Vape-Associated lncRNA Transcript 1 (VALT1) Amplifies the Tumorigenic Effects of e-Cigarette Vapor in Lung Epithelial Cells

Abstract

Background/Objectives: Lung cancer remains a major global health burden, largely driven by cigarette use. Although electronic cigarettes (e-cigarettes) are viewed as safer alternatives due to their reduced chemical load, growing evidence shows their vapor can disrupt cellular transcriptomes, including long noncoding RNAs (lncRNAs). In this study, we examined the regulation and function of vape-associated lncRNA transcript 1 (VALT1), a novel transcript upregulated in the oral transcriptomes of e-cigarette users and similarly elevated in non-small-cell lung cancer (NSCLC) tumors. Methods: Publicly available RNA-seq datasets were analyzed, and VALT1 was identified as an e-cigarette-responsive lncRNA. Its dose-dependent induction by e-cigarette smoke extract (eCSE) and cytoplasmic localization were confirmed via RT-qPCR. Its effects on cancer-associated phenotypes including proliferation, ROS detoxification, resistance to apoptosis, migration, cytoskeletal disorganization, and nuclear remodeling were assessed through overexpression and siRNA-mediated knockdown in A549 and BEAS-2B cells. Results: Acute eCSE exposure induced a biphasic, dose-dependent increase in VALT1 expression, accompanied by enhanced proliferation, ROS detoxification, apoptosis resistance, migration, cytoskeletal disorganization, and nuclear remodeling in A549 cells. VALT1 overexpression reproduced these phenotypes in both cell lines without eCSE treatment, whereas knockdown attenuated them. VALT1 promoted survival under cytotoxic stress in A549 but not BEAS-2B cells. Conclusions: These findings support an active role for VALT1 as an e-cigarette vapor-upregulated transcript that contributes to its phenotypic readout and enhances cellular survival under extracellular chemical stress—thereby aggravating tumorigenic phenotypes even in the absence of mutations that contribute to malignant transformation.

1. Introduction

Lung cancer, the second most prevalent cancer type, is the leading cause of cancer mortality worldwide [1]. Chronic exposure to tobacco cigarette smoke, whether via direct usage or side stream inhalation, is a significant risk factor for lung oncogenesis [2,3]. Exogenous carcinogens in cigarette smoke are known to potentiate genetic perturbations, particularly DNA adduct formation and subsequent mutagenesis [4]. Further mutational diversification while maintaining key driver mutations in vital cellular pathways drives the transformation of a healthy cellular population into a malignant neoplasm [5,6].

While tobacco-induced genetic disruptions have been established as a crucial factor in lung oncogenesis, other factors have gained increasing attention due to their emerging roles in maintaining cellular homeostasis. Noncoding RNAs (ncRNAs), such as microRNAs (miRNAs), circular RNAs (circRNAs), and long noncoding RNAs (lncRNAs), are aberrantly expressed in multiple tumor types. Furthermore, exogenous environmental stimuli, including tobacco smoke exposure, have been shown to dysregulate the ncRNA landscape of cells independent of the genetic lesions it induces. In many instances, these stimuli are translated into transient intracellular cues through the binding of ligands onto their respective extracellular signaling receptors. For example, the binding of nicotine, a key component in tobacco and e-cigarette formulations, onto the homomeric α7 nicotinic acetylcholine receptor (α7-nAChR) promotes transient and reversible effects on various cancer hallmarks such as proliferation and inhibition of apoptosis [7,8]. Pertinent cellular phenotypes are expressed through the acute cross-activation of signaling cascades, including the PI3K/Akt and Ras/Raf/Mek/Erk pathways [9]. This intersection of environmental stress and transcriptomic plasticity reinforces the role of ncRNAs as crucial mediators of cancer progression. Indeed, many lncRNAs dysregulated by transient exogenous stimuli have been demonstrated to aggravate in vitro and in vivo oncogenic phenotypes. For instance, the cigarette smoke-upregulated lncRNA lung cancer progression-associated transcript 1 (LCPAT1) [10] and smoke and cancer-associated lncRNA 1 (SCAL1) [11] are implicated in the attenuation of the DNA damage response (DDR) pathway and intracellular reactive oxygen species (ROS) detoxification, respectively, to promote tumor progression in lung epithelia.

The emergence of these epigenetic factors has raised concerns about the purported safety of electronic cigarettes or e-cigarettes. While likely safer than tobacco due to their simpler chemical composition, e-cigarettes are not risk-free. The potential risk posed by potent tumor promoters like nicotine—a highly enriched ingredient in e-cigarette fluid formulations and a precursor for the formation of tobacco-specific nitrosamines [12]—and other suspected carcinogens in e-cigarette smoke cannot be ignored [13]. Likewise, the unique chemical interactions that arise from the vaporization of e-cigarette fluid, a complex mixture that includes nicotine, solvent carriers (propylene glycol and glycerol), flavorings, heavy metals, and volatile organic compounds, open an avenue to examine the potential cytological fingerprint of e-cigarettes, which may be distinct from the known effects of nicotine treatment or cigarette smoke exposure alone [14]. While in vivo correlations between chronic e-cigarette smoke exposure and lung tumorigenesis remain highly lacking due to the relative novelty of e-cigarettes, preliminary evidence from multiple studies has already demonstrated that e-cigarette smoke exposure can induce global transcriptomic alterations that translate to significant changes in cellular phenotypes. For instance, transcriptomic studies led by Tommasi and colleagues [15] have noted the preponderance of dysregulated ncRNAs, including lncRNAs, in the oral epithelia of e-cigarette users. Furthermore, in vitro studies on e-cigarette smoke exposure to respiratory epithelia have demonstrated that it promotes tumorigenic phenotypes, including proliferation [16], migration [17], and epithelial-to-mesenchymal transition [18].

Here, we characterize VALT1 (Vape-Associated LncRNA Transcript 1), a novel transcript previously annotated as AC016773.1 [15] that is significantly upregulated in the oral mucosa of e-cigarette users. The VALT1 locus encodes an 802 nt long intergenic noncoding RNA (lincRNA) upregulated in the oral epithelial cells of e-cigarette users when compared to nonsmoker controls by a factor of 5.92 [15]. Interestingly, multiple studies have demonstrated its parallel upregulation in other types of cancer, such as hepatocellular carcinoma [19], clear cell renal cell carcinoma [20,21,22,23], bladder cancer [24], and cervical cancer [25]. Studies also show that VALT1 expression levels negatively correlate with overall patient outcomes [22]. More recently, preliminary in vitro studies in multiple myeloma [26] and clear cell renal cell carcinoma [27], as well as in vivo studies in prostate cancer [28], have pointed toward the physiological relevance of VALT1 as a phenotypic driver of tumorigenesis. To further define its functional significance, we employed both exogenous overexpression and siRNA-mediated knockdown in two complementary models of human lung epithelia, A549 lung adenocarcinoma cells and BEAS-2B normal bronchial epithelial cells, and interrogated how VALT1 modulation influences the phenotypic outcomes associated with e-cigarette vapor-driven tumorigenicity.

2. Results

2.1. VALT1 Is Upregulated in e-Cigarette Smoker Transcriptomes and NSCLC Tumors

Analysis of RNA-seq data from Tommasi et al. [15] confirmed VALT1 (GenBank Accession No. AC016773.1) as the most upregulated long intergenic noncoding RNA (lincRNA), exhibiting a 5.92-fold increase in expression in the oral epithelia of e-cigarette smokers relative to the nonsmoking cohort (Figure 1a). Genomically, the VALT1 locus spans an 802 bp region on the short arm of chromosome 4 and bears hallmarks of active transcriptional activity, including DNase I hypersensitivity and H3K27Ac enrichment (Figure S1a). Additionally, cap analysis of gene expression (CAGE) from the FANTOM5 project revealed a transcription start site (TSS) immediately proximal to the VALT1 locus (Figure S1b).

Publicly available RNA-seq datasets were accessed from 32 cancer types cataloged in The Cancer Genome Atlas (TCGA). Analysis through the DeepBase v3.0 database [29] revealed that VALT1 is significantly upregulated in 9 of the 24 TCGA cancer types eligible for differential gene expression analysis (those with at least one matched normal tissue) (Figure 1b). Notably, VALT1 shows marked overexpression in multiple epithelial carcinomas, including lung adenocarcinoma (LUAD) (Figure 1c) and lung squamous cell carcinoma (LUSC) (Figure 1d), both of which represent non-small-cell lung cancer (NSCLC) subtypes (Figure 1b). Other cancer types where VALT1 is significantly upregulated include breast invasive carcinoma (BRCA), colon adenocarcinoma (COAD), head and neck squamous carcinoma (HNSC), kidney renal clear cell carcinoma (KIRC), kidney renal papillary cell carcinoma (KIRP), liver hepatocellular carcinoma (LIHC), prostate adenocarcinoma (PRAD), and stomach adenocarcinoma (STAD) (Figure S2). Even when only accounting for matched adjacent normal and tumor samples from the same patient, the upregulation of VALT1 remains statistically significant in both LUAD (Figure 1e) and LUSC (Figure 1f), suggesting that VALT1 upregulation can be directly linked to NSCLC pathogenesis. Lastly, VALT1 lncRNA upregulation is sustained in both LUAD (Figure 1g) and LUSC across cancer stages (Figure 1h), further hinting toward the role of VALT1 in maintaining tumorigenic phenotypes.

2.2. VALT1 Is a Cytoplasmic lncRNA Upregulated in eCSE-Treated A549 Cells

To validate if acute eCSE exposure upregulates VALT1 in A549 lung adenocarcinoma cells, a standard model cell line for studying the oncogenic effects of cigarette smoke extract (CSE) and eCSE in NSCLC, different dilutions of eCSE extract in reduced serum were used to treat cells. As shown in Figure 2a, higher concentration of eCSE elevated VALT1 expression up to the 10-fold dilution. However, higher concentrations (0.5× and 1× eCSE-DMEM) led to reduced expression, indicating a biphasic dose-dependent response in which VALT1 is upregulated at lower to intermediate concentrations but downregulated once a higher threshold is surpassed. To further validate the transient nature of eCSE-dependent regulation of VALT1, cells were treated with 0.1× eCSE-DMEM for 6 h and then allowed to recover in eCSE-free reduced serum medium for 18 h. VALT1 expression declined upon withdrawal compared to cells continuously exposed for 24 h (Figure 2b), though both conditions remained elevated relative to the untreated baseline. The recovery of VALT1 levels upon eCSE withdrawal demonstrates eCSE as a transient and reversible inducer of VALT1 expression.

Given the immediate genomic proximity of VALT1 to its parent gene, SLBP, an established carcinogenic driver that is also upregulated in NSCLC (Figure S3), the potential for cis-regulation via a shared promoter region was examined. As shown in Figure S4, eCSE does not alter SLBP expression up to the twofold dilution titer. Furthermore, eCSE-modulated VALT1 and SLBP levels are not significantly correlated (Figure 2c). VALT1 expression, though significantly positively correlated with SLBP expression in LUAD and LUSC TCGA datasets (Figure 2d,e), does not account for a substantial proportion of SLBP expression (R² < 0.1). Taken together, these results suggest that potential cis-regulation of SLBP expression by VALT1 is insufficient to explain the potential phenotypic effects of eCSE-mediated VALT1 dysregulation.

To gain insights into where and how VALT1 mediates its phenotypic effects, its cellular localization was interrogated via RNA fractionation. Using MALAT1 and NEAT1 lncRNAs as nuclear RNA controls and CypA mRNA as the cytoplasmic RNA control, the cytoplasmic localization of VALT1 was established in both A549 (Figure 2f) and BEAS-2B (Figure 2g), an immortalized normal human bronchial cell line, consistent with earlier reports of its cytoplasmic localization [26,28].

2.3. VALT1 Expression Phenocopies eCSE-Induced Proliferative Phenotypes in A549 and BEAS-2B Cells

To compare the effects of eCSE treatment and VALT1 expression on proliferative phenotypes, the lung cancer cell line A549 and its normal counterpart BEAS-2B were either treated with eCSE, transfected with VALT1 expression construct, or transfected with VALT1-specific siRNA. Two orthogonal proliferation assays, one measuring nuclear DNA content and another measuring collective metabolic activity, were done. Experiments confirming overexpression and knockdown of VALT1 in transfected A549 and BEAS-2B cells are shown in Figure S5. Experiments confirming acceptable transfection efficiency in A549 and BEAS-2B via parallel transfection of a GFP-encoding plasmid and Texas Red-conjugated oligonucleotides as transfection controls for plasmids and siRNAs, respectively, are shown in Figure S6. Lastly, as shown in Figure S7, VALT1 overexpression and knockdown do not affect endogenous SLBP expression, indicating that the phenotypic readout of cells upon VALT1 expression and knockdown can be decoupled from any potential cis-regulation of the adjacent SLBP locus and its complementary phenotypic effects.

A549 cells treated with eCSE for 48 h and 72 h demonstrated a marked increase in proliferative capacity in metabolic MTS assays (Figure 3a,b). To isolate the effect of VALT1 expression from other transcripts that could be dysregulated by eCSE treatment, A549 and BEAS-2B cells were transfected with pTargeT™-VALT1 construct in the absence of eCSE. Overexpression of VALT1 alone increased proliferation in both A549 (Figure 3c,d) and BEAS-2B (Figure 3e,f) cells at 48–72 h post-transfection compared to the empty vector control setup.

To further demonstrate that endogenous VALT1 alone sufficiently contributes to observed proliferative phenotypes, endogenous VALT1 transcripts were knocked down by transfecting two VALT1-specific siRNAs (designated as siVALT1-1 and siVALT1-2) in A549 and BEAS-2B cells. Depletion of endogenous VALT1 attenuated proliferation in A549 (Figure 3g,h) and BEAS-2B cells (Figure 3i,j) 48–72 h post-transfection.

To further validate the results of MTS assays, DNA-based proliferation measurements using the CyQUANT assay were done for an orthogonal readout. CyQUANT reproduced the trends observed with MTS (Figure 4), indicating that both eCSE exposure and VALT1 expression consistently enhance proliferative capacity irrespective of metabolic influences. Taken together, these experiments corroborate the potential of VALT1 lncRNA to amplify tumorigenic phenotypes by promoting proliferation, with or without eCSE stimulation.

2.4. VALT1 Facilitates LPS-Induced ROS Detoxification in A549 Cells but Not BEAS-2B Cells

To determine if eCSE and VALT1 can activate intrinsic detoxification mechanisms, intracellular ROS levels in eCSE-treated, VALT1-overexpressing, and VALT1-depleted setups were visualized using DCFDA staining after lipopolysaccharide (LPS) exposure (Figure 5a–e). Because LPS drives maximal ROS responses, subsequent decreases reflect a shift toward antioxidant activity, as transformed cells maintain elevated redox-buffering capacity to withstand persistent oxidative and electrophilic stress.

Indeed, LPS-induced intracellular ROS accumulation was attenuated by low to intermediate concentrations of eCSE in a dose-dependent manner, after which ROS levels rose again, closely paralleling the biphasic pattern of endogenous VALT1 expression upon eCSE treatment (Figure 5f). In A549 cells, VALT1 overexpression without eCSE treatment attenuated LPS-induced ROS accumulation (Figure 5g), decoupling its effects from other transcripts that may be dysregulated by eCSE, while VALT1 knockdown showed opposite effects and increased ROS accumulation (Figure 5h). Interestingly, VALT1 overexpression (Figure 5i) and knockdown (Figure 5j) did not have a significant effect on ROS levels in the non-transformed background of BEAS-2B cells.

2.5. VALT1 Rescues A549 and BEAS-2B Cells from MSB-Induced Cell Death

The capacity of cancer cells to activate antioxidant defenses limits ROS-mediated induction of apoptosis via the intrinsic pathway. This pathway is highly regulated by mitochondrial permeability, which can be visualized using TMRM, a dye that selectively accumulates in mitochondria with intact membrane potential (ΔΨm), in A549 (Figure 6a–c) and BEAS-2B (Figure 6d,e).

eCSE was shown to attenuate menadione sodium bisulfite (MSB)-induced loss of ΔΨm among adherent A549 cells in a concentration-dependent manner up to the 10-fold dilution, after which ΔΨm begins to decline (Figure 6a,f). These trends correlate well with eCSE-mediated VALT1 induction in Figure 2a. Notably, ectopic overexpression of VALT1 in the absence of eCSE recapitulates the effects of lower eCSE concentrations in this assay (Figure 6b,d), whereas VALT1 knockdown mirrors the effects of higher eCSE concentrations (Figure 6c,e) in both A549 and BEAS-2B cells.

These observations have been verified through the staining of adherent A549 cells with the CellEvent™ Caspase-3/7 Green Detection Reagent, which only selectively stains apoptotic cells (Figure 7a–c), with corresponding analyses shown in Figure 7d–f. MSB-treated BEAS-2B cells, on the other hand, were not imaged with CellEvent™ Caspase-3/7 Green Detection Reagent due to apoptotic detachment.

To examine whether changes in ΔΨm translate to increased resistance to MSB-induced cytotoxicity, flow cytometry was utilized to discriminate between apoptotic and necrotic cell death, which provides the added benefit of capturing non-adherent cells. The utilization of a short incubation time in MSB (3 h) ensured that apoptotic cells will mostly be in their early stages and will only stain positive in annexin V (AV) conjugated with AF488 but not propidium iodide (PI) in AV/PI staining via flow cytometry. Flow cytometry results largely recapitulated the TMRM findings in A549 cells (Figure 8a–c) and BEAS-2B (Figure 8d,e). In A549, treatment of eCSE increased cell viability up to the 10-fold dilution titer, as reflected by higher proportions of AV–/PI– cells, after which viability declined (Figure 8f). Since necrotic populations (PI+) remained relatively constant across titers, these data suggest that eCSE specifically suppresses apoptotic cell death, a non-inflammatory mode of cell death, in A549 cells. In the same assay, exogenous VALT1 overexpression alone conferred protection against MSB-induced apoptosis in A549 cells (Figure 8g), whereas VALT1 knockdown marginally enhanced apoptotic susceptibility (Figure 8h), as reflected by a slight increase in AV+/PI– populations and a decrease in AV–/PI– cells among siVALT1-transfected setups. This suggests that eCSE-upregulated VALT1 contributes to the cytoprotective phenotype of eCSE exposure in A549 cells, even in the presence of cytotoxic factors like eCSE.

Interestingly, in BEAS-2B cells, VALT1 overexpression also mitigated MSB-induced cytotoxicity (Figure 8i) but primarily by reducing necrotic rather than apoptotic cell death, whereas knockdown attenuated necrosis (Figure 8j), suggesting overwhelming oxidative injury. The observed trends in BEAS-2B cytotoxicity align with the pattern of DCFDA assay results, which revealed higher ROS accumulation in BEAS-2B cells compared to A549 upon ROS saturation using LPS.

2.6. VALT1 Promotes Migration of A549 and BEAS-2B Cells

The effects of eCSE treatment and VALT1 expression on the migration of A549 and BEAS-2B cells were investigated via the wound healing assay (Figure 9a–e). Migration rates of A549 cells exposed to varying concentrations of eCSE followed a similar trend to the RT-qPCR results of VALT1 expression in Figure 2a, where migration, as indicated by percent wound closure, increased up to the 10-fold dilution, then progressively declined at higher concentrations of eCSE (Figure 9f).

Furthermore, in A549 cells, VALT1 overexpression significantly enhanced migratory capacity (Figure 9g), whereas VALT1 knockdown markedly reduced wound closure rates (Figure 9h). These results are consistent with the observed upregulation of VALT1 in response to eCSE exposure and suggest that VALT1 promotes migration in lung cancer cells. In BEAS-2B cells, a similar trend was observed, with VALT1 overexpression increasing (Figure 9i) and VALT1 knockdown impairing migratory ability (Figure 9j). Taken together, these findings indicate that VALT1 promotes cellular migration in both cancerous and noncancerous lung epithelial cells, supporting a broader role for VALT1 in regulating lung epithelial cell motility.

2.7. VALT1 Promotes Nuclear Deformation Secondary to Extensive Actin Cytoskeletal Remodeling

To assess the impact of eCSE exposure and VALT1 expression on the invasive behavior of A549 cells, the actin cytoskeleton was stained with phalloidin followed by fluorescence microscopy.

Cytoskeletal alterations became progressively evident with increasing eCSE concentrations and were detectable even at the lowest titer (0.01× eCSE), as shown in Figure 10a. Tunneling nanotube (TNT)-like structures were present in all setups, consistent with the transformed nature of A549 cells. Nuclear aberrations were also observed, even in the untreated baseline, which is a documented feature for most cancerous cell lines. Nonetheless, untreated A549 cells maintained a relatively stable phenotype, with polygonal morphology, strong adherent junctions, and highly fibrillar stress fibers that tend to culminate at focal adhesions. Meanwhile, A549 cells treated with lower and intermediate titers (0.01× to 0.1× eCSE) demonstrated increasingly apparent perturbations in cellular morphology, particularly in the cortical actin cytoskeleton. There was increasing prominence of lamellipodia, invadopodia, filopodia, TNT-like projections, and membrane ruffling, which are structures that tend to underlie invasive and migratory cell behavior. While also present throughout all treatments, nuclear abnormalities, including multinucleation, become increasingly evident in higher concentrations of eCSE, which are often reflective of hyperproliferative states and cytogenetic instability.

VALT1 overexpression alone (Figure 10b) was shown to recapitulate many aggressive phenotypes observed with eCSE treatment in A549 cells, suggesting that VALT1 predisposes the formation of these transient cytoskeletal features. Conversely, while VALT1 knockdown did not abrogate the formation of the aforementioned structures, the depletion of endogenous VALT1 resulted in the formation of highly fibrillar actin cytoskeletal structures that suggest stabilization of the actin cytoskeleton (Figure 10c). Collectively, these results suggest VALT1 as a contributor of eCSE-induced changes in cellular morphology that bias cells toward tumorigenicity.

On the other hand, BEAS-2B cells displayed relatively stable cytoskeletal architecture marked by well-aligned stress fibers and highly anisotropic cortical actin arrangements consistent with their non-transformed state. VALT1 manipulation yielded a similar set of results for BEAS-2B cells. Stress fibers appeared more prominent in control setups when compared to the VALT1-overexpressing setups (Figure 11a). Furthermore, VALT1-overexpressing cells displayed phenotypic hallmarks consistent with tumorigenic differentiation, which include a general shift toward a more squamous, A549-like morphology and the flattening of the cell body. This was further supported by the loss of highly anisotropic stress fibers that anchor the cell to the substrate. Nuclear aberrations, including multinucleation, also became more apparent. On the other hand, endogenous VALT1 depletion shifted BEAS-2B toward a spindle-like morphology that is often indicative of cellular stress (Figure 11b).

Overall trends in cytoskeletal and nuclear remodeling were corroborated by quantifying actin anisotropy and nuclear area. Decreasing anisotropy, which is indicative of increasing cytoskeletal disorganization, was observed with increasing eCSE up to 0.5× (Figure 12a). The decrease in actin anisotropy coincided with increasing punctation of the cytoplasm and crisscrossing of contractile actin fibers, which reduce fibrillar directionality. Likewise, the average nuclear area increased until the 0.5× dilution (Figure 12b), likely reflecting reduced rigidity and dorsoventral flattening typical of transformed cells on a two-dimensional substrate. Notably, nuclear eccentricity concurrently decreased, indicating rounder nuclei (Figure 12c). As VALT1 expression correlates with morphological changes associated with eCSE exposure, these findings suggest the synergistic role of VALT1 in promoting the disorganization and nuclear remodeling that facilitate neoplastic migration.

To determine whether VALT1 contributes to eCSE-driven changes in actin organization and nuclear morphology, cells were transfected with VALT1 expression constructs in the absence of eCSE. Ectopic VALT1 expression caused a decrease in actin anisotropy (Figure 12d), an increase in nuclear area (Figure 12e), and a tendency toward eccentric nuclei (Figure 12f), whereas VALT1 knockdown produced the opposite phenotype marked by highly anisotropic stress fibers (Figure 12g), smaller nuclear areas (Figure 12h), and a shift toward circular nuclei (Figure 12i). While parallel VALT1 overexpression experiments in BEAS-2B confirmed similar shifts in actin anisotropy (Figure 12j) and nuclear area (Figure 12k), observed changes in nuclear eccentricity, which trended toward eccentric nuclei (Figure 12l), diverged from observations made in A549 cells. Conversely, VALT1 knockdown in BEAS-2B increased fibrillar anisotropy (Figure 12m) and decreased nuclear area (Figure 12n), while also promoting nuclear roundedness (Figure 12o). Collectively, these results suggest that VALT1 modulates cytoskeletal remodeling and cell shape dynamics in a cell type-specific manner.

When interpreted alongside the emergence of migratory and invasive cytoskeletal structures, the combined shifts in actin anisotropy and nuclear morphology are suggestive of a shift toward more migratory phenotypes consistent with the results of wound healing assays in both cells.

3. Discussion

LncRNAs constitute the majority of human transcriptomes [30]. Their dynamic expression patterns across tissues and developmental stages as well as their aberrant expression in pathological states suggest a role in both normal physiology and disease pathogenesis [31]. More recently, however, lncRNAs are being investigated for their response to environmental stressors. Changes in their expression are being linked to the perturbation of key transcriptional networks that can drive phenotypic changes in cells [32].

E-cigarette-mediated pathogenicity has been a recent topic of discussion due to mounting evidence that it can significantly alter the transcriptome of respiratory epithelia—this is against its purported safety relative to traditional tobacco cigarettes. One of the earliest pre-clinical studies comparing the differential effects of cigarette and e-cigarette smoke using in vitro transcriptomic profiling of respiratory epithelial models revealed that pathways related to DNA damage response, cell cycle regulation, antioxidant defense, and cellular adhesion were significantly dysregulated following e-cigarette exposure, yet without inducing the same degree of cytotoxicity observed with mainstream tobacco smoke [14]. While not suggestive of e-cigarette-mediated carcinogenicity by itself, more recent clinical studies point toward a concerning trend: e-cigarette usage seems to be higher among those diagnosed with lung cancer compared to the known prevalence in the general population [33]. However, definitive associations have yet to be demonstrated thoroughly through longitudinal and cross-sectional studies, which have reported mixed results so far [34]. The lack of definitive evidence linking e-cigarettes to lung cancer risk is not surprising given the recency of vape use and the known latency period of lung oncogenesis among established risk factors such as tobacco smoking, which can take as long as 30 years [35].

Tommasi and colleagues [15] identified differentially expressed transcripts in the oral epithelia of nonsmokers, cigarette smokers, and e-cigarette users. Molecular pathway and functional network analyses revealed that cancer is the top disease associated with the deregulated genes among e-cigarette users. Notably, the noncoding transcriptome of e-cigarette users was more dysregulated when compared to that of cigarette smokers when measured relative to the overall transcriptome: 26% for e-cigarette users and 17% for cigarette smokers when compared to the nonsmoking cohort. Taken together, these findings suggest that e-cigarette exposure induces a distinct pattern of transcriptomic reprogramming, particularly in the noncoding RNA landscape, when compared to cigarette smoke exposure alone.

Among the dysregulated transcripts in the oral transcriptome of e-cigarette users is a previously unnamed and uncharacterized transcript, originally annotated as AC016773.1 (now under AC016773.2) and provisionally named “RP11-572O17.1”. The genomic locus encodes for an 802 nt intergenic lncRNA in the short arm of chromosome 4, which was subsequently found to be upregulated in the oral transcriptome of e-cigarette users by a factor of 5.92 (p < 0.0005). It ranks as the third most upregulated lncRNA and the most upregulated lincRNA overall (Table S1). Publicly available RNA-seq data from TCGA accessed via DeepBase v3.0 likewise reveal that AC016773.1 is significantly overexpressed in various carcinomas, including NSCLC subtypes LUAD and LUSC, when compared to matched normal tissues, with expression correlating with cancer stage. Hence, publicly available RNA-seq data reveal that AC016773.1 expression in respiratory malignancies and oral epithelia is coincident.

To validate the potential regulatory relationship between eCSE exposure and VALT1 expression, an in vitro approach was used in this study to allow precise concentration-dependent manipulation of exogenous stimuli. The transformed background of the A549 lung adenocarcinoma cell line provided a model by which to interrogate the effects of e-cigarette smoke-mediated tumorigenicity. Moreover, A549 cells share notable molecular similarities with oral cancers, which are frequently employed as practical surrogates for lung cancers in clinical settings [36]. A wide range of eCSE titers was used, as literature suggests a certain threshold of CSE flips the readout from cytoprotective to cytotoxic [16,37]. The same observation has been made for nicotine treatment alone [38]. Interestingly, flavorant compounds [39] and carriers found in e-cigarette formulations [40] have been found to exert a general cytotoxic effect. Due to these opposing factors, using a wide range of eCSE titers would be more instructive for both AC016773.1 expression patterns and the phenotypic readout of e-cigarette treatment, especially given its diverse chemical composition. Furthermore, employing a broad spectrum of eCSE concentrations helps model differences in e-cigarette consumption among individuals, which may differ by as much as two orders of magnitude between the least and most frequent users [41]. Utilizing this approach, a concentration-dependent expression pattern of AC016773.1 until the 10-fold dilution of concentrated eCSE extract was observed. The biphasic pattern of regulation of VALT1, brought about by its downregulation in higher eCSE titers, may be due to negative compensatory effects or broader changes in RNA expression brought about by cytotoxic or oxidative eCSE components [42]. Although 6 h of eCSE exposure induced only a modest twofold increase, expression persisted long after stimulus withdrawal, implying that chronic exposure may stabilize its upregulation. Based on these observations, AC016773.1 was renamed as vape-associated lncRNA transcript 1 (VALT1) to highlight its specific regulation by acute e-cigarette exposure.

However, the close genomic proximity of VALT1 to the adjacent protein-coding SLBP locus, previously implicated in increased proliferation and epithelial–mesenchymal transition (EMT) within respiratory epithelia [43], raises the possibility that phenotypic effects attributed to VALT1 manipulation may be influenced by cis-regulatory interactions. For instance, modulation of VALT1 expression could conceivably alter chromatin accessibility within the shared locus, thereby indirectly affecting SLBP transcription [44]. Although TCGA datasets reveal a marginal yet statistically detectable correlation between SLBP and VALT1 transcript levels at the population level, eCSE exposure experiments demonstrate that SLBP mRNA expression remains stable across most concentrations, with concurrent downregulation observed only at the highest titer. Importantly, the absence of a robust correlation across treatment conditions argues against strict reciprocal coregulation and instead suggests parallel responsiveness to shared upstream regulatory inputs that may be coordinately activated by eCSE treatment at higher concentrations. Taken together, these observations support the interpretation that eCSE-mediated VALT1 modulation and its phenotypic readouts are not solely attributable to cis-regulation of the adjacent SLBP locus.

The upregulation of VALT1 in carcinomas, including LUAD and LUSC, suggests that it may confer a survival advantage to transformed cells. Overexpressing VALT1 reproduced the key phenotypes seen with eCSE exposure, whereas silencing VALT1 markedly alleviated them. These effects were consistent in both A549 lung adenocarcinoma cells and BEAS-2B normal bronchial cells, suggesting that the phenotypic changes reflect VALT1 function rather than cell-type–dependent readouts. The phenotypic assays revealed that eCSE treatment and its subsequent upregulation of VALT1 promote proliferation, ROS detoxification, resistance to apoptosis, migration, and extensive actin cytoskeletal remodeling in vitro. This behavior largely mirrors the effects of CSE and nicotine treatment alone, as previously described. Furthermore, these observations are consistent with reported bidirectional phenotypic outcomes upon eCSE treatment, likely reflecting competing effects by cytoprotective components such as nicotine [38], toxicants such as e-cigarette flavorants [39], and carriers such as propylene glycol [40]. Importantly, these phenotypes are unlikely to result from indirect trans-regulatory effects of VALT1 perturbation on the SLBP locus, as both exogenous overexpression and siRNA-mediated knockdown are insulated from positional effects of genomic loci.

Moreover, the experiments described in this study reveal that exogenous overexpression of VALT1 phenocopies the entirety of the tested cancer hallmarks induced by eCSE in A549 cells and, in part, in BEAS-2B. Evidence has generally been mixed toward the inflammatory role of e-cigarettes, with some reporting immunosuppression [45] and some reporting pro-inflammatory effects [40] in cellular models, which is not surprising given the chemical diversity of commercially available e-cigarette formulations. However, there is a consensus that these anti-inflammatory redox responses tend to be activated upon e-cigarette smoke exposure as an adaptive mechanism to detoxify cells from otherwise deleterious stressors. These responses include those mediated by early growth response 1 (EGR1) [45] and nuclear factor erythroid 2-related factor 2 (NRF2) via its downstream effectors, such as NAD(P)H quinone dehydrogenase (NQO1) [46]. This has been demonstrated with the attenuation of cisplatin-induced ROS induction by nicotine, a prominent e-cigarette component [47], and the attenuation of MSB-induced ROS induction by SCAL1 [37]. In like manner, this study showed that nicotine-rich eCSE and its concomitant VALT1 expression can attenuate LPS-induced ROS formation in a concentration-dependent manner. Furthermore, the divergence of these phenotypes in normal and cancer cells implies that VALT1 might, in part, contribute to the differential survival of cancer cells over normal cells in spite of continued cytotoxic insults.

Lastly, a preponderance of actin cytoskeletal structures associated with invasiveness—including filopodia, lamellipodia, and invadopodia, in eCSE-treated setups—was observed. The enrichment of these structures in both eCSE-treated and VALT1-overexpressing cells may provide mechanistic insights into how these conditions promote migratory behavior, which is especially pronounced in many cancers, including NSCLC tumors, as these structures are closely linked to directional motility. Lamellipodia, which are veil-shaped membranous protrusions characterized by highly isotropic arrangements of actin [47], are firmly established as the driving force for polarized cell migration [48]. Lamellipodial functions are complemented by filopodia, which are thin actin protrusions that emanate from the cell periphery and are often associated with chemosensation [49]. The formation of filopodia has since been shown to be closely associated with lamellipodial F-actin organization [50]. The emergence of these structures is also consistent with observed peripheral membrane ruffling, a positive indicator of lamellipodial formation and negative for cellular adhesion [51]. The results of wound healing assays concur with these observed changes. Invadopodia, prominent protrusions highly rich in F-actin that drive ECM degradation [52], were also observed.

In both cell models, cytoskeletal rearrangements were observed alongside quantitative changes in nuclear morphology, including dorsoventral (DV) flattening, which reflects a loss of nuclear rigidity and enhanced nuclear deformability in both cell models—changes that enable migration. The opposite phenotype, as observed in BEAS-2B cells, could be suggestive of dynamic forces that support directional motility [53,54]. Collectively, these changes in nuclear dynamics can predispose cells toward nuclear aberrations, such as multinucleation, typical of transformed malignancies [55].

The potential functional role of VALT1 in mediating these phenotypes can be seen in BEAS-2B cells overexpressing the lncRNA. The shift in BEAS-2B toward an A549-like phenotype is suggestive of squamous differentiation, a preneoplastic marker often seen in lung tumors [56]. Alongside its observed pro-proliferative effects, this suggests that VALT1 may predispose normal cells toward transformation-associated traits that parallel pathological processes in NSCLC, even in the absence of genetic insults that may be associated with e-cigarettes.

The results described in this study support VALT1 as a functional lncRNA. Currently, VALT1 is being investigated as a functional driver for other cancer types, including multiple myeloma [26], clear cell renal cell carcinoma [27], and prostate cancer [28]. This study adds to the growing body of knowledge on lncRNA dysregulation as a stimulus-responsive driver of disease by showing that VALT1 expression can be induced by acute e-cigarette vapor exposure while being unusually persistent, remaining elevated even after withdrawal of the stimulus. While mechanistic studies remain limited, the results of these phenotypic studies hint toward dysregulation of universal cellular processes. Various bioinformatic studies posit VALT1 as a competitive endogenous RNA (ceRNA) [57,58], consistent with findings from argonaute RNA immunoprecipitation (AGO-RIP) in prostate cancer [28] and compartment-specific CRISPR–Cas13d screens in multiple myeloma [26]. Indeed, RNA localization experiments in this study confirm cytoplasmic localization of VALT1, also consistent with a potential ceRNA role in NSCLC. However, given the pronounced tissue- and context-specificity of miRNA expression and activity [59], further work is required to define which miRNAs are functionally sequestered by VALT1 in an NSCLC context. This should include confirmation of candidate miRNA expression via RT-qPCR, direct binding and target engagement via dual-luciferase reporter assays, site-directed abrogation of predicted miRNA response elements (MREs) within VALT1, and steric blockade of specific miRNA–target interactions using target protectors. This is particularly important given that miRNA binding, especially during imperfect complementarity, may not necessarily result in effective post-transcriptional degradation through RNA interference. It is also plausible that VALT1 exerts its phenotypic effects through the summative buffering of multiple miRNAs simultaneously, producing a net shift in post-transcriptional regulatory tone rather than a single dominant miRNA axis.

Overall, this study adds to the growing body of evidence supporting VALT1 as a functional contributor to tumorigenesis, particularly in NSCLC, where its mechanistic role remains incompletely defined despite reports of parallel upregulation in the oral transcriptomes of e-cigarette users and in NSCLC tumors. The study demonstrates that e-cigarette exposure is associated with a dose-dependent induction of VALT1 in vitro, although the magnitude does not fully mirror that observed among users. Notably, VALT1 upregulation appears specific to e-cigarette exposure and is not significantly elevated in conventional cigarette smoker transcriptomes, which necessitates chemical characterization studies to identify the specific constituents responsible for its induction. Moreover, while eCSE is established here as an upstream stimulus, the intermediary signaling modules governing VALT1 transcription remain undefined. The presence of conserved E2F-binding motifs upstream of VALT1, together with chromatin immunoprecipitation (ChIP) evidence supporting E2F1 occupancy at its promoter region (ENCSR000EVJ) [60], provides a plausible mechanistic basis for its tumorigenic and e-cigarette–responsive behavior, given the established roles of E2F family transcription factors in driving cell cycle progression, proliferation, and migration [61]. Additionally, NRF2—recognized as a master regulator of oxidative stress responses and apoptotic resistance [62]—has also been shown to bind upstream of VALT1 (ENCSR584GHV) [63], offering a potential explanation for its cytoprotective and ROS-modulating effects following e-cigarette exposure. Nevertheless, because both E2F [64] and NRF2 [65] are broadly implicated in cellular responses to conventional cigarette smoke, it is likely that additional upstream regulatory modules contribute to the selective induction of VALT1.

Lastly, this study has its own limitations. Given the heterogeneity of e-cigarette formulations and the lack of complete chemical characterization of the utilized eCSE formulation, the extent to which these findings can be extrapolated to e-cigarette use more broadly is limited. Although a ceRNA-based mechanism remains likely, the documented cellular specificity of miRNA expression profiles precludes direct extrapolation of candidate miRNA interactors and ceRNA networks identified in other cancer types to the NSCLC context. Future work will be necessary to elucidate the precise mechanisms by which VALT1 functions. Nonetheless, the present work provides preliminary data linking e-cigarette exposure to lncRNA-mediated tumorigenic phenotypes and highlights VALT1 as a relevant molecular node warranting deeper investigation.

4. Materials and Methods

4.1. Analysis of VALT1 lncRNA Expression Levels from Publicly Available RNA-Seq Datasets

To explore the expression patterns of VALT1 in human cancers, publicly available RNA-sequencing data from cancerous tumors and matched non-tumorigenic tissues obtained from deepBase v3.0—which integrates large-scale datasets from ENCODE, TCGA, ICGC, and GTEx—were analyzed. Transcript abundance across samples was normalized as log₂(FPKM + 1), where FPKM denotes fragments per kilobase of transcript per million mapped reads. The scripts and computational workflows used for TCGA and deepBase v3.0 data analyses are publicly available at https://github.com/darmirador (accessed on 4 March 2026).

4.2. Culture, Maintenance, and eCSE Treatment of A549 Cells

A549 adenocarcinomic alveolar epithelial cells (ATCC^®, Manassas, VA, USA, Cat. No. CCL-185) were cultured in T-75 culture flasks in Dulbecco’s Modified Eagle Medium (DMEM; Gibco^®, Thermo Fisher Scientific, Inc., Waltham, MA, USA, Cat. No. 12100-038) supplemented with 10% fetal bovine serum (FBS; Gibco^®, Cat. No. 10500-064) [DMEM + 10% FBS] in controlled environmental conditions (37 °C, 5% CO₂). Cancer hallmark assays on A549 cells were done under reduced serum conditions (DMEM + 4% FBS). A popular flavored e-cigarette formulation containing nicotine (strawberry cheesecake-flavored; 0.6 mg/mL) was used to simulate both nicotinic and aldehydic load borne from e-cigarette flavorants and heat-induced aerosolization during normal usage. Preparation of e-cigarette smoke extract (eCSE) was performed by connecting a 50 mL syringe filled with reduced serum medium to the e-cigarette (22W power output, 950 mAh) atomizer set at the highest power level using a protocol modified from Gilpin et al. (2019) [66]. Briefly, during each cycle, vapor generated from the atomizer was drawn into the syringe, allowed to dissolve briefly (approximately 5 s) in the medium, and expelled as spent vapor. This procedure was repeated 25 times, after which the medium was filter-sterilized using a 0.22 µm polyethersulfone (PES) syringe filter, yielding the concentrated 1× eCSE-DMEM. To account for inter-batch variation across batches of eCSE, standardized pumping parameters and dissolution protocols were strictly maintained to ensure batch-to-batch consistency.

For eCSE dose–response experiments, VALT1 overexpression, and siRNA-mediated VALT1 knockdown, A549 cells in maintenance medium were seeded on assay plates and incubated for at least 24 h to facilitate cellular attachment to the substrate. The spent medium was then replaced with eCSE-DMEM at varying concentrations: undiluted (1×), twofold dilution (0.5×), 10-fold dilution (0.1×), 50-fold dilution (0.02×), and 100-fold dilution (0.01×). Untreated control setups (0×) consisting of eCSE-free reduced serum medium were maintained in parallel with eCSE-treated setups as baseline controls.

4.3. Culture and Maintenance of BEAS-2B Cells

Immortalized BEAS-2B normal human bronchial epithelial cells (ATCC; Cat. No. CRL-3588) were maintained and cultured in T-75 culture flasks pre-coated with a 3 mL fibronectin-collagen coating solution consisting of 0.01 mg/mL human plasma fibronectin (Gibco^®, Cat. No. 33016-015), 0.03 mg/mL bovine collagen I (Gibco^®, Cat. No. A10644-01), and 0.01 mg/mL bovine serum albumin (BSA; Sigma-Aldrich, St. Louis, MO, USA, Cat. No. A9418) in serum-free bronchial epithelial basal medium (BEBM; Lonza Bioscience, Walkersville, MD, USA, Cat. No. CC-3171) per ATCC guidelines. Serum-free LHC-9 medium (Gibco^®, Cat. No. 12680-013) was used for the maintenance of BEAS-2B cells under controlled environmental conditions to prevent squamous differentiation (37 °C, 5% CO₂).

4.4. RNA Extraction and First-Strand cDNA Synthesis

The RNeasy^® Mini Kit (QIAGEN Sciences, Inc., Germantown, MD, USA, Cat. No. 74106) was used to extract total RNA from A549 and BEAS-2B cells seeded at a density of 300,000 cells and 200,000 cells per well, respectively, on a 6-well plate. RNA concentration and purity were determined through UV-Vis spectrophotometry (λmax  =  260 nm) using the Nanodrop 2000c spectrophotometer (Thermo Fisher Scientific, Inc.). RNA was immediately utilized for first-strand complementary DNA (cDNA) synthesis using M-MLV reverse transcriptase (Promega^®, Madison, WI, USA, Cat. No. 28025-013). Briefly, 200 U of M-MLV reverse transcriptase was used to reverse-transcribe 2000 ng of extracted RNA. A 15 µL annealing reaction mixture containing 50 pmol 15-mer oligo-dTs, 50 pmol random hexamers (Invitrogen™, Waltham, MA, USA, Cat. No. N8080127), and 2000 ng RNA sample was incubated at 70 °C for 5 min. For RNA localization experiments, 50 pmol 15-mer oligo-dTs was removed from the reaction to avoid biasing against nuclear MALAT1 and NEAT1 transcripts, which lack canonical poly(A) tails. The 5× M-MLV reverse transcriptase buffer (Promega^®, Cat. No. M531A) was subsequently added to the annealing reaction mixture to a final 1× concentration. In addition, 25 U of RNase inhibitor (Invitrogen™, Cat. No. AM2682), 15 pmol of deoxynucleoside triphosphate (dNTPs), and 200 U of M-MLV reverse transcriptase (Promega^®, Cat. No. M170B) were added to the solution, yielding a final reaction volume of 25 µL. The components were then incubated at 37 °C for 1 h. cDNA was then stored at 4 °C prior to downstream applications.

4.5. Generation of pTargeT™-VALT1 lncRNA Expression Construct

The predicted full-length VALT1 lncRNA transcript (GenBank ID: AC016773.1 or AC016773.2; GENCODE ID: ENST00000605571.1; NONCODE ID: NONHSAT094735.2) was acquired from the University of California, Santa Cruz (UCSC) Genome Browser Basic Gene Annotation Set from GENCODE Version 37 (Ensembl 103) [67]. Linearized EcoRI-digested pTargeT™ mammalian expression vector (Promega^®, Cat. No. A140A) was used as a backbone for cloning. Primers for Gibson Assembly^® cloning, shown in

Table 1. Primers used in Gibson Assembly^® cloning of the full-length VALT1 transcript. Backbone-derived sequences are highlighted in italics.

Primer	Primer Sequence (5′–3′)
Forward	ATTATAATACGACTCACTATAGGGCGACCCGGAACCCGGAACCCG
Reverse	GGTGACACGATAGAATACTCAAGCTTGGGCCCTGCCAGGATTTTATTTTTAG

, were based on the terminal sequences of the VALT1 transcript and the 3′ overhangs produced by EcoRI digestion (Promega^®, Cat. No. R601J) in Buffer H (Promega^®, Cat. No. R008A) flanking both ends of the multiple cloning site (MCS) of the pTargeT™ mammalian expression vector.

Purified EcoRI-digested linearized pTargeT™ vector and VALT1 amplicon appended with backbone-derived sequences were combined in a 2:1 insert-to-vector molar ratio (15 ng insert with 50 ng linearized vector) using the NEBuilder^® HiFi DNA Assembly (New England Biolabs, Inc., Ipswich, MA, USA, Cat. No. M5520A) reaction protocol following the manufacturer’s instructions. Chemically competent (ultracompetent) DH5α E. coli cells prepared using the Inoue Method [68] stored at −80 °C were used for heat shock bacterial transformation of the recombinant pTargeT-VALT1 expression constructs based on the NEBuilder HiFi DNA Assembly Transformation Protocol. Purified pTargeT-VALT1 plasmids were sequence-verified prior to downstream cellular applications.

4.6. Transfection and Transient Expression of VALT1 in A549 and BEAS-2B Cells

In a 6-well plate, 300,000 A549 cells were transfected with the plasmid construct using Lipofectamine™ 3000 Transfection Reagent (Invitrogen™, Cat. No. 100022052) and P3000 reagent (Invitrogen™, Cat. No. 100022058) 24 h post-seeding. DNA-Lipofectamine™ complexes for transfection were prepared by adding 2000 ng of the desired construct, 6 µL Lipofectamine™ 3000 Transfection Reagent (0.3% v/v), 4 µL P3000 Reagent (1 µL reagent per 20 ng transfected construct), and 100 µL serum-free Opti-MEM (Gibco^®, Cat. No. 31985-070) after combining the DNA solution (2000 ng of the construct in P3000 diluted to 50 µL with Opti-MEM) and Lipofectamine™ solution (6 µL Lipofectamine™ 3000 Transfection Reagent diluted to 50 µL with Opti-MEM). Upon mixing, the resultant solutions were allowed to incubate for at least 5 min to facilitate complex formation. The formed lipofection complexes were pipetted onto adherent cells. After 24 h, the spent medium was replaced with DMEM + 4% FBS.

For BEAS-2B cells, at least 24 h before seeding, 6-well plates were pre-coated with the aforementioned fibronectin-collagen coating solution. Cultured on coated plates, 200,000 BEAS-2B cells were transfected with the plasmid construct using Lipofectamine™ 2000 Transfection Reagent (Invitrogen™, Cat. No. 11668019) 48 h post-seeding. DNA-Lipofectamine™ complexes were prepared by adding 2000 ng of the construct, 8 µL Lipofectamine™ 2000 Transfection Reagent (0.4% v/v), and 100 µL serum-free Opti-MEM after combining the DNA solution (2000 ng of the construct diluted to 50 µL Opti-MEM) and Lipofectamine™ solution (8 µL Lipofectamine™ 2000 Transfection Reagent diluted to 50 µL Opti-MEM). Upon mixing, the resultant solutions were also allowed to incubate for at least 5 min to facilitate complex formation. The formed lipofection complexes were pipetted onto BEAS-2B cells in BEBM + 4% FBS. After 6 h, the spent medium was replaced with LHC-9. In both cell lines, parallel transfection of the pmR-ZsGreen1 plasmid was done to assess transfection efficiency in plasmid-transfected setups.

For downstream assays utilizing 96-well formats, 20-fold scaled-down transfection setups were utilized. An empty vector control, consisting of recircularized pTargeT™ at EcoRI ends, was utilized as a baseline control for setups transfected with pTargeT-VALT1 to account for cellular stress associated with plasmid transfection.

4.7. siRNA-Mediated VALT1 Knockdown in A549 and BEAS-2B Cells

Two independent short interfering RNAs (siRNAs), whose sequences are shown in

Table 2. siRNA target sequences for siRNA-mediated VALT1 knockdown.

siRNA	siRNA Target Sequence
siVALT1-1 (GeneGlobe ID SIC0057023)	5′-GCTTAAAGTTTTGGAGTAACC-3′
siVALT1-2 (GeneGlobe ID SIC0057024)	5′-GGGTTTGTATCCTGAAGAATC-3′

, were designed to specifically target VALT1 lncRNA transcripts while accounting for potential sequence-specific effects. Both the guide and the passenger strands were appended with 3′ uridine dinucleotides to facilitate siRNA bioactivity.

Following the same post-seeding incubation periods used in lncRNA overexpression experiments, 40 pmol of siRNA was transfected into each well containing 300,000 A549 cells and 200,000 BEAS-2B cells using Lipofectamine™ RNAiMAX Transfection Reagent (Invitrogen™). SiRNA-Lipofectamine™ complexes were prepared by adding 40 pmol siRNA and 6 µL Lipofectamine™ 2000 Transfection Reagent (0.3% v/v) diluted to 100 µL with Opti-MEM after combining the siRNA solution (40 pmol siRNA diluted to 50 µL Opti-MEM) and Lipofectamine™ solution (6 µL Lipofectamine™ RNAiMAX Transfection Reagent diluted to 50 µL with Opti-MEM). The resultant solutions were allowed to incubate for at least 5 min to facilitate complex formation. The formed lipofection complexes were pipetted onto A549 cells in DMEM + 4% FBS and BEAS-2B cells in LHC-9 medium.

All subsequent downstream experiments were performed at least 48 h post-transfection. A baseline control transfected with AllStars negative control siRNA (siNEG; QIAGEN Sciences, Inc., Cat. No. SI03650318) was also included. The negative control siRNA is a proprietary siRNA that has no established homology to any mammalian gene, making it the most robust baseline control that can account for cellular stress associated with transfection procedures. Meanwhile, transfection using Texas Red-conjugated oligonucleotides (System Biosciences, Palo Alto, CA, USA; Cat. No.: XMIR-POS) was done as a transfection control to assess transfection efficiency in siRNA-treated setups.

4.8. Quantification of Normalized VALT1 and SLBP Expression via RT-qPCR

To establish the dose-dependent regulation of VALT1 upon exposure to eCSE, 300,000 A549 cells were seeded onto 6-well plates and allowed to incubate for 48 h post-seeding. Spent maintenance medium was replaced with reduced serum medium containing various eCSE titers and incubated for 6 h in eCSE in DMEM + 4% FBS prior to RNA extraction and RT-qPCR.

Recovery experiments were also done to ascertain the transient effects of eCSE exposure on VALT1 transcript levels. For eCSE recovery experiments, 300,000 A549 cells were seeded onto 6-well plates and allowed to incubate for 48 h post-seeding. Spent maintenance medium was replaced with reduced serum medium corresponding to the appropriate treatment. Two eCSE treatment setups were subsequently maintained in parallel alongside an untreated control. One eCSE treatment setup, designated as the ‘Recovery’ treatment setup, was allowed to incubate in 0.1× eCSE-DMEM for 6 h followed by an 18 h recovery period in DMEM + 4% FBS without eCSE. Another eCSE treatment setup, designated as the ‘Full’ treatment setup, was allowed to incubate continuously for 24 h in 0.1× eCSE-DMEM. Lastly, to verify the overexpression and siRNA knockdown efficiency of VALT1 transcripts, A549 and BEAS-2B cells plated in 6-well plates were transfected appropriately as previously described.

Following the aforementioned total RNA extraction and reverse transcription procedure, quantitative reverse transcription PCR (RT-qPCR) was performed using PowerUp™ SYBR™ Green Master Mix (Thermo Fisher Scientific, Inc.; Cat. No. A25742) following the manufacturer’s recommended protocol. Briefly, 5 µL 2× SYBR™ Green PCR Master Mix was mixed with 2.7 pmol of both forward and reverse primers (3 µL primer, each diluted to a concentration of 900 nM) whose sequences are highlighted in

Table 3. Primers designed for qPCR analysis of VALT1, SLBP, and CypA transcript levels.

Primer	Sequence (5′–3′)
VALT1 qPCR Forward Primer	TGACAACTACACTGAGTCTTTCC
VALT1 qPCR Reverse Primer	GGATACAAACCCAGTCAACTCC
SLBP qPCR Forward Primer	CAGTCTTGCCACAACTTCAATC
SLBP qPCR Reverse Primer	ATGGAGCCGATTATGAGAACAC
CypA qPCR Forward Primer	CCTAAAGCATACGGGTCCTGGCATC
CypA qPCR Reverse Primer	GTGGAGGGGTGCTCTCCTGAGCTAC

and cDNA template corresponding to 200 ng total RNA (2 µL cDNA, diluted 10-fold), filled to a final working volume of 10 µL. Transcripts for cyclophilin A (CypA) were used as a housekeeping control. RT-qPCR experiments were done in technical quadruplicates. Sequences for CypA qPCR primers were adapted from Cruz et al. [37], while SLBP qPCR primers were adapted from Brocato et al. [69]. For the relative quantification of VALT1 levels of the experimental setups relative to a baseline control, the ΔΔCT method was used. VALT1 lncRNA and SLBP mRNA levels were normalized with respect to CypA mRNA levels.

4.9. RNA Fractionation and Identification of VALT1 Localization via RT-qPCR

To establish the localization of VALT1, RNA from the cytoplasmic and nuclear fractions was serially extracted. Using a method adapted from Jahn et al. (2023) [70], 6,000,000 A549 or BEAS-2B cells cultured to confluence in T-25 culture flasks were harvested. One-third of the cells were delegated for total RNA extraction, using protocols described previously. The remaining cells were subjected to hypotonic lysis using buffer containing 50 mM Tris-Cl (pH 8.0; Sigma-Aldrich, Cat. No. 93352), 100 mM NaCl (RCI Labscan, Bangkok, Thailand, Cat. No. AR1167), 5 mM MgCl₂ (HiMedia, Mumbai, India, Cat. No. GRM686), and 0.1% Triton X-100 (Sigma-Aldrich, Cat. No. T8787) to separate cytoplasmic and nuclear compartments.

The cytoplasmic fraction (supernatant) and nuclear fraction (pellet) were subsequently processed for RNA extraction by adding 1050 µL and 600 µL RLT lysis buffer (RNeasy^® Mini Kit, QIAGEN Sciences Inc., Cat. No. 79216), respectively. Ethanol was added to each fraction (750 µL of 90% ethanol to the cytoplasmic fraction and 600 µL of 70% ethanol to the nuclear fraction) prior to loading onto RNeasy Mini Spin Columns. All subsequent purification steps were performed according to the manufacturer’s instructions.

First-strand cDNA synthesis and RT-qPCR for each fraction were done, as described previously. CypA mRNA served as the cytoplasmic control, whereas MALAT1 and NEAT1 lncRNAs served as nuclear lncRNA controls. Sequences for MALAT1 qPCR primers were adapted from Jiao et al. (2014) [71], whereas that of NEAT1 was adapted from Lee et al. (2016) [72]. Primer sequences for MALAT1 and NEAT1 are shown in

Table 4. Primers designed for qPCR analysis of MALAT1 and NEAT1 transcript levels in RNA fractionation experiments.

Primer	Sequence (5′–3′)
MALAT1 qPCR Forward Primer	GAATTGCGTCATTTAAAGCCTAGTT
MALAT1 qPCR Reverse Primer	GTTTCATCCTACCACTCCCAATTAAT
NEAT1 qPCR Forward Primer	TCGGGTATGCTGTTGTGAAA
NEAT1 qPCR Reverse Primer	TGACGTAACAGAATTAGTTCTTACCA

.

4.10. MTS and CyQUANT^® Cell Proliferation Assays

After the corresponding incubation period post-seeding, as described previously, 2000 A549 cells and 4000 BEAS-2B cells suspended in 100 µL were seeded onto the appropriate 96-well plates. For eCSE treatment experiments, spent maintenance medium was replaced with reduced serum medium containing various titers of eCSE. For transfection and knockdown experiments, a 20-fold scaled-down lipofection complex mixture was pipetted onto each well upon replacement of spent maintenance medium with the appropriate medium. Proliferation was quantified 48 h and 72 h post-treatment.

Prior to quantification, the spent medium was replaced with 100 µL fresh medium. Per well, 10 µL of the MTS dye-based CellTiter 96^® AQueous One Solution Reagent (Promega^®, Cat. No. G3581) was added. Plates were allowed to incubate for 2 h. Absorbance at 490 nm (optical density at 490 nm or OD490) was subsequently measured using CLARIOstar^® Plus Microplate Reader (BMG LABTECH, Ortenberg, Germany). In a separate experiment, 0.4 µL of CyQUANT^® Direct Red (Invitrogen™, Cat. No. C35013) and 2 µL CyQUANT™ Direct background suppressor (Invitrogen™, Cat. No. C35013A) in 97.6 µL fresh medium corresponding to a 2× staining solution was added on top of the freshly replaced medium. Fluorescence readings were then acquired after an hour-long incubation period. To acquire fluorescence values, a red fluorescent filter was used (λex/λem = 614/653 nm). Optical density measurements were blanked using the corresponding medium with the appropriate amounts of dye in the absence of cells. Five technical replicates were done per trial.

4.11. Caspase 3/7 Apoptosis and TMRM Mitochondrial Permeability Assays

In the appropriate 96-well plates, 8000 A549 cells were seeded for eCSE treatment experiments, and 4000 A549 or BEAS-2B cells were seeded for downstream transfection. After the designated incubation and treatment periods, intrinsic apoptosis was induced in A549 and BEAS-2B cells using 100 µM menadione sodium bisulfite (MSB; Sigma-Aldrich, Cat. No. M5750). The proportion of caspase 3/7+ cells was quantified 3 h post-MSB induction using 5 µM CellEvent™ Caspase-3/7 Green Detection Reagent (Invitrogen™, Cat. No. C10423).

In a separate experiment, mitochondrial integrity was quantified using 200 nM Image-iT™ tetramethylrhodamine, methyl ester (TMRM; Invitrogen™). For both experiments, stains were visualized against a nuclear counterstain (10 µg/mL Hoechst 33342, trihydrochloride trihydrate; Invitrogen™, Cat. No. H1399). Fluorescent images were obtained with the GE IN Cell Analyzer 6000 high-content imager using a Nikon (Tokyo, Japan) 10×/0.45, Plan Apo, CFI/60 objective lens. For each experimental condition, five randomly selected fields per well were imaged across five independent wells, and all captured cells were included in the subsequent quantitative analysis. Quantification of fluorescent signals was performed using the IN Cell Developer Toolbox v1.6 (GE Healthcare Life Sciences, Marlborough, MA, USA). The following fluorescent filters were utilized: Caspase-3/7 (blue excitation laser, λex = 488 nm; FITC emission filter, λem = 525/20 nm); TMRM (green excitation laser, λex = 561 nm; dsRed emission filter, λem = 605/52 nm); and Hoechst 33342 (UV excitation laser, λex = 405 nm; DAPI emission filter λem = 455/50 nm).

4.12. Annexin V (AV)/Propidium Iodide (PI) Flow Cytometry Analysis

For both A549 and BEAS-2B cells, 80,000 cells were seeded in the appropriate 6-well plates. At 48 h post-transfection or 6 h post-eCSE treatment, intrinsic apoptosis was induced in A549 and BEAS-2B cells using 100 µM MSB. Alexa Fluor 488 (AF488) conjugated to Annexin V (AV) and propidium iodide (PI) were utilized as early apoptotic and viability markers, respectively. The proportions of live cells (AV–/PI–), early apoptotic cells (AV+/PI–), and late apoptotic or necrotic cells (PI+) were quantified 3 h post-induction. The total cell fraction, including both adherent and non-adherent cells, was collected via centrifugation. Afterward, the cell populations were observed using the AF488 Annexin V/Dead Cell Apoptosis Kit (Thermo Fisher Scientific, Inc., Cat. No. V13245) following the manufacturer’s instructions, and using the Attune™ NxT Flow Cytometer (Invitrogen™). The blue excitation laser (488 nm) was used, whereas the BL1 (530/30 nm) and BL2 (574/26 nm) emission filters were used to quantify AF488-AV and PI fluorescence, respectively. Single-stained controls were used as compensation controls and fluorescence-minus-one (FMO) controls in setting quadrant gates.

4.13. DCFDA Staining for Intracellular Reactive Oxygen Species Quantification

Following the same seeding, incubation, and treatment procedures of Caspase 3/7 activity and mitochondrial permeability setups in 96-well plates, lipopolysaccharide (LPS) challenge was induced among cells to saturate the levels of intracellular reactive oxygen species (ROS). Cells were treated to a concentration of 80 μg/mL LPS (Sigma-Aldrich, Cat. No. L9023) for 24 h. After LPS treatment, cells were stained with 25 µM DCFDA stain. The assay plate was incubated at 37 °C for 20 min, followed by a 5 min nuclear counterstain with 10 µg/mL Hoechst 33342, trihydrochloride trihydrate (Sigma-Aldrich, Cat. No. D6883) before observation and imaging with the GE IN Cell Analyzer 6000 high-content imager. Quantification of cellular fluorescent signals was performed using the IN Cell Developer Toolbox v1.6 (GE Healthcare Life Sciences) using a Nikon 20×/0.45, Plan Fluor, ELWD, Corr Collar 0-2.0, CFI/60 objective lens. Five randomly placed fields per well were imaged across five independent wells, and all captured cells were included in the subsequent quantitative analysis. The following lasers and fluorescent filters were utilized: DCFDA (blue excitation laser, λex = 488 nm; FITC emission filter, λem = 525/20 nm) and Hoechst 33342 (UV excitation laser, λex = 405 nm; DAPI emission filter λem = 455/50 nm).

4.14. Scratch Wound 2D Migration Assay

To promote the formation of a confluent monolayer of adherent cells, 20,000 A549 cells and BEAS-2B cells were seeded onto appropriate 96-well plates. Following the aforementioned incubation and treatment procedures, a sterile white pipette tip was then used to scratch the well in a straight line to create an open wound in the confluent cell monolayer. Wells were then washed twice with 1× PBS, followed by incubation in reduced serum medium (A549) or LHC-9 medium (BEAS-2B). Cells were imaged immediately after seeding (0 h) and 16 h thereafter. Each setup was briefly stained with calcein AM (Invitrogen™, Cat. No. C3100MP) to a final concentration of 2 µg/mL prior to imaging with the IN Cell Developer Toolbox v1.6 (GE Healthcare) using a Nikon 4×/0.20, Plan Apo, CFI/60 objective lens. A single field was captured per well, and the same field was imaged post-scratching. The blue excitation laser (λex = 488 nm) and the FITC emission filter λem = 525/20 nm) set were utilized for visualizing calcein AM-stained cells. Wound closure was quantified through the ImageJ software (version 1.53k), accompanied by the Wound Size Healing Tool plugin [73].

4.15. Phalloidin Staining for Visualizing Actin Cytoskeletal Reorganization

The actin cytoskeleton of A549 and BEAS-2B cells seeded onto appropriate 96-well plates was visualized using Alexa Fluor™ 488 phalloidin (Invitrogen™, Cat. No. A12379) with a nuclear counterstain (Hoechst 33342). After seeding and treatment, wells were washed with 1× PBS prior to a brief period of sample fixation with 4% paraformaldehyde in 1× PBS (PFA; ChemCruz™, Santa Cruz, CA, USA, Cat. No. NC0238527). Cells were washed twice with ice-cold 1× PBS to remove excess 4% PFA. Samples were stored at 4 °C prior to staining.

PFA-fixed cells were permeabilized with 0.1% Triton™ X-100 for 10 min. Upon washing, fixation was followed by a blocking step, which involved the addition of 10 mg/mL BSA, incubated for 1 h with continuous mixing (60 rpm). Subsequent steps were taken in the dark. Without washing, permeabilized cells were treated with 13 nM Alexa Fluor™ 488 phalloidin in 1 mg/mL BSA + 1× PBS for 30 min. Afterward, Alexa Fluor™ 488 phalloidin-stained cells were washed twice with ice-cold 1× PBS prior to adding 10 µg/mL Hoechst 33342, trihydrochloride trihydrate. Cells were once again washed twice with 1× PBS prior to mounting with SlowFade Diamond Antifade Mountant (Invitrogen™, Cat. No. S36972). Fluorescent images were obtained with the GE IN Cell Analyzer 6000 high-content imager. Images were captured using a Nikon 40×/0.60, Plan Fluor, ELWD, Corr Collar 0-2.0, CFI/60 objective lens. Twelve randomly placed fields were imaged across five independent wells, and all captured cells were included in downstream analysis. The following filters were used: AF488 phalloidin (blue excitation laser, λex = 488 nm; FITC emission filter, λem = 525/20 nm) to visualize F-actin and Hoechst 33342 (UV excitation laser, λex = 405 nm; DAPI emission filter λem = 455/50 nm) to visualize the nuclei. Anisotropy was measured by randomly placing a uniform region of interest (ROI) within the cytoplasm, defined as AF488+ regions, of visible cells using the FibrilTool ImageJ plug-in [74]. One ROI, representing a data point, was placed per field, for a total of n = 60 fields per setup. Nuclear eccentricity and area, on the other hand, were quantified by thresholding the Hoechst 33342-stained channel to create a binary mask, followed by nuclear segmentation and measurement using the ‘Analyze Particles’ tool in ImageJ. All captured nuclei per field were quantified, and the mean data across each well, representing a data point, was utilized for a total of n = 5 data points.

4.16. Statistical Analyses

The appropriate statistical tests are presented on a per-figure basis. Each experiment was done in three independent trials. Figures are presented as the mean  ±  standard error of the mean (SEM). A cutoff of p  <  0.05 was used to denote statistical significance.