Triton™ X-100, Dulbecco’s Phosphate-Buffered Saline (DPBS), transepithelial electric resistance (TEER) Buffer-GLC, and the MTT kit were provided by MatTek Corporation (Ashland, MA). Phorbol 12-myristate 13-acetate (PMA), ionomycin, Human IL-8/CXCL8 DuoSet ELISA, and Human IL-6 DuoSet ELISA were purchased from R&D Systems (Minneapolis, MN). The LDH Cytotoxicity Detection Kit, Butylated Hydroxytoluene (BHT), Bovine Serum Albumin (BSA), Urea, and Gastric Mucin III were acquired from Sigma-Aldrich (St. Louis, MO). The 8-isoprostane ELISA kit was obtained from Cayman Chemical (Ann Arbor, MI). Potassium Chloride was purchased from JT Baker (Radnor, PA). Sodium Chloride was obtained from Research Products International (Mount Prospect, IL). Calcium Chloride, Disodium Hydrogen Phosphate, and D-(+)-Glucose were purchased from Spectrum Chemical MFG Corp (New Brunswick, NJ). Magnesium Chloride Hexahydrate was acquired from Fisher Scientific (Waltham, MA). Sodium Hydroxide was sourced from Reagents (Charlotte, NC). Hydrochloric Acid (HCl) was purchased from LabChem (Pittsburgh, PA). Alpha Amylase and Lysozyme were acquired from MP Biomedicals (Solon, OH). Acid Phosphatase was obtained from Tokyo Chemical Industry (TCI; Tokyo, Japan).

Preparation of AS, CAS, and CRP1.1 extract as well as extract characterization was carried out by McKinney Specialty Labs (Richmond, VA) in accordance with the International Organization for Standardization - International Electrotechnical Commission (ISO-IEC) 17025:2017 standards. The AS (Supplementary Table S1) was prepared following the procedures outlined in the German standard DIN V Test Method 53160-1 2002-10 (German Institute for Standardization, 2002). The preparation of CAS (Supplementary Table S2) was based on the method described by Chou and Hee (1994), with two modifications. Specifically, 0.68 g/L disodium hydrogen phosphate (Na₂HPO₄) was used instead of 0.58 g/L sodium dihydrogen phosphate (NaH₂PO₄); this brought the total Na⁺ concentration slightly higher to 33.5 mM. The lysozyme concentration was 17,550 U/L, which falls within the reported range for human saliva (Tomašovský et al., 2025).

The CORESTA Swedish-style snus reference product, CRP1.1 (North Carolina State University, Raleigh, NC) was used as a representative oral tobacco product to optimize extraction procedures (Figure 2) and evaluate the relationship between extract concentrations and extraction efficiency (Figure 3). To refine the extraction procedures, two cutting methods were tested (Figure 2A). In the first method, each pouch was cut in half, while in the second method, each pouch was fragmented into pieces ≤4 mm. CRP1.1 (both the pouch content and wrapping material) was then extracted in freshly prepared AS at a concentration of 30% w/v. The concentrations of nicotine and TSNAs were quantified and compared across the two methods.

In addition to optimizing the pouch cutting method, the extraction duration was also refined (Figure 2B). For this purpose, CRP1.1 pouches, prepared according to the results of the cutting method optimization experiment, were suspended in 10 mL of freshly prepared CAS at a concentration of 30% w/v. The suspension was extracted for either 2 h or 4 h, followed by centrifugation and filtration. Longer extraction duration was not further tested based on previous findings where no difference in extraction efficiency was shown between 2 h and 24 h (Zhang J. et al., 2019).

Extracts of CRP1.1 were prepared in either AS or CAS at concentrations of 10%, 20%, and 30% w/v, with pouch preparation and extraction times based on the results of the optimization experiments. Briefly, the suspensions were shaken at 250 ± 25 rpm at 37 °C for 2 h, followed by centrifugation at 4,000 rpm (3,080 × g using an SX4400 rotor) (Beckman Coulter, Indianapolis, IN) for 10 min. The resulting supernatant was filtered once through a sterile 0.2-μm pore-size filter. For the CAS extracts at 20% and 30% w/v, two filters were required during the initial filtration. The combined filtrates were then filter-sterilized a second time using a new 0.2-μm filter unit. The final extracts were aliquoted into smaller volumes and immediately analyzed for nicotine and TSNAs.

Nicotine quantification was conducted using a low-nicotine method developed and validated by McKinney Specialty Labs (2022). In brief, a 100-µL aliquot of CRP1.1 extract was combined with an equal volume of 2 N sodium hydroxide in an amber screw-top bottle and allowed to stand for 15 min. Subsequently, 100 µL of nicotine-d4 (used as the internal standard) was added, followed by addition of 10 mL of methyl tert-butyl ether (MTBE). The mixture was briefly vortexed, and the layers were allowed to separate. Once separation occurred, a 100-µL aliquot of the upper organic layer was transferred to an autosampler vial and diluted with 900 µL of MTBE diluent solution. The final mixture was analyzed using gas chromatography-mass spectrometry (GC-MS; Agilent 7890 Gas Chromatograph coupled with an Agilent 5975 Mass Selective Detector; Santa Clara, CA). Nicotine separation was achieved using an Agilent J&W CAM analytical column (30 m × 0.25 mm ID; 0.25 µm df).

The quantification of TSNAs was conducted following the CORESTA CRM 72 guideline (CORESTA, 2021). A 1.0-mL aliquot of the extract was mixed with 19 mL of 100 mM ammonium acetate solution and shaken mechanically at 210 rpm to facilitate TSNA extraction. The resulting solution was filtered and analyzed using a Waters liquid chromatography system coupled with a tandem mass spectrometer (LC-MS/MS; Milford, MA). TSNA separation was achieved using a Waters Acquity BEH C18 analytical column (2.1 mm × 100 mm; 1.7 µm). Each TSNA in the sample was quantified using deuterated analogues of the analytes as internal standards, including N-Nitrosoanabasine-d4 (d4-NAB-d4), N-Nitrosoanatabine-d4 (d4-NAT), 4-(N-Nitrosomethylamino)-1-(3-pyridyl)-1-butanone-d4 (d4-NNK), and N-Nitrosonornicotine-d4 (d4-NNN).

Normal human gingival fibroblasts (NHGFs) and oral epithelial cells (NHOEs) were isolated from the same donor (donor ID# G29) by MatTek Corporation and cultured as 2D monolayer cells. Cryopreserved NHGFs were recovered at passage 7, cultured in DMEM supplemented with 10% Fetal Bovine Serum (FBS) and maintained at 37 °C and 5% CO₂ until they reached confluence. The day before the exposure, 15,000 cells were seeded into each well of a 96-well plate and incubated overnight at 37 °C and 5% CO₂. AS and CAS were diluted to 5%, 10%, 15%, and 20% v/v in low serum media comprising DMEM and 1% FBS. Cells were exposed to the AS and CAS dilutions for 24 h at 37 °C and 5% CO₂.

Cryopreserved NHOEs were recovered at passage 3 and cultured in EpiLife™ Medium (Gibco, Waltham, MA) and maintained at 37 °C and 5% CO₂ until they reached confluence. The day before the exposure, 25,000 cells were seeded into each well of a 96-well plate and incubated overnight at 37 °C and 5% CO₂. AS and CAS were diluted to 5%, 10%, 15%, and 20% v/v in EpiLife™ Medium. Cells were exposed to the AS and CAS dilutions for 24 h at 37 °C and 5% CO₂.

ORL-200 and ORL-300-FT were produced by MatTek using NHGFs and NHOEs from the same donor as those used in the 2D experiments. ORL-200 is a partial thickness oral epithelial model containing only NHOEs, while ORL-300-FT is a full thickness oral model comprising NHOEs cultured over a collagen matrix embedded with NHGFs. Both models were produced and cultured in accordance with MatTek’s internal Standard Operating Procedures. Prior to treatment, tissues were equilibrated in their respective media (i.e., ORL-200-ASY for ORL-200 and ORL-300-FT-ASY for ORL-300-FT) for approximately 1 h at 37 °C and 5% CO₂. AS and CAS were either undiluted (100%) or diluted to 20%, 50%, and 80% v/v in TEER-Buffer-GLC (DPBS with CaCl₂, MgCl₂, and 3.6 mg/mL glucose; MatTek). Following the equilibration, tissues were transferred to a new 6-well plate containing 0.9 mL of fresh media. A total of 100 µL of either diluted or undiluted AS or CAS was applied to the apical surface of each tissue. Tissues were then treated for approximately 24 h.

Appropriate control groups were also included. For NHGFs and NHOEs, vehicle controls consisted of cultures maintained in the respective media. For the organotypic tissue models, controls included cultures with the apical side exposed to air and cultures exposed to TEER Buffer-GLC from the apical side. Positive controls included cultures exposed to 1% Triton X-100 (positive control for tissue death) and cultures treated with 1 μg/mL PMA/50 ng/mL ionomycin (positive control for 8-isoprostane). A 1% Triton X-100 solution is widely employed as a positive control in cell viability and cell cytotoxicity assays. It has been used by MatTek to establish the ET-50 range for EpiOral quality control (Klausner et al., 2007). The concentrations of Triton X-100 and PMA/ionomycin were selected based on the methodology reported by Keyser (2022). All control groups were treated in a similar manner and for the same duration as the AS- and CAS-exposed groups.

The MTT assay was conducted in accordance with MatTek’s internal SOP (Supplementary Figure S1). Following exposure, media from all samples were collected and stored at −70 °C for downstream analysis, including 8-isoprostane, IL-6, and IL-8 ELISA. The viability of NHGFs and NHOEs was subsequently assessed using the MTT assay, which measures the ability of metabolically active (viable) cells to reduce MTT dye to formazan crystals. Briefly, pre-warmed MTT working reagent (150 µL) was added to each well of the 96-well plates, and the cells were incubated with MTT reagent for 3 h at 37 °C and 5% CO₂. At the end of the incubation, the MTT reagent was removed, and the cells were rinsed once with DPBS. Next, 150 µL of MTT extractant solution (i.e., isopropanol) was added to each well to dissolve the formazan crystals. Extraction proceeded for 2 h on a shaker at room temperature in the dark. The absorbance of the extracted samples was measured at 570 nm using a SpectraMax^® M2 microplate reader (Molecular Devices, San Jose, CA). Two independent experiments were performed for NHGFs and three for NHOEs, with each independent experiment comprising six technical replicates.

Following exposure, the ORL-200 and ORL-300-FT tissues were rinsed with DPBS three times, and the apical surface of the tissues was carefully dried with a sterile swab. The tissues were then transferred to a new plate containing 300 µL of pre-warmed MTT working reagent and incubated for 3 h. The tissues were blotted dry and moved to a fresh 24-well plate. MTT extractant solution (2 mL of isopropanol) was added into the apical compartment and allowed to overflow into the basolateral compartment. The tissues were extracted overnight at room temperature in the dark. At the end of the extraction period, the extractant solution from both the apical and basolateral compartments was combined and thoroughly mixed. A 200-µL aliquot of each of the extracted samples was transferred to a 96-well plate, and absorbance measured at 570 nm using a SpectraMax^® M2 microplate reader. Two independent experiments were performed for both ORL-200 and ORL-300-FT, with each independent experiment comprising four technical replicates.

Cell viability for treated groups was expressed as a percentage relative to the respective vehicle controls (i.e., the media- or TEER Buffer-GLC-treated groups).

Media was analyzed for 8-isoprostane secretion within 10 days of collection using an 8-isoprostane ELISA kit (Cayman Chemical, Ann Arbor, MI) by following the manufacturer recommended protocol. Upon medium collection, 55 µL of the culture media from each sample was mixed with 1 µL of 2.5 mg/mL BHT to prevent oxidative formation of 8-isoprostane during storage. An eight-point standard curve of known 8-isoprostane concentrations was prepared. Then, 50 µL of each standard and culture media sample was added to a Mouse Anti-Rabbit IgG-coated ELISA plate and incubated with 50 µL of 8-isoprostane-AChE Tracer and 50 µL of 8-isoprostane ELISA Antiserum for 18 h at 4 °C. After incubation, the ELISA plate was washed and developed with Ellman’s reagent for approximately 90 min. Absorbance was measured at 405 nm using a SpectraMax^® M2 microplate reader. A four-parameter logistic fit was applied to the standard curve to calculate 8-isoprostane concentrations in the unknown samples. Two independent experiments were performed in NHGFs, ORL-200, and ORL-300 FT, and three for NHOEs. Each 2D cell independent experiment included six technical replicates, while each 3D tissue model experiment comprised four technical replicates.

Secretion levels of 8-isoprostane in treated groups were expressed as fold changes relative to the respective vehicle controls (i.e., the media- or TEER Buffer-GLC-treated groups), providing a normalized measure of the relative effects.

Following exposure, IL-6 and IL-8 concentrations in culture media were quantified using the Human IL-6 ELISA kit and Human IL-8/CXCL8 ELISA kits (R&D Systems, Minneapolis, MN) according to the manufacturer’s recommended protocol. Samples were diluted in Assay Reagent Diluent as needed prior to analysis. Briefly, 96-well microplates were coated with capture antibody overnight at room temperature. The plates were then washed and blocked with 1% BSA in PBS for 1 h at room temperature. After blocking, the plates were washed again, and 100 µL of the eight-point standard curves of known IL-6 and IL-8 concentrations, along with the unknown samples, were added to the plates and incubated for 2 h at room temperature. Following incubation, the plates were washed, and detection antibody was added for 1 h at room temperature, followed by washing and the addition of Streptavidin-HRP for 20 min. After a final wash, 100 µL of TMB Substrate Solution was added for 20 min, followed by 50 µL of 2 N H₂SO₄ as a stop solution. Absorbance was measured at 450 nm with a 570 nm correction using a SpectraMax^® M2 microplate reader. A four-parameter logistic fit was applied to the standard curve to calculate the IL-6 and IL-8 concentrations in the unknown samples. For diluted samples, the calculated concentrations were multiplied by the dilution factor prior to analysis. Two independent experiments were performed in NHGFs, ORL-200, and ORL-300 FT, and three for NHOEs. Each 2D cell independent experiment included three technical replicates, while each 3D tissue model experiment comprised four technical replicates.

Secretion levels of IL-6 and IL-8 in treated groups were expressed as fold changes relative to their respective vehicle controls (i.e., the media- or TEER Buffer-GLC-treated groups), providing a normalized measure of the relative effects.

The data are presented as means ± standard deviation (SD). Normality of distribution and homogeneity of variance were assessed using JMP® Pro (version 17.0.0). Normality was evaluated within the Fit Distributions platform using the Shapiro-Wilk Goodness-of-Fit Test. Homogeneity of variance was assessed with ANOM for Variances (Levene’s test based on absolute deviation from the median, ADM) within the Oneway Analysis platform. This method is robust to outliers and non-normality, enabling comparison of group variances by evaluating group ADM means against the overall ADM mean.

Treatment-related effects were analyzed using GraphPad Prism (version 10.4.3, Lo Jolla, CA). For dataset with a normal distribution, one-way ANOVA followed by Dunnett’s multiple comparison test was performed. For dataset with a non-normal distribution, the Kruskal–Wallis test followed by Dunn’s multiple comparison test was applied. Comparisons were made against the respective vehicle control, i.e., the medium-treated group for NHGFs and NHOEs, and TEER Buffer-GLC for ORL-200 and ORL-300-FT.

Before extracting CRP1.1 at various concentrations, we evaluated the effects of two key factors, i.e., cutting techniques (in AS) and extraction time (in CAS), on extraction efficiency. To examine the impact of cutting techniques, CRP1.1 pouches were either halved or fragmented into pieces of ≤4 mm before being extracted at a concentration of 30% w/v in AS. Both methods demonstrated comparable extraction efficiencies, with nicotine and TSNAs concentrations differing by less than 5%. Specifically, nicotine extraction efficiency was approximately 77%, and TSNAs extraction efficiencies ranged from 60% to 77% (NAB: ∼60%, NAT: ∼64%, NNK: ∼74%, and NNN: ∼77%). Notably, the TSNA extraction efficiency observed in this experiment was significantly higher than those in the subsequent experiments. This difference is likely due to the TSNA analytical method not being fully optimized at the time. As a result, this data is considered valid only for intra-experiment comparisons. Volume recovery was similar between the two methods, averaging around 80% (Supplementary Table S3). These results suggest that fragmenting pouches into smaller pieces does not improve extraction efficiency further. Halving the pouch, therefore, was selected for subsequent studies.

To investigate the impact of extraction time, CRP1.1 cut in half was suspended in CAS at a concentration of 30% w/v and extracted for either 2 h or 4 h. Both extraction durations produced similar extraction efficiencies, with nicotine and TSNAs concentrations differing by less than 8%. Nicotine extraction efficiency was approximately 70%, and TSNAs extraction efficiency ranged from 47% to 67% (NAB: ∼50%, NAT: ∼47%, NNK: ∼50%, and NNN: ∼67%) between the two time points. Volume recovery was also comparable between the two durations, averaging around 40% (Supplementary Table S4). However, the volume recovery rate was lower than that observed in the other two experiments. This reduction may be attributed to two factors. At the end of the extraction, a significant amount of extract was retained by the filter disc. Additionally, the extraction was conducted in a relatively smaller volume (10 mL). Despite the lower recovery rate, the data are considered valid for intra-experiment comparison. These findings indicate that extending the extraction duration does not improve extraction efficiency further. Therefore, the 2-h extraction duration was selected for subsequent studies.

The difference in extraction efficiency between AS and CAS was examined by extracting CRP1.1 in both solvent vehicles at concentrations of 10%, 20%, and 30% w/v using optimized procedures as described above. Overall, a decrease in nicotine and TSNA extraction efficiency was observed in both AS and CAS extracts as extract concentrations increased (Table 1). Nicotine extraction efficiency decreased from approximately 83% at 10% w/v to 66% at 30% w/v in both AS and CAS. However, the reduction in TSNA extraction efficiencies was more pronounced in AS (dropping from approximately 85%–90% to around 50%) compared to in CAS (declining from approximately 60%–67% to 45%–60%). Notably, TSNA extraction efficiencies in AS extracts were consistently higher at 10% and 20% w/v concentrations compared to those in CAS extracts. However, both solvent vehicles performed similarly at 30% w/v. In addition, volume recovery was comparable between AS and CAS extracts, with both solvent vehicles exhibiting a decreasing trend from 88% at 10% w/v to 74% at 30% w/v.

Cell viability was assessed using the MTT assay 24 h after treatment (Figure 4). The viability of AS-treated NHGFs and NHOEs consistently remained around 100% (Figure 4A, grey bars). While a statistically significant increase in viability was observed from 5% to 15% v/v of AS in NHOEs, the change was marginal (less than 10%) and therefore not considered biologically relevant. Similarly, the viability changes in CAS-treated NHGFs and NHOEs were also within 10%, except at 5% v/v of CAS in NHOEs (Figure 4A, black bars). However, none of these responses reached statistical significance.

After 24 h of exposure, the viability of AS-exposed ORL-200 and ORL-300-FT tissue models remained above 90% at all test concentrations (Figure 4B, grey bars). Similarly, the viability of both organotypic tissue models exposed to CAS was above 90%, except for the ORL-300-FT model at 100% CAS (Figure 4B, black bars). The reduction in viability observed at 100% CAS in this model was statistically significant. However, this decrease was less than 15%, suggesting that it may not have a meaningful biological impact.

Oxidative stress induction was assessed by measuring 8-isoprostane secretion in the culture media 24 h post-exposure. To facilitate comparisons across different in vitro models, 8-isoprostane levels were normalized to their respective vehicle controls. In NHGFs and NHOEs, exposure to AS did not alter 8-isoprostane levels (Figure 5A, grey bars). In contrast, CAS exposure increased 8-isoprostane secretion by approximately 15% at the two highest concentrations in NHGFs (Figure 5A, left panel, black bars). In NHOEs, CAS exposure resulted in an approximately 20% increase in 8-isoprostane release at the two lowest concentrations (Figure 5A, right panel, black bars). These CAS-induced responses in both cell type were statistically significant.

In ORL-200, neither AS nor CAS exhibited a dose-dependent effect on 8-isoprostane release, although CAS induced a statistically significant increase of approximately 50% at a concentration of 100% (Figure 5B, left panel). However, in the ORL-300-FT model, AS exposure led to an approximately 50% increase in 8-isoprostane secretion at a concentration of 50% v/v (Figure 5B, right panel, grey bars). Given the absence of a dose-response relationship, this increase is likely attributable to intrinsic variability within the organotypic tissue models. In contrast, CAS exposure induced a significant increase of approximately 1.8-fold in 8-isoprostane secretion at a concentration of 100% v/v in the ORL-300-FT model (Figure 5B, right panel, black bars).

The modulation of inflammatory responses by AS and CAS was evaluated by measuring the release of IL-6 and IL-8 into the media during a 24-h exposure. The effects of AS and CAS on IL-6 secretion differed between the two cell types. In NHOEs, AS and CAS had opposing effects: AS significantly increased IL-6 secretion, while CAS inhibited it (Figure 6A, right panel). In NHGFs, CAS elevated IL-6 release in a concentration-dependent manner (Figure 6A, left panel, black bars). In contrast, in ORL-300 FT, both AS and CAS significantly suppressed IL-6 secretion, suggesting potential anti-inflammatory effects (Figure 6B, right panel). Neither AS nor CAS induced significant changes in ORL-200 (Figure 6B, left panel).

The impacts of AS and CAS exposure on IL-8 secretion also varied across the tested models. AS markedly increased IL-8 release in NHOEs by up to 15-fold in a concentration-dependent manner (Figure 7A, right panel, grey bars). CAS significantly stimulated IL-8 secretion in both NHGFs and NHOEs, with a more pronounced effect observed in NHOEs compared to NHGFs (Figure 7A, left panel vs. right panel, black bars). However, neither AS nor CAS significantly modulated IL-8 secretion in the organotypic tissue models under the tested conditions (Figure 7B). Even though some responses reached statistical significance, the changes are likely due to intrinsic variability within the organotypic tissue models, given the absence of a dose-response relationship.