Oxidative stress-mediated epidermal growth factor receptor activation by cigarette smoke or heated tobacco aerosol in human primary bronchial epithelial cells from multiple donors

The epidermal growth factor receptor (EGFR) signaling pathway has essential roles in maintaining homeostasis of various tissues by regulating cell proliferation and differentiation. Deregulation of the EGFR signaling pathway is associated with various chronic diseases including chronic obstructive pulmonary disease. Cigarette smoke (CS) is known to activate EGFR, which is linked to chronic obstructive pulmonary disease. The biological sequence from CS exposure to EGFR activation is initiated by oxidative stress caused by intracellular reactive oxygen species (ROS) and the depletion of glutathione, which led to EGFR ligand secretion and EGFR activation. We hypothesized that reducing exposure to CS constituents contributes to preventing CS-inducible EGFR activation. Therefore, we examined the aerosol from heated tobacco products (HTPs) because the aerosol contains fewer chemical constituents at lower levels than CS. We exposed primary human bronchial epithelial cells from four donors to the extracts of CS from a 1R6F reference cigarette or HTP aerosol from three in-market products, including our DT3.0a. The biological sequence from ROS to EGFR activation was assessed. CS induced all the tested endpoints although inter-donor differences were observed, whereas HTPs elicited most of the biological events at higher concentrations; however, EGFR phosphorylation was not observed even at fivefold higher concentration than CS. Overall, our results indicate that HTPs are less effective than CS to elicit ROS-induced EGFR activation. The reduced-risk potential of HTPs on EGFR-related diseases should be investigated further. In addition, testing with multiple donors is warranted when considering the individual differences in responses of primary cells to stimuli.


| INTRODUCTION
Epidermal growth factor receptor (EGFR), which is a member of the tyrosine kinase family, has key roles in tissue homeostasis by regulating cell proliferation and differentiation. Hundreds of genes are regulated by the EGFR signaling pathway, and the proper regulation of these gene expression levels is critical to maintaining tissue functions (Chen et al., 2016). The overexpression and abnormal activation of EGFR are associated with the malfunction of various tissues and organs such as the skin, kidney, and lung (Harskamp et al., 2016;Holcmann & Sibilia, 2015;Vallath et al., 2014). EGF, heparin binding-EGF-like growth factor (HB-EGF), transforming growth factor-alpha (TGF-α), and amphiregulin (AREG) are EGFR ligands that mediate their effects after binding via the autophosphorylation of EGFR and then transduce the downstream signaling pathway to express EGFR-regulated genes (Singh et al., 2016). Luettich et al. (2017) proposed an EGFR-mediated simplistic biological pathway that led to a decrease in lung function (i.e., adverse outcome pathway [AOP] for "lung function decrease"), which is elicited by reactive oxygen species (ROS). ROS, as well as oxidative stress, has been suggested to induce the expression and secretion of EGFR ligands, which activate EGFR (Miyazaki et al., 1996) in a ligand-dependent manner. Of note, EGFR is also activated via a ligand-independent mechanism (Paulsen et al., 2011;Shao & Nadel, 2005;Sheng et al., 2020). EGFR has prominent functions, including tissue development and repair in the airway epithelium. Furthermore, the activation of EGFR leads to the expression of mucins and goblet cell metaplasia, which are features seen in hypersecretory diseases of the airways (Casalino-Matsuda et al., 2006).
One of known causes of lung disease is cigarette smoking.
Cigarette smoke (CS) contains thousands of chemicals, some of which can elicit intracellular ROS and oxidative stress. In addition, Khan et al. (2008) reported that CS exposure elicited EGFR activation. Therefore, oxidative stress-mediated pathogenesis of lung diseases might be avoidable by reducing exposure to such chemical substances. The U.S. Food and Drug Administration (U.S. FDA, 2012) has proposed guidance for assessing the harm reduction potential of a "modified risk tobacco product" (MRTP) based on the possibility that decreased toxicant exposure may lead to health benefits. One of the expected MRTPs is heated tobacco products (HTPs), which generate vapor by heating but not combusting tobacco leaves; thus, the aerosol contains lower levels of harmful constituents than CS (Hirn et al., 2020;Poget et al., 2021).
Several previous studies showed that the aerosol of HTP elicited less biological effects in in vitro and in vivo studies, as well as clinical testing (Ito et al., 2019;Phillips et al., 2016;Thorne et al., 2020;Yuki et al., 2022). Moreover, the U.S. FDA (2020) has approved one HTP as an MRTP (i.e., reduced-exposure product) after rigorous examination of scientific evidence not limited to exposure reduction but also the reduction of biological impact and expectation of health benefit for the population as a whole. Nevertheless, further evidence that augments our understanding regarding emissions from HTPs shows that reduced biological responses when compared with those of CS are required because CS elicits a wide range of biological effects. Although clinical and epidemiological studies have provided the most relevant data regarding the health effects in humans, in vitro studies using human cells might help infer the potential toxicity and underlying mechanisms of test products in humans and uncover their associated mechanisms (Gohlsch et al., 2019). Although HTPs induced lower levels of oxidative stress responses than cigarettes (Munakata et al., 2018;Taylor et al., 2018), the effects on the downstream events of oxidative stress after HTP exposure remain unclear. An evaluation of the sequence of biological events from oxidative stress to EGFR activation would thus be beneficial for assessing the reduced-risk potential of HTPs.
The application of in vitro tests for risk assessment is in line with the recent growing trend toward the reduction, replacement, and refinement (3R principles) of animal testing. An AOP framework is such a methodology, which depicts simplistic biological pathways leading to adverse outcomes based on accumulated knowledge. Thus, AOP-based risk assessment may provide mechanistic reasoning for toxicity and risk. In this study, we partially applied the AOP proposed by Luettich et al. (2017), in which the molecular initiating event is intracellular ROS. Although the adverse outcome of the proposed AOP is "lung function decrease," we focused on the biological sequence from ROS to elicitation of EGFR activation because the biological events are related to various lung diseases as mentioned above.
To investigate the reduced-risk potential of HTPs using the AOP framework, we exposed the aerosol collected mass (ACM) of our proprietary Direct heating Tobacco System Platform 3 Generation 3 version a (DT3.0a), and other in-market HTPs, and the total particulate matter (TPM) of combustible cigarettes. This information should help us understand the reduced-risk potential of HTPs.

| Test products
Three commercially available HTP products were used in this study: DT3.0a, tobacco heating system (THS), and tobacco heating product (THP). All these HTPs were purchased from the Japanese market. The THS heats tobacco stick directly via a heating blade inside the system, whereas THP heats tobacco stick peripherally. DT3.0a employs peripheral heating with convection flow. The heating temperature and the result of representative chemical analysis were reported previously (Eaton et al., 2018;Forster et al., 2018;Jaccard et al., 2017;Mallock et al., 2019). A regular tobacco-flavored heating stick for each product was used. The 1R6F reference cigarette (University of Kentucky, Lexington, KY, USA) was used as a representative conventional cigarette.

| Preparation of CS-TPM and heated tobacco ACM
CS-TPM and HTP-ACMs were prepared in accordance with the International Organization for Standardization Intense smoking regimen (a 55-mL puff taken over 2 s, repeated every 30 s) (ISO 20778, 2018).
The ventilation holes of the 1R6F were blocked and it was then applied to a Borgwaldt RM20H smoking machine (Borgwaldt KC GmbH, Hamburg, Germany). The TPM was collected on a 45-mmdiameter Cambridge filter pad and extracted using dimethyl sulfoxide (DMSO; Sigma-Aldrich, St. Louis, MO, USA).
HTP-ACMs was generated in the same manner as for the 1R6F cigarettes with a LM5E smoking machine (Borgwaldt) but without vent blocking (ISO 20778, 2018). The initial concentrations of CS-TPM and HTP-ACMs were adjusted to 40 and 200 mg/mL, respectively, with DMSO. The CS-TPM and HTP-ACMs in DMSO solution were then immediately aliquoted for single use and stored at À80 C until use. Each aliquot was thawed just before each medium change and diluted with the culture medium for exposure.

| Nicotine concentrations in CS-TPM and HTP-ACMs
The nicotine concentration in each test product was determined by gas chromatography using an Agilent 7890A GC system (Agilent Technologies, Santa Clara, CA, USA) with flame ionization detection. The limit of detection and limit of quantification were 12.58 and 42 μg/mL, respectively.

| Cell culture and exposure conditions
Primary normal human bronchial epithelial (NHBE) cells derived from four different donors were purchased from LONZA (Basel, Switzerland) and cultured in Airway Epithelial Cell Growth medium (PromoCell, Germany) supplemented with growth factors at 37 C in a 5% CO 2 atmosphere. Donor information is summarized in Table 1.
Exposure to the test products was conducted under the same conditions and untreated cells were considered controls.
The maximum concentration of each test substance was selected to not exceed 1% DMSO. These maximum concentrations of TPM and ACMs were 400 and 2000 μg/mL, respectively. We confirmed that exposure to the TPM or ACM at the highest concentration did not induce severe cytotoxicity ( Figure S1) in a 1-h exposure.

| Glutathione assay
The glutathione (GSH) level was determined using a monochlorobimane (MCB) (Sigma-Aldrich) dye: 96-well cell culture plates were initially seeded with 10,000 NHBE cells per well followed by 24 h of incubation at 37 C. NHBE cells were exposed to the test product CS-

| Measurement of EGFR ligands
First, 96-well cell culture plates were seeded with 10,000 NHBE cells per well followed by 24 h of incubation at 37 C and then exposed to the test products CS-TPM and HTP-ACMs for 1 h. The supernatants of the NHBE cells were collected and the concentrations of TGF-α and AREG were determined by a Luminex assay (R&D Systems) using the Bio-plex 200 (Bio-Rad, Hercules, CA, USA) and a Quantikine Human Amphiregulin ELISA Kit (R&D Systems), respectively. The experiment was performed on three independent occasions per donor. Because the culture medium contained 10 ng/mL of EGF as a supplement, the dynamic range of EGFR phosphorylation was also examined by adding 20 ng/mL of EGF to the culture medium for 1 h.
As a result, we confirmed approximately fourfold induction of EGFR phosphorylation ( Figure S2).

| Calculation of a benchmark dose
The results of a dose response for each assay were analyzed by benchmark dose (BMD) modeling using the U.S. Environmental Pro- To interpret the results conservatively, some donors that did not show a response were eliminated from the calculation in this study.
The data calculated with all donors are shown in Table S1.
We assumed dose-dependent increases in ROS, EGF ligands, and EGFR activation and a dose-dependent decrease in GSH. The BMD calculation was not performed when a clear dose-dependent change was not observed.

| Statistical analysis
Dunnett's multiple comparison was performed to examine the statistically significant difference between CS-TPM and HTP-ACMs at the comparable concentration of 400 μg/mL, using the pooled donor data. A p-value < 0.05 was considered statistically significant.

| Nicotine concentrations in test products
We quantified the nicotine concentrations in the undiluted CS-TPM and HTP-ACMs. to CS-TPM and HTP-ACMs at different concentration ranges: CS-TPM (0-400 μg/mL) and HTP-ACMs (0-2000 μg/mL). The concentrations of nicotine at the highest doses were approximately three times higher in HTP-ACMs than in CS-TPM.

| Oxidative stress responses
The values were consistently higher for HTPs than CS, and no significant differences between HTPs were detected.

| Secretion of EGF ligands
Levels of epidermal growth factor ligands, TGF-α and AREG, secreted into the cell culture medium were measured after exposure to all tested products for 1 h. The concentrations of TGF-α and AREG were substantially higher after exposure to CS-TPM than HTP-ACMs for Donors A, B, and D. Significant changes in the EGF ligands were not observed for Donor C after CS-TPM and HTP-ACMs exposure (Figures 3 and 4). A statistically significant difference was observed between the comparable concentrations of CS-TPM and HTP-ACMs (400 μg/mL) in TGF-α secretion ( p-value < 0.0001), whereas that in AREG was not statistically significant. The calculated BMD (μg/mL) values were consistently higher in HTP-ACMs than CS-TPM. Because of the weak dose response of AREG secretion by HTP-ACMs exposure, the BMD values were generally greater than the highest concentration of each test product (Table 3). To interpret the results conservatively, we eliminated Donor C from the calculation of these events. The results of all donors are shown in Figure S3.

| Phosphorylation of EGFR
Levels of pEGFR were measured after exposure to all tested products for 1 h. CS-TPM showed dose-dependent responses related to pEGFR levels ( Figure 5). However, none of the HTP-ACMs increase the pEGFR levels. A statistically significant difference in EGFR phosphorylation between the comparable concentrations of CS-TPM and HTP-ACMs (400 μg/mL) was observed (p-value < 0.0001). The BMD calculation was therefore not performed for HTP-ACMs. The magnitude of EGFR phosphorylation varied between donors, from approximately 1.2-to 3.5-fold increases at the highest concentration F I G U R E 1 Intracellular reactive oxygen species in each donor normal human bronchial epithelial cells exposed to cigarette smoke-total particulate matter (TPM) or heated tobacco products-aerosol collected mass (ACM). Fold changes in intracellular reactive oxygen species generation were calculated as a ratio to the non-treatment control of each donor (A-D, Donor A through Donor D  Figure S3. F I G U R E 2 Glutathione depletion in each donor normal human bronchial epithelial cells exposed to cigarette smoke-total particulate matter (TPM) or heated tobacco products-aerosol collected mass (ACM). Fold changes in glutathione depletion were calculated as a ratio to the nontreatment control of each donor (A-D, Donor A through Donor D). The data are the means and standard deviations of triplicate measurements. Each fold change was then summarized as an all-donor result (E). 1R6F, Kentucky reference cigarette; DT3.0a, Direct heating Tobacco System Platform 3 Generation 3 version a; THP, tobacco heating product; THS, tobacco heating system.
F I G U R E 3 Secretion of transforming growth factor-alpha (TGF-α) in each normal human bronchial epithelial cells donor exposed to cigarette smoke-total particulate matter (TPM) or heated tobacco products-aerosol collected mass (ACM). The data are the means and standard deviations of triplicate measurements (A-D, Donor A through Donor D). Each fold change was then summarized as an all-donor result (E). Donor C was eliminated from the calculation because the donor cells did not respond to the exposure. 1R6F, Kentucky reference cigarette; DT3.0a, Direct heating Tobacco System Platform 3 Generation 3 version a; THP, tobacco heating product; THS, tobacco heating system.

| DISCUSSION
This study assessed the sequence of biological responses from oxidative stress to EGFR activation (i.e., phosphorylation) based on the AOP proposed by Luettich et al. (2017) in four different NHBE donors and compared them with ACMs from three commercially available HTPs (DT3.0a, THP, and THS) and the TPM from 1R6F reference cigarettes. EGFR activation is a molecular initiating event for some lung diseases and might be an indicator of the potential risk of developing these diseases. Exposure concentrations of HTP-ACMs were selected up to five times higher than those of CS-TPM because HTPs were reported to contain lower levels of harmful and potentially harmful constituents (HPHCs); thus, we expected that higher concentrations of HTP-ACMs would be required to observe biological responses (Margham et al., 2016;Munakata et al., 2018;Sakaguchi et al., 2014;Schaller et al., 2016). The concentrations of HTPs used in this study achieved up to three times higher nicotine levels than CS-TPM.
As expected, CS induced oxidative stress-related endpoints and their downstream events, EGFR ligand expression, and EGFR activation. HTP-ACMs elicited only oxidative stress (i.e., increased ROS and GSH depletion), the uppermost event of the biological sequence that we tested, whereas higher concentrations were needed to obtain similar response levels after CS-TPM exposure. The occurrence of downstream biological events was weaker or negligible when compared with CS-TPM at the same concentrations. Therefore, this suggested that HTP-ACMs have a limited impact on activation of the EGFR signaling pathway.
Although most of the results in our study were consistent, donor-to-donor variation was observed. Donor C did not elicit EGFR ligand secretion after CS-TPM and HTP-ACMs exposure ( Figures 3C   and 4C). EGFR ligand secretion is regulated by their expression and shedding by proteinases such as ADAM-17 (Blobel, 2005;Kasina et al., 2009). Therefore, donor-to-donor variation and lacking responsiveness in this study may be dependent on the activity of such proteinases. However, Donor C showed a marked increase in EGFR phosphorylation ( Figure 5C). This implies that CS also induces ligand-independent phosphorylation of EGFR or secretion of other F I G U R E 4 Secretion of amphiregulin in each normal human bronchial epithelial cells donor exposed to cigarette smoke-total particulate matter (TPM) or heated tobacco products-aerosol collected mass (ACM). The data are the means and standard deviations of triplicate measurements (A-D, Donor A through Donor D). Each fold change was then summarized as an all-donor result (E). Donor C was eliminated from the calculation because the donor cells did not respond to the exposure. 1R6F, Kentucky reference cigarette; DT3.0a, Direct heating Tobacco System Platform 3 Generation 3 version a; THP, tobacco heating product; THS, tobacco heating system.
T A B L E 3 BMD (μg/mL) for each endpoint. EGF ligands. In addition, Donor A did not elicit significant EGFR activation even when there was an increase in EGFR ligand secretion.
These results suggest donor-dependent differences in the threshold of EGFR activation via ligand binding, time-course of ligand binding to receptor activation, or ligand-independent phosphorylation (Filosto et al., 2012). The variations observed in primary cell cultures might be explained by the original characteristics of the donors (Bovard et al., 2020;Mori et al., 2022;Rayner et al., 2019). Theoretically, not all individuals have the same transcriptomic, proteomic, and epigenetic characteristics (Jackson et al., 2020;Stefanowicz et al., 2012); therefore, the capacity of the antioxidant system might also vary. F I G U R E 5 Phosphorylation of epidermal growth factor receptor in each normal human bronchial epithelial cells donor exposed to cigarette smoke-total particulate matter (TPM) or heated tobacco products-aerosol collected mass (ACM). Fold changes of epidermal growth factor receptor phosphorylation levels were calculated as a ratio to the non-treatment control of each donor. The data are the means and standard deviations of triplicate measurements (A-D, Donor A through Donor D). Each fold change was then summarized as an all-donor result (E). Donor A was eliminated from the calculation because the donor cells did not respond to the exposure. 1R6F, Kentucky reference cigarette; DT3.0a, Direct heating Tobacco System Platform 3 Generation 3 version a; THP, tobacco heating product; THS, tobacco heating system.
Additionally, our preliminary experiment showed that deterioration of ROS by antioxidant treatment (i.e., N-acetyl-cysteine) effectively reduced EGFR activation by CS ( Figure S4). Therefore, the reduction and elimination of ROS are also considered to be effective to prevent EGFR activation.
EGFR activation is known to be associated with mucin overexpression, which is often seen in chronic obstructive pulmonary disease (COPD) and asthma patients (Burgel & Nadel, 2008;Takeyama et al., 1999). EGFR and its signaling pathway are therefore therapeutic targets for lung diseases (Lai & Rogers, 2010;Vallath et al., 2014).
Taken together, prevention of EGFR activation may contribute to reducing the risk. Although further investigations are necessary to draw a conclusion, our results suggest a reduced-risk potential of HTPs on EGFR-related diseases. Bentley et al. (2020) reported that most chemical constituents in HTP aerosol (i.e., THS2.2) overlapped with those in CS, and only a few chemicals were present at higher levels than in CS. This type of study will help us understand the reduced-risk potential and toxicity of HTP aerosol. Taken together with the in vitro test results of HTPs reported to date, high levels of chemical constituents in HTP aerosol compared with CS are thought to not have acute effects on elicitation of oxidative stress and inflammation. However, limited studies have reported the potential risk of chronic exposure to HTP aerosol, and therefore, the effects of chronic exposure to such aerosol constituents remain unclear. The advancement of in vitro cell culture platforms allows us to perform repeated exposure studies, could therefore help to investigate the effects of HTP aerosol in chronic exposure (Bovard et al., 2020;Dye et al., 2015).
We demonstrated the reduced-risk potential of HTPs in a single exposure of NHBE to aerosol based on some of the AOP proposed by Luettich et al. However, enhancement of the capability of in vitro testing with chronic exposure enabled us to investigate the remaining biological sequence to approximate adverse outcomes or disease.
Therefore, investigation of the chronic effects of exposure to HTP aerosol might provide deeper insights into its health effects. Moreover, although we used the BMD calculation to infer the quantitative difference between HTPs and CS, we can improve the link between the results and real-world situations. The combination of in vitro studies and in silico analyses, such as mathematical modeling of AOP (i.e., quantitative AOP) and simulating aerosol deposition in real-world use scenarios, will be a future direction.