ADAR1 affects gastric cancer cell metastasis and reverses cisplatin resistance through AZIN1

Abstract

Adenosine deaminases acting on RNA1 (ADAR1) are involved in the occurrence and development of cancers. Although the role of ADAR1 in gastric cancer metastasis has been reported, the role of ADAR1 in the mechanism of cisplatin resistance in gastric cancer is not clear. In this study, human gastric cancer tissue specimens were used to construct cisplatin-resistant gastric cancer cells; the results indicated that the mechanism underlying the inhibition of gastric cancer metastasis and reversal of cisplatin-resistant gastric cancer by ADAR1 inhibits gastric cancer occurs through the antizyme inhibitor 1 (AZIN1) pathway. We assessed ADAR1 and AZIN1 expression in the tissues of patients with low to moderately differentiated gastric cancer. Gastric cancer cells (human gastric adenocarcinoma cell line [AGS] and HGC-27 cells) and gastric cancer cisplatin-resistant cells (AGS^CDDP and HGC-27^CDDP) were selected, and the protein expression of ADAR1 and AZIN1 was detected using immunocytochemistry and immunocytofluorescence. The effects of ADAR1 small interfering RNA (siRNA) on the invasion, migration and proliferation of cisplatin-resistant gastric cancer cells were investigated. Western blot assays were used to assess the protein expression levels of ADAR1, AZIN1 and epithelial–mesenchymal transition (EMT)-related markers. In-vivo experiments, a subcutaneous tumor formation model of nude mice was established, and the effects of ADAR1 on tumor growth and AZIN1 expression level were detected by hematoxylin and eosin, immunohistochemistry and western blot. The expression of ADAR1 and AZIN1 in human gastric cancer tissue was significantly higher than that in paracancerous tissues. The colocalization of ADAR1, AZIN1 and E-cadherin expression in immunofluorescence assays indicated a significant correlation between the three. In in-vitro experiments, ADAR1 knockout not only reduced the invasion and migration ability of AGS and HGC-27 cells but also reduced that of cisplatin-resistant gastric cancer cells. ADAR1 siRNA inhibited the proliferation and decreased the colony number of cisplatin-resistant gastric cancer cells. ADAR1 siRNA decreased the expression of AZIN1 and EMT-related marker proteins (vimentin, N-cadherin, β-catenin, MMP9, MMP2 and TWIST). The effect of ADAR1 siRNA combined with AZIN1 siRNA was more significant. In-vivo, the knockdown of ADAR1 significantly inhibited tumor growth and AZIN1 expression. ADAR1 and AZIN1 are antimetastatic targets of gastric cancer, and AZIN1 is a downstream regulatory target of ADAR1. ADAR1 knockout can inhibit the metastasis of gastric cancer cells and reverse the cisplatin resistance of gastric cancer cells by downregulating the expression of AZIN1, potentially resulting in increased treatment efficacy.

Introduction

Gastric cancer is a gastrointestinal tumor for which there has been a gradual increase in incidence. Radical resection is the preferred treatment approach in clinical practice. When radical resection is combined with a basic chemotherapy drug treatment regimen, the survival rate for gastric cancer patients improves [¹]; however, the high mortality rate of gastric cancer is still a substantial challenge for clinicians. In recent years, the development of targeted drugs has made up for the shortcomings of traditional treatment regimens and has led to continuous improvements in the survival rate for cancer patients [²]. Exploring novel targets and elucidating the role of targets in drug resistance mechanisms have become a focus of cancer treatment research.

Adenosine deaminases acting on RNA1 (ADAR1) is an adenosine dehydrogenase that acts on RNA and has been shown to be involved in the occurrence and progression of various cancers, including breast cancer, colon cancer and gastric cancer [³]. ADAR1 has two subtypes of promoters, that is, the p110 and p150 subtypes, which have different roles in cancer development [⁴]. In melanoma, silencing the ADAR1-p150 subtype abolishes IFN-γ-driven T-cell migration, and silencing and overexpressing the ADAR1-p110 subtype reduces and increases T-cell migration, respectively [⁵]. ADAR1 regulates the peritoneal implantation and metastasis of gastric cancer cells through Wnt/β-catenin signaling in gastric cancer tissue [⁶]. In endometrial cancer, ADAR1 affects endometrial cancer cell metastasis by regulating antizyme inhibitor 1 (AZIN1), indicating that high ADAR1 and AZIN1 expression may be a predictive factor for a poor prognosis for individuals with endometrial cancer [⁷]. ADAR1 and AZIN1 overexpression is an important prognostic biomarker for gastric cancer [⁸].

Understanding drug resistance mechanisms in cancer treatment is key to successfully treating cancer patients. Although ADAR1 and AZIN1 have been shown to have important diagnostic and prognostic value in the diagnosis and treatment of gastric cancer, the role of ADAR1 and AZIN1 in the mechanism of drug resistance in gastric cancer is currently unclear. We used a cisplatin-resistant gastric cancer cell line and knocked out ADAR1 and AZIN1 expression, revealing that ADAR1 plays a role as a target for reversing cisplatin resistance in gastric cancer by regulating AZIN1 expression.

Materials and methods

Specimen of human pathological tissue

The clinical specimens of 75 patients with moderately or poorly differentiated gastric cancer were collected from the Pathology Department of People’s Hospital of Ningxia Hui Autonomous Region. None of the patients had an experience of chemotherapy or radiotherapy. The project was approved by the Ethical Committee of People’s Hospital of Ningxia Hui Autonomous Region (No.2020-NZR-100). The research has been carried out in accordance with the World Medical Association Declaration of Helsinki. All methods were carried out in accordance with relevant guidelines and regulations. All patients provided written informed consent.

Cell culture and chemical

Human gastric cancer cells human gastric adenocarcinoma cell line (AGS) and HGC-27 were purchased from the National Collection of Authenticated Cell Cultures (https://www.cellbank.org.cn/). Cisplatin resistance gastric cancer cell lines (AGS^CDDP and HGC-27^CDDP) were obtained according to a previous study [⁹]. Both AGS, HGC-27, AGS^CDDP and HGC-27^CDDP were maintained in RPMI-1640 medium with 10% fetal bovine serum and 1% penicillin/streptomycin, at an incubator of 37 ℃ and 5% CO₂.

Transwell

The cells were first cultured in a serum-free medium for 24 h to further remove the influence of serum. A volume of 200 µl of cell suspension was added to the upper chamber. A complete medium of 600 µl was added to the lower chamber of the 24-well plate. Cells were then cultured at 37 ℃ for 48 h. The cells were fixed with 4% paraformaldehyde for 10 min followed by dyeing for 15 min using crystal violet. The number of cells in the lower chamber was counted under a microscope.

Wound-healing scratch assay

Inoculate the cells into a 6-well plate, when the cells converge reached above 90% and line the central area of cell growth with 200 μl of the spear tip, about every 1 cm of a line. The cells were washed with PBS three times and the dropping cells were removed. Images were collected under a microscope (0 h). Then different treatments were performed following the design of the experiment. After being incubated in an incubator at 37 ℃ for 48 h, images were collected under a microscope again (48 h). Cell migration was analyzed using Image-Pro Plus software, Media Cybernetics, USA.

Immunohistochemistry and immunofluorescence

Tissue samples (peritumoral and gastric carcinoma) were embedded in paraffin; cells like AGS, HGC-27 and Cisplatin resistance gastric cancer cell lines (AGS^CDDP and HGC-27^CDDP) were fixed with 4% paraformaldehyde. Dewaxing is performed first, followed by antigen repair. Then permeabilized the sections by 0.1% triton X-100 for 20 min at 37 ℃ and peroxidase was used to block for 15 min. The section was then incubated with ADAR1 (1 : 20; cat no.sc-73408; Santa Cruz Biotechnology, Dallas, Texas, USA), AZIN1 (1 : 100; cat no.11548-1-AP; Proteintech, Chicago, Illinois, USA) and E-cadherin (1 : 100, cat no.ab40772; Abcam, Cambridge, UK) as primary antibodies at 4 ℃ overnight. On the following day, for immunofluorescence, secondary antibodies Alexa Fluor 488 Goat Anti-Rabbit IgG (cat no. ZF-0511; 1 : 100; ZSGB-BIO, Beijing, China) and Alexa Fluor 594 Goat Anti-mouse IgG (cat no. ZF-0513; 1 : 100; ZSGB-BIO, Beijing, China) were then used to incubate the sections for 1 h at 37 ℃. The cells were then stained with DAPI. For immunohistochemistry, a secondary antibody (PV6000, ZSGB-BIO, Beijing, China) was used to incubate the sections for 20 min at 37 ℃ followed by staining the sections with DAB and hematoxylin respectively.

Synthesis validation of small interference sequences

Three small interfering RNA (siRNA) oligonucleotides against ADAR1 (Sangon Biotech, Shanghai, China) were as follows:

siADAR1#1 : 5′-CAUCAAAUGCCUCAAAUAA-3′ (sense), 5′-UAAAUGCUGUGCUAAUUGA-3′ (antisense);

siADAR1#2 : 5′-GCCTCAAATAACATGGTAACC-3′ (sense), 5′-CCATGAACCTCGATTTAAATT-3′ (antisense);

siADAR1#3 : 5′-CCUUCUACAGUCAUGGCUUTT-3′ (sense), 5′-AAGCCAUGACUGUAGAAGGTT-3′ (antisense);

si-negative control: 5′-UUCUCCGAACGUGUCACGUTT-3′ (sense), 5′-ACGUGACACGUUCGGAGAATT-3′ (antisense);

siAZIN1 : 5′-CGGAUUUGCUUGUUCCAGUAATT-3′ (sense), 5′-UUACUGGAACAAGCAAAUCCGTT-3′ (antisense);

The oligonucleotides were transfected via Lipofectamine 3000 (#11668019, Invitrogen, Carlsbad, California, USA) in line with the manufacturer’s instructions.

Cell clone formation experiment

The cell suspensions were seeded in petri dishes and cultured in a 37 ℃ incubator for 24 h. The CDDP treatment and ADAR1 siRNA transfection were made after 24 h incubation. The culture was then continued in an incubator for 2–3 weeks and observed frequently. When visible clones appeared in the petri dish, the culture was terminated. The cells were fixed with 4% paraformaldehyde for 10 min then paraformaldehyde was discarded and 0.1% crystal violet was dyed for 20 min.

Western blot assay

Total protein extraction and quantification were respectively using Total Protein Extraction Kit (KGP2100; KeyGENBioTECH, Nanjing, China) and BCA protein quantification Kit (PT0001; LEAGENE, Beijing, China). The protein was separated by 9% SDS-PAGE and transferred to PVDF membranes for 90 min. TBST (TBS with Tween-20) and 5% milk powder were used to prepare blocking solution, which was then used to incubate transferred membranes for 1 h at room temperature (RT). The primary antibodies were prepared as follows: ADAR1 (cat no.sc-73408; 1 : 200; Santa Cruz Biotechnology, Dallas, Texas, USA), β-actin (cat no.20536-1-AP; 1 : 2000; Proteintech, Chicago, Illinois, USA), AZIN1 (cat no.11548-1-AP; 1 : 1000; Proteintech, Chicago, Illinois, USA), Vimentin (cat no.bs-8533R; 1 : 1000; Bioss, Beijing, China), N-cadherin (cat no. bs-1172R; 1 : 1000; Bioss, Beijing, China), GAPDH (cat no.bs-2188R; 1 : 4000; Bioss, Beijing, China), β-catenin (cat no.bs-1165R;1 : 500; Bioss, Beijing, China), MMP2 (cat no.10373-2-AP;1 : 1000; Proteintech, Chicago, Illinois, USA), MMP9 (cat no.10375-2-AP;1 : 1000; Proteintech, Chicago, Illinois, USA), TWIST (cat no.bs-2441R; 1 : 2000; Bioss, Beijing, China). Following the incubation overnight at 4 ℃, the secondary antibody Horseradish-labeled Goat anti-rabbit IgG (1 : 5000; ZB-2301, ZSGB-BIO, Beijing, China) and Horseradish-labeled Goat anti-mouse IgG (1 : 4000; ZB-2305, ZSGB-BIO, Beijing, China) was used to incubate the PVDF membranes for 1 h at RT. Finally, PVDF membranes detected signals with electrochemiluminescence solution.

Nude mice subcutaneous tumor model

BALB/cnude mice were purchased from Beijing WeitongLihua Laboratory Animal Technology Co., Ltd (https://www.vitalriver.com), age, 4–5 weeks; weight, 18–20 g. The mice were randomly divided into two groups (n = 5/group). We subcutaneously injected mice with HGC-27 cells transfected with either siRNA-negative control or ADAR1 siRNA (5 × 106 ml). After being fed for 21 days, the nude mice were euthanized, we removed their tumors, and the tumor volume was measured [(length × width2)/2]. All animal experiments complied with«Guidelines for the care and use of laboratory animals», and were approved by the Ethical Committee of People’s Hospital of Ningxia Hui Autonomous Region.

Hematoxylin and eosin staining

The tissue samples were fixed in 4% paraformaldehyde and dehydrated through graded ethanol series. After embedding with paraffin, the tissue samples were cut into 5 mm-thick sections. The sections were deparaffinized and dehydrated, followed by stained with hematoxylin and washed under running water. Then the sections were put in the acetic acid for 3 s, washed with running water for 2 min, and then stained with eosin. Finally, the sections were sealed with neutral gelatin after drying. The images were captured under an inverted light microscope (Olympus, Tokyo, Japan).

Statistical analysis

Experimental data are expressed as mean ± SE. T-test was used to compare the means of two samples, and one-way analysis of variance was used to compare the means of multiple samples. The histogram is represented by GraphPad Prism 8.0 software, USA. Significance is indicated as *P < 0.05; **P < 0.01; ***P < 0.001.

Results

Expression and correlation of adenosine deaminases acting on RNA1 and antizyinhibitor 1 in gastric cancer tissue and paracancerous tissue

We used immunohistochemistry to assess the expression levels of ADAR1 and AZIN1 in gastric cancer tissue and paracancerous tissue. ADAR1 and AZIN1 expression were significantly higher in gastric cancer tissue than in paracancerous tissue (Fig. 1a–c). Immunofluorescence revealed that ADAR1 colocalized with AZIN1 and E-cadherin, that ADAR1 was mainly expressed in the nucleus, and that AZIN1 was mainly expressed in the cytoplasm. A small portion of the two overlapped in the nucleus. We also selected the epithelial–mesenchymal transition (EMT) marker E-cadherin to determine if it colocalized with ADAR1. E-cadherin was mainly expressed in the cell membrane, and ADAR1 was expressed in the nucleus, as also noted above (Fig. 1d–g). To study the regulatory roles of ADAR1 and AZIN1, we constructed a cell model in which ADAR1 expression was knocked out. Lip3000 was used as the control vector, and the interference sequence with the best effect, as determined by screening via western blot, was used for subsequent experiments (Fig. 1h–k).

Fig. 1: ADAR1 and AZIN1 were highly expressed in gastric cancer tissue; construction of ADAR1 siRNA cell models. (a–c) Immunohistochemistry method was used to detect the expression level of ADAR1 and AZIN1 in gastric cancer tissue and peritumoral (**P < 0.01, Scale bar, 20 μm, 400×). (d–g) Immunofluorescence method was used to detect the expression level and the location of ADAR1, AZIN1 and E-cadherin (**P < 0.01, Scale bar, 20 μm, 400×). (h and i) AGS cells were transfected with ADAR1 siRNA; (j and k) HGC-27 cells were transfected with ADAR1 siRNA. Western blot was used to select the best siRNA sequences for the following experiments (***P < 0.001). ADAR1, adenosine deaminases acting on RNA1; AGS, human gastric adenocarcinoma cell line; AZIN1, antizyinhibitor 1; siRNA, small interfering RNA.

Analysis of the relationship between ADAR1 and the clinicopathological characteristics of gastric cancer patients indicated that ADAR1 was closely associated with invasion, tumor-node-metastasis stage, peritoneal metastasis and lymph node metastasis (Table 1). The results of the correlation analysis indicated thatADAR1 expression was positively correlated with AZIN1 expression and negatively correlated with E-cadherin expression, further indicating the possible regulatory role of ADAR1 and AZIN1 in the progression of gastric cancer and indicating that ADAR1 overexpression was related to the expression of EMT-related markers (Table 2).

Table 1 - Correlation between adenosine deaminases acting on RNA1, antizyinhibitor 1 expression and clinicopathologic features across gastric cancer patients

Clinical parameters	Total (n = 83)	ADAR1		P value	AZIN1		P value
		Positive (n = 62)	Negative (n = 21)		Positive (n = 58)	Negative (n = 25)
Sex
Male	43	33	10	0.657	31	12	0.649
Female	40	29	11		27	13
Age
<60	32	25	7	0.570	26	6	0.074
≥60	51	37	14		32	19
Lymph metastasis
No	56	38	18	0.039*	35	21	0.035*
Yes	27	24	3		23	4
Tumor-node-metastasis stage
I–II	29	16	13	0.003**	16	13	0.032*
III–IV	54	46	8		42	12
Peritoneal metastasis
No	59	48	11	0.029*	37	22	0.026*
Yes	24	14	10		21	3
Depth of invasion
T1/T2	39	25	14	0.037*	20	19	0.001***
T3/T4	44	37	7		38	6

ADAR1, adenosine deaminases acting on RNA1

*P < 0.05; **P < 0.01; ***P < 0.0001.

Table 2 - Correlation between adenosine deaminases acting on RNA1, antizyinhibitor 1 and E-cadherin

Factor	Pearson correlation	P value
ADAR1, AZIN1	0.556	0.000
ADAR1, E-cadherin	−0.992	0.000
AZIN1, E-cadherin	−0.646	0.000

ADAR1, adenosine deaminases acting on RNA1; AZIN1, antizyinhibitor 1

Adenosine deaminases acting on RNA1 regulates antizyinhibitor 1 expression in gastric cancer cells and drug-resistant cells

Immunohistochemistry and immunocytofluorescence revealed that the expression levels of ADAR1 and AZIN1 in cisplatin-resistant gastric cancer cells (HGC-27^CDDP and AGS ^CDDP) cells were significantly higher than those in Lip3000 cells; however, ADAR1 knockout significantly reduced the expression levels of AZIN1 in HGC-27, AGS cells and HGC-27^CDDP and AGS^CDDP cells. These results indicated that ADAR1 knockout decreased the expression of AZIN1 in cisplatin-resistant gastric cancer cells (Figs. 2 and 3).

Fig. 2: Immunocytochemistry method was used to detect the expression level of ADAR1 and AZIN1 in AGS, HGC-27 (transfected with ADAR1 siRNA) and Cisplatin resistance gastric cancer cell lines (AGS^CDDP and HGC-27^CDDP transfected with or without ADAR1 siRNA). (a–c) The expression level of ADAR1 and AZIN1 in AGS and AGS^CDDP. (d–f) The expression level of ADAR1 and AZIN1 in HGC-27 and HGC-27^CDDP. **P < 0.01, ***P < 0.001, Scale bar, 20 μm, 400×. ADAR1, adenosine deaminases acting on RNA1; AGS, human gastric adenocarcinoma cell line; AZIN1, antizyinhibitor 1; siRNA, small interfering RNA.

Fig. 3: Immunofluorescence staining analyzed the expression of ADAR1 and AZIN1 in AGS, HGC-27 (transfected with ADAR1 siRNA) and Cisplatin resistance gastric cancer cell lines (AGS^CDDP and HGC-27^CDDP transfected with or without ADAR1 siRNA). (a, c, d) The expression level of ADAR1 and AZIN1 in AGS and AGS^CDDP. (b, e, f) The expression level of ADAR1 and AZIN1 in HGC-27 and HGC-27^CDDP. **P < 0.01, ***P < 0.001, Scale bar, 20 μm, 400×. ADAR1, adenosine deaminases acting on RNA1; AGS, human gastric adenocarcinoma cell line; AZIN1, antizyinhibitor 1; siRNA, small interfering RNA.

Adenosine deaminases acting on RNA1 regulate the proliferation, invasion and migration of gastric cancer cells by targeting antizyinhibitor 1

The interaction between ADAR1 and AZIN1 was confirmed. Next, we attempted to explore the effect of ADAR1 siRNA on the invasion and migration capacity of HGC-27^CDDP and AGS^CDDP cells. The results of Transwell (Fig. 4a–c) and cell scratch assays indicated that the invasion and migration abilities of cisplatin-resistant gastric cancer cell lines were enhanced and that ADAR1 knockout inhibited the invasion and migration abilities of cisplatin-resistant gastric cancer cell lines (Fig. 4d–g). In addition, cell cloning experiments revealed that the number of proliferating colonies of HGC-27 ^CDDP and AGS^CDDP cells significantly increased; ADAR1 siRNA inhibited the number of cell colonies (Fig. 5a–c).

Fig. 4: Silence of ADAR1 inhibited the invasion and migration of gastric cancer cells and cisplatin-resistant gastric cancer cells. (a–c) Transwell assays showed cells invasion quantity (48 h) of AGS, HGC-27 cells and Cisplatin resistance gastric cancer cells (AGS^CDDP and HGC-27^CDDP), ***P < 0.001, Scale bar, 50 μm, 200×. (d, f) Wound-Healing Scratch Assay showed cells migration ability (48 h) of AGS and AGS^CDDP cells, *P < 0.05, **P < 0.01, ***P < 0.001, Scale bar, 100 μm, 200×. (e, g) Wound-Healing Scratch Assay showed cells migration ability (48 h) of HGC-27 and HGC-27^CDDP cells,**P < 0.01, ***P < 0.001, Scale bar, 100 μm, 200×. ADAR1, adenosine deaminases acting on RNA1; AGS, human gastric adenocarcinoma cell line.

Fig. 5: Silence of ADAR1 inhibited cell proliferation ability of Cisplatin resistance gastric cancer cells. (a) Cell clone formation experiment showed that ADAR1 siRNA inhibits the cell proliferation of AGS^CDDP and HGC-27^CDDP cells. (b and c) Cell clone formation analysis showed that ADAR1 siRNA inhibits the cell proliferation of AGS^CDDP and HGC-27^CDDP cells,***P < 0.001. ADAR1, adenosine deaminases acting on RNA1; AGS, human gastric adenocarcinoma cell line; siRNA, small interfering RNA.

Adenosine deaminases acting on RNA1 exert an antitumor effect and reverse cisplatin resistance in gastric cancer through antizyinhibitor 1

Next, we used western blot to assess the regulatory relationship between ADAR1 and AZIN1. ADAR1 siRNA significantly reduced the expression of AZIN1 and EMT-related indicators, including Vimentin, N-cadherin (Fig. 6), β-catenin, MMP9, MMP2 and TWIST in cisplatin-resistant and nonresistant gastric cancer cells (Fig. 7a–j).

Fig. 6: Western blot assay was used to determine the ADAR1, AZIN1, Vimentin and N-cadherin protein expression levels in gastric cancer cells and Cisplatin resistance gastric cancer cells. (a–e) The ADAR1, AZIN1, Vimentin and N-cadherin protein expression levels in AGS and AGS^CDDP. (f–j) The ADAR1, AZIN1, Vimentin and N-cadherin protein expression levels in HGC-27 and HGC-27^CDDP. *P < 0.05, **P < 0.01, ***P < 0.001. ADAR1, adenosine deaminases acting on RNA1; AGS, human gastric adenocarcinoma cell line; AZIN1, antizyinhibitor 1.

Fig. 7: Western blot assay was used to determine the β-catenin, MMP2, MMP9 and TWIST protein expression levels in gastric cancer cells and Cisplatin resistance gastric cancer cells; construction of AZIN1 siRNA cell models. (a–e) Protein expression levels in AGS and AGS^CDDP. (b) β-catenin, (c) MMP2, (d) MMP9, (e) TWIST. (f–j) Protein expression level in HGC-27 and HGC-27^CDDP. (g) β-catenin, (h) MMP2, (i) MMP9, (j) TWIST. (k–n) Construction of AZIN1 siRNA cell models. AZIN1 protein expression level in (k and l) AGS and (m and n) HCG-27 cells.**P < 0.01, ***P < 0.001. ; AGS, human gastric adenocarcinoma cell line; AZIN1, antizyinhibitor 1; siRNA, small interfering RNA.

To further determine the regulatory role of ADAR1 and AZIN1 in gastric cancer cells and to determine whether they produce opposite effects, we designed two siRNAs that specifically targeted AZIN1 (Fig. 7k–n).

The results showed that ADAR1 expression was not significantly reduced in cells treated with AZIN1 siRNA. Interestingly, the expression of EMT-related indicators, including Vimentin, N-cadherin, β-catenin, MMP9 and MMP2 decreased (Fig. 8a–p). The combined application of ADAR1 siRNA and AZIN1 siRNA had a more significant regulatory effect on EMT indicators. These results indicated that ADAR1 exerted an antitumor effect and reversed cisplatin resistance in gastric cancer by regulating the expression of AZIN1, thereby enhancing the therapeutic effect.

Fig. 8: Western blot assay was performed to detect the ADAR1, AZIN1, β-catenin, MMP2, MMP9, Vimentin, N-cadherin protein expression level in AGS and HGC-27 cells under the transfection of ADAR1 siRNA. (a) Protein expression level in AGS cells. (c) ADAR1, (d) AZIN1, (e) β-catenin, (f) MMP2, (g) MMP9, (h) Vimentin, (i) N-cadherin; (b) protein expression level in HGC-27 cells. (j) ADAR1, (k) AZIN1, (l) β-catenin, (m) MMP2, (n) MMP9, (o) Vimentin, (p) N-cadherin. **P < 0.01, ***P < 0.001. ADAR1, adenosine deaminases acting on RNA1; AGS, human gastric adenocarcinoma cell line; AZIN1, antizyinhibitor 1; siRNA, small interfering RNA.

Knockdown of adenosine deaminases acting on RNA1 inhibited the growth of subcutaneous implantation tumor and expression of antizyinhibitor 1 in nude mice

Knockdown of ADAR1 significantly inhibited the growth of subcutaneous xenograft in nude mice (Fig. 9a). Hematoxylin and eosin staining showed that the tumor nuclei in the siRNA-negative control group were dark and clustered tightly. Compared with the siRNA-negative control group, ADAR1 siRNA group had a loose structure of tumor cells, some nuclear pyknosis and scattered distribution (Fig. 9b). Immunohistochemical staining results showed that the expressions of ADAR1 and AZIN1 in the ADAR1 siRNA group were significantly decreased compared with the siRNA-negative control group (Fig. 9c–e). The results of the western blot were consistent with those of immunohistochemistry (Fig. 9f–h). In conclusion, the knockdown of ADAR1 expression can significantly inhibit tumor growth and may play an antitumor role through AZIN1.

Fig. 9: Knockdown of ADAR1 inhibited the growth and AZIN1 expression of subcutaneous allograft tumor in nude mice. (a) Knockout of ADAR1 inhibited the volume of subcutaneous allograft tumor in nude mice. (b) HE stain of subcutaneous allograft tumor. (c–e) The expression levels of ADAR1 and AZIN1 in subcutaneous allograft tumor were detected by immunohistochemistry. (f–h) The expression levels of ADAR1 and AZIN1 in the subcutaneous allograft tumor were determined by western blot. ***P < 0.001. ADAR1, adenosine deaminases acting on RNA1; AZIN1, antizyinhibitor 1; HE, hematoxylin and eosin.

Discussion

Cisplatin resistance is a key challenge associated with basic chemotherapy regimens for gastric cancer patients and is also a major factor affecting the prognosis of gastric cancer patients. Therefore, there is an urgent need for a stable biomarker to predict and reverse cisplatin resistance, explore potential mechanisms of action, and suggest potential prevention and treatment targets to improve the efficacy of cisplatin in the treatment of gastric cancer [⁹]. Our findings revealed the relationship between ADAR1 and AZIN1 expression in gastric cancer tissue, indicating that they are stable biomarkers in the process of cisplatin resistance in gastric cancer. Although many studies have shown that ADAR1 has important research significance in the occurrence, development and metastasis of gastric cancer, the results from our study suggest an interaction between ADAR1 and AZIN1 in cisplatin-resistant gastric cancer cells, thereby providing new insights into drug resistance mechanisms.

The dysregulation of ADAR1 expression leads to abnormal RNA editing, which can drive phenotypic changes in cancer cells. Phenotypic changes are common in many cancers, such as liver cancer, lung cancer, breast cancer and esophageal cancer. In most cases, phenotypic changes promote the occurrence and development of cancer [¹⁰]. In triple-negative breast cancer, ADAR1 is a key oncogene, and the deletion of ADAR1 inhibits the biological effects, such as proliferation and metastasis, of breast cancer cells [¹¹]. ADAR1 and AZIN1 have also shown strong specificity in lower gastrointestinal tumors. In particular, ADAR1 and AZIN1 were significantly higher in colon cancer tissue than in normal intestinal mucosa and regulated vimentin protein expression and fibroblast activation to promote metastasis. Additionally, ADAR1 and AZIN1 editing enhances the infiltration of fibroblasts and alters cancer cell metastasis, providing an ideal tumor microenvironment [¹²]. Our study revealed phenotypic changes in ADAR1 and AZIN1 in gastric cancer tissue and paracancerous tissue, findings that are consistent with the results of published studies; however, regarding localization, ADAR1 and AZIN1 expression overlapped to a small extent in the nucleus. Correlation analysis further demonstrated the potential relationship between the two; however, the cellular immunohistochemistry results were somewhat different from the tissue expression characteristics, especially the expression of AZIN1. The expression of AZIN1 in tissue is mainly located in the cytoplasm; however, in gastric cancer cells, there is strong expression in the cytoplasm and nucleus, especially, in the nucleus of cisplatin-resistant gastric cancer cells; however, after ADAR1 interference, ADAR1 expression in the nucleus was significantly reduced in AGS and HGC-27 cells. Interestingly, AZIN1 expression in the nucleus was also significantly reduced. We speculate that ADAR1 knockout blocked the nuclear translocation of AZIN1 and reversed the development of cisplatin resistance.

EMT is involved in many changes in the cellular environment during cancer treatment, for example, changes in cell phenotype, loss of cell polarity, and remodeling of the extracellular matrix. These changes are also core factors for the emergence of drug resistance [¹³]. EMT-related markers also play a key role in the formation of drug resistance. Cellular matrix metalloproteins are key factors in cancer cell metastasis and drug resistance. In prostate cancer studies, MMP2 and MMP9 were found to be involved in prostate cancer invasion [¹⁴], and both were involved in the development of drug resistance in prostate cancer [¹⁵]. In liver cancer studies, the upregulation of MMP2 and MMP9 expression enhanced the migration and proliferation ability of liver cancer cells [¹⁶]. In drug resistance mechanisms in gliomas, the reduced expression of MMP2 and MMP9 provides the necessary conditions for reversing drug resistance [¹⁷]. Our experimental results indicated that in cisplatin-resistant gastric cancer cells, interference with ADAR1 or AZIN1 expression reduced MMP2 and MMP9 expression; the effect was more significant after the combined ADAR1 and AZIN1 interference, indirectly indicating that ADAR1 regulates MMP2 and MMP9 expression through AZIN1 to play a role in the inhibition of gastric cancer cell metastasis and reversal of cisplatin resistance.

ADAR1 overexpression alters the expression of some key indicators of the canonical Wnt pathway, such as β-catenin, which promotes cell invasion and angiogenesis in breast cancer [¹⁸]. ADAR1 knockout inhibits the proliferation of acute myeloid leukemia cells by regulating the Wnt/β-catenin pathway, inhibits the growth of subcutaneous tumors in nude mice, and reduces the expression of the Wnt effector β-catenin [¹⁹]. In our study, ADAR1 and AZIN1 knockout decreased the expression of β-catenin, a finding that is similar to the results of a study by Li et al. [⁶]; however, we identified another regulatory mechanism. ADAR1 may regulate β-catenin expression through AZIN1, thereby affecting gastric cancer cell metastasis. In the field of gastric cancer research, researchers have used human gastric cancer tissue specimens to demonstrate the diagnostic and prognostic value of ADAR1 and AZIN1 in gastric cancer patients, but there is no relevant study involving mechanisms of drug resistance [²⁰]. In this study, in cisplatin-resistant gastric cancer cells, the knockout of ADAR1 expression significantly inhibited AZIN1, β-catenin, vimentin and N-cadherin expression in cisplatin-resistant gastric cancer cells. These results indicate that ADAR1 regulates the expression of EMT-related markers through AZIN1, thereby playing a target role in gastric cancer metastasis and cisplatin resistance.

In-vivo experiments, we found that knocking down the expression of ADAR1 significantly inhibited the growth of gastric cancer cells subcutaneously in nude mice, and inhibited the expression of AZIN1. This is similar to the results of Li et al. [⁶] and consistent with the results of cell experiments in this study, which further indicates that knockdown of ADAR1 may inhibit gastric cancer cell metastasis through AZIN1 signal, and also provides therapeutic reference for the prevention and treatment of gastric cancer metastasis.

In addition, we further confirmed the regulatory roles of ADAR1 and AZIN1 in the metastasis and proliferation of gastric cancer, revealing that ADAR1 knockout inhibits EMT-related indicators (vimentin, N-cadherin, β-catenin, MMP2, MMP9 and TWIST) through the downregulation of AZIN1 expression, thereby reversing cisplatin resistance in gastric cancer and improving the efficacy of gastric cancer treatment.

Acknowledgements

This work was supported by [Natural Science Foundation of Ningxia Hui Autonomous Region] [Grant numbers (2021AAC03298)] and (Cultivation and Revitalization Project of People’s Hospital of Ningxia Hui Autonomous Region) [Grant numbers (202014)].

Conflicts of interest

There are no conflicts of interest.