IL17F Expression as an Early Sign of Oxidative Stress-Induced Cytotoxicity/Apoptosis

Interleukin 17F (IL17F) has been found to be involved in various inflammatory pathologies and has recently become a target for therapeutic purposes. In contrast to IL17F secreted by immune cells, the focus of this study is to describe the triggers of IL17F release in non-immune cells with a particular focus on IL17F-induced fibrosis. IL17F induction was examined in human lung epithelial (BEAS-2B) and myeloid cell lines as well as in peripheral blood mononuclear cells after in vitro exposure to aqueous cigarette smoke extract (CSE), inorganic mercury, cadmium or the apoptosis inducer brefeldin A. Fibrosis was examined in vitro, evaluating the transition of human primary dermal fibroblasts to myofibroblasts. We observed that all stressors were able to induce IL17F gene expression regardless of cell type. Interestingly, its induction was associated with cytotoxic/apoptotic signs. Inhibiting oxidative stress by N-acetylcysteine abrogated CSE-induced cytotoxic and IL17F-inducing effects. The induction of IL17F was accompanied by IL17F protein expression. The transition of fibroblasts into myofibroblasts was not influenced by either recombinant IL17F or supernatants of CSE-exposed BEAS-2B. In addition to IL17F secretion by specialized or activated immune cells, we underscored the cell type-independent induction of IL17F by mechanisms of inhibitable oxidative stress-induced cytotoxicity. However, IL17F was not involved in dermal fibrosis under the conditions used in this study.


Introduction
Interleukin 17F (IL17F) [1] belongs to the IL17 family of cytokines, which consists of the six members IL17A, -B, -C, -D, -E and -F. Expression of IL17F is not restricted to a single cell type. In addition to specific T helper (Th)17 cells [2], it is produced in other activated immune cells such as mast cells, basophils and epithelial cells [3,4]. It can act as a heterodimer with IL17A or as a homodimer via an IL17RA/RC or IL17RC/RC receptor complex, respectively, to induce inflammatory cytokines, chemokines and metalloproteinases via several signaling pathways [5]. IL17F is a pro-inflammatory cytokine involved in autoimmune and inflammatory diseases [3,6].
Under certain inflammatory conditions, the expression of IL17F is somewhat increased, irrespective of tissue type. Higher levels of IL17F have been found in the sputum of patients with cystic fibrosis (CF) undergoing pulmonary exacerbation [7], in saliva of patients with periodontitis [8], in psoriatic skin lesions [9] or in sera of patients with atopic asthma or systemic sclerosis [10,11]. The cellular origin of IL17F under these pathological conditions is rather unknown.
For tissues with a barrier function, such as skin or intestine, there is a strong dichotomy of IL17 effects. Outcomes of IL17 signaling can be beneficial or detrimental depending on the micromilieu of the responding cells [12]. Under normal conditions, IL17 is more responsible for host defense. By contrast, in autoimmune diseases, IL17 contributes to pathogenic inflammation [13]. Currently, IL17F is used as a target for clinical intervention. Dual inhibitors and bispecific antibodies simultaneously targeting IL17A and IL17F are under investigation in different clinical studies to treat psoriasis and psoriatic arthritis [14], or to prevent pathological bone formation, which may be helpful in the treatment of ankylosing spondylitis [15].
In addition to IL17F secreted by activated immune cells, we here describe the close relationship between strong IL17F induction and cytotoxic/apoptotic stress in vitro. Cellular oxidative stress was induced in three different human cell types by different xenobiotics such as the composed mixture of aqueous cigarette smoke extract (CSE), the metals cadmium and mercury as well as the apoptosis inducer brefeldin A. Since the majority of patients (up to 70%) suffering from idiopathic pulmonary fibrosis smoked in early life [16] and elevated serum concentration of IL17F has been shown to correlate with the severity of fibrosis-causing autoimmune diseases [11], we sought to examine whether IL17F also plays a role in the induction of fibrosis under experimental in vitro conditions.

Chemicals Used for Exposure Experiments
For histochemical staining, monensin from Sigma-Aldrich (Munich, Germany) and Hoechst 33,342 from Invitrogen (Carlsbad, CA, USA) were used. For exposure experiments, cadmium chloride, mercury chloride, brefeldin A (17.8 µM) and N-acetylcysteine (NAC) were used from Sigma-Aldrich (Munich, Germany). Annexin V-FITC was purchased from Biolegend (Koblenz, Germany) and used as recommended by the supplier.
Cigarette smoke extract (CSE) was freshly prepared according to the protocol described by Adenuga and coworkers [17]. Briefly, research-grade reference cigarettes (1R6F) from the University of Kentucky (Tobacco Health Research, Lexington, KY, USA) were used to prepare CSE by bubbling smoke from one cigarette into 2 mL of RPMI 1640 without supplements at a rate of 1 cigarette/minute. Afterwards CSE was sterile-filtered through a 0.22 µm filter (Sartorius, Göttingen, Germany). The control was obtained by using air instead of cigarette smoke by the same procedure.

Cell Culture Experiments
Human adherent lung epithelial cell line, BEAS-2B, and human monocytic leukemia cell line, THP-1, were obtained from American Type Culture Collection (ATCC, Wesel, Germany). Peripheral blood mononuclear cells (PBMC) were isolated from blood of three pseudonymous healthy volunteers obtained from the blood bank at the University of Leipzig. PBMCs were obtained by density-gradient centrifugation using Ficoll-Paque (GE Healthcare, Solingen, Germany). For the use of human blood samples, the study received approval from the Ethics Committee of the University of Leipzig (reference number 079-15-09032015). Cells were kept in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS), 0.05 mM 2-mercaptoethanol, 100 units/mL penicillin and 100 µg/mL streptomycin, and cultured at 37 • C under 5% CO 2 in a humidified atmosphere. THP-1 cells growing in suspension were seeded for 2 days in the presence of 10 ng/mL phorbol-12-myristat-13-acetat (PMA, Sigma-Aldrich, Munich, Germany) to obtain adherence. Afterwards, cells were kept in medium without PMA for another 24 h before exposure experiments started.
In vitro experiments to analyze fibrosis were performed on primary human dermal fibroblasts (dFb) isolated from healthy breast skin as previously described [18]. Experiments were approved by the local ethics committee (065-2009) and conducted according to the Declaration of Helsinki principles (1975). Cells were cultured with Dulbecco's Modified Eagle Medium (DMEM, anprotec, Bruckberg, Germany) supplemented with 10% FBS (PAN-Biotech, Aidenbach, Germany) and 1% penicillin/streptomycin (anprotec) at 37 • C, 5% CO 2 until confluence. Exposure of dFb was performed between passages 2-4. In comparison to TGF-β1 (5 ng/mL) as an inducer of transition of fibroblasts to myofibroblasts (FMT, induction of fibrosis), dFb was exposed to IL17F (100 ng/mL) and supernatants of CSEexposed BEAS-2B. Supernatants of BEAS-2B were acquired after exposure to CSE for 2 h, and subsequent incubation for another 24 h in normal media.
Plasma membrane damage was measured by the release of cytosolic glucose 6phosphate dehydrogenase (G6PD) into medium with the Vybrant TM Cytotoxicity Assay Kit (Molecular Probes, Eugene, OR, USA), according to manufacturer's instructions. The morphological changes of adherent epithelial cells as a part of transition from polygonal toward round cell shape (illustrated in Supplementary Figure S3) were estimated by quantification of cell-free area using hybrid cell count option for phase contrast pictures of the Fluorescence Microscope BZ-X800 (Keyence, Neu-Isenburg, Germany).

Gene Expression by Quantitative PCR
Total RNA of samples (in triplicates) was prepared by using TRIzol reagent (Ther-moFisher Scientific, Göttingen, Germany) according to the manufacturer's instructions. The cDNA synthesis was carried out with 1 µg of RNA by using the ImProm-II TM Reverse Transcription System (Promega, Mannheim, Germany). Intron-spanning primers were designed and appropriate UPL probes (Supplementary Table S1) were selected by the Universal Probe Library Assay Design Center (http://qpcr.probefinder.com/organism.jsp (accessed until 31 December 2019)). Comprehensive gene expression was measured using the 96.96 or 48.48 Dynamic Array (Fluidigm, San Francisco, CA, USA). Genes associated to fibrosis were measured using the Roche Lightcycler 480 (Roche, Mannheim, Germany). Gene expression values were determined by using the 2 −∆∆CT method [20].

Statistical Analysis
Significant differences were analyzed with analysis of variance (ANOVA). Test of normality was performed using the Shapiro-Wilks test. All p-values < 0.05 were considered to be statistically significant. For multiple testing, the Bonferroni correction was considered. If not

Induction of IL17F Gene Expression by CSE in Different Cell Types
Pilot experiments analyzing the effect of different stressors on immune cell-related gene expression using the myeloid cell line THP-1 indicated a specific and strong induction of IL17F when exposed to aqueous cigarette smoke extract (CSE, Supplementary Figure S2). To show that this induction of IL17F by CSE was not dependent on cell type, we analyzed the effect of CSE on three different human cell types. We investigated IL17F expression together with five other members of the IL17 family (IL17A, -B, -C, -D and -E) and three IL17 receptors (IL17RA, -RB and RC). In the lung epithelial cell line BEAS-2B, IL17F was by far the most induced gene (about 120-fold at 4% CSE compared to the control, p < 0.001) at an early timepoint of 4 h, in contrast to the majority of the other IL17-related genes ( Figure 1A). Expression was further increased after 24 h (up to 770-fold, p < 0.001, Table 1, Figure 1A). Similarly, in myeloid THP-1 ( Figure 1B) and blood PBMCs ( Figure 1C in One Fluorescence Microscope BZ-X800 (Keyence, Neu-Isenburg, Germany).

Statistical Analysis
Significant differences were analyzed with analysis of variance (ANOVA). Test of normality was performed using the Shapiro-Wilks test. All p-values < 0.05 were considered to be statistically significant. For multiple testing, the Bonferroni correction was considered. If not otherwise indicated, statistical calculations were performed with Statistica for Windows version 10 (StatSoft Inc. Europe, Hamburg, Germany). Heatmaps of correlation analysis and principal component analysis (PCA) were performed in R framework (R3.6.1, https://cran.r-project.org (accessed on 30 June 2021)) for Windows using the opensource solution for statistical computing, the SINGuLAR Analysis Toolset (Fluidigm, San Francisco, CA, USA).

Induction of IL17F Gene Expression by CSE in Different Cell Types
Pilot experiments analyzing the effect of different stressors on immune cell-related gene expression using the myeloid cell line THP-1 indicated a specific and strong induction of IL17F when exposed to aqueous cigarette smoke extract (CSE, Supplementary Figure S2). To show that this induction of IL17F by CSE was not dependent on cell type, we analyzed the effect of CSE on three different human cell types. We investigated IL17F expression together with five other members of the IL17 family (IL17A, -B, -C, -D and -E) and three IL17 receptors (IL17RA, -RB and RC). In the lung epithelial cell line BEAS-2B, IL17F was by far the most induced gene (about 120-fold at 4% CSE compared to the control, p < 0.001) at an early timepoint of 4 h, in contrast to the majority of the other IL17related genes ( Figure 1A). Expression was further increased after 24 h (up to 770-fold, p < 0.001, Table 1, Figure 1A). Similarly, in myeloid THP-1 ( Figure 1B   With respect to the adverse effect of CSE on viability at higher concentrations, it b came evident that IL17F induction was related to cytotoxicity ( Figure 2). Although ce membrane integrity (   With respect to the adverse effect of CSE on viability at higher concentration came evident that IL17F induction was related to cytotoxicity ( Figure 2). Althou membrane integrity (Figure 2A) was not affected, depressed viability (EC75, MTT was reached with 4% CSE ( Figure 2B). Additionally, at high concentrations of C thelial BEAS-2B lost their polygonal shape and became irreversibly more rounded 2C, Supplementary Figure S3a-e). Table 1. The family in three different human cell types (epithelial lung cell line BEAS-2B, cell line THP-1 and peripheral blood mononuclear cells PBMC) exposed to 4% aqueous smoke extract for 24 h compared to control (related to Figure 1). IL17F shows cell type-inde With respect to the adverse effect of CSE on viability at higher c came evident that IL17F induction was related to cytotoxicity (Figu membrane integrity (Figure 2A) was not affected, depressed viability was reached with 4% CSE ( Figure 2B). Additionally, at high concent thelial BEAS-2B lost their polygonal shape and became irreversibly m 2C, Supplementary Figure S3a-e). Table 1. The family in three different human cell types (epithelial lung cell l cell line THP-1 and peripheral blood mononuclear cells PBMC) exposed to smoke extract for 24 h compared to control (related to Figure 1). IL17F shows the strongest induction. Gene expression was normalized to median of four re  With respect to the adverse e came evident that IL17F inductio membrane integrity (Figure 2A   With respect to the adverse effect of CSE on viability at higher co came evident that IL17F induction was related to cytotoxicity (Figur membrane integrity (Figure 2A) was not affected, depressed viability was reached with 4% CSE ( Figure 2B). Additionally, at high concentr thelial BEAS-2B lost their polygonal shape and became irreversibly mo 2C, Supplementary Figure S3a-e). Table 1. The family in three different human cell types (epithelial lung cell lin cell line THP-1 and peripheral blood mononuclear cells PBMC) exposed to 4 smoke extract for 24 h compared to control (related to Figure 1). IL17F shows c the strongest induction. Gene expression was normalized to median of four ref GUSB, PGK1 and PPIA). *, ** significant differences with p < 0.01 or p < 0.001, r ], red line) showed the strongest increase over time after 4 h or 24 h of exposure. In ordinate, the semi-quantitative frequency of transcripts to each other is considered, where IL17F shows the lowest basal transcript level. Gene expression was normalized to median of four reference genes (GAPD, GUSB, PGK1 and PPIA). Mean and 95% confidence interval (whiskers) are depicted. Table 1. The family in three different human cell types (epithelial lung cell line BEAS-2B, myeloid cell line THP-1 and peripheral blood mononuclear cells PBMC) exposed to 4% aqueous cigarette smoke extract for 24 h compared to control (related to Figure 1). IL17F shows cell type-independent the strongest induction. Gene expression was normalized to median of four reference genes (GAPD, GUSB, PGK1 and PPIA). *, ** significant differences with p < 0.01 or p < 0.001, respectively. With respect to the adverse effect of CSE on viability at higher concentrations, it became evident that IL17F induction was related to cytotoxicity ( Figure 2). Although cell membrane integrity (Figure 2A) was not affected, depressed viability (EC75, MTT assay) was reached with 4% CSE ( Figure 2B). Additionally, at high concentrations of CSE, epithelial BEAS-2B lost their polygonal shape and became irreversibly more rounded ( Figure 2C

Effect of NAC on CSE-Induced Gene Expression
To determine whether reactive oxygen species (ROS) are involved in early CSE-induced IL17F, the epithelial cell line BEAS-2B was co-treated with CSE and NAC for 4 h. NAC strongly blocked the early effect of CSE on IL17F (Figure 3). By contrast, the early CSE-induced HMOX1 (NRF2-mediated) and CYP1A1 (AhR-mediated) was not eliminated by NAC; rather, the reduction in induction at high CSE concentration for HMOX1 was abrogated. The cell-protecting effect of NAC against oxidative stress was apparent exclusively in the concomitant presence of NAC and CSE and was microscopically visible by the absent transition from polygonal to round cells at higher CSE concentrations (Supplementary Figure S3f-i). The effect of NAC also remained during a longer exposure up to

Effect of NAC on CSE-Induced Gene Expression
To determine whether reactive oxygen species (ROS) are involved in early CSEinduced IL17F, the epithelial cell line BEAS-2B was co-treated with CSE and NAC for 4 h. NAC strongly blocked the early effect of CSE on IL17F (Figure 3). By contrast, the early CSE-induced HMOX1 (NRF2-mediated) and CYP1A1 (AhR-mediated) was not eliminated by NAC; rather, the reduction in induction at high CSE concentration for HMOX1 was abrogated. The cell-protecting effect of NAC against oxidative stress was apparent exclusively in the concomitant presence of NAC and CSE and was microscopically visible by the absent transition from polygonal to round cells at higher CSE concentrations (Supplementary Figure S3f-i). The effect of NAC also remained during a longer exposure up to 24 h (Supplementary Figure S7).

Effect of NAC on CSE-Induced Gene Expression
To determine whether reactive oxygen species (ROS) are involved in early CSE-induced IL17F, the epithelial cell line BEAS-2B was co-treated with CSE and NAC for 4 h. NAC strongly blocked the early effect of CSE on IL17F (Figure 3). By contrast, the early CSE-induced HMOX1 (NRF2-mediated) and CYP1A1 (AhR-mediated) was not eliminated by NAC; rather, the reduction in induction at high CSE concentration for HMOX1 was abrogated. The cell-protecting effect of NAC against oxidative stress was apparent exclusively in the concomitant presence of NAC and CSE and was microscopically visible by the absent transition from polygonal to round cells at higher CSE concentrations (Supplementary Figure S3f-i). The effect of NAC also remained during a longer exposure up to 24 h (Supplementary Figure S7).

Relationship between IL17F Induction and Cytotoxicity
Because of the effect of NAC and the close relationship between IL17F induction and cytotoxicity, we explored the effect of further inducers of oxidative stress, in particular cadmium (Cd 2+ , Figure 4A) and mercury (Hg 2+ , Figure 4B). The cytotoxic effect of both components on BEAS-2B after 24 h exposure was accompanied by loss of polygonal shape (shown for Hg 2+ in Supplementary Figure S3j-m). Regardless of the different strengths of the cytotoxic effect (MTT assay), IL17F was induced at cytotoxic concentrations for both Cd 2+ and Hg 2+ , with the highest IL17F expression at the concentration at which BEAS-2B completely lost its polygonal shape. For Hg 2+ , the expression pattern of others included IL17 genes was similar to CSE exposure ( Figure 4C). Because of the effect of NAC and the close relationship between IL17F induction and cytotoxicity, we explored the effect of further inducers of oxidative stress, in particular cadmium (Cd 2+ , Figure 4A) and mercury (Hg 2+ , Figure 4B). The cytotoxic effect of both components on BEAS-2B after 24 h exposure was accompanied by loss of polygonal shape (shown for Hg 2+ in Supplementary Figure S3j-m). Regardless of the different strengths of the cytotoxic effect (MTT assay), IL17F was induced at cytotoxic concentrations for both Cd 2+ and Hg 2+ , with the highest IL17F expression at the concentration at which BEAS-2B completely lost its polygonal shape. For Hg 2+ , the expression pattern of others included IL17 genes was similar to CSE exposure ( Figure 4C).  to concentration-dependent cytotoxicity (decreased viability in MTT-assay) and loss of polygonal shape (cell-free area). IL17F-induction starts with cytotoxicity and loss of polygonal shape of adherent cells. Highest gene expression was set to 100%. The expression pattern of members of the IL17 family including IL17F for Hg 2+ is similar to CSE exposure (C), see Figure 1. Gene expression was normalized to median of four reference genes (GAPD, GUSB, PGK1 and PPIA).

IL17F Protein Induction by CSE and Brefeldin A
To verify whether IL17F gene expression is translated into protein release and could be induced in cells that are undergoing apoptosis, IL17F protein expression was examined by immune fluorescence after treatment with the apoptosis inducer brefeldin A in comparison to CSE in BEAS-2B ( Figure 5). In contrast to an exposure of 4 h, IL17F protein expression was apparent after 24 h by a cytoplasmic granular pattern for both brefeldin A and CSE. Concomitantly, the frequency of cells positive for the apoptosis stain Annexin V was increased (Supplementary Figure S4), indicating the involvement of apoptosis in the cytotoxic effect of CSE. Figure 6 schematically illustrates the cell type-independent induction of IL17F under oxidative stress-induced cytotoxic/apoptotic conditions in vitro, compared to IL17F secretion by specialized Th17 immune cells.  5 µM, b,c). Cytoplasmic granular pattern of IL17F was detectable for brefeldin A (e,f) and for CSE (3%, h,i) at 24 h exposure. Nuclear staining was performed with Hoechst 33,342 (a,d,g).  5 µM, b,c). Cytoplasmic granular pattern of IL17F was detectable for brefeldin A (e,f) and for CSE (3%, h,i) at 24 h exposure. Nuclear staining was performed with Hoechst 33,342 (a,d,g). Figure 5. Induction of IL17F protein in lung epithelial cell line BEAS-2B by the brefeldin A and the aqueous cigarette smoke extract (CSE). IL17F staining is absent after 4 h exposure shown for brefeldin A (17.5 µM, b,c). Cytoplasmic granular pattern of IL17F was detectable for brefeldin A (e,f) and for CSE (3%, h,i) at 24 h exposure. Nuclear staining was performed with Hoechst 33,342 (a,d,g).

Figure 6.
Summary of IL17F inducing conditions. In addition to IL17F from specialized Th17 immune cells [21], IL17F is also inducible by cytotoxic/apoptotic conditions in non-immune cells where the oxidative stress is decisively involved in IL17F induction.

Comprehensive Analysis of Exposure-Induced Gene Expression
To gain more insight into the mechanism of exposure-induced IL17F, we analyzed transcripts of genes related to stress, apoptosis and inflammation and those that are involved in IL17F transcript release (Supplementary Table S2). By comparing 39 investigated genes, it became evident that IL17F was the only gene showing strong induction independent of cell type and stressor (about 15 up to 700-fold). In a similar manner, Figure 6. Summary of IL17F inducing conditions. In addition to IL17F from specialized Th17 immune cells [21], IL17F is also inducible by cytotoxic/apoptotic conditions in non-immune cells where the oxidative stress is decisively involved in IL17F induction.

Comprehensive Analysis of Exposure-Induced Gene Expression
To gain more insight into the mechanism of exposure-induced IL17F, we analyzed transcripts of genes related to stress, apoptosis and inflammation and those that are involved in IL17F transcript release (Supplementary Table S2). By comparing 39 investigated genes, it became evident that IL17F was the only gene showing strong induction independent of cell type and stressor (about 15 up to 700-fold). In a similar manner, DDIT3, a marker of ER stress, HMOX1, indicating oxidative stress, and CDKN1A, related to senescence, were induced, but in contrast to IL17F, this occurred at lower concentrations of stressors and, in some cases, at a much lower degree of induction. The best correlation for IL17F was found to be IL17B in CSE-induced BEAS-2B (Supplementary Figure S5A). A heatmap of the correlation analysis shows a strong increase in most of the genes as CSE concentration increases. Concentration-dependent stratification was confirmed by principal component analysis (PCA, Supplementary Figure S5B). Due to the inhibition of protein synthesis by brefeldin A, the pattern of its induced gene expression differed from that of CSE and metals. However, a pronounced dose-dependent transcript induction was present for the stress-indicating heat shock proteins HSPA1A and HSPA2, pro-inflammatory TNF and IL22 and pro-apoptotic FASLG. Gene expression of proteins involved in IL17F transcript regulation such as RORC, SOCS3 and IL23A was inconspicuous (Supplementary Table S2). Furthermore, no stratification to the functionality of genes was apparent.

Influence of IL-17F on In Vitro Fibrosis Using Primary Dermal Fibroblasts
To verify a potential involvement of CSE-induced IL17F in fibrosis, we examined in vitro the transition of fibroblasts to myofibroblasts (FMT) using primary human dermal fibroblasts. Although recombinant IL17F (rIL17F) and the inducer of FMT TGF-β1 (rTGF-β1), both induced IL6 in fibroblasts (Supplementary Figure S6), neither rIL17F alone nor supernatants from lung epithelial cell line BEAS-2B, exposed to CSE, induced FMT (Figure 7). The FMT was evaluated by increased gene expression of the four FMT-associated genes α smooth muscle actin (ACTA2), collagen type I (COL1A1), palladin (PALLD) and ED-A fibronectin (FN1). In contrast to rIL17F and supernatants from CSE-exposed BEAS-2B, rTGF-β1 induced FMT independently of the presence of supernatants from CSE-exposed BEAS-2B.

Discussion
This study aims to understand the ability of CSE to induce IL17F production and whether this is associated with fibrosis. This idea arose from results of preliminary experiments using THP-1 cells that showed a strong and seemingly highly specific induction of IL17F by CSE, compared to five other stimulants and their combinations and compared to 47 other genes. In contrast to IL17F induction in specific immune cells by chemokines, the CSE-induced IL17F seems to rely on another mechanism of IL17F induction, which we wanted to investigate in more detail. In order to exclude a cell-type specific effect, we replicated experiments on different human cell types (epithelial BEAS-2B, myeloid THP-1, and primary PBMC). In order to exclude CSE-specific effects, we included inorganic mercury, cadmium and the apoptosis inducer brefeldin A.
First of all we compared exposure-induced IL17F induction to other members of the IL17 family as well as reported genes related to be involved in IL17 induction. The exposure-induced IL17F induction was unique, correlating neither with other IL17 members such as IL17A, which can form heterodimers, nor with that of IL17 receptors IL17RA or RC, which can bind IL17F as a heterodimer or homodimer, respectively [22]. Interestingly IL22, a further Th17 effector cytokine, was concomitantly induced under these cytotoxic/apoptotic conditions, assuming a common transcriptional activation for both cytokines. Since genes associated with IL17-inducing pathways such as RORC, IL23/STAT3, SOCS3 and IL6/RORA [23,24] remained unchanged, an IL17F-inducing mechanism specific to cytotoxic/apoptotic conditions is suggested. The expression of other members of the IL17 family under these conditions differed depending on the cell type. For example, in contrast to the myeloid THP-1 cell line, in the lung epithelial BEAS-2B cell line, IL17B correlated well with IL17F, possibly for its role in neutrophil attraction under stress conditions [25].
CSE-induced IL17F expression was associated with cytotoxicity/apoptosis. To discover the reason for the chemical-induced cytotoxic/apoptotic signs, we hypothesized that oxidative stress could play a decisive role. Thus, we examined (i) results after using inducers of oxidative stress, (ii) a marker gene indicating oxidative stress and (iii) the effects of an inhibitor of oxidative stress. As inducers of oxidative stress, we used cadmium (Cd) and mercury (Hg), metals which are also found in cigarette smoke. Cadmium is potent to induce oxidative stress, although by itself it is not redox-active. It is unable to generate reactive oxygen species (ROS) directly [26]. In contrast, Hg-induced oxidative stress may occur due to both prooxidant action of the metal and a decrease in antioxidant enzymes. Despite the absence of direct indications, it can be proposed that mercury may induce endoplasmic reticulum (ER) stress [27]. Indeed, both metals induced IL17F. ER stress could be also involved in cytotoxic mechanisms, since transcripts of the ER stress marker DNA damage inducible transcript 3 (DDIT3), were strongly induced.
With respect to CSE, we propose that antioxidant proteins such as HMOX1, becoming activated via antioxidant response elements by a nuclear factor erythroid 2-related factor 2 (NRF2), should be induced. Indeed, HMOX1 was augmented upon CSE exposure but its induction could not be reduced by NAC. Therefore, it can be assumed that despite the cytoprotective effect of NAC, CSE-exposed cells are still under stress, and HMOX1 exerts its protective effect via heat shock proteins or anti-inflammatory properties [28]. In addition, the cytoprotective effect of NAC is caused less by ROS scavenging and more by intracellular generation of protein-protecting sulfane sulfur species from NAC [29]. Besides the lack of effect of NAC on CSE-induced HMOX1 expression, other CSE-induced genes remained unchanged. Among them was CYP1A1, a gene induced by the aryl hydrocarbon receptor (AhR). Since AhR agonists also induced oxidative stress via the formation of ROS by AhR-induced cytochrome P450 (CYP) enzymes [30], cytoprotection through NAC could also rely on the inhibition of AhR-triggered ROS. However, neither AhR antagonist (α-naphtoflavone) nor AhR agonist (benz(a)pyrene) impaired CSE-induced IL17F (Supplementary Figure S2). Thus, ROS induction via an AhR-dependent pathway does not seem to be responsible for the IL17F induction. Nevertheless, (i) the effects of metals, (ii) the induction of HMOX, (iii) the effect of NAC, all together, give rise to assume a general role of oxidative stress in induction of IL17F.
The presence of apoptosis signs elicited by CSE both at the gene expression level and in histological Annexin V staining suggests a prominent role of apoptosis in IL17F induction. Similar apoptotic signs have also been described in airway epithelial cells from cigarette smoking-induced COPD patients [31]. The impact of apoptosis on IL17F was further substantiated by brefeldin A-inducing IL17F as well as protein expression. Brefeldin A as an endoplasmic reticulum (ER)-Golgi transport inhibitor causes accumulation of proteins in the ER, thus causing ER stress accompanied by the formation of reactive oxygen species and depletion of GSH, subsequently causing apoptosis [32]. Because of its inhibition of protein synthesis, it influences transcription of genes. Thus, the brefeldin A-induced gene expression pattern did not fit well with that of CSE or metals even when newly formed proteins would impair transcription. Nevertheless, IL17F, DDIT3, and the proapoptotic FASLG were inducible. This indicated that transcription of these genes is less under control of exposure-triggered newly-synthesized proteins, which might explain the early transcriptional activation of IL17F.
The putative role of IL17F in apoptosis induction remains speculative. We have not examined this in detail. However, since IL17-secreting Th17 cells are resistant to glucocorticoid-induced apoptosis [33], an apoptosis-promoting effect of IL17F seems rather unlikely. A somewhat anti-apoptotic effect was shown by the inhibition of Fas-induced cell death in Fas-sensitive T cells exposed to recombinant IL17F [34].
With the examined cytotoxicity/apoptosis-inducing chemicals we could confirm reports of IL17F induction by similar acting agents. Indeed, this has been reported for the African swine fever virus (ASFV) in macrophages, for UV-A in fibroblasts and for apigenin in pancreatic cancer cells. In swine macrophages, the virus upregulated IL17F and cytokines of the TNF family including FASLG and TNF [35]. In skin fibroblasts, in a similar manner UV-A induced IL17F and proteins involved in apoptosis [36]. Interestingly, similar to the apoptosis-preventing effect of NAC, the effect of UV-A was restored by the antioxidant rutin. Furthermore, in pancreatic cancer cells, the citrus fruit bioactive flavonoid apigenin strongly upregulated IL17F (>100-fold) and induced cell death through inhibition of the glycogen synthase kinase-3β (GSK3β)/nuclear factor kappa B (NFKB) signaling pathway [37]. Recently it was suggested that CSE is also able to inactivate GSK-3β via activation of the PI3K/Akt pathway [38]. The question of whether GSK-3β has a central role in apoptosis remains elusive, since deactivation of GSK-3β in cardiomyocytes protected against doxorubicin-induced apoptosis [39].
To clarify whether IL17F release from cells undergoing cytotoxicity/apoptosis can cause fibrosis, we analyzed FMT in human primary dermal fibroblasts. This is used as an indicator of fibrosis. However, we could not detect pro-fibrotic effects of CSE, either of rIL17F or of supernatants from CSE-exposed epithelial cells in which apoptosis-related IL17F was induced. Due to easier availability, dermal fibroblasts were used primarily assuming similar conditions to lung fibroblasts with respect to IL17RA. A functional IL17RA is essential for IL17A-induced FMT in lung fibroblasts [40] and IL17RA is also found together with IL17RC on dermal fibroblasts [41]. According to Levy et al. [42], we confirmed the functionality of IL17RA by induction of IL-6 by IL17F in dermal fibroblasts. However, since we focused solely on dermal fibroblasts, we cannot exclude that IL17Finduced fibrotic effects may occur in other tissue-specific fibroblast types. The putative role of cytotoxicity/apoptosis-associated induction of IL17F could, therefore, not be definitively clarified in our experiments.

Conclusions
In addition to IL17F secretion by specialized or activated immune cells under physiological or pathological conditions, we have now revealed a cell type-independent induction of IL17F under oxidative stress-induced cytotoxicity in vitro. However, this IL17F does not appear to play a role in dermal fibrosis. Because of its strong induction and association with cytotoxicity/apoptosis, it could be of interest to establish whether IL17F could serve as a diagnostic marker to follow therapeutic interventions or whether its cellular origin has an impact on therapeutic interventions.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/genes13101739/s1. Figure S1: Immunohistological controls considered for detection of IL17F protein in lung epithelial cell line BEAS-2B. IgG2b isotype and 24 h unexposed with CSE culture control show only a marginal nuclear background staining but clear absent cytoplasmic granular staining. Figure S2: Preliminary data from a pilot experiment analyzing the effect of single and multiple stressors on expression of 48 immune cell-related genes using the myeloid cell line THP-1. Highlighted are expression data from three independent experiments (bars in light grey, dark grey or striped) for the genes IL1B, TNFA, IL17A and IL17F. IL17F was the only gene showing a highly specific and pronounced induction when exposed to CSE (red frame of the bars). Figure S3: Microscopic cytotoxicity pattern of CSE and mercury on lung epithelial cell line BEAS-2B after 24 h exposure is affected by N-acetylcysteine (NAC). Both CSE (a-e) and mercury (Hg 2+ , j-m) induces a concentration-dependent loss in polygonal shape toward round shape of epithelial cells. NAC (0.5 mM) prevents the cytotoxic effect of CSE (f-i) as well as Hg 2+ (n-q). Figure S4: Apoptosis induction by brefeldin A and CSE in lung epithelial cell line BEAS-2B. Frequency of Annexin V stained cells are higher in Brefeldin A (e) or CSE (h) exposed cells compared to control (b). Figure S5: Heatmap (A) and principal component analysis plot (PCA plot, B) for gene expression of CSE-exposed (24 h) lung epithelial cell line BEAS-2B from three independent experiments. As expected, genes (n, 39) for stress, apoptosis and inflammation were increased with increasing concentration of CSE (A). However, a strong correlation between IL17F and these genes was not found. For both the heatmap and PCA plot a clear concentration-dependent discrimination of exposed cells can be achieved (A, B). Expression data were log2-scaled. Figure S6: Recombinant IL17F-induced IL6 expression in primary human dermal fibroblasts (dFb) after 48 h exposure as a sign of rIL17F-activity in dFb. Gene expression was normalized to median of three reference genes (GAPD, PGK1 and PPIA). ***, p < 0.001, paired t-test compared to control. Figure S7: Microscopic cytotoxicity pattern of 4% CSE on lung epithelial cell line BEAS-2B after 24 h exposure is affected by N-acetylcysteine (NAC). CSE (a) induces a loss in polygonal shape toward round shape of epithelial cells. NAC (0.5 mM) prevents the cytotoxic effect of CSE (c). Table S1: Primer sequences and universal probe library probes (UPL) used for qPCR. Table S2: Change in gene expression (mean, x-fold) compared to control after 24 h of exposure.