIF-1α knockdown (Figs. 3A, 4A). HIF-1α is involved in VEGF gene expression only in 3D spheroids and not in 2D cells (Fig. 3C). As ARNT regulates VEGF gene expression by complexing with HIF-1α,16) we expected that VEGF gene expression under ARNT knockdown would be reduced in 3D spheroids and remain unchanged in 2D cells. However, VEGF gene expression was reduced in both 3D spheroids and 2D cells following ARNT knockdown (Fig. 4B). These results indicated that HIF-1α and ARNT are involved in the regulation of VEGF gene expression in 3D spheroids. However, the results differed between 2D cells and 3D spheroids, with only ARNT involved in VEGF gene expression in 2D cells.
(A) The data, derived by Western blotting indicates the ARNT protein levels in 2D and 3D cultures after 72 h treatment with siRNA for ARNT knockdown. N: none; C: addition of control RNA; and i: addition of siRNA for ARNT. (B) Gene expression levels of VEGF under ARNT knockdown. Statistical analysis was performed using Tukey’s multiple comparison test and p < 0.05 is considered to indicate statistical significance.
The spread of cancer cells to tissues or organs to secondary sites is called metastasis, and the formation of a new tumor is one of the events that lead to death in cancer patients.19) Tumor growth and metastasis depend on the angiogenesis in the tumors during the growth stage,20) and tumors undergo apoptosis when angiogenesis is suppressed.21,22) VEGF expression is induced by hypoxia from inadequate oxygen supply as a result of the proliferation of cancer cell spheroids.23) In addition, VEGF expression in 2D cells depends on HIF-1α.2–4) HIF-1α is rapidly degraded in a normoxia environment, usually in about 5 min.5) HIF-1α accumulates intracellularly under hypoxic conditions, accumulated HIF-1α translocates to the nucleus and forms a complex with ARNT to induce VEGF.5)
The results of in vitro analysis of cancer cells differ from those observed in vivo.7) The 3D spheroids gene expression showed a similar pattern to that in the in vivo systems.11) Therefore, 3D spheroids analysis is important to understand the mechanism of gene induction in vivo.
We hypothesized that the VEGF gene expression regulatory pathway is different in 3D spheroids, where the cell population structure is different from 2D cells. As a result, the VEGF gene expression pattern differed from that observed in previous studies. VEGF gene expression was analyzed assuming that it is induced by HIF-1α and ARNT. The culture schedule is as shown (Figs. 1A, 3A). The culture schedule is different from previous research because 3D spheroids are pre-cultured in a normoxic condition to construct and stabilize the spheroids. The 2D cells culture schedule used the 3D spheroids culture schedule because of comparison under the same conditions as 3D spheroids. As a result, VEGF gene expression by HIF-1α knockdown changed similarly under normoxia and hypoxia. In a previous study by Hänze et al. using A549, culture cells were incubated from the start of culture with knockdown and under hypoxia for 24 h. The incubation time under hypoxia is the same in this study and Hänze et al. but the total incubation time is different. We focused on the total incubation time from the start of incubation to sample collection rather than the hypoxic incubation time. Although data was not shown, we confirmed that knockdown of HIF-1α suppressed VEGF gene expression in 2D cells using the same culture schedule as Hänze et al. This suggests that prolonged suppression of HIF-1α in 2D cells might be converted to a VEGF gene induced pathway that is not mediated by HIF-1α. 3D spheroids show that HIF-1α regulates VEGF gene expression regardless of suppression time. This revealed that the VEGF gene expression pathway differed between 2D cells and 3D spheroids. VEGF gene expression is regulated by the same molecule in A549 and H1703 but it is not clear whether the same results can be obtained in other cell lines. We will analyze whether these results can be applied to other cell lines.
We reported that CYP gene expression patterns and induction pathways differ between 2D cells and 3D spheroids in human lung cancer cells and human liver cancer cells.12,13) The results of this study suggest that VEGF gene expression is induced by ARNT and an unknown molecule in 2D cells. We suspect that the unknown molecule is one of the molecules that form a complex with ARNT. AhR is a candidate for an unknown molecule and known to form a complex with ARNT. AhR is known to form a complex with ARNT and is involved in the induction of VEGF under glucose deprivation in hepatocellular carcinoma cells.24) AhR possibly regulates the expression of the VEGF gene in 2D lung cancer cells.
The study revealed that the VEGF gene expression pathways differed between the 2D cells and 3D spheroids. The molecules involved in VEGF gene expression also differed between hypoxic 2D cells and normoxic 3D spheroids. Therefore, differences in VEGF gene expression cannot be attributed to hypoxia during 3D spheroids formation. The construction of 3D spheroids potentially alters stress response pathways. To the best of our knowledge, this is the first report on the differences in the VEGF gene induction pathway between 2D cells and 3D spheroids and is expected to provide useful insights into the mechanism of angiogenesis induction in cancer cells.
This research was supported by a Grant from Sumitomo Electric Industries, Ltd.
The authors declare no conflict of interest.
