ROS/TNF-α Crosstalk Triggers the Expression of IL-8 and MCP-1 in Human Monocytic THP-1 Cells via the NF-κB and ERK1/2 Mediated Signaling

IL-8/MCP-1 act as neutrophil/monocyte chemoattractants, respectively. Oxidative stress emerges as a key player in the pathophysiology of obesity. However, it remains unclear whether the TNF-α/oxidative stress interplay can trigger IL-8/MCP-1 expression and, if so, by which mechanism(s). IL-8/MCP-1 adipose expression was detected in lean, overweight, and obese individuals, 15 each, using immunohistochemistry. To detect the role of reactive oxygen species (ROS)/TNF-α synergy as a chemokine driver, THP-1 cells were stimulated with TNF-α, with/without H2O2 or hypoxia. Target gene expression was measured by qRT-PCR, proteins by flow cytometry/confocal microscopy, ROS by DCFH-DA assay, and signaling pathways by immunoblotting. IL-8/MCP-1 adipose expression was significantly higher in obese/overweight. Furthermore, IL-8/MCP-1 mRNA/protein was amplified in monocytic cells following stimulation with TNF-α in the presence of H2O2 or hypoxia (p ˂ 0.0001). Synergistic chemokine upregulation was related to the ROS levels, while pre-treatments with NAC suppressed this chemokine elevation (p ≤ 0.01). The ROS/TNF-α crosstalk involved upregulation of CHOP, ERN1, HIF1A, and NF-κB/ERK-1,2 mediated signaling. In conclusion, IL-8/MCP-1 adipose expression is elevated in obesity. Mechanistically, ROS/TNF-α crosstalk may drive expression of these chemokines in monocytic cells by inducing ER stress, HIF1A stabilization, and signaling via NF-κB/ERK-1,2. NAC had inhibitory effect on oxidative stress-driven IL-8/MCP-1 expression, which may have therapeutic significance regarding meta-inflammation.


Introduction
Chemotactic cytokines called chemokines play a role in driving the metabolic inflammation associated with obesity, which later leads to the development of type 2 diabetes (T2D), or results in metabolic syndrome. Therefore, it is important to study the factors and mechanisms of expression of obesity-related chemokines, since these are the early pathophysiologic changes that could later lead to more serious metabolic disorders, such as T2D, atherosclerosis, cardiovascular disease, hepatic steatosis, and metabolic syndrome. These small proteins act as "directional signals" which regulate the trafficking of effector leukocytes to the sites of inflammation. The human chemokine system comprises of about compared to their lean counterparts (BMI: 22.86 ± 1.89 kg/m 2 ), 15 each. Their clinicodemographic characteristics are summarized in Supplementary Table S1. To this end, as determined by immunohistochemistry (IHC), protein expression of both IL-8 ( Figure 1A,B; p < 0.0001) and MCP-1 ( Figure 1C,D; p < 0.0001) was found to be significantly upregulated in overweight and obese individuals as compared to lean individuals. As expected, a direct association was found between the protein and mRNA expression of IL-8 (r = 0.60 p = 0.004, Supplementary Figure S1A  Consistent with the IHC data, increased protein expression of IL-8 and MCP-1 was further validated by confocal microscopy ( Supplementary Figures S2 and S3, respectively). Overall, these data suggest that adipose expression of IL-8 and MCP-1 is significantly upregulated in obesity.

Oxidative Stress Triggers the Expression of IL-8 and MCP-1 in Human Monocytic THP-1 Cells
In obesity, increased proinflammatory responses and oxidative stress contribute to the development of immuno-metabolic complications [21]. We asked whether the co-exposure to oxidative stress and TNF-α could promote the expression of IL-8 and MCP-1 in human monocytic cells. To this end, THP-1 cells were stimulated with TNF-α, in the presence or absence of H 2 O 2 , which is a standard method used to induce ROS-mediated oxidative stress in cells [22,23]. The data show that, as compared to vehicle control, the relative mRNA expression of IL8 was 11-fold increased with TNF-α (p < 0.0001) and 4-fold increased with H 2 O 2 (p = 0.002) ( Figure 2A); whereas the relative mRNA expression of MCP1 was 7-fold increased with TNF-α (p < 0.0001) and 4-fold increased with H 2 O 2 (p = 0.0001) ( Figure 2B). However, co-stimulation with TNF-α and H 2 O 2 cooperatively upregulated the IL8 and MCP1 gene expression by 49-fold and 15-fold, respectively (p < 0.0001) (Figure 2A,B).
We further investigated if 1% hypoxia could also induce a similar effect in these cells. To this effect, as expected, we found that 1% hypoxia also elevated the gene expression of IL8 and MCP1 by 3-fold (p = 0.02) ( Figure 2C) and by 4-fold (p < 0.05) ( Figure 2D), respectively, as compared to normoxia. However, as expected, TNF-α stimulation under 1% hypoxia, induced significantly high gene expression of both IL8 (23-fold) and MCP1 (70-fold) as compared to normoxia control (p < 0.0001) ( Figure 2C,D). Since H 2 O 2 and hypoxia yielded comparable results, we mostly used H 2 O 2 treatments for inducing oxidative stress in the subsequent experiments since it did not require longer incubation of cells in hypoxia chambers under the complex gaseous mixture containing 1% O 2 , 5% CO 2 , and 94% N 2 .
Moreover, the elevated protein expression of IL-8 and MCP-1 following co-stimulation of THP-1 cells with TNF-α and H 2 O 2 , as compared to stimulation with either TNF-α or H 2 O 2 , was further confirmed by using confocal microscopy. In corroboration with flow cytometry data, the confocal microscopy also revealed significantly increased expression of IL-8 ( Figure 4A,B; p < 0.0001) and MCP-1 ( Figure 4C,D; p = 0.031) in cells that were co-stimulated with TNF-α and H 2 O 2 , as compared to stimulation with TNF-α alone. As expected, strong agreement was found between the transcripts (fold change) and protein expression (MFI) of both IL-8 (r = 0.93, p < 0.0001) and MCP-1 (r = 0.96, p < 0.0001) in THP-1 monocytic cells (Supplementary Figure S4). Altogether, these data support that oxidative stress drives the expression of IL-8 and MCP-1 at both the transcriptional and translational levels in human monocytic THP-1 cells.
as compared to normoxia control (p < 0.0001) ( Figure 2C,D). Since H yielded comparable results, we mostly used H2O2 treatments for inducin in the subsequent experiments since it did not require longer incubation o chambers under the complex gaseous mixture containing 1% O2, 5% CO Figure 2. IL8 and MCP1 gene expression in human monocytic THP-1 cells stim and/or oxidative stress. THP-1 cells were stimulated, in triplicate, with TNF-α H2O2 (10 mM) for 24 h while control cells were treated with vehicle only. Likewis also stimulated, in triplicate, with TNF-α under oxidative stress by 1% hypox collected from cells for measuring gene expression of IL8 and MCP1 using qRT IL8 and MCP1 gene expression in human monocytic THP-1 cells stimulated with TNF-α and/or oxidative stress. THP-1 cells were stimulated, in triplicate, with TNF-α (10 ng/mL) and/or H 2 O 2 (10 mM) for 24 h while control cells were treated with vehicle only. Likewise, THP-1 cells were also stimulated, in triplicate, with TNF-α under oxidative stress by 1% hypoxia. Total RNA was collected from cells for measuring gene expression of IL8 and MCP1 using qRT-PCR as detailed in methods. One-way ANOVA (Tukey's multiple comparisons test) was used to calculate group differences and p-values less than 0.05 were considered as significant. The representative data (mean ± SEM) obtained from three independent experiments with similar results show the upregulated transcripts of (A) IL8 and (B) MCP1 in cells co-stimulated with TNF-α and H 2 O 2 , compared to TNF-α stimulation alone (p ≤ 0.0001). Similarly, the representative data (mean ± SEM) from three independent experiments with similar results show increased transcripts expression of (C) IL8 and (D) MCP1 in the cells stimulated with TNF-α under 1% hypoxia, compared to TNF-α stimulation under normoxia (20% O 2 ) (p < 0.0001).
shown by flow cytometry images for IL-8 ( Figure 3A-C) and MCP-1 ( Figure 3D-F co-stimulation with TNF-α and H2O2 induced significantly higher protein express IL-8 (MFI: 14194 ± 256, p < 0.0001) ( Figure 3G) and MCP-1 (MFI: 12,650 ± 176, p < 0 ( Figure 3H) comparing, respectively, to cell stimulations with either TNF-α (MFIIL-± 132.4, MFIMCP-1: 3129 ± 136) or H2O2 (MFIIL-8: 6601 ± 151, MFIMCP-1: 3098 ± 201). . THP-1 cells were stimulated, in triplicate, with TNF-α (10 ng/mL) H2O2 (10 mM) for 24 h, while control cells were treated with vehicle only. IL-8 and MCP-1 p expression was measured using flow cytometry as described in materials and methods. On ANOVA (Tukey's multiple comparisons test) was used to calculate group differences and pless than 0.05 were considered as significant. The representative histograms from three indep experiments with similar results show (A-C) expression of IL-8 and (D-F) expression of MC cells stimulated with TNF-α and/or H2O2 as compared to control. Quantitative analysis of th (mean ± SEM) shows elevated protein expression expressed as mean fluorescence intensity (M THP-1 cells were stimulated, in triplicate, with TNF-α (10 ng/mL) and/or H 2 O 2 (10 mM) for 24 h, while control cells were treated with vehicle only. IL-8 and MCP-1 protein expression was measured using flow cytometry as described in materials and methods. One-way ANOVA (Tukey's multiple comparisons test) was used to calculate group differences and p-values less than 0.05 were considered as significant. The representative histograms from three independent experiments with similar results show (A-C) expression of IL-8 and (D-F) expression of MCP-1 in cells stimulated with TNF-α and/or H 2 O 2 as compared to control. Quantitative analysis of the data (mean ± SEM) shows elevated protein expression expressed as mean fluorescence intensity (MFI) of (G) IL-8 and (H) MCP-1 in the cells co-stimulated with TNF-α and H 2 O 2 as compared to those stimulated with TNF-α or H 2 O 2 alone (p < 0.0001). pected, strong agreement was found between the transcripts (fold change) and protein expression (MFI) of both IL-8 (r = 0.93, p < 0.0001) and MCP-1 (r = 0.96, p < 0.0001) in THP-1 monocytic cells (Supplementary Figure S4). Altogether, these data support that oxidative stress drives the expression of IL-8 and MCP-1 at both the transcriptional and translational levels in human monocytic THP-1 cells. Figure 4. IL-8 and MCP-1 protein expression were determined by confocal microscopy. THP-1 cells were stimulated, in triplicate, with TNF-α (10 ng/mL) and/or H2O2 (10 mM) for 24 h, while control cells were treated with vehicle only. IL-8 and MCP-1 protein expression (fluorescence assessed as integrated density) was determined by confocal microscopy as detailed in materials and methods. For IL-8 and MCP-1 protein quantification, 10 random images (40× magnification) were captured for each treatment. The fluorescent channels were split using image J software (NIH, USA). The fluorescence intensity of IL-8 and MCP-1 was detected in 10 fields of each image and mean integrated density was calculated for each image. One-way ANOVA (Tukey's multiple comparisons test) was used to calculate group differences and p-values less than 0.05 were considered as signifi- . IL-8 and MCP-1 protein expression were determined by confocal microscopy. THP-1 cells were stimulated, in triplicate, with TNF-α (10 ng/mL) and/or H 2 O 2 (10 mM) for 24 h, while control cells were treated with vehicle only. IL-8 and MCP-1 protein expression (fluorescence assessed as integrated density) was determined by confocal microscopy as detailed in materials and methods. For IL-8 and MCP-1 protein quantification, 10 random images (40× magnification) were captured for each treatment. The fluorescent channels were split using image J software (NIH, USA). The fluorescence intensity of IL-8 and MCP-1 was detected in 10 fields of each image and mean integrated density was calculated for each image. One-way ANOVA (Tukey's multiple comparisons test) was used to calculate group differences and p-values less than 0.05 were considered as significant. The representative photomicrography images (40× magnification) obtained from three independent determinations with similar results show elevated expression of (A,B) IL-8 (p < 0.0001) and (C,D) MCP-1 (p = 0.031) in the cells co-stimulated with TNF-α and H 2 O 2 as compared with cells stimulated with TNF-α or H 2 O 2 alone.
cant. The representative photomicrography images (40× magnification) obtained from three pendent determinations with similar results show elevated expression of (A,B) IL-8 (p < 0.0001 (C,D) MCP-1 (p = 0.031) in the cells co-stimulated with TNF-α and H2O2 as compared with stimulated with TNF-α or H2O2 alone.

TNF-α/H2O2 Cooperativity Triggers ROS, and the ROS Scavenging Suppresses IL-8 an MCP-1 Expression
Next, we determined whether the TNF-α and H2O2 induced chemokine expre was pivoted to an increase in the intracellular ROS expression, which was measure dichloro-dihydro-fluorescein diacetate (DCFH-DA) assay. To this end, flow cytometr ages ( Figure 5 A-C) represent a significant increase of the intracelular ROS in the cells were co-stimulated with TNF-α and H2O2 (MFI: 31,779 ± 743.60), as compared to t stimulated with either TNF-α (MFI: 1342 ± 51.19) or H2O2 (MFI: 1650 ± 25.32) (p < 0.0 ( Figure 5D  TNF-α/H2O2 cooperativity promotes the intracellular ROS expression. THP-1 cells stimulated, in triplicate, with TNF-α (10 ng/mL) and/or H2O2 (10 mM), while control cells treated with vehicle only. ROS expression was determined by using DCFH-DA assay and DC pression, shown as mean fluorescence intensity (MFI), was measured by flow cytometry a scribed in materials and methods. One-way ANOVA (Tukey's multiple comparisons test) was to calculate group differences and p-values less than 0.05 were considered as significant. The r sentative histograms from three independent experiments with similar results show (A-C) com ative ROS expression in cells stimulated with TNF-α and/or H2O2 versus control. (D) Quanti analysis of flow cytometry data (MFI: mean ± SEM) shows significantly elevated ROS express THP-1 cells co-stimulated with TNF-α and H2O2, compared to stimulation with either TNF H2O2 (p ˂ 0.0001).
To further see if ROS scavenging could suppress the IL8 and MCP1 gene expres THP-1 cells were treated with an antioxidant N-acetyl cysteine (NAC) or a ROS-inhi curcumin, before stimulation with TNF-α and/or H2O2. To this end, our data show Figure 5. TNF-α/H 2 O 2 cooperativity promotes the intracellular ROS expression. THP-1 cells were stimulated, in triplicate, with TNF-α (10 ng/mL) and/or H 2 O 2 (10 mM), while control cells were treated with vehicle only. ROS expression was determined by using DCFH-DA assay and DCF expression, shown as mean fluorescence intensity (MFI), was measured by flow cytometry as described in materials and methods. One-way ANOVA (Tukey's multiple comparisons test) was used to calculate group differences and p-values less than 0.05 were considered as significant. The representative histograms from three independent experiments with similar results show (A-C) comparative ROS expression in cells stimulated with TNF-α and/or H 2 O 2 versus control. (D) Quantitative analysis of flow cytometry data (MFI: mean ± SEM) shows significantly elevated ROS expression in THP-1 cells co-stimulated with TNF-α and H 2 O 2 , compared to stimulation with either TNF-α or H 2 O 2 (p < 0.0001).
To further see if ROS scavenging could suppress the IL8 and MCP1 gene expression, THP-1 cells were treated with an antioxidant N-acetyl cysteine (NAC) or a ROS-inhibitor curcumin, before stimulation with TNF-α and/or H 2 O 2 . To this end, our data show that IL8 gene expression induced by TNF-α and H 2 O 2 co-stimulation was suppressed significantly by NAC (7-fold, p < 0.0001), but not by curcumin (120-fold, p = 0.52), as compared to vehicle control (127-fold) ( Figure 6A). However, MCP1 gene expression induced by TNF-α and H 2 O 2 co-stimulation was suppressed significantly (p ≤ 0.01) by both NAC (16-fold) and curcumin (23-fold) as compared to vehicle control (56-fold) ( Figure 6B). Collectively, these findings present the evidence that ROS-driven IL8 and MCP1 upregulation in THP-1 cells could be effectively counteracted by NAC.
Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEW 10 IL8 gene expression induced by TNF-α and H2O2 co-stimulation was suppressed sig cantly by NAC (7-fold, p < 0.0001), but not by curcumin (120-fold, p = 0.52), as comp to vehicle control (127-fold) ( Figure 6A). However, MCP1 gene expression induce TNF-α and H2O2 co-stimulation was suppressed significantly (p ≤ 0.01) by both NAC fold) and curcumin (23-fold) as compared to vehicle control (56-fold) ( Figure 6B). Co tively, these findings present the evidence that ROS-driven IL8 and MCP1 upregul in THP-1 cells could be effectively counteracted by NAC. Figure 6. ROS inhibition suppresses the IL8 and MCP1 transcripts expression in monocytic cel stimulated with TNF-α and H2O2. THP-1 cells were stimulated, in triplicate, with TNF-α (10 ng and/or H2O2 (10 mM) while control cells were treated with vehicle only. In ROS inhibition as cells were preincubated for 30 min with either NAC (1 mM) or curcumin (10 μM) before stimul with TNF-α and/or H2O2. Total RNA was extracted and IL8/MCP1 gene expression was assess qRT-PCR as described in methods. Two-way ANOVA was used to calculate group difference p-values less than 0.05 were considered as significant. The representative data (mean ± SEM tained from three independent experiments with similar results show effective suppression o IL8 mRNA by NAC (p ˂ 0.0001) and (B) MCP1 mRNA by both NAC and curcumin (p ≤ 0.01).

TNF-α/H2O2 Cooperativity Drives the ER Stress and Stabilizes Hypoxia-Inducible Fact (HIF)-1α
We asked if oxidative stress driving IL-8 and MCP-1 chemokine expression in m ocytic cells induced the ER stress and stabilized expression of the oxygen sensor HIF To this effect, the data show that TNF-α/H2O2 cooperativity induced significantly hi gene expression of both CHOP (9-fold) and ERN1 (6-fold) in THP-1 cells compared to stimulation with H2O2 alone (CHOP: 6-fold, p = 0.007; ERN1: 4-fold, p = 0.01) (Figure A direct regulatory role of ROS in HIF1A expression and stabilization has been repo before [24,25]; however, it remains controversial [26]. Therefore, we next asked wh ROS induction by H2O2 via a non-canonical mechanism or by 1% hypoxia via a cano mechanism could both induce the HIF1A gene expression in THP-1 cells. To this eff significant increase in HIF1A expression was observed after co-stimulation with TN and H2O2 (˃2-fold), as compared to stimulation with either TNF-α (˂2-fold) or H2O2 fold) (p ≤ 0.004, Figure 7B). Similarly, and as expected, stimulation with TNF-α unde hypoxia significantly upregulated HIF1A gene expression (40-fold), as compared to ulation with TNF-α under normoxia (˂2-fold) or exposure to 1% hypoxia alone (˂2-(p < 0.0001, Figure 7C). Taken together, these data indicate that THP-1 cell stimul with TNF-α in the presence of H2O2 or under 1% hypoxia induces the ER stress and H stabilization. . ROS inhibition suppresses the IL8 and MCP1 transcripts expression in monocytic cells costimulated with TNF-α and H 2 O 2 . THP-1 cells were stimulated, in triplicate, with TNF-α (10 ng/mL) and/or H 2 O 2 (10 mM) while control cells were treated with vehicle only. In ROS inhibition assays, cells were preincubated for 30 min with either NAC (1 mM) or curcumin (10 µM) before stimulation with TNF-α and/or H 2 O 2 . Total RNA was extracted and IL8/MCP1 gene expression was assessed by qRT-PCR as described in methods. Two-way ANOVA was used to calculate group differences and p-values less than 0.05 were considered as significant. The representative data (mean ± SEM) obtained from three independent experiments with similar results show effective suppression of (A) IL8 mRNA by NAC (p < 0.0001) and (B) MCP1 mRNA by both NAC and curcumin (p ≤ 0.01).

TNF-α/H 2 O 2 Cooperativity Drives the ER Stress and Stabilizes Hypoxia-Inducible Factor (HIF)-1α
We asked if oxidative stress driving IL-8 and MCP-1 chemokine expression in monocytic cells induced the ER stress and stabilized expression of the oxygen sensor HIF-1α. To this effect, the data show that TNF-α/H 2 O 2 cooperativity induced significantly higher gene expression of both CHOP (9-fold) and ERN1 (6-fold) in THP-1 cells compared to cell stimulation with H 2 O 2 alone (CHOP: 6-fold, p = 0.007; ERN1: 4-fold, p = 0.01) ( Figure 7A). A direct regulatory role of ROS in HIF1A expression and stabilization has been reported before [24,25]; however, it remains controversial [26]. Therefore, we next asked whether ROS induction by H 2 O 2 via a non-canonical mechanism or by 1% hypoxia via a canonical mechanism could both induce the HIF1A gene expression in THP-1 cells. To this effect, a significant increase in HIF1A expression was observed after co-stimulation with TNF-α and H 2 O 2 (>2-fold), as compared to stimulation with either TNF-α (<2-fold) or Figure 7B). Similarly, and as expected, stimulation with TNF-α under 1% hypoxia significantly upregulated HIF1A gene expression (40-fold), as compared to stimulation with TNF-α under normoxia (<2-fold) or exposure to 1% hypoxia alone (<2-fold) (p < 0.0001, Figure 7C). Taken together, these data indicate that THP-1 cell stimulation with TNF-α in the presence of H 2 O 2 or under 1% hypoxia induces the ER stress and HIF1A stabilization. . TNF-α/ROS cooperativity induces the ER stress and HIF1A stabilization. THP-1 cells were stimulated, in triplicate, with TNF-α (10 ng/mL), H2O2 (10 mM), and/or 1% hypoxia, while contro cells were treated with vehicle only. Total RNA was purified for assessing gene expression of ER stress markers (CHOP and ERN1) and oxygen sensor (HIF1A) using qRT-PCR as detailed in meth odology. As appropriate, one-way ANOVA (Tukey's multiple comparisons test) or two-way ANOVA was used to calculate group differences and p-values less than 0.05 were considered as significant. The representative data (mean ± SEM) from two independent experiments with simila results show increased expression of: (A) CHOP and ERN1 in cells co-stimulated with TNF-α and H2O2 compared to cells treated with H2O2 alone (p ≤ 0.01); (B) HIF1A in cells co-stimulated with TNF-α and H2O2 compared to treatments with TNF-α or H2O2 alone (p ≤ 0.004); and (C) HIF1A in cells stimulated with TNF-α under hypoxia as compared to TNF-α treatment under normoxia o induction by hypoxia alone (p < 0.0001).
We further demonstrate by flow cytometry that consistent with the Western blot data . TNF-α/ROS cooperativity induces the ER stress and HIF1A stabilization. THP-1 cells were stimulated, in triplicate, with TNF-α (10 ng/mL), H 2 O 2 (10 mM), and/or 1% hypoxia, while control cells were treated with vehicle only. Total RNA was purified for assessing gene expression of ER stress markers (CHOP and ERN1) and oxygen sensor (HIF1A) using qRT-PCR as detailed in methodology. As appropriate, one-way ANOVA (Tukey's multiple comparisons test) or two-way ANOVA was used to calculate group differences and p-values less than 0.05 were considered as significant. The representative data (mean ± SEM) from two independent experiments with similar results show increased expression of:

Discussion
The mechanisms addressing inflammatory gene reprogramming under oxidative stress remain poorly understood. Chronic low-grade inflammation called metabolic inflammation and adipose oxidative stress are the hallmarks of obesity. Increasing evidence points to a potential link between inflammation and oxidative stress in metabolic conditions. In this study, we present the evidence, for the first time to our knowledge, that ROS/TNF-α crosstalk drives the expression of IL-8 and MCP-1 in monocytic cells.
First, we present the clinical evidence that α/β chemokines (IL-8 and MCP-1) adipose expression is significantly high in overweight and obese individuals as compared to their lean counterparts. In obesity, cellular changes in the adipose compartment occur, which include M1 polarization and colonization of adipose tissue by proinflammatory M1 macrophages. These effectors secrete the proinflammatory cytokines/ chemokines and adipokines, which are pivotal to sustaining an inflammatory milieu and inducing insulin resistance [27]. Consistent with our findings of the elevated IL-8/MCP-1 expression in obese states and its association with the systemic inflammation (CRP), Kim et al. found the increased circulating levels of IL-8 and MCP-1 and their positive association with obesityrelated parameters including BMI, HOMA, and CRP [14].
Since the increased TNF-α expression and oxidative stress are two primal level pathophysiological triggers in obesity [28][29][30][31], it was intriguing to further explore whether these two agents could modulate IL-8 and MCP-1 expression in monocytic cells. Using THP-1 monocytic cells experimental model, we next showed that ROS/TNF-α crosstalk promotes the expression of IL-8 and MCP-1 in monocytic cells. In this model, the cells were treated with H2O2 to induce oxidative stress. This agent is commonly employed to study the redox-regulated processes as it is relatively stable intracellularly as compared to other ROS [32]. IL-8 and MCP-1 mainly act as neutrophil and monocyte chemoattractants, respectively. As recruited to site(s) of inflammation, the innate phagocytes such as

Discussion
The mechanisms addressing inflammatory gene reprogramming under oxidative stress remain poorly understood. Chronic low-grade inflammation called metabolic inflammation and adipose oxidative stress are the hallmarks of obesity. Increasing evidence points to a potential link between inflammation and oxidative stress in metabolic conditions. In this study, we present the evidence, for the first time to our knowledge, that ROS/TNF-α crosstalk drives the expression of IL-8 and MCP-1 in monocytic cells.
First, we present the clinical evidence that α/β chemokines (IL-8 and MCP-1) adipose expression is significantly high in overweight and obese individuals as compared to their lean counterparts. In obesity, cellular changes in the adipose compartment occur, which include M1 polarization and colonization of adipose tissue by proinflammatory M1 macrophages. These effectors secrete the proinflammatory cytokines/ chemokines and adipokines, which are pivotal to sustaining an inflammatory milieu and inducing insulin resistance [27]. Consistent with our findings of the elevated IL-8/MCP-1 expression in obese states and its association with the systemic inflammation (CRP), Kim et al. found the increased circulating levels of IL-8 and MCP-1 and their positive association with obesity-related parameters including BMI, HOMA, and CRP [14].
Since the increased TNF-α expression and oxidative stress are two primal level pathophysiological triggers in obesity [28][29][30][31], it was intriguing to further explore whether these two agents could modulate IL-8 and MCP-1 expression in monocytic cells. Using THP-1 monocytic cells experimental model, we next showed that ROS/TNF-α crosstalk promotes the expression of IL-8 and MCP-1 in monocytic cells. In this model, the cells were treated with H 2 O 2 to induce oxidative stress. This agent is commonly employed to study the redox-regulated processes as it is relatively stable intracellularly as compared to other ROS [32]. IL-8 and MCP-1 mainly act as neutrophil and monocyte chemoattractants, respec-tively. As recruited to site(s) of inflammation, the innate phagocytes such as neutrophils and monocytes/macrophages readily produce the effector molecules such as cytokines, chemokines, granular proteins, and oxidant molecules. Our data showing IL-8 and MCP-1 expression in monocytic cells after stimulation with TNF-α or H 2 O 2 are partially corroborated by other reports reporting expression of these chemokines in other perspectives, such as IL-8 production and regulation in human adipose tissue shown by Bruun et al. and TNFα-induced MCP1 gene expression in endothelial cells reported by Chen et al. [33,34].
Importantly, in regard to the novelty aspects of the study, we report the cooperativity between the proinflammatory (TNF-α) and oxidative stress (H 2 O 2 ) stimuli, leading to the amplification of these two chemokines in THP-1 cells. We observed a similar amplification pattern of IL-8 and MCP-1 co-expression as the oxidative stress was induced by incubating cells under 1% hypoxia. Nonetheless, comparing levels of transcripts expression, TNFα/H 2 O 2 interaction induced more IL8 and TNF-α/hypoxia interaction induced more MCP1 expression in THP-1 cells. This differential expression of IL8 and MCP1 in monocytic cells by TNF-α, in the presence of ROS induction by non-canonical and canonical mechanisms, is an interesting observation and merits further studies involving different cell types and proinflammatory agents other than TNF-α.
Our data show that THP-1 cell co-stimulation with TNF-α and H 2 O 2 induces increased intracellular ROS synthesis as compared to stimulation with either TNF-α or H 2 O 2 . Indeed, the inflammatory and oxidative stress processes are inter-twined. Upon activation, many immune cells generate free radicals, which, in turn, promote the ROS synthesis and induce inflammation [35]. TNF-α binding and signaling through its receptor activates the inflammatory signaling pathways and ROS synthesis [36]. TNF-α promotes ROS by facilitating electron transfer to oxygen and thereby generating superoxide anions [37]. In monocytes, several sources may contribute ROS, such as mitochondria, NADPH, and xanthine oxidase [38]. To maintain the cellular redox homeostasis, H 2 O 2 , which is the main component of ROS, is normally scavenged by antioxidant enzymes such as superoxide dismutase, catalase, and glutathione peroxidase. However, during in vitro stimulation, exogenous H 2 O 2 diffuses through cell membranes in a process regulated by membrane lipids and channel proteins [39], rendering ROS synthesis through the Fenton reaction [40]. ROS, in turn, induces the expression of a wide variety of molecules including inflammatory mediators, adhesion molecules, and growth factors through activation of the redox-sensitive NF-κB/NOX pathways [41]. Our results showing ROS elevation by TNF-α or H 2 O 2 are in line with other reports [32,42].
Moreover, we found strong associations of ROS with the induction of IL8 (r = 0.98) and MCP1 (r = 0.92) gene expression (p < 0.0001). To verify the ROS-dependence of the expression of both these chemokines, cells were treated with two ROS scavengers or inhibitors including NAC and curcumin before exposure to TNF-α and/or H 2 O 2 . The cooperative expression of IL8 was effectively suppressed by NAC and that of MCP1 was suppressed by both NAC and curcumin. NAC is a prodrug to L-cysteine, which is a precursor of GSH, and it acts as a free radical scavenger by replenishing protective levels of GSH and by producing sulfhydryl groups that directly eliminate ROS [43]. Curcumin is the primary active agent in turmeric, Curcuma longa Linn, and is known to have phytonutrient and bioprotective properties [44]. Curcumin inhibits ROS by inhibiting protein kinase C activity and Ca 2+ influx [45]. Our data showing effective suppression of IL8 and MCP1 by NAC and/or curcumin are consistent with the available literature [46][47][48][49].
Next, our data show that H 2 O 2 treatment of THP-1 cells leads to the increased expression of ER stress markers CHOP and ERN1, which is further enhanced manifold following co-stimulation with TNF-α and H 2 O 2 . ER is the primary site for protein folding and transportation, and it plays a key role in the regulation of intracellular Ca 2+ and the synthesis of secretory proteins. Unfolded protein accumulation and intracellular Ca 2+ depletion cause ER stress, resulting in ROS accumulation, mitochondrial dysfunction, and proinflammatory cytokine expression [50]. Consistent with our results showing the increased expression of CHOP/ERN1 and IL8, Vij et al. showed that CHOP regulated IL-8 promoter activity and cyclooxygenase (COX)-2 inhibition led to CHOP and IL8 suppression [51]. Interestingly, Kodama et al. [52] demonstrated that the increase in CHOP expression reduced the MCP1 promoter activity induced by PI3-kinase, while the overexpression of CHOP alone upregulated MCP1 promoter activity in a dose-dependent manner, the latter corroborating the increased CHOP and MCP-1 expression in our experimental model. ERN1, also called IRE1α, is the most prominent and highly conserved ER stress sensing protein and immune cells with strong secretory function have the increased ERN1 activity [53]. In line with our findings of the increased ERN1 and MCP1 expression in THP-1 cells, Tufanli et al. showed that ERN1 inhibition led to downregulation of cytokines/chemokines genes in monocytes/macrophages [54]. Our results showing elevated ERN1 and IL8 expression in monocytic cells are in line with the study reporting IRE1-dependent proinflammatory signaling, leading to increased expression of proinflammatory chemokines [55].
HIF-1 is an oxygen-regulated transcriptional activator that coordinates gene expression during hypoxia. In addition to upregulation by hypoxia, our data show the nonhypoxic upregulation of HIF1A by H 2 O 2 , which is in partial agreement with another report demonstrating HIF-1α stabilization by exogenous H 2 O 2 [56]. We also found that TNFα stimulation under hypoxia was a much potent trigger of HIF1A, which conforms in principle with the report linking HIF-1α as a key molecule with the TNF-α expression in macrophages under hypoxia [57]. Our findings of HIF1A stabilization together with IL-8 and MCP-1 expression under hypoxic stress are in line, at least partially, with other reports [58,59].
Next, we show that expression of IL-8 and MCP-1 in THP-1 monocytic cells costimulated with TNF-α and H 2 O 2 involves increased phosphorylation of NF-κB and ERK1/2. As expected, increased NF-κB phosphorylation in cells was paralleled with the IκBα degradation as well as an increase in the phospho-IκBα/ total-IκBα ratio. These observations are in agreement with other studies of chemokine expression in different cell types or their induction by different stress stimuli [60][61][62]. Elliott et al. showed that NF-κB binding site was essential for the up-regulation of IL8 gene expression in different uterine cell types [63]. Other studies also support NF-κB binding to the IL-8 or MCP-1 promoter [64][65][66][67]. Of note, we also found that IL-8 and MCP-1 expression was significantly diminished as the cells were pre-incubated with NDGA or PD98059 to block signaling via the NF-κB or ERK-1/2, respectively, which indicates that these pathways may be critical to the oxidative stress-induced expression of IL-8 and MCP-1 in monocytic cells.
Nonetheless, our study involves certain caveats that warrant caution while interpreting the results. First, the experimental data were obtained using THP-1 monocytic cell line model. Second, the clinical data represent the subcutaneous adipose tissue changes and, therefore, further studies will be required to validate these findings in primary monocytes/macrophages, as well as in visceral adipocytes, together with determining changes in proinflammatory cytokines/chemokines other than IL-8 and MCP-1.
In conclusion, first, we present the clinical evidence that adipose expression of IL-8 and MCP-1 is elevated in individuals with obesity. Next, the experimental in vitro data support that ROS/TNF-α crosstalk triggers the expression of IL-8 and MCP-1 in THP-1 monocytic cells through the mechanisms involving ER stress, HIF1A stabilization, and NF-κB/ERK1-2 mediated signaling. Importantly, we also show that NAC has an inhibitory effect on oxidative stress-driven IL-8 and MCP-1 expression, which may have therapeutic significance in meta-inflammation.

Participants, Anthropometry, and Collection of Subcutaneous Adipose Tissue Samples
To determine the adipose tissue expression of IL-8 and MCP-1, 45 individuals were recruited in this study, classified based on their BMI (ratio of body weight in kg to height in m 2 ) [68] as lean (22.86  Institute, Kuwait (Protocol #: RA 2010-003); written informed consent was obtained from each participant. Waist circumference was measured using constant tension tape; height was measured using portable inflexible height measuring bars and weight was measured by using calibrated portable electronic weighing scales.
Adipose tissue biopsies (~0.5 g) were collected from the abdominal subcutaneous fat pad lateral to the umbilicus using the standard sterile technique. Briefly, periumbilical area was disinfected using ethanol swab and locally anesthetized using 2% lidocaine (2 mL, Fresenius Kabi, LLC, Lake Zurich, IL, USA). Fat tissue was collected making a small 0.5 cm superficial skin incision, incised into smaller pieces (~50-100 mg), put in RNAlater (Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany) and stored at −80 • C until use.

Immunohistochemistry (IHC)
Paraffin-embedded adipose tissue sections (4 µm thick) were deparaffinized in xylene and rehydrated using descending grades of ethanol (100%, 95%, and 75%) to water. Antigen retrieval was carried out using target retrieval solution (pH 6.0; Dako, Glostrup, Denmark) by pressure cooker boiling for 8 min and cooling for 15 min. After PBS washing, endogenous peroxidase activity was blocked with 3% H 2 O 2 for 30 min and non-specific antibody binding was blocked with 5% nonfat milk (1 h), followed by 1% bovine serum albumin (BSA) solution (1 h). The samples were incubated overnight at room temperature using primary rabbit polyclonal antibodies against IL-8 (1:200 dilution, Abcam ® ab106350, pH 6.0, Cambridge, MA, USA) and MCP-1 (1:400 dilution, Abcam ® ab9669, pH 6.0). After washing with PBS (0.5% Tween), samples were incubated for 1 h with secondary, horseradish peroxidase (HRP)-conjugated goat anti-rabbit antibody (EnVision Kit, Dako, Glostrup, Denmark) and color was developed using chromogenic 3,3 -diaminobenzidine (DAB) substrate. Samples were washed in running tap water, lightly counterstained with Harris hematoxylin, dehydrated using ascending grades of ethanol (75%, 95%, and 100%), cleared in xylene, and mounted in dibutylphthalate xylene (DPX). For analysis, digital photomicrographs of adipose tissue sections [20×; PanoramicScan II, 3DHistech, Hungary. URL: https://www.3dhistech.com/products-and-software/hardware/pannoramicdigital-slide-scanners/pannoramic-scan-2/ (accessed on 10 August 2021)] were used to quantify the staining in ten different regions and assess the regional heterogeneity in the tissue samples. The regions were outlined using Aperio ImageScope software [Aperio Vista, CA, USA. URL: https://aperio-imagescope.software.informer.com/9.0/ (accessed on 10 August 2021)]. Aperio-positive pixel count algorithm (version 9) integrated into Imagescope Software was used to quantify the intensity of specific staining in the region. The number of positive pixels was normalized to the number of total pixels (positive and negative) to account for variations in the size of the region sampled. Color and intensity thresholds were set to detect the immuno-staining as positive and the background as negative pixels. Once set, all slides were analyzed using the same parameters. The resultant color markup of the analysis was confirmed for each slide.
For culturing cells under hypoxia (1% O 2 ), 10 6 cells per mL were seeded in triplicate wells of 12-well culture plates and treated with TNF-α (10 ng/mL) in designated wells. To induce hypoxic stress, plates were kept in a hypoxia chamber (StemCell Technologies) and purged with a gaseous mixture containing 1% O 2 , 5% CO 2 , and 94% N 2 for 20 min (at a flow rate of 10 L/min) before sealing. The sealed chamber was placed in an incubator at 37 • C for 36 h. For normoxia control in parallel, similarly prepared culture plates were incubated at 37 • C under 5% CO 2 (i.e., in 20% O 2 ) for 36 h. Cells were collected, harvested by centrifugation, and lysed in RLT buffer (cat# 1015762, Qiagen, GmbH, Germany) for total RNA extraction using RNeasy kit (Qiagen, Valencia, CA, USA), following the manufacturers' instructions.
In assays involving pathway inhibitors, 10 6 cells per mL in triplicate wells were preincubated for 1 h with NF-κB inhibitor NDGA (30 µM) and ERK1/2 inhibitor PD98059 (10 µM), followed by stimulation with TNF-α (10 ng/mL) for 10 min and/or H 2 O 2 (10 mM) for 30 min. Cell lysates were resolved by 12% SDS-PAGE and blots were probed with primary and secondary antibodies for detection of phosphorylated and total NF-κB (65 KDa) and ERK1/2 (42, 44 KDa) by Western blotting. In similar experiments, THP-1 cells were pre-incubated with the pathway inhibitors as above, followed by stimulation with TNF-α (10 ng/mL) and/or H 2 O 2 (10 mM) for 24 h, while controls were treated with vehicle only. IL-8 and MCP-1 protein expression (MFI) was measured using flow cytometry in cells with and without inhibitor treatment.

Quantitative, Real-Time Reverse-Transcription Polymerase Chain Reaction (qRT-PCR)
Total RNA was extracted using RNeasy kit (Qiagen, Valencia, CA, USA) and following the manufacturer's instructions. RNA was quantified using Epoch™ Spectrophotometer System (BioTek, Winooski, USA). Each RNA sample (1 µg) was reverse transcribed into cDNA using random hexamer primers and TaqMan reverse transcription reagents (High-Capacity cDNA Reverse Transcription Kit; Applied Biosystems, CA, USA). For realtime RT-PCR, each cDNA sample (50 ng) was amplified using TaqMan ® Gene Expression MasterMix (Applied Biosystems, CA, USA) and target gene-specific TaqMan Gene Expression Assay (Applied Biosystems, CA, USA) products including IL-8 (Hs00174103_m1), MCP-1 (Hs00234140_m1), CHOP (Hs00358796_g1), ERN1 (Hs00980095_m1) and GAPDH (Hs03929097_g1) containing forward and reverse gene-specific primers and a target-specific TaqMan ® MGB probe labeled with FAM dye at 5 -end and NFQ-MGB at 3 -end of the probe, with 40 cycles of PCR amplification using TaqMan ® Gene Expression Master Mix (Applied Biosystems, Foster city, CA, USA) in 7500 Fast Real-Time PCR System (Applied Biosystems, CA, USA). Each cycle included denaturation (95 • C for 15 s) and annealing/extension (60 • C for 1 min), after uracil DNA glycosylase (UDG) (50 • C for 2 min) and AmpliTaq Gold enzyme (95 • C for 10 min) activation. Target gene expression relative to control was calculated using the comparative 2 −∆∆CT method. Results were normalized to GAPDH expression and expressed as fold change over average control gene expression taken as 1.

Flow Cytometry
THP-1 cells, suspended as 10 6 cells per mL, were incubated at 37 • C with BSA as the vehicle control as well as TNF-α (10 ng/mL) and/or H 2 O 2 (10 mM/mL) for 24 h. Brefeldin A was added during the last 6h of incubation. Later, cells were harvested by centrifugation, washed twice with permeabilization buffer, and then stained by incubating separately with PE-conjugated mouse anti-human IL-8 (cat# 554720, BD Biosciences, San Jose, CA, USA) and MCP-1 (cat# 554666, BD Biosciences, San Jose, CA, USA) antibodies for 20 min in a fixation and permeabilization solution (cat# 554714, BD Cytofix/Cytoperm BD biosciences, San Jose, CA, USA). Cells were then washed with permeabilization buffer and resuspended in PBS supplemented with 2% paraformaldehyde for FACS analysis (FACSCanto II; BD Bioscience, San Jose, USA). MFI data were analyzed using BD FACSDiva TM Software 8 (BD Biosciences, San Jose, CA, USA).

Confocal Microscopy
Cultured THP-1 cells were harvested from treatment and control wells (in triplicate), fixed with 4% formaldehyde for 10 min and washed thrice with PBS. Cells were permeabilized using 0.25% Triton X-100 in PBS for 10 min and then incubated with BSA for 1 h. After blocking, cells were incubated with primary rabbit polyclonal antibodies against IL-8 (1:100 dilution, Abcam ® ab106350, pH 6.0, Cambridge, MA, USA) and MCP-1 (1:100 dilution, Abcam ® ab9669, pH 6.0, Cambridge, MA, USA) at room temperature for 1 h. Cells were washed 3 times using PBS-Tris and incubated for 1 h at room temperature with goat anti-rabbit secondary antibody conjugated with Alexa Fluor-488 (Abcam ® ab150077) or with goat anti-rabbit secondary antibody conjugated with Alexa Fluor-647 (Abcam ® ab150079). The nuclear counterstaining was performed by curing specimens with 4 ,6-diamidino-2-phenylindole (DAPI) (cat# H1500, Vectashield hard-set mounting medium, Vector Laboratories Inc, Burlingame, CA, USA) at room temperature for 15 min and hardened completely at 4 • C overnight. Confocal images were obtained using inverted Zeiss LSM710 Spectral confocal microscope (Carl Zeiss, Gottingen, Germany) and EC Plan-Neofluar 40×/1.30 oil DIC M27 objective lens. After sample excitation using 405 nm and 488 nm lines of an argon-ion laser, optimized emission detection bandwidths were configured using Zeiss Zen 2010 control software. For IL-8 and MCP-1 protein quantification, 10 random images (40× magnification) of stained cells were captured for each treatment. The fluorescent channels were split using Image J software (NIH, USA). The fluorescence intensity for respective markers i.e., IL-8 and MCP-1 was detected in 10 randomly selected fields from the entire tissue section (at 40× magnification) and subtracted from the background of the image. Mean integrated density was calculated for each image.