Introduction

An electronic cigarette (e-cigarette) is a device designed to simulate the feeling of traditional cigarette smoking to vaporize a liquid (often flavored) to generate aerosols for inhalation1. The e-liquid typically contains propylene glycol and/or vegetable glycerin as well as nicotine and flavoring agents/chemicals2. According to the 2023 National Youth Tobacco Survey, 10.0% of high school students and 4.6% of middle school students reported current use of e-cigarettes3. E-cigarette delivers nicotine to users, leading to addiction comparable to or even exceeding that of conventional cigarettes4. While e-cigarettes appear to reduce or eliminate certain carcinogens found in combustible cigarettes, the toxicological effects of inhaling e-cigarette aerosols containing flavorings remain unclear5,6,7.

Animal and cell line studies indicate that e-cigarette aerosols can trigger oxidative stress, cause DNA damage, and induce inflammatory responses in both human lung epithelial cells and mouse lungs8,9,10,11,12,13. Several cross-sectional national survey studies have indicated an association between e-cigarette use and respiratory symptoms and diseases such as wheezing, asthma, and chronic obstructive pulmonary disease (COPD)14,15,16,17,18. Additionally, e-cigarette use is associated with an increased risk of hypertension and cardiovascular diseases, including myocardial infarction, stroke, and coronary artery disease19,20,21,22,23. Mental health problems, depressions, and cognitive impairments have also been linked to e-cigarette use23,24,25,26.

Exosomes are small extracellular vesicles (EVs) released by various cell types, including immune and cancer cell, and are present in biofluids such as blood, urine, and other body fluids27,28. Exosomes play a crucial role in mediating cell-to-cell communication and influencing numerous physiological processes29,30,31,32. Exosomes are enriched with surface proteins, regulatory proteins, mRNAs, microRNAs and long non-coding RNAs33,34. Plasma exosomes, in particular, hold potential as biomarkers for immune dysfunction in cancer35.

MicroRNAs (miRNAs), a class of small non-coding RNAs consisting of 21–25 nucleotides, play vital regulatory roles in various cellular and biological process, including immune regulation and inflammatory responses36,37,38,39. miRNAs have been reported to play pivotal roles in the development of chronic inflammatory lung diseases and lung cancers40,41,42,43,44,45. Exosomal miRNAs, aside from revealing novel toxicological subtypes, show promise as potential biomarkers for flavor-related toxicities. Previous research has demonstrated that tobacco consumption modulates the expression levels of miRNAs through signaling pathways such as apoptosis, angiogenesis and inflammatory pathways46,47,48. These pathways play significant roles in the development of various human diseases, including lung disease, cardiovascular disease, liver disease, and cancer49,50,51. Exosomal microRNAs has emerged as diagnostic biomarkers and therapeutic targets for these diseases. Currently, there is a lack of information on the patterns of genome-wide and locus-specific miRNA expression in plasma samples when the subjects are exposed to e-cigarettes.

This study utilized plasma samples obtained from participants in the Population Assessment of Tobacco and Health (PATH) Study Wave 1 (2013–2014), including 15 exclusive e-cigarette users and 15 non-users52. Our aim is to identify plasma exosomal miRNAs that are differentially expressed between participants who exclusively use e-cigarettes and participants who don’t use any tobacco products. We seek to understand the functions of identified significant miRNAs in various signaling pathways. The identified plasma exosomal miRNAs and their associated targeted genes will significantly contribute to our understanding of the impact of e-cigarettes on cellular signaling pathways.

Materials and methods

Study participants

The PATH study is a nationally representative longitudinal cohort study initiated in September 2013 with 32,320 adults and 13,651 youth (12 years and older) participants52. In addition to collecting epidemiological information on tobacco use behaviors, attitudes, beliefs, exposures, and health, the PATH study includes biospecimens from a subsample of 11,500 adult participants from the Wave 1 study. Blood specimens were exclusively collected from PATH Wave 1 adults and aged-up adults from Wave 1 youth participants.

We obtained plasma samples through the PATH Study Biospecimen Access Program from a cohort of 15 current adult (aged 18 years or older) established exclusive e-cigarette users and 15 adult non-users who were not established any tobacco product users. Current established exclusive e-cig users are defined as individuals who have ever used an e-cigarette, have used regularly, and use every day or some days. Furthermore, these current established exclusive e-cig users do not regularly use any other tobacco products.

The study has been reviewed by the University of Rochester Research Subjects Review Board and was determined as not research involving human subjects (Study ID: STUDY00005750). All biospecimens obtained from the PATH biospecimen program were deidentified, and the linkage ID connecting demographic information with the PATH biospecimens were disclosed only after the completion of sequencing experiments on all biospecimens. All research in this study were performed in accordance with relevant guidelines and regulations.

Sample collection

Biospecimens, including urine and blood, are acquired from PATH Study participants who have provided informed consent, following a protocol reviewed and approved by the Westat Institutional Review Board (IRB) using validated PATH Study standard operating procedures (SOPs)52. The collected biospecimens were transported from the field to the PATH Study biorepository and subsequently processed into aliquots. The PATH Wave 1 blood sample was collected between 2013 and 2014, processed into plasma aliquots with EDTA additives, and has been stored at the PATH Study biorepository since then. Following approval from the PATH Study Biospecimen Access program, we received 30 plasma aliquots (200ul each) via mail in October 2021.

Exosome isolation and RNA extraction

The plasma samples were processed together with library prepared in one batch. Exosome RNA was isolated from 175ul of plasma using Norgen’s plasma exosome purification and RNA Isolation Kit (Cat. 58300). The kit was used according to the manufacturer’s instructions. RNA quantification was then done on an Agilent Bioanalyzer, using the Agilent RNA 6000 Pico Kit kit (cat. 5067 − 1513) according to the manufacturer’s instructions. Small RNA library prep was done using Norgen’s Small RNA Library Prep Kit for Illumina (cat. 63620) according to the manufacturer’s instructions. Library QC was done on an Agilent Bioanalyzer, using the Agilent High Sensitivity DNA Kit (cat. 5067 − 4626).

Small RNA sequencing

The small RNA sequencing analysis was carried out using the Illumina NextSeq 500 sequencing platform with 6ul RNA from each sample. Finally, samples were pooled and sequenced on the Illumina NextSeq 550 platform, using the NextSeq 500/550 High Output Kit v2.5 (75 Cycles). The number of reads from the plasma exosomal miRNAs for the 30 participants ranged from 3,010 to 243,141 in the small RNA sequencing experiment.

Statistical analysis

Unique participant IDs were used to establish links between PATH demographic data and the sequencing outcomes of plasma exosomal miRNAs. Fisher’s exact tests were used to identify potential confounding variables for the e-cigarette user group. miRNAs with a total count below 10 across all samples were excluded from subsequent analyses. The DESeq2 package in Bioconductor (version 3.19) was used to identify statistically significant miRNAs differentially expressed between exclusive e-cigarette users and non-users, while accounting for significant confounding variable. miRNAs with a p-value less than 0.05 were considered statistically significant. The BonEV method was used to control the false discovery rate at 5%53.

To visually represent the results, enhanced volcano plot was used to show the fold change in log2 scales and p-values of these significant miRNAs. Dot plots were used to show the distributions of miRNAs in the e-cigarette and non-user groups. The pheatmap function in the pheatmap package in R (version 4.4.0) was used to generate the heatmap of significant miRNAs. Additionally, a receiver operating characteristic (ROC) curve was constructed using the results of a logistic regression model with predictor variables including significant miRNAs and covariate, employing the rocit function in the ROCit package in R version 4.4.0 (R Core Team, 2019). The area under curve (AUC) of the ROC curve was calculated using the auc function in the pROC package in R (version 4.4.0). For molecular experiments the statistical analysis was performed using Student’s t-test. Data were analyzed using Graphpad prism software (version 7.02).

Gene enrichment analysis

Predictions of gene targets for differentially expressed miRNAs were conducted using the EVmiRNA (http://bioinfo.life.hust.edu.cn/EVmiRNA)54, miRDB (http://mirdb.org/)55,56, and ExoCarta (http://exocarta.org)57,58,59database. The target genes predicted from these databases were subsequently imported into the enrichrichment function within the enrichR package in R60. This facilitated gene enrichment analysis using the Gene Ontology (GO) 2021 and the Kyoto Encyclopedia of Genes and Genomes (KEGG) 2021 databases61,62,63.

Cell culture

Human bronchial epithelial cells (BEAS-2B, ATCC Cat#: CRL-9609) were obtained from the American Type Culture Collection (ATCC) and maintained in DMEM/F12K medium supplemented with 5% fetal bovine serum (FBS) and 1% Penicillin-Streptomycin-Glutamine (Thermo Fisher Scientific Cat#: 103–78016). The cells were incubated at 5% CO₂ and 95% humidity.

BEAS-2B cells (PASSAGE 11–13) were seeded at a density of 30,000 cells/cm² in 6-well plates and allowed to adhere for 24 h. Following this, the cells were starved in 1% FBS medium overnight. After starvation, transfection was performed using Lipofectamine RNAiMAX reagent (Invitrogen, Carlsbad, CA, USA). The treatments applied were Normal/negative control (NC), miR-125b-5p mimics (Cat#: 4464066; Thermo Fisher Scientific), miR-125b-5p inhibitors (Cat#: AM17000; Thermo Fisher Scientific). For miR-125b-5p treatment, the concentrations used were Negative Control: 25 µL Lipofectamine, miRNA Mimic Negative Control: 25 ng/mL in nuclease-free water, miR-125b-5p mimics: 25 ng/mL and 50 ng/mL. After 24 h, cells were exposed to Eliquid (Eliq), PG/VG (50:50), Eliq + miR-125 Mimic, Eliq + miR-125 Inhibitors, PG/VG + miR-125 Mimic, PG/VG + miR-125 Inhibitors. After exposure, cells were either lysed for RNA isolation or fixed for flow-based assays. The dose was referred from literature64,65 .

E-liquid

The e-liquid used in this study was procured from a local vendor and consisted of a PG/VG mixture in a 50:50 ratio. A concentration of 0.25% of the e-liquid was applied to the cells for treatment. This in vitro exposure model aims to investigate the effects of acute exposure on miR-125 and its role in wound healing dysregulation.

Wound healing assay

Confluent BEAS-2B cells were “scratched” by dragging a 200 µL micropipette tip across the center of each well in a single motion. Only wells with cells displaying well-defined edges were included in the study. Immediately following the scratch, images were captured using the Cytation 5 imaging system. The cells were then exposed to e-liquid, PG/VG, and the respective groups of miR-125b-5p mimics and inhibitors. The same scratched area was imaged again at 24 h post-exposure. The wound or scratch area was calculated using ImageJ software.

DNA damage assay

The DNA damage assay was conducted using the FlowCellect™ DNA Damage Histone H2A.X Dual Detection Kit (Cat#: FCCS025153, Merck) according to the manufacturer’s instructions. DNA damage was analyzed using the Luminex Millipore Guava EasyCyte flow cytometer.

Results

Demographic characteristics of the participants providing plasma samples

Table 1 displays the demographic characteristics of the participants who contributed plasma samples for analysis. Ethnicity, gender, and age groups are balanced between the exclusive e-cigarette user and the non-user group with similar frequency distributions. However, there is a significantly difference in the frequency distribution of race, with exclusive e-cigarette users having a higher proportion of White individuals than the non-user group (86% vs. 67%). Consequently, race was included as a covariate in the generalized linear model for miRNA differential analyses.

Table 1 Demographic characteristics and tobacco use history of participants providing plasma samples with P-values from Fisher’s exact tests.
Full size table

We also checked the status of other tobacco products used by all 30 participants. No participants had used any other tobacco products (e.g., cigars, cigarillos, pipes filled with tobacco, snus pouches, dissolvable tobacco) except two participants who used hookah for some days. However, the distribution of hookah use was not significantly different between the exclusive e-cigarette users and non-users (P-value = 0.483). The frequency distribution of subjects who ever smoked a cigarette, even one or two puffs, was not significantly different between exclusive e-cigarette users and non-users (P-value = 0.682). Secondhand smoking exposure was also not significantly different between exclusive e-cigarette users and non-users, including those living with a smoker during childhood (P-value = 0.272) and currently living with a smoker (P-value = 1.000).

Identification of differentially expressed miRNAs in exclusive adult electronic cigarette users

Utilizing a cutoff of two times fold change and a p-value less than 0.05, we identified four significant miRNAs that were differentially expressed between exclusive e-cigarette users and non-users, as illustrated in the volcano plot (Fig. 1). Further examination of the distributions of these four significant miRNAs revealed significantly higher expression levels in the exclusive e-cigarette user group than the non-user group (Fig. 2). The heatmap of the four significant miRNAs, differentially expressed between exclusive e-cigarette users and non-users, highlighted their upregulation in the exclusive e-cigarette user group (Fig. 3). The identified significant miRNAs include hsa-miR-100-5p, hsa-miR-125a-5p, hsa-miR-125b-5p, and hsa-miR-99a-5p. After adjusting for the multiplicity using the BonEV method53 to control for the false discovery rate (FDR), no significant miRNAs remain significant at the 5% level.

Fig. 1
figure 1

Volcano plot showing significant differences in miRNA expression between exclusive e-cigarette users and non-users. The grey dots mean no significant difference (NS) between exclusive e-cigarette users and non-users. The green dots indicate the log2 fold change is beyond the cutoff value (−1 or 1), so the fold change in miRNA expression levels is either larger than 2 or smaller than 0.5. The red dots indicate the miRNA expression levels between exclusive e-cigarette users and non-users are both significant (p-value < 0.05) and have a folder change smaller than 0.5 when comparing non-users with exclusive e-cigarette users. Four miRNAs (hsa-miR-100-5p, hsa-miR-125a-5p, hsa-miR-125b-5p, and hsa-miR-99a-5p) were identified based on the two-fold change and p-value less than 0.05 cutoff.

Full size image
Fig. 2
figure 2

Distribution plot of significant miRNAs differentially expressed between exclusive e-cigarette users and non-users. The four miRNAs (hsa-miR-100-5p, hsa-miR-125a-5p, hsa-miR-125b-5p, and hsa-miR-99a-5p) were all overexpressed in PATH W1 exclusive e-cigarette users.

Full size image
Fig. 3
figure 3

Heatmap of significant miRNAs differentially expressed between exclusive e-cigarette users and non-users. This heatmap highlighted the overexpression of the four miRNAs (hsa-miR-100-5p, hsa-miR-125a-5p, hsa-miR-125b-5p, and hsa-miR-99a-5p) in PATH W1 exclusive e-cigarette users.

Full size image

Identification of differentially expressed miRNAs in exclusive White adult electronic cigarette users

Considering the majority of the participants in our study are White, we conducted a sensitivity analysis to identify differentially expressed miRNAs in exclusive White adult electronic cigarette users. With the same cutoff of two times fold change and a p-value less than 0.05, we identified four significant miRNAs (hsa-miR-100-5p, hsa-miR-125b-5p, hsa-miR-200b-3p, and hsa-miR-99a-5p) that were differentially expressed between exclusive e-cigarette users and non-users in White adults, as illustrated in the volcano plot (S1 Fig). All four miRNAs had higher expression levels in the exclusive White adult electronic cigarette users group, with three of the identified miRNAs overlapping with our previously identified miRNAs using all participants (S2 Fig). S3 Fig shows the heatmap of the four significant miRNAs. One miRNA (hsa-miR-200b-3p) remains significant after the FDR is controlled for at 5% using the BonEV method (adjusted P-value = 0.008).

Sensitivity analysis of miRNA expression profiles between exclusive electronic cigarette users and non-users

We observed that two participants (P000055469 and P000436666) showed much higher miRNA expression levels than others in Fig. 3 and S3, suggesting that they might be potential outliers. To ensure the reliability of our findings, we conducted a sensitivity analysis of the miRNA expression profiles between exclusive electronic cigarette users and non-users, excluding these two participants. In the analysis involving all races, we found three significant miRNAs (hsa-miR-100-5p, hsa-miR-125a-5p, hsa-miR-125b-5p). However, none of them remained significant at the 5% level after adjusting for multiplicity using the BonEV method to control the false discovery rate (FDR). When analyzing only White participants, we identified four significant miRNAs (hsa-miR-100-5p, hsa-miR-125a-5p, hsa-miR-125b-5p, hsa-miR-200b-3p). After controlling for the FDR at 5% using the BonEV method, the miRNA hsa-miR-200b-3p remained significant (adjusted P-value = 0.024). Therefore, we arrived at the same conclusion as when we included the two participants (P000055469 and P000436666), ensuring the robustness of our findings.

Gene enrichment analysis of differentially expressed miRNAs in exosomes of exclusive adult electronic cigarette users

The target genes of the four significantly upregulated miRNAs in exclusive e-cigarette users were predicted using the EVmiRNA (http://bioinfo.life.hust.edu.cn/EVmiRNA), miRDB (http://mirdb.org/), and ExoCarta (http://exocarta.org) databases. The Venn diagrams illustrate the overlap of target genes identified by all three methods for each significant miRNA (Fig. 4). Notably, variations of the number of target genes were observed among different miRNA target gene prediction methods, as well as variations of the number of target genes regulated by different miRNAs. EVmiRNA identified a greater number of gene targets than the other two methods. The union of target genes from all three methods was used for gene enrichment analysis.

Fig. 4
figure 4

Target genes of significant miRNAs differentially expressed between exclusive e-cigarette users and non-users predicted by three different miRNA target prediction methods (EVmiRNA, miRDB, ExoCarta). (A) has-miR-100-5p; (B) has-miR-125a-5p; (C) has-miR-125b-5p; (D) has-miR-99a-5p; (E) has-miR-200b-3p.

Full size image

Figure 5 showed the gene enrichment analysis results from the GO functional enrichment analysis of significant miRNA target genes and the systematic analysis of target gene functions using the KEGG 2021 Human database. The GO molecular function results highlighted the involvement of significant miRNA target genes in sequence-specific DNA binding, RNA-binding, ubiquitin protein ligase binding, and protein kinase binding. Genes related to intrinsic and integral component of mitochondrial and Golgi membrane were implicated in the GO cellular component analysis. The GO biological process revealed involvement in processes such as extrinsic apoptotic signaling pathway, miRNA catabolic process, miRNA metabolic process, embryonic appendage morphogenesis, and more. The KEGG analysis results identified target genes associated with various cancer pathways, including non-small cell lung cancer, small cell lung cancer, hepatocellular carcinoma, pancreatic cancer, Hippo signaling pathway, HIF-1 signaling pathway, p53 signaling pathway, and MAPK signaling pathway.

Fig. 5
figure 5

Gene enrichment analysis results using identified common target genes of significant miRNAs differentially expressed between exclusive e-cigarette users and non-users predicted by at least one of the three different miRNA target prediction methods (EVmiRNA, miRDB, ExoCarta). (A) GO Molecular Function 2021 database; (B) GO Cellular Component 2021 database; (C) GO Biological Process 2021 database; (D) KEGG 2021 Human database.

Full size image

Gene enrichment analysis of the significant miRNA has-miR-200b-3p also showed similar GO molecular function (e.g., transcription regulatory region nucleic acid binding, GDP, mRNA, RNA, DNA binding, protein kinase binding and activities), GO cellular components (e.g., intracellular membrane-bounded organelle, adherence junction, cell-cell junction, cell-substrate junction, cytoskeleton), GO biological process (regulation of transcription and gene expression, cellular protein modification process, regulation of intracellular signal transduction and cell differentiation, and protein phosphorylation), and various disease and cancer pathways from the KEGG analysis results (Fig. 6). The miRNA has-miR-200b-3p regulated genes involved in a neurotrophin signaling pathway that regulates neural cell survival, growth, and differentiation, ErbB signaling pathway that regulates cell proliferation, differentiation, migration, apoptosis, and motility, renal cell carcinoma, PI3K-Akt, FoxO, Hippo, MAPK, and TGF-beta signaling pathway, longevity regulating pathway, small cell lung cancer, and prostate cancer.

Fig. 6
figure 6

Gene enrichment analysis results using identified common target genes of significant miRNA hsa-miR-200b-3p differentially expressed between exclusive e-cigarette users and non-users predicted by at least one of the three different miRNA target prediction methods (EVmiRNA, miRDB, ExoCarta). (A) GO Molecular Function 2021 database; (B) GO Cellular Component 2021 database; (C) GO Biological Process 2021 database; (D) KEGG 2021 Human database.

Full size image

Inhibiting miR-125b-5p impairs the wound-healing ability of bronchial epithelial cells upon exposure to e-liquid

To investigate whether inhibiting miR-125b-5p could impair the wound-healing ability of bronchial epithelial cells upon exposure to e-liquid, we treated the cells with e-liquid and PGVG (control for e-liquid) in combination with the miR-125b mimic and inhibitors. A schematic representation of the workflow is depicted in Fig. 7A. To understand the mechanism of action of the miR-125b mimic, we assessed cell viability. We observed reduced viability following treatment with the miR-125b mimic and the combination of e-liquid and the miR-125b mimic, indicating the effectiveness of the miR-125b mimic (Fig. 7B). Next, we performed a wound healing assay, summarized in Fig. 7C. After transfections, we generated a scratch and treated the cells with e-liquid, PGVG, and their combinations. The combination of miR-125b mimic treatment and e-cig treatment significantly slowed down the migration of bronchial epithelial cells into the wounded area. In contrast, a significant increase in migration was observed following treatment with the inhibitor and PGVG (Fig. 7D).

Fig. 7
figure 7

Inhibition of miR-125b-5p Impairs Wound Healing Mediated by Bronchial Epithelial Cells. A: Schematic representation of the workflow for the treatment of bronchial epithelial cells. Cells were seeded in six-well plates and transfected with miR-125b-5p mimic, inhibitors, or a nonspecific control. Following transfection, the cells were treated with e-liquid and PGVG in combination with the mimic and inhibitor. B: Cell number and viability were assessed using AO/PI staining. C: Schematic of the wound healing assay. D: The scratched wounds were monitored by capturing images under a microscope at 0 h and 24 h post-scratch.Statistical significance was determined as follows: ***p < 0.001 was considered significant compared to the negative control; #p < 0.05 and ##p < 0.001 were considered significant compared to the miR-125b-5p mimic treatment.

Full size image

Overexpression of miR-125b-5p promotes DNA damage in bronchial epithelial cells

miR-125 is a microRNA known to play a significant role in various cellular processes, including the response to DNA damage. When DNA damage occurs, the expression of certain microRNAs, including miR-125, can be altered as part of the cellular stress response66,67. To assess the effect of the miR-125 mimic during DNA damage, we measured the levels of the DNA damage marker γH2AX following treatments. An enhanced level of DNA damage was observed with the miR-125 mimic and e-liquid treatments, which was alleviated upon treatment with the miR-125 inhibitor. No significant effects were noted in the control group (PGVG) with either the mimic or inhibitor treatment (Fig. 8).

Fig. 8
figure 8

Overexpression of miR-125b-5p promotes DNA Damage in Bronchial Epithelial Cells.

Full size image

Receiver operating characteristic (ROC) curve for differentiation of exclusive e-cigarette users from non-users

A ROC curve was constructed using the four significant miRNAs differentially expressed between exclusive e-cigarette users and non-users, along with race, to predict whether a participant falls into the category of exclusive e-cigarette users or non-users (Fig. 9). At the optimal Youden Index point, the sensitivity of the four significant miRNA markers and race reached 100%, with a specificity of 60%, effectively distinguishing exclusive e-cigarette users from non-users. The calculation of the Area Under Curve (AUC) yielded a value of 0.8571, indicating a highly reliable predictive capability for determining the e-cigarette use status using the four significant miRNAs and race. Using the four significant miRNAs differentially expressed between exclusive e-cigarette users and non-users within White adults, we found the e-cigarette use status can be predicted reliably from the expression levels of the four miRNAs with AUC value of 0.9167 (S6 Fig).

Fig. 9
figure 9

ROC curve of the four significant miRNAs differentially expressed between exclusive e-cigarette users and non-users and race for prediction of whether the participant is exclusive e-cigarette users or non-users. ROC curve showed that the sensitivity of the four miRNA markers and race for distinguishing non-users from exclusive e-cigarette users can reach 100% and the specificity can reach to 60% based on the optimal Youden Index point. The Area Under Curve (AUC) of the ROC is 0.8571, signifying excellent predictive capability when utilizing the four significant miRNAs in conjunction with race. TRP: true positive rate; FPR: false positive rate.

Full size image

Discussion

Exosomal miRNAs play pivotal roles in intercellular communication and various biological processes68,69. Dysregulation of exosomal miRNAs could link to tumorigenesis through regulation changes in central cellular process including cell proliferation, cell survival, and apoptosis. Current research has emphasized the potential of exosomal miRNAs as prognostic and diagnostic biomarkers for diseases70. While short-term adverse effects of e-cigarettes on human health have been well-documented, the long-term health effects of e-cigarette use remains unknown due to the relatively short period these products have been on the market71,72. Few studies have examined the impact of e-cigarette use, particularly exclusive e-cigarette use, on exosomal miRNAs.

Using plasma samples from 30 participants in the PATH Wave 1 study, we investigated differences in exosomal miRNA expression levels between exclusive e-cigarette users and non-users. We identified four significant exosomal miRNAs (hsa-miR-100-5p, hsa-miR-125a-5p, hsa-miR-125b-5p, and hsa-miR-99a-5p) that were upregulated in exclusive e-cigarette users compared to non-users, after adjusting the confounding effect from race. However, none of them remain statistically significant after we control the FDR at 5%. A subgroup analysis on White adult participants identified four significant miRNAs (hsa-miR-100-5p, hsa-miR-125b-5p, hsa-miR-200b-3p, and hsa-miR-99a-5p) between exclusive e-cigarette users and non-users and one (hsa-miR-200b-3p) remain statistically significant after controlling the FDR at 5%. Further gene enrichment analysis of predicted target genes regulated by these five significant miRNAs revealed their involvement in various cancer pathways. Our results suggest a pro-carcinogenic cellular signaling pathways affected by exclusive e-cigarette use.

The involvement of hsa-miR-100-5p in numerous pathological processes associated with diseases has been documented. Overexpression of hsa-miR-100-5p in human ovarian endometriotic stromal cells has been linked to the promotion of invasion, contributing to the pathogenesis of endometriosis73. Additionally, hsa-miR-100-5p has been identified as an independent risk factor and a prognostic signature for patients with stomach adenocarcinoma74. Elevated expression levels of hsa-miR-100-5p have been associated with extracapsular extension and poorer survival in patients with oral squamous cell carcinoma compared to their normal counterparts75. In Alzheimer’s disease, hsa-miR-100-5p is known to regulate neuron survival by targeting the Mammalian Target of Rapamycin (mTOR) pathway, a central player in regulating many fundamental cell processes and a critical factor in tumor metabolism76,77,78. The upregulation of hsa-miR-100-5p in exclusive e-cigarette users, when compared to non-users, indicates potential elevated risks of various cancers.

The miRNA has-miR-125a-5p has been recognized as a tumor suppressor in various malignancies, including those affecting the breast, ovary, lung, and central nervous system79,80. It plays a role in cell proliferation through cell cycle regulation and has potential as a therapeutic target for treating squamous cell carcinoma of the head and neck79,80. Prior study has indicated that has-miR-125a-5p can induce apoptosis through a p53-dependent pathway in human lung cancer cells81. Furthermore, its associated with the pathological stage or lymph node metastasis in non-small cell lung cancer has been reported82.

The miRNA hsa-miR-125b-5p has been identified as a regulator of inflammatory genes, targeting MAPKs and NF-kB signaling pathways in human osteoarthritic chondrocytes83,84. Additionally, miR-125b-5p inhibits the expression of TNFR2, demonstrating immunosuppressive activity to enhance the antitumor efficacy in human colon adenocarcinoma patients85. In liver cancer, miR-125b-5p has been shown to inhibit cell proliferation, migration, and invasion86. The miRNA hsa-miR-99a-5p, recognized as a tumor suppressor in tumors like bladder cancer and breast cancer, suppresses cell proliferation, migration, and invasion87. Interestingly, a previous study found that the hsa-miR-99a-5p expression level in breast cancer tissues were significantly lower than healthy breast tissue, while the expression level in plasma samples were significantly higher in breast cancer patients than in healthy controls88.

Although has-miR-125a-5p, miR-125b-5p, and hsa-miR-99a-5p are known tumor suppressors, it remains uncertain whether the upregulation of these exosomal miRNAs in plasma samples indicates an elevated risk in exclusive e-cigarette users. The experimental validation assay showed that overexpression of miRNA hsa-miR-125b-5p along with e-Liquid treatments, slowed wound healing and contributed to DNA damage in bronchial epithelial cells. miR-125b is one miRNA among the freely circulating microRNAs and no significant differences were found between plasma miRNAs and plasma-derived exosomal miRNAs89. Further research is necessary to evaluate the role of exosomal miRNAs in the association between exclusive e-cigarette use and adverse health effects.

Sensitivity analysis of the miRNA expression levels between exclusive e-cigarette users and non-users within White adults identified three overlapped miRNAs (hsa-miR-100-5p, hsa-miR-125b-5p, and hsa-miR-99a-5p). This indicated the importance of these three miRNAs for all races, although they are not statistically significant after controlling the FDR at 5% due to the relatively small sample size. The extra miRNA identified in the White adult only (hsa-miR-200b-3p) remains significant after adjusting the multiple testing error rate to control the FDR at 5%. Gene enrichment analysis of all identified miRNAs showed their involvement in cancer-related pathways, which indicates the potential harm of exclusive e-cigarette use on human health.

Limited research has explored the impact of e-cigarette use on exosomal miRNA profiles. A prior study in 2019 investigated plasma exosomal miRNA expression levels, revealing a different set of miRNAs significantly differentially expressed between exclusive e-cigarette users and non-smokers47. This variation may stem from differences in e-liquid and e-cigarette devices. Our samples were collected during the PATH Wave 1 study from 2013 to 2014, aligning with the era of the first generation of e-cigarette devices90. During this period, most e-cigarettes were usually disposable single units containing natural nicotine in the e-liquid. In contrast, the predominant e-cigarette devices on the market in 2019 were fourth generation devices, largely disposable with nicotine salt in the e-liquid to mitigate nicotine harshness90. Given the large differences in e-liquid and e-cigarette devices, it is highly plausible that distinct exosomal miRNAs are influenced by e-cigarette use.

The activity of microRNA 125b and 100 was found to be differentially expressed in the participant cohort of e-cigarette users. We wanted to study the impact of these microRNAs on bronchial epithelial cells, which are considered to be one of the main targets when inhaling e-cigarettes. Prolonged inhalation of e-cigarettes may directly or indirectly cause dysfunction in the airway epithelial barrier, which could then potentially allow respiratory pathogens to invade, leading to the development and progression of respiratory comorbidities. This may also activate an indirect mechanism of signaling cascades, such as mediating paracrine and endocrine pathways. Our future research will focus on investigating the mechanisms of how these microRNAs interact with their targets to affect pro-carcinogenic cellular signaling pathways leading to damaging/injurious responses in exclusive e-cigarette users and subjects who have been exposed to e-cigarette use.

Our current study has several limitations. First, no biomarker of exposure (BOE) data is available for the 30 participants. The absence of BOE data limited our ability to verify the accuracy of the self-reported tobacco use history data, which might bring potential bias to our analysis results. Second, the relatively small sample size of 30 participants limited the power to identify more significant miRNAs associated with exclusive e-cigarette use. Third, our study lacked a positive control group for cigarette use, and we could not assess the expression profiles of those significant miRNAs in smokers. Considering the ample evidence of less toxicant exposure and potential adverse health effects of e-cigarettes compared to cigarette smoking, we anticipate that the changes in miRNA expression levels from exclusive e-cigarette users will be less pronounced than those in smokers, as e-cigarette users have a significantly lower risk than smokers91,92,93,94. Finally, with the evolution of e-cigarette products, results from our current study might not be applicable to current e-cigarette users, given the differences in nicotine form and devices.

Conclusions

In summary, our analysis of blood samples collected during the PATH Wave 1 study from 2013 to 2014 identified five plasma exosomal miRNAs including hsa-miR-100-5p, hsa-miR-125a-5p, hsa-miR-125b-5p, hsa-miR-200b-3p, and hsa-miR-99a-5p. All five miRNAs were upregulated in exclusive e-cigarette users compared to non-users. Gene enrichment analysis of predicted target genes regulated by these five significant miRNAs uncovered numerous target genes associated with cancer-related signaling pathways. Overexpression of miRNA hsa-miR-125b-5p can promote DNA damage in bronchial epithelial cells, suggesting an activation of pro-carcinogenic signaling pathways. To validate these results, further investigations into the impact of exclusive e-cigarette use on plasma exosomal miRNAs and their correlation with cancer risks are warranted. The AUC of the ROC curve indicated a robust predictive ability of the four significant miRNAs when combined with race in distinguishing exclusive e-cigarette users from non-users.