519 articles were extracted based on our search strategy, of which 266 were from PubMed, 170 from Embase, 52 from Scopus, and 31 from Web of Science. After removing duplicate articles by Endnote software, 217 studies remained and were selected for initial screening. Finally, 43 in-vivo (animal) and 15 in-vitro studies were included in the qualitative and quantitative analysis. The results of the different stages of screening are shown in the PRISM flowchart in Figure 2.
Characteristics of Articles
The characteristics of the articles included in the study are listed in Table 3, divided into in-vivo animal and in vitro sections.
Table 3-A includes 43 in-vivo (animal) studies, including 28 studies measuring NOX4 protein and 16 studies measuring NOX4 mRNA. Protein measurement was conducted using Western blot (WB) tests in 15 studies, immunohistochemistry (IHC) in 8 studies, flow cytometry (FC) in 6 studies, and enzyme-linked immunosorbent assay (ELISA) in one study. In addition, mRNA measurement was conducted using real-time PCR (qRT PCR) in 16 studies. NOX4 expression was measured in various organs, including the lungs in 14 studies, bone marrow in 11 studies, cardiovascular system in 7 studies, kidney in 3 studies, salivary gland in 2 studies, and 8 other organs each in one study, comprising the brain, liver, small intestine, spleen, thymus, prostate, parotid gland, and ovarian. The studies employed administered doses of IR that varied from 1 to 75 Gy. Out of the total, 18 studies focused on whole-body irradiation, while 21 studies investigated localized radiation treatment. The radiation sources utilized were: proton beam in one study, electron beam in one study, X-ray (KV) in 12 studies, X-ray (MV) in 9 studies, γ-ray (60Co) in 5 studies, and γ-ray (137Cs) in 15 studies. The time intervals post-irradiation employed for evaluating NOX4 expression varied from 24 hours to 22 months.
Table 3-B includes 15 in-vitro studies, with 10 measuring NOX4 protein and 10 measuring NOX4 mRNA, with some studies measuring both. The measurement of protein used WB tests in 8 studies and FC in 2 studies. For mRNA assessment, qRT-PCR was the method used in all 10 studies. NOX4 expression was evaluated in various cell types, including human cells in 11 studies, rat cells in 3 studies, and mouse cells in 2 studies. The doses of IR used in these studies ranged from 3 cGy to 16 Gy, with X-ray used in 10 studies and γ-ray in 5 studies. The follow-up period for NOX4 analysis post-irradiation varied from 15 minutes to 5 days. Of these, 13 studies utilized normal cells, while 2 studies employed cancer cells.
The following studies were excluded from the meta-analysis; an in vivo study from Jiyoung Park 2022 (38), and in vitro studies from Jiwon Choi 2022 (39), Urbain Weyemi 2015 (40), Yingchun Zhang 2012 (92), Yingchun Zhou 2022 (93), Sung Hyo Park 2020 (48), Sarah Park-2010 (96), Eun Joo Chung 2019 (97), and XiaoHong Yang 2017 (98) because they did not contain complete analysis, the results were not clearly stated, or were not in line with the inclusion criteria. Based on the search results, there were no clinical studies specifically investigating the effect of IR on NOX4 expression in humans.
Meta-analysis (Animal studies)
Funnel plot
The results of the funnel plot analysis and Egger’s test on the animal studies are reported in Figure 2, indicating possible publication bias for the outcome measures including mRNA expression (Egger's p < 0.05) (Figure 3-A), and protein expression (Egger's p < 0.05) (Figure 3-B).
Overall Analysis
It was found that 16 studies (containing 19 separate experiments) reported a noteworthy increase in NOX4 mRNA expression in the irradiated animals, with a SMD of 5.070 and CI95% ranging from 3.374 to 6.766;(p < 0.001; I2=91.8%). Furthermore, 28 studies (containing separate 38 experiments) reported a significant increase in NOX4 protein expression post-irradiation, with an SMD of 3.452, CI95% of 2.807 to 4.097, (p < 0.001, and; I2=74.5%) (Figure 4).
Sensitivity Analysis
The leave-one-out sensitivity analysis indicated that no single study had a significant effect on the overall findings of NOX4 mRNA and protein expression after IR, suggesting that the results were robust and not unduly influenced by any individual study.
Subgroup Analysis
Subgroup analysis was conducted based on various factors including the type of tests used to measure NOX4 expression, specific organs studied, dose of IR administered, source of radiation, and the follow-up time after exposure to IR.
Subgroup analysis of tests measuring NOX4 expression post-irradiation show varying results. The use of WB (SMD=2.886; P < 0.001), IHC (SMD=6.165; P < 0.001), FC (SMD=2.862; P < 0.001), and qRT-PCR (SMD=5.070; P < 0.001) all provided significant results, while ELISA was only utilized in a single study (Table 5-a).
Eight organs (brain, liver, ovaries, parotid gland, prostate, small intestine, spleen, and thymus) were limited to a single study (Table 5-b).
Subgroup analysis focusing on different organs where NOX4 expression was assessed showed notable increases following IR exposure in specific organs. Increased NOX4 protein expression was observed in the lungs (SMD=4.391; P < 0.001), bone marrow (SMD=2.862; P < 0.001), cardiovascular system (SMD=2.206; P < 0.01), salivary gland (SMD=3.714; P < 0.001), and kidney (SMD=1.502; P < 0.01) post-irradiation. Moreover, increased NOX4 mRNA expression was observed in the bone marrow (SMD=9.371; P < 0.01) and cardiovascular system (SMD=2.716; P < 0.01) post-irradiation.
Subgroup analysis based on the dose of IR showed a substantial increase in NOX4 protein expression across different dose levels: 3-6 Gy (SMD=2.631; P < 0.001); 6-12 Gy (SMD=4.236; P < 0.001); 12-24 Gy (SMD=2.941; P < 0.001); and 75 Gy (SMD=8.303; P < 0.05) (Table 5c). Subgroup analysis based on the dose of IR showed also a substantial increase in NOX4 mRNA expression across different dose levels: 1-3 Gy (SMD=13.904; P < 0.05); 3-6 Gy (SMD=11.505; P < 0.05); 6-12 Gy (SMD=2.706; P < 0.01); and 12-24 Gy (SMD=4.405; P < 0.01) (Table 5-c).
Moreover, subgroup analysis based on the irradiation method revealed that both whole body irradiation (SMD=3.907; P < 0.001) as well as local irradiation (SMD=3.581; P < 0.001) had a significant effect on increasing NOX4 protein expression (Table 5d). The same results were observed for NOX4 mRNA expression: whole body irradiation (SMD=6.784; P < 0.001) and local irradiation (SMD=2.736; P < 0.01) (Table 5-d).
Subgroup analysis based on the radiation source, showed that various radiation sources, including X-ray (KV) (SMD=2.725; P value < 0.001); X-ray (MV) (SMD=3.162; P value < 0.001); γ-ray (137Cs) (SMD=3.724; P value < 0.001); γ-ray (60Co) (SMD=16.507; P value < 0.05); and electron beam (SMD=1.997; P value < 0.05) all had a significant effect on increasing NOX4 protein expression (Table 5-e). Similarly, for NOX4 mRNA expression similar results were observed: X-ray (KV) (SMD=1.288; P value < 0.01); X-ray (MV) (SMD=17.133; P value < 0.05); γ-ray (137Cs) (SMD=4.823; P value < 0.001); γ-ray (60Co) (SMD=9.211; P value < 0.05); and proton beam (SMD=15.014; P value < 0.01) (Table 5e). Proton beam and electron beam were each only used in one study.
Finally, subgroup analysis based on the follow-up time after IR exposure showed a consistent increase in NOX4 protein expression across all follow-up times. This included times below one week (SMD=2.455; P < 0.001), between one and two weeks (SMD=3.588; P < 0.001), between two weeks and one month (SMD=3.169; P < 0.001), and more than one month (SMD=4.653; P < 0.001) (Table 5f). Similarly, for NOX4 mRNA expression similar results were observed: below one week (SMD=8.110; P < 0.001), between one and two weeks (SMD=9.347; P < 0.01), between two weeks and one month (SMD=5.861; P < 0.01), and more than one month (SMD=3.916; P < 0.001) (Table 5-f).
Meta-analysis (in-vitro studies)
Funnel plot
The results of the funnel plot analysis and Egger’s test of the in-vitro studies are reported in (Figure 5), indicating possible publication bias for the outcome measure of mRNA expression in normal cells (Egger's p < 0.05) (Figure 5-A) and protein expression in normal cells (Egger's p < 0.05) (Figure 5-B), but showed no evidence of publication bias for the outcome measure of protein expression in cancer cells (Egger's p > 0.05) (Figure 5-C).
Analysis of normal and cancer cells
General analysis of the normal cell studies revealed that 9 studies (10 independent experiments) had a significant increase in NOX4 (mRNA) expression in the irradiated cells (SMD = 3.563; CI95% 2.297 to 4.829; P < 0.001; I2=69 %). Additionally, 8 studies reported a significant increase in NOX4 protein expression after IR (SMD = 2.689; CI95% 1.314 to 4.064; P < 0.001; I2=79.5 %) (Figure 6-A). These findings are in agreement with the meta-analysis of the animal studies. In addition, the studies included in Table 3B utilized various methods to assess the effects of IR on NOX4 expression, with the majority of studies showing a significant increase in NOX4 expression in normal cells after irradiation.
The general analysis of the cancer cell studies revealed that 1 study observed a significant increase in NOX4 (mRNA) expression in irradiated cells (SMD = 3.280; CI 95% 1.865 to 4.694; P < 0.001). In addition, 2 studies (3 independent experiments) reported a significant increase in NOX4 (protein) expression after IR (SMD = 2.159; CI 95% 0.202 to 4.116; P < 0.05; I2=81.5 %) (Figure 6-B).
Sensitivity Analysis
The leave-one-out sensitivity analysis indicated that no single study had a significant effect on the results of NOX4 mRNA and protein expression in normal cells after IR, suggesting that the results were robust and not heavily influenced by any individual study.
The estimated pooled SMD was significant in the original analysis of NOX4 protein expression in cancer cells (P < 0.05). However, the leave-one-out sensitivity analysis revealed that the statistical significance was changed by the removal of the experiments included in Jiyoung Park 2022-3 study (38): 4T1-Luc mouse breast cancer cells independent experiment (SMD = 2.064; CI95% (-1.638 to 5.766); P > 0.05) or MDA-MB-231 human breast cancer cells independent experiment (SMD = 1.319; CI95% (-0.786 to 3.424); P > 0.05).
Subgroup Analysis of Normal Cells
Subgroup analysis of normal cells was performed based on the tests that measured NOX4 expression after radiation, including WB (SMD=1.945; P value < 0.01), FC (SMD=4.193; P < 0.001), and qRT-PCR (SMD=3.563; P < 0.001) (Table 6-a).
Following the initial analysis, further investigation was carried out focusing on the source of the normal cells in which NOX4 expression was assessed. The findings indicated a significant rise in NOX4 protein expression levels after IR in human cells (SMD=2.491; P < 0.01) (Table 6-b). Also, the findings indicated a significant rise in NOX4 mRNA expression levels after IR in human cells (SMD=3.412; P < 0.01), rat cells (SMD=5.652; P < 0.001), and mouse cells (SMD=2.878; P < 0.01) (Table 6-b).
Additionally, subgroup analysis was undertaken based on the radiation dose, radiation source, and duration of follow-up. The subgroup analysis concerning the dose of IR revealed a noteworthy impact on the upregulation of NOX4 protein expression in normal cells at two dose levels; 3-50 cGy (SMD=4.193; P < 0.001) and 8-16 Gy (SMD=2.516; P < 0.05) (Table 6c). NOX4 mRNA expression in normal cells was upregulated at all dose levels; 3-50 cGy (SMD=4.149; P < 0.001), 2-6 Gy (SMD=3.173; P < 0.001), and 8-16 Gy (SMD=2.987; P < 0.05) (Table 6-c).
Moreover, the subgroup analysis, focusing on the radiation source, showed that both X-ray (SMD=2.601; P < 0.01) and γ-ray (SMD=3.531; P < 0.01) sources had a significant impact on increasing NOX4 protein expression in normal cells (Table 6d). Similarly, for NOX4 mRNA expression in normal cells, the same results were observed: X-ray (SMD=3.020; P < 0.001) and γ-ray (SMD=4.224; P < 0.001) (Table 6-d).
Lastly, the subgroup analysis based on the follow-up time post-IR exposure indicated an increase in NOX4 protein expression in normal cells at all follow-up time points: 1 hour (SMD=3.601; P < 0.05), 2-24 hours (SMD=2.590; P < 0.01), and 2-5 days (SMD=2.306; P < 0.001) (Table 6-e). Similarly, for NOX4 mRNA expression in normal cells, the same results were observed: 1 hour (SMD=4.112; P < 0.001), 2-24 hours (SMD=3.843; P < 0.001), and 2-5 days (SMD=2.612; P < 0.01) (Table 6-e).