Divergent Effects of Arsenic on NF-κB Signaling in Different Cells or Tissues: A Systematic Review and Meta-Analysis

Arsenic is ubiquitously present in human lives, including in the environment and organisms, and has divergent effects between different cells and tissues and between different exposure times and doses. These observed effects have been attributed to the nuclear transcription factor kappa B(NF-κB) signaling pathway. Herein, a meta-analysis was performed by independently searching databases including the Cochrane Library, PubMed, Springer, Embase, and China National Knowledge Infrastructure, to analyze effects of arsenic exposure on NF-κB signaling. Compared to controls, in the exposed group, p-IκB levels were found to be 8.13-fold higher (95% CI, 2.40–13.85; Z = 2.78; p = 0.005), IκB levels were 16.19-fold lower (95% CI, −27.44–−4.94; Z = 2.78; p = 0.005), and NF-κBp65 levels were 0.77-fold higher (95% CI, 0.13–1.42; Z = 2.34; p = 0.02) for normal cells and tissue, while NF-κBp65 levels were 4.90-fold lower (95% CI, −8.49–1.31; Z = 2.62; p = 0.009), NF-κB activity was 2.45-fold lower (95% CI, −3.66–1.25; Z = 4.00; p < 0.0001), and DNA-binding activity of NF-κB was 9.75-fold lower (95% CI, −18.66–4.54; Z = 2.15; p = 0.03) for abnormal cells and tissue. Short exposure to high arsenic doses activated the NF-κB signaling pathway, while long exposure to low arsenic doses suppressed NF-κB signaling pathway activation. These findings may provide a theoretical basis for injurious and therapeutic mechanisms of divergent effects of arsenic.


Introduction
Arsenic exists ubiquitously on earth including in the earth's crust, in soil, in water, in air, in food, and in organisms [1,2]. The International Agency for Research on Cancer (IARC) has classified arsenic and its compounds as Group 1 carcinogens. Nearly 200 million people worldwide are threatened by arsenic poisoning, and the number of people suffering from arsenic poisoning due to water contamination alone is more than 3 million in China [3]. Epidemic investigations and studies have shown that inorganic arsenic may increase the risk of many cancers including bladder, kidney, liver, lung, prostate, and skin cancer [4][5][6][7][8]. The risks of other diseases including cardiovascular disease [9], hypertension [10], and diabetes [11] are also increased by arsenic. Nevertheless, arsenic was recently used as an experimental anti-tumor drug in clinical treatment [12]. Investigations into arsenic poison effects and related molecular mechanisms as well as those of the therapeutic effects of arsenic have

Outcome Indicators
Nuclear transcription factor kappa B(NF-κB), inhibitor of kappa B(IκB), phosphorylase inhibitor of kappa B(p-IκB), inhibitor of kappa B kinase (IKK), nuclear transcription factor kappa B activity (NF-κB activity), DNA-binding activity of the nuclear transcription factor kappa B (DNA-binding activity of NF-κB), nuclear transcription factor kappa B p65 (NF-κBp65), and nuclear transcription factor kappa B mRNA(NF-κB mRNA) were included as outcome indicators.

Data Extraction
Two reviewers (Shugang Li and Meng Wei) independently screened full-length articles. The following information was extracted from the complete manuscripts of each qualified study: publication characteristics (title of the study, first author, publication date, and journal/magazine title), baseline data (n, mean˘standard deviation [SD]) for the experimental and control groups, subject characteristics (source of cells and tissue, arsenic doses and exposure times), outcome indicators, and the source of indicator estimates. When the two reviewers' opinions differed, Shuxia Guo, who teaches meta-analysis at Shihezi University School of Medicine, was asked to make the final decision regarding the results.

Data Analysis
Twenty-seven articles were analyzed in Review Manager Version 5.2 (The Nordic Cochrane Centre, The Cochrane Collaboration, 2012, Portland Oregon, OR, USA) and Stata 12.0 (StataCorp., College Station, Texas, TX, USA). Significant heterogeneity was detected (p < 0.05, I 2 > 75%) and a random effects model was therefore applied for the meta-analysis. A multivariate meta-regression analysis was performed to determine the source of heterogeneity and continuous variables were estimated as standardized mean differences (SMDs) with 95% confidence intervals (CIs) between the arsenic treated groups and control groups. All reported p-values are two-sided and a significance level of 0.05 was used. For additional insight, subgroup analyses were performed based on exposure dose (ď5 µmol/L or ą5 µmol/L in vivo; ď5 mg/kg or ą5 mg/kg in vitro) and exposure time (ď24 h or ą24 h), based on the median of the indexes reported in the papers, to determine the factors associated with differences in the outcome indicators across different studies. Small-study effects were explored using funnel plots and Egger's tests and studysensitivities were assessed.

Study Characteristics
Using the search strategy described in Section 2, 573 relevant articles were identified (Figure 1), of which 27 were used for the meta-analysis based on the eligibility and exclusion criteria (Table 1). Various cell lines and animals were used as arsenic model groups in these studies, and in each study, the effect of arsenic on the NF-κB signaling pathway was assessed. The arsenic model groups were primarily cell lines and animals treated with arsenicin various forms (e.g., arsenite, As 2 O 3 , and dimethylarsinic acid), and the control models were blankcontrols. Arsenic exposure time varied among the studies and was categorized as ď24 h (n = 15) or >24 h (n = 12). Arsenic dose also varied among the studies, and was categorized as either low dose (n = 16) or high dose (n = 11). NF-κB signaling pathway indices (i.e., NF-κB, IκB, p-IκB, NF-κB mRNA, NF-κB activity, IKK, DNA-binding activity of NF-κB, and NF-κBp65) were examined either in normal cells or tissues (n = 9) or in inflammatory and cancer cells or tissue (n = 18).

Data Analysis
Twenty-seven articles were analyzed in Review Manager Version 5.2 (The Nordic Cochrane Centre, The Cochrane Collaboration, 2012, Portland Oregon, OR, USA) and Stata 12.0 (StataCorp., College Station, Texas, TX, USA). Significant heterogeneity was detected (p < 0.05, I 2 > 75%) and a random effects model was therefore applied for the meta-analysis. A multivariate meta-regression analysis was performed to determine the source of heterogeneity and continuous variables were estimated as standardized mean differences (SMDs) with 95% confidence intervals (CIs) between the arsenic treated groups and control groups. All reported p-values are two-sided and a significance level of 0.05 was used. For additional insight, subgroup analyses were performed based on exposure dose (≤5 μmol/L or ˃5 μmol/L in vivo; ≤5 mg/kg or ˃5 mg/kg in vitro) and exposure time (≤24 h or ˃24 h), based on the median of the indexes reported in the papers, to determine the factors associated with differences in the outcome indicators across different studies. Small-study effects were explored using funnel plots and Egger's tests and studysensitivities were assessed.

Study Characteristics
Using the search strategy described in Section 2, 573 relevant articles were identified (Figure 1), of which 27 were used for the meta-analysis based on the eligibility and exclusion criteria (Table 1). Various cell lines and animals were used as arsenic model groups in these studies, and in each study, the effect of arsenic on the NF-κB signaling pathway was assessed. The arsenic model groups were primarily cell lines and animals treated with arsenicin various forms (e.g., arsenite, As2O3, and dimethylarsinic acid), and the control models were blankcontrols. Arsenic exposure time varied among the studies and was categorized as ≤24 h (n = 15) or >24 h (n = 12). Arsenic dose also varied among the studies, and was categorized as either low dose (n = 16) or high dose (n = 11). NF-κB signaling pathway indices (i.e., NF-κB, IκB, p-IκB, NF-κB mRNA, NF-κB activity, IKK, DNA-binding activity of NF-κB, and NF-κBp65) were examined either in normal cells or tissues (n = 9) or in inflammatory and cancer cells or tissue (n = 18).   Note: n = number of experimental cells or animals group; 1 = IKK, 2 = p-IκB, 3 = IκB, 4 = NF-κB mRNA, 5 = NF-κBp65, 6 = NF-κBactivity, 7 = DNA-binding activity of NF-κB and 8 = NF-κB.

Effects of Arsenic Exposure on p-IκB
A total of six studies assessed p-IκB levels. A pooled analysis showed that p-IκB levels were 8.13-fold higher in the exposed group than in the control group (95% CI, 2.40-13.85; Z = 2.78; p = 0.005) for normal cells and tissue, with no significant heterogeneity (p = 0.65; I 2 = 0%; Figure 2). Pooled analysis furthermore showed that p-IκB levels were 3.69-fold lower in the exposed group than in the control group (95% CI,´12.59-5.21; Z = 0.81; p = 0.42) for abnormal cells and tissue, with significant heterogeneity (p = 0.007; I 2 = 76%; Figure 2).

Effects of Arsenic Exposure on p-IκB
A total of six studies assessed p-IκB levels. A pooled analysis showed that p-IκB levels were 8.13-fold higher in the exposed group than in the control group (95% CI, 2.40-13.85; Z = 2.78; p = 0.005) for normal cells and tissue, with no significant heterogeneity (p = 0.65; I 2 = 0%; Figure 2). Pooled analysis furthermore showed that p-IκB levels were 3.69-fold lower in the exposed group than in the control group (95% CI, −12.59-5.21; Z = 0.81; p = 0.42) for abnormal cells and tissue, with significant heterogeneity (p = 0.007; I 2 = 76%; Figure 2).

Effects of Arsenic Exposure on IκB
A total of nine studies assessed IκB levels. A pooled analysis showed that IκB levels were 16.19-fold lower in the exposed group than in the control group (95% CI, −27.44-−4.94; Z = 2.78; p = 0.005) for normal cells and tissue, with no significant heterogeneity (p = 0.65; I 2 = 0%; Figure 3). Pooled analysis furthermore showed that IκB levels were 0.28-fold lower in the exposed group than in the control group (95% CI, −3.14-2.57; Z = 0.19; p = 0.85) for abnormal cells and tissue, with significant heterogeneity (p < 0.0001; I 2 = 91%; Figure 3).

Effects of Arsenic Exposure on IκB
A total of nine studies assessed IκB levels. A pooled analysis showed that IκB levels were 16.19-fold lower in the exposed group than in the control group (95% CI,´27.44-´4.94; Z = 2.78; p = 0.005) for normal cells and tissue, with no significant heterogeneity (p = 0.65; I 2 = 0%; Figure 3). Pooled analysis furthermore showed that IκB levels were 0.28-fold lower in the exposed group than in the control group (95% CI,´3.14-2.57; Z = 0.19; p = 0.85) for abnormal cells and tissue, with significant heterogeneity (p < 0.0001; I 2 = 91%; Figure 3).

Effects of Arsenic Exposure on p-IκB
A total of six studies assessed p-IκB levels. A pooled analysis showed that p-IκB levels were 8.13-fold higher in the exposed group than in the control group (95% CI, 2.40-13.85; Z = 2.78; p = 0.005) for normal cells and tissue, with no significant heterogeneity (p = 0.65; I 2 = 0%; Figure 2). Pooled analysis furthermore showed that p-IκB levels were 3.69-fold lower in the exposed group than in the control group (95% CI, −12.59-5.21; Z = 0.81; p = 0.42) for abnormal cells and tissue, with significant heterogeneity (p = 0.007; I 2 = 76%; Figure 2).

Effects of Arsenic Exposure on IκB
A total of nine studies assessed IκB levels. A pooled analysis showed that IκB levels were 16.19-fold lower in the exposed group than in the control group (95% CI, −27.44-−4.94; Z = 2.78; p = 0.005) for normal cells and tissue, with no significant heterogeneity (p = 0.65; I 2 = 0%; Figure 3). Pooled analysis furthermore showed that IκB levels were 0.28-fold lower in the exposed group than in the control group (95% CI, −3.14-2.57; Z = 0.19; p = 0.85) for abnormal cells and tissue, with significant heterogeneity (p < 0.0001; I 2 = 91%; Figure 3).

Effects of Arsenic Exposure on NF-κBp65
A total of six studies assessed NF-κBp65 levels. A pooled analysis showed that NF-κBp65 levels were 0.77-fold higher in the exposed group than in the control group (95% CI, 0.13-1.42; Z = 2.34; p = 0.02) for normal cells and tissue (Figure 4), while pooled analysis showed that NF-κBp65 levels were 4.90-fold lower in the exposed group than in the control group (95% CI,´8.49-1.31; Z = 2.62; p = 0.009) for abnormal cells and tissue, with significant heterogeneity (p < 0.0001; I 2 = 90%; Figure 4).

Effects of Arsenic Exposure on NF-κBp65
A total of six studies assessed NF-κBp65 levels. A pooled analysis showed that NF-κBp65 levels were 0.77-fold higher in the exposed group than in the control group (95% CI, 0.13-1.42; Z = 2.34; p = 0.02) for normal cells and tissue (Figure 4), while pooled analysis showed that NF-κBp65 levels were 4.90-fold lower in the exposed group than in the control group (95% CI, −8.49-1.31; Z = 2.62; p = 0.009) for abnormal cells and tissue, with significant heterogeneity (p < 0.0001; I 2 = 90%; Figure 4).

Effects of Arsenic Exposure on NF-κBp65
A total of six studies assessed NF-κBp65 levels. A pooled analysis showed that NF-κBp65 levels were 0.77-fold higher in the exposed group than in the control group (95% CI, 0.13-1.42; Z = 2.34; p = 0.02) for normal cells and tissue (Figure 4), while pooled analysis showed that NF-κBp65 levels were 4.90-fold lower in the exposed group than in the control group (95% CI, −8.49-1.31; Z = 2.62; p = 0.009) for abnormal cells and tissue, with significant heterogeneity (p < 0.0001; I 2 = 90%; Figure 4).

Effects of Arsenic Exposure on IKK
A total of 11 studies assessed IKK levels. A pooled analysis showed that IKK levels were 38.11-fold higher in the exposed group than in the control group (95% CI, −38.82-115.04; Z = 0.97; p = 0.33) for normal cells and tissue, with significant heterogeneity (p < 0.0001; I 2 = 94%; Figure 6). Pooled analysis furthermore showed that IKK levels were 3.81-fold lower in the exposed group than in the control group (95% CI, −7.98-0.35; Z = 1.79; p = 0.07) for abnormal cells and tissue, with significant heterogeneity (p < 0.0001; I 2 = 91%; Figure 6).

Effects of Arsenic Exposure on NF-κB Activity
A total of four studies assessed NF-κB activity. A pooled analysis showed that NF-κB activity was 1.08-fold higher in the exposed group than in the control group (95% CI, −0.83-2.99; Z = 1.11; p = 0.27) for normal cells and tissue (Figure 7). Pooled analysis furthermore showed that NF-κB activity was 2.45-fold lower in the exposed group than in the control group (95% CI, −3.66-1.25; Z = 4.00; p < 0.0001) for abnormal cells and tissue, with no significant heterogeneity (p = 0.27; I 2 = 24%; Figure 7).

Effects of Arsenic Exposure on NF-κB Activity
A total of four studies assessed NF-κB activity. A pooled analysis showed that NF-κB activity was 1.08-fold higher in the exposed group than in the control group (95% CI,´0.83-2.99; Z = 1.11; p = 0.27) for normal cells and tissue (Figure 7). Pooled analysis furthermore showed that NF-κB activity was 2.45-fold lower in the exposed group than in the control group (95% CI,´3.66-1.25; Z = 4.00; p < 0.0001) for abnormal cells and tissue, with no significant heterogeneity (p = 0.27; I 2 = 24%; Figure 7). Health 2016, 13, 163

Effects of Arsenic Exposure on IKK
A total of 11 studies assessed IKK levels. A pooled analysis showed that IKK levels were 38.11-fold higher in the exposed group than in the control group (95% CI, −38.82-115.04; Z = 0.97; p = 0.33) for normal cells and tissue, with significant heterogeneity (p < 0.0001; I 2 = 94%; Figure 6). Pooled analysis furthermore showed that IKK levels were 3.81-fold lower in the exposed group than in the control group (95% CI, −7.98-0.35; Z = 1.79; p = 0.07) for abnormal cells and tissue, with significant heterogeneity (p < 0.0001; I 2 = 91%; Figure 6).

Effects of Arsenic Exposure on NF-κB Activity
A total of four studies assessed NF-κB activity. A pooled analysis showed that NF-κB activity was 1.08-fold higher in the exposed group than in the control group (95% CI, −0.83-2.99; Z = 1.11; p = 0.27) for normal cells and tissue (Figure 7). Pooled analysis furthermore showed that NF-κB activity was 2.45-fold lower in the exposed group than in the control group (95% CI, −3.66-1.25; Z = 4.00; p < 0.0001) for abnormal cells and tissue, with no significant heterogeneity (p = 0.27; I 2 = 24%; Figure 7).

Effects of Arsenic Exposure on DNA-Binding Activity of NF-κB
A total of six studies assessed the DNA-binding activity of NF-κB. A pooled analysis showed that the DNA-binding activity of NF-κB was 4.52-fold higher in the exposed group than in the control group (95% CI,´2.04-11.09; Z = 1.35; p = 0.18) for normal cells and tissue, with no significant heterogeneity (p = 0.07; I 2 = 62%; Figure 8). Pooled analysis furthermore showed that the DNA-binding activity of NF-κB was 9.75-fold lower in the exposed group than in the control group (95% CI,´18.66-4.54; Z = 2.15; p = 0.03) for abnormal cells and tissue, with significant heterogeneity (p = 0.0005; I 2 = 87%; Figure 8). Health 2016, 13, 163

Effects of Arsenic Exposure on DNA-Binding Activity of NF-κB
A total of six studies assessed the DNA-binding activity of NF-κB. A pooled analysis showed that the DNA-binding activity of NF-κB was 4.52-fold higher in the exposed group than in the control group (95% CI, −2.04-11.09; Z = 1.35; p = 0.18) for normal cells and tissue, with no significant heterogeneity (p = 0.07; I 2 = 62%; Figure 8). Pooled analysis furthermore showed that the DNA-binding activity of NF-κB was 9.75-fold lower in the exposed group than in the control group (95% CI, −18.66-4.54; Z = 2.15; p = 0.03) for abnormal cells and tissue, with significant heterogeneity (p = 0.0005; I 2 = 87%; Figure 8).

Effects of Arsenic Exposure on NF-κB mRNA
A total of four studies assessed NF-κB mRNA levels. A pooled analysis showed that NF-κB mRNA levels were 1.61-fold higher in the exposed group than in the control group (95% CI,´1.57-4.79; Z = 0.99; p = 0.32) for normal cells and tissue, with significant heterogeneity (p = 0.002; I 2 = 89%; Figure 9). Pooled analysis furthermore showed that NF-κB mRNA levels were 2.60-fold lower in the exposed group than in the control group (95% CI,´8.85-3.65; Z = 2.15; p = 0.41) for abnormal cells and tissue, with significant heterogeneity (p < 0.0001; I 2 = 96%; Figure 9). Health 2016, 13, 163 3.2.7. Effects of Arsenic Exposure on DNA-Binding Activity of NF-κB A total of six studies assessed the DNA-binding activity of NF-κB. A pooled analysis showed that the DNA-binding activity of NF-κB was 4.52-fold higher in the exposed group than in the control group (95% CI, −2.04-11.09; Z = 1.35; p = 0.18) for normal cells and tissue, with no significant heterogeneity (p = 0.07; I 2 = 62%; Figure 8). Pooled analysis furthermore showed that the DNA-binding activity of NF-κB was 9.75-fold lower in the exposed group than in the control group (95% CI, −18.66-4.54; Z = 2.15; p = 0.03) for abnormal cells and tissue, with significant heterogeneity (p = 0.0005; I 2 = 87%; Figure 8).

Small-Study Effect Evaluation
Visual inspection of the funnel plot and Egger's test results showed no evidence of significant small-study effects (p = 0.441; Figure 12).

Small-Study Effect Evaluation
Visual inspection of the funnel plot and Egger's test results showed no evidence of significant small-study effects (p = 0.441; Figure 12).

Sensitivity Analysis
A sensitivity analysis was conducted for NF-κB. As shown in Figure 13, the results for all studies were stabilized, and thus, no individual study was found to influence the combined results.

Meta-Regression Analysis of Arsenic Exposure Effects
A multivariate meta-regression analysis of the retrieved data revealed that the effects of arsenic exposure time (p = 0.013) and dose (p = 0.022) were significantly associated with differences in NF-κB levels.

Discussion
The findings of this meta-analysis reveal that arsenic activates the NF-κB signaling pathway in normal cells or tissues; however, the antitumor and anti-inflammatory effects of arsenic are

Sensitivity Analysis
A sensitivity analysis was conducted for NF-κB. As shown in Figure 13, the results for all studies were stabilized, and thus, no individual study was found to influence the combined results.

Sensitivity Analysis
A sensitivity analysis was conducted for NF-κB. As shown in Figure 13, the results for all studies were stabilized, and thus, no individual study was found to influence the combined results.

Meta-Regression Analysis of Arsenic Exposure Effects
A multivariate meta-regression analysis of the retrieved data revealed that the effects of arsenic exposure time (p = 0.013) and dose (p = 0.022) were significantly associated with differences in NF-κB levels.

Discussion
The findings of this meta-analysis reveal that arsenic activates the NF-κB signaling pathway in normal cells or tissues; however, the antitumor and anti-inflammatory effects of arsenic are

Meta-Regression Analysis of Arsenic Exposure Effects
A multivariate meta-regression analysis of the retrieved data revealed that the effects of arsenic exposure time (p = 0.013) and dose (p = 0.022) were significantly associated with differences in NF-κB levels.

Discussion
The findings of this meta-analysis reveal that arsenic activates the NF-κB signaling pathway in normal cells or tissues; however, the antitumor and anti-inflammatory effects of arsenic are accompanied by partial suppression of the NF-κB signaling pathway. Short arsenic exposure may promote the activation of NF-κB signaling, while long exposure may suppress the NF-κB signaling pathway. Compared with high doses, low doses of arsenic were shown to weaken NF-κB signaling pathway expression. These findings provide a theoretical foundation for understanding the contrasting toxic and therapeutic effects of arsenic.
Arsenic has been reported to damage the human body via activation of the NF-κB signaling pathway [18][19][20]. Here, we show that at high doses, the action of IKK enhances IκB phosphorylation, degrades IκB, reduces the content of IκB in the cytoplasm, and increases the DNA-binding activity of NF-κB in normal tissues or cells, thereby enhancing the expression of NF-κBp65 and NF-κB ( Figure 14A). Arsenic may therefore increase cytoplasmic IKK levels, thereby enhancing IKK phosphorylation, which, in turn, results in an increase in IKK activity, resulting in IκB phosphorylation and subsequent cytoplasmic degradation of IκB [13,14]. Elsewhere, arsenic has been shown to boost the DNA-binding activity of NF-κB and to enhance NF-κB mRNA expression, thereby promoting the expression of NF-κB [21]. These effects may result from arsenic-induced increases in cellular reactive oxygen species, which, in turn, promote NF-κB expression. Arsenic can also decrease NF-κB levels in the cytoplasm and increase the nuclear transcription of NF-κB [22]. It was recently also shown that arsenic has a therapeutic effect on tumors and inflammation, and one of the main mechanisms by which arsenic exerts such effects is by inhibiting the activation of NF-κB signaling [23][24][25][26].
Int. J. Environ. Res. Public Health 2016, 13,163 accompanied by partial suppression of the NF-κB signaling pathway. Short arsenic exposure may promote the activation of NF-κB signaling, while long exposure may suppress the NF-κB signaling pathway. Compared with high doses, low doses of arsenic were shown to weaken NF-κB signaling pathway expression. These findings provide a theoretical foundation for understanding the contrasting toxic and therapeutic effects of arsenic.
Arsenic has been reported to damage the human body via activation of the NF-κB signaling pathway [18][19][20]. Here, we show that at high doses, the action of IKK enhances IκB phosphorylation, degrades IκB, reduces the content of IκB in the cytoplasm, and increases the DNA-binding activity of NF-κB in normal tissues or cells, thereby enhancing the expression of NF-κBp65 and NF-κB ( Figure 14A). Arsenic may therefore increase cytoplasmic IKK levels, thereby enhancing IKK phosphorylation, which, in turn, results in an increase in IKK activity, resulting in IκB phosphorylation and subsequent cytoplasmic degradation of IκB [13,14]. Elsewhere, arsenic has been shown to boost the DNA-binding activity of NF-κB and to enhance NF-κB mRNA expression, thereby promoting the expression of NF-κB [21]. These effects may result from arsenic-induced increases in cellular reactive oxygen species, which, in turn, promote NF-κB expression. Arsenic can also decrease NF-κB levels in the cytoplasm and increase the nuclear transcription of NF-κB [22]. It was recently also shown that arsenic has a therapeutic effect on tumors and inflammation, and one of the main mechanisms by which arsenic exerts such effects is by inhibiting the activation of NF-κB signaling [23][24][25][26]. Arsenic has been shown to inhibit the activation of NF-κB signaling in tumors and inflammatory cells [27][28][29][30][31]. The findings of our analysis indicate that, in inflammatory or tumor Figure 14. The NF-κB signaling pathway. (A) shows that at high doses, the action of IKK enhances IκB phosphorylation, degrades IκB, reduces the content of IκB in the cytoplasm, and increases the DNA-binding activity of NF-κB in normal tissues or cells, thereby enhancing the expression of NF-κBp65 and NF-κB; (B) indicates that, in inflammatory or tumor cells, arsenic reduces cytoplasmic IKK levels, suppresses IκB phosphorylation, attenuates IκB degradation, increases cytoplasmic IκB levels, and weakens the DNA-binding activity of NF-κB, thereby decreasing the expression levels of NF-κB and NF-κBp65.
Arsenic has been shown to inhibit the activation of NF-κB signaling in tumors and inflammatory cells [27][28][29][30][31]. The findings of our analysis indicate that, in inflammatory or tumor cells, arsenic reduces cytoplasmic IKK levels, suppresses IκB phosphorylation, attenuates IκB degradation, increases cytoplasmic IκB levels, and weakens the DNA-binding activity of NF-κB, thereby decreasing the expression levels of NF-κB and NF-κBp65 ( Figure 14B). On the one hand, the inhibition of IKK activity in the presence of arsenic may be a result of arsenic acting on and affecting the lysosome, thereby diminishing IKK activity [32], reducing IκB phosphorylation, and inhibiting the degradation of IκB [33]. On the other hand, arsenic may act by reducing the NF-κB connected with DNA activity and inhibiting the translocation of NF-κB into the nucleus, thereby suppressing NF-κB mRNA and then protein expression [34][35][36]. Some researchers believe that arsenic exposure leads to elevatedreactive oxygen species levels which inhibit NF-κB activity, resulting in decreased expression of NF-κB [37][38][39].
Different arsenic concentrations and exposure times may account for the different effects of arsenic on NF-κB activity in the same cells. Huand colleagues [22] found that short (3-24 h) periods of As 2 O 3 administration (<25 µmol/L) in GM847 cells activates NF-κB signaling, while prolonged exposure (10 weeks) to arsenic (0.1~0.5 µmol/L) inhibits NF-κB expression in these same cells. Liao and colleagues [40] showed that 105 µmol/L doses of As 2 O 3 for 48 h in HEK cells significantly increased NF-κB activity, while exposure to 5 µmol/L As 2 O 3 induced significant inhibition of NF-κB activity accompanied by an increase in the apoptosis ratio. These findings are consistent with our meta-analysis results: high dose exposure and short exposure may lead to NF-κB signaling pathway activation, while low dose exposure and long exposure suppresses the activation of NF-κB signaling.
Arsenic is widely distributed in nature and may therefore have a significant impact on long-term human development. Long-term chronic arsenic exposure can affect human health, specifically by affecting the cardiovascular [41], respiratory [42], gastrointestinal [43] systems among others. In recent years, arsenic has been shown to also play important antitumor and anti-inflammatory roles in disease systems including breast cancer [44], leukemia [45], and rheumatoid arthritis [39], exhibiting curative effects in these diseases. At present, some scholars consider the NF-κB signaling pathway to be one of the important mechanisms underlying the poison effects as well as the antitumor and anti-inflammatory effects of arsenic. In the application of arsenic as a potential treatment for cancer, therefore, the dose and treatment time of choice is particularly important, considering that inappropriate dosage and exposure time may increase the extent of the damage to the body.

Conclusions
As described in Sections 3 and 4 the present study demonstrates that arsenic may cause toxic injury via NF-κB signaling pathway activation in normal cells or tissue; however, in cancers or inflammatory cells and tissues, arsenic may have certain therapeutic effects which result from suppression of the NF-κB signaling pathway. These findings contribute to the potential development of therapeutic arsenic and provide a theoretical basis for the implementation of such therapies.