DR SIMON A FOX (Orcid ID : 0000-0003-2042-1809)

Accepted Article PROFESSOR CAMILE S FARAH (Orcid ID : 0000-0002-1642-6204)

Article type : Original Article

Transcriptome changes induced in vitro by alcohol-containing mouthwashes in normal and

dysplastic oral keratinocytes

Simon A Fox1, Sean S Currie2, Andrew J Dalley2, Camile S Farah1,2

1
Australian Centre for Oral Oncology Research & Education, UWA Dental School, University
of Western Australia, Nedlands, WA, Australia.
2
University of Queensland Centre for Clinical Research, The University of Queensland,
Herston, Queensland, Australia.

Corresponding Author:

Professor Camile Farah

UWA Dental School, University of Western Australia

17 Monash Ave, Nedlands WA 6009

T: 08 9346 7636

E: camile.farah@uwa.edu.au

Abstract
Background

The role of alcohol-containing mouthwash as a risk factor for the development of oral
cancer is a subject of conflicting epidemiological evidence in the literature despite alcohol
being a recognised carcinogen. The aim of this study was to use in vitro models to
investigate mechanistic and global gene expression effects of exposure to alcohol-
containing mouthwash.

This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process, which may
lead to differences between this version and the Version of Record. Please cite this article as
doi: 10.1111/jop.12704
This article is protected by copyright. All rights reserved.
 Methods
Accepted Article
Two brands of alcohol-containing mouthwash and their alcohol-free counterparts were used
to treat two oral cell lines derived from normal (OKF6-TERT) and dysplastic (DOK) tissues.
Genotoxicity was determined by Comet assay. RNA-seq was performed using the Ion
Torrent platform. Bioinformatics analysis used R/Bioconductor packages with differential
expression using DEseq2. Pathway enrichment analysis used EnrichR with the Wikipathways
and Kegg databases.

Results

Both cell lines displayed dose dependent DNA damage in response to acute exposure to
ethanol and alcohol-containing mouthwashes as well as alcohol-free mouthwashes
reconstituted with ethanol as shown by Comet assay. The transcriptomic effects of alcohol-
containing mouthwash exposure were more complex with significant differential gene
expression ranging from >2000 genes in dysplastic (DOK) cells to <100 genes in normal
(OKF6-TERT) cells. Pathway enrichment analysis in DOK cells revealed alcohol-containing
mouthwashes showed common features between the two brands used including DNA
damage response as well as cancer associated pathways. In OKF6-TERT cells the most
significantly enriched pathways involved inflammatory signalling.

Conclusions

Alcohol-containing mouthwashes are genotoxic in vitro to normal and dysplastic oral

keratinocytes and induce widespread changes in gene expression. Dysplastic cells are more
susceptible to the transcriptomic effects of mouthwash.

This article is protected by copyright. All rights reserved.

 INTRODUCTION
Accepted Article Oral squamous cell carcinoma (OSCC) exhibits a classical multistep carcinogenesis whereby
cells undergo accumulated acquisition of dysregulated oncogenic signalling and evasion of
homeostatic mechanisms1. OSCC can arise through a process of transition from normal
epithelium through potentially malignant precursor disorders to carcinoma1. The main risk
factors for OSCC development are tobacco and alcohol use, which are reported to account
for the majority of oral cancers worldwide1 . Furthermore, epidemiological evidence has
indicated a synergistic interaction between the two exposures2.

Ethanol itself is not directly mutagenic but exerts its carcinogenic effects through
downstream metabolic products3. Following ethanol consumption, acetaldehyde
accumulates in the upper digestive tract through oxidation of alcohol by microbes, parotid
glands and mucosal cells4. Acetaldehyde is a recognised human carcinogen and has been
shown to form mutagenic DNA adducts as well as interfere with DNA repair mechanisms5.
Polymorphisms in genes involved in acetaldehyde metabolism modulate the cancer risk
associated with ethanol exposure6. Induction of CYP2E1 by ethanol has a number of
carcinogenic effects including the activation of pro-carcinogens (notably tobacco smoke
derived compounds), depletion of retinoic acid and most significantly, the generation of
reactive oxygen species (ROS)3. Formation of ROS leads to a number of genotoxic effects
including DNA damage as well as generation of mutagenic lipid peroxidation products7.
Alcohol exposure may also synergise with other carcinogens including tobacco products by
enhancing the permeability of the mucosa8.

Ethanol is a significant component of a number of commercially available mouthwashes and

this has led to the scrutiny of these products as potential contributors to cancer risk. Indeed,
some mouthwashes contain as much as 26% v/v ethanol. Most studies of mouthwash and
oral cancer have been epidemiological, however, several groups have demonstrated that
use of an alcohol-containing mouthwash results in the generation of detectable oral
acetaldehyde levels for up to 10 minutes9,10. The question of whether this translates into
increased oral cancer risk has been investigated in a number of case-control studies11–15 and

This article is protected by copyright. All rights reserved.

 meta analyses16 with overall contradictory conclusions. More recently, the International
Head and Neck Cancer Epidemiology (INHANCE) consortium has reported a pooled analysis
Accepted Article
of mouthwash use in a large sample (8981 cases and 10090 controls) derived from 12 case-
control studies17. This analysis supported increased risk of head and neck cancer in long-
term and frequent users of mouthwash within the limitations of the retrospective study
design which was unable to evaluate effects in non-users of tobacco and alcohol17. The
literature is deficient in specific investigations of the cellular and molecular effects of
repeated exposure to alcohol containing mouthwashes. Apart from demonstration of oral
acetaldehyde generation9, one in vivo investigation found increased nuclear abnormalities
in exfoliated buccal cells of volunteers after twice-daily use of alcohol-containing
mouthwash for a month18. Rodrigues et al19 investigated mouthwash genotoxicity using a
Drosophila wing-spot test and found evidence that increased ethanol concentration
contributed to genetic instability.

We aimed to address deficiencies in the literature regarding the mechanistic effects of

alcohol containing mouthwashes using in vitro culture models. We used cultured cell lines
derived from normal and dysplastic oral mucosa in order to examine differences in response
to two commercial alcohol containing mouthwashes and their alcohol-free counterparts.

MATERIALS AND METHODS

Cell Culture

The cell lines DOK (human mild dysplastic oral keratinocytes) and OKF6-TERT2 (TERT
transfection immortalized normal human keratinocytes) were cultured as previously
reported20. For the transcriptome experiments, the standardised culture conditions were
DMEM/F-12 containing 10% foetal bovine serum (FBS) and 1x Antibiotic-Antimycotic. All
culture media and reagents were from Thermofisher (MA, USA) and Sigma-Aldrich (St. Louis,
USA).

This article is protected by copyright. All rights reserved.

 Mouthwash preparations
Accepted Article
Two brands of commercially available mouthwashes were used: Brand A and Brand B. For
each brand there was an alcohol-containing mouthwash (A+ or B+) and a non-alcohol
mouthwash (A- or B-). The compositions were as follows: A+ (Chlorhexidine Gluconate 1.28
mg/mL + Ethanol 96 mg/mL), A- (Chlorhexidine Gluconate 2 mg/mL), B+ (Ethanol 213.03
mg/mL + Benzoic Acid 1.5 mg/mL + Thymol 0.64 mg/mL + Eucalyptol 0.92 mg/mL) and B-
(Sodium Fluoride 0.22 mg/mL + Methyl Salicylate + Thymol + Eucalyptol). For some
experiments solutions of ethanol in distilled water were also used. To control for and
examine the role of compounds in each mouthwash, several preparations of each were
used. Firstly, there was the unaltered mouthwash (e.g. A+, A-, B+, B-). In addition,
mouthwash was dried using a centrifugal evaporator to remove volatile compounds
including ethanol and reconstituted to its original weight with either distilled water (e.g.
A+W, A-W, B+W, B-W) or an appropriate solution of ethanol in distilled water (e.g. A+E, A-E,
B+E, B-E) to restore the original ethanol concentration. The various treatment preparations
and identifying codes are given in Table 1.

Comet Assay

Cells were seeded at 2x105 cells/well in 12-well plates, incubated overnight, exposed to
treatments or controls in RPMI media for 5 min at 37°C and washed twice with PBS.
Following cell harvesting with TrypLE Express (Thermofisher) the comet assay was
performed essentially as described by Speit and Hartmann21 with the exception that the
lysis buffer included 1% sarkosyl. Duplicate slides were prepared for each sample and
immediately upon staining 50-100 comets per slide were captured using a Zeiss Axiovision
M1 fluorescence microscope (Carl Zeiss AG, Jena, Germany). Comets were analysed using
CometScore™ (TriTek , VA, USA). Percentage tail DNA migration (% Tail intensity) was used
as the statistic of DNA damage evaluation. Representative comet images of OKF6-TERT cells
demonstrating increasing amounts of tail DNA migration indicative of DNA damage are
shown in Figure 1. Controls were RPMI media alone (negative) and 0.3% hydrogen peroxide
(positive).

This article is protected by copyright. All rights reserved.

 Cell Treatment and RNA sequencing
Accepted Article
Treatments were diluted to 0.5% v/v concentration in culture media. Cells were exposed for
10 mins, twice daily for 7 days. Following the final exposure, cells were allowed to recover
for 2 hr before harvesting. All groups were photographed on an Incucyte™ kinetic imager
(Essen BioScience, UK) both at Day 0 and after the final exposure to control for cytotoxic
effects. At the completion of the culture experiment, cells were washed twice with PBS and
harvested with TrypLE Express (Thermofisher). RNA was extracted using an AllPrep
DNA/RNA/miRNA Universal kit (Qiagen, Netherlands) and quantity and quality assessed
using NanoDrop spectrophotometer (Thermofisher), Qubit fluorometer (Thermofisher) and
2100 Bioanalyzer (Agilent Technologies, CA, USA). The RiboMinus™ Eukaryote System v2 kit
(Thermofisher) was used to deplete ribosomal RNA (rRNA). 50ng of rRNA depleted RNA
input was used for library preparation with an Ion Total RNA-Seq Kit v2 (Thermofisher),
barcoded with the Ion Xpress™ RNA-Seq Barcode 1-16 Kit (Thermofisher). Templating and
chip loading was performed with 110pM of each library using the Ion PI IC kit
(ThermoFisher) and an Ion Chef system (ThermoFisher) using Ion PIv3 sequencing chips.
Loaded chips were sequenced on a Proton sequencer (ThermoFisher) with data collection
and adaptor trimming using Torrent Suite v4.2 software.

Sequencing data filtering and bioinformatics

Prior to further analysis raw reads were quality checked with fastQC R tool and pre-
processed to remove low quality sequence using trimmomatic R tool. The processed reads
were mapped to reference genome hg19 using Burrows-Wheeler Aligner (BWA-MEM)22.
Following the mapping step, the alignment files were analysed with RNA-seQC23 for
descriptive quality control (QC) of the mapped reads. Read count matrices of summarised
data for each gene were generated using HTseq24. Differential gene expression analysis was
performed with the R/Bioconductor package DESeq2 which uses a negative binomial
model25. Analysis was performed using standard parameters with the independent filtering
function enabled to filter genes with low mean normalized counts. Adjusted p-values (Padj)
for multiple testing, using Benjamini-Hochberg to estimate the false discovery rate (FDR),
were calculated for final estimation of DE significance using a 0.05 cutoff.

This article is protected by copyright. All rights reserved.

 Gene set enrichment analysis for biological pathways and gene ontology terms
Accepted Article
Overrepresentation analysis (ORA) of differentially expressed genes for pathways and gene
ontology terms was performed using the web based platform EnrichR
(http://amp.pharm.mssm.edu/Enrichr) which permits interrogation of multiple databases 26.
Up and down regulated gene lists were evaluated for significant enrichment against the
following gene set libraries: GO Biological Process, GO Molecular Function (all from
http://www.geneontology.org), the Wikipathways database
(http://www.wikipathways.org/) and KEGG (http://www.kegg.jp/kegg) (all current as of
December 2017). Enriched annotations/pathways were selected and ranked based upon the
combined score which was calculated by the EnrichR platform following Z-score
permutation background correction on the Fischer Exact Test p-value 26.

RESULTS

DNA damage in normal and dysplastic keratinocyte cultures

Initially, we performed a preliminary characterisation of the cellular effects of alcohol-

containing mouthwashes and their alcohol-free counterparts in culture models of normal
and dysplastic oral keratinocytes. We used an initial dose finding toxicity analysis to
establish suitable sub-cytotoxic exposures for subsequent experiments (data not shown).
We found that an acute exposure of 1% v/v for 5 minutes and a chronic exposure of 0.5%
v/v for 10 minutes twice daily for 7 days were not growth inhibitory or cytotoxic for any of
the preparations used. We next examined the genotoxic effects of acute exposure to the
mouthwash preparations and ethanol alone using a single cell electrophoresis assay (Comet
assay). Figure 2A shows the effects of ethanol upon DNA strand breakage in OKF6-TERT and
DOK cells with acute exposure to concentrations greater the 0.115% ethanol resulting in a
dose dependant increase in DNA damage. Overall the dysplastic DOK cells demonstrated a
higher baseline level of DNA strand breaks than the normal OKF6-TERT cells.

We similarly investigated the genotoxic effects of various alcohol-containing and alcohol-

free mouthwash preparations upon these cell lines. Since these commercial products
contain a complex mixture of components we prepared various treatment solutions to

This article is protected by copyright. All rights reserved.

 control for the differences in composition and allow interrogation of the effects of ethanol.
As described in the Methods, we essentially evaporated the volatile compounds in the
Accepted Article
mouthwash products and then reconstituted either with distilled water or restored to the
original ethanol concentration. The full listing of these solutions is provided in Table 1 and
all preparations were used at 1% v/v final concentration for acute cell culture treatment. For
both brands of mouthwash, we consistently observed in both cell lines the genotoxic effects
of the alcohol-containing mouthwash in comparison to the alcohol-free counterpart (Figure
2B and 2C). Evaporation and reconstitution with water (A+W and B+W) abrogated this
genotoxic effect so that the levels were consistent with the alcohol-free solutions (A- and B-
). Conversely, when the alcohol-free mouthwashes were dried and reconstituted with
ethanol (A-E and B-E) there was an increase in the DNA damage to a level broadly congruous
with alcohol-containing mouthwash (A+ and B+). These results are consistent with alcohol-
containing mouthwashes exerting genotoxic effects and suggest that alcohol is the
component contributing to this effect. Brand B+ was slightly more genotoxic than Brand A+,
but this was not significantly different (Figure 2D).

Identification of mouthwash induced genes in normal and dysplastic keratinocyte cultures

To monitor gene expression changes in response to chronic mouthwash exposure DOK and
OKF6-TERT cells were exposed to 0.5% v/v mouthwash preparations for 10 minutes twice
daily for 7 days. Treatments for each cell line included a negative control, the alcohol-free
and alcohol-containing mouthwash of each brand. Three independent experiments were
carried out and RNA extracted to perform transcriptome wide expression profiling by
RNAseq. We performed RNAseq on 30 samples with an average read depth of 26.5 million
reads. Pre-processed reads were mapped against the hg19 reference genome with an
average efficiency of 92.6% (range 87.4%-96.6%) with detection of around 30,000
transcripts overall per library. We then performed a pairwise differential expression analysis
between the treatment groups with an overview of differentially expressed gene numbers
and the subsets of up and down regulated genes shown in Table 2. As can be seen there was
a striking difference in the transcriptomic effects of the mouthwashes between the normal
and dysplastic cultures. With the exception of the alcohol-free brand B (B-), the effects were

This article is protected by copyright. All rights reserved.

 much greater in the DOK (dysplastic) cell line with more than 1000 genes significantly
regulated by the treatments. Although a large number of genes were found regulated, the
Accepted Article
magnitude of the effects were modest with the maximum fold change of 20 and average
across the treatments of around 1.5-fold. To further explore the regulated genes we used
Venn analysis to identify sets of genes uniquely regulated by the alcohol containing
mouthwashes as well as commonly regulated genes between the mouthwashes (Figure 3 A-
B). This analysis showed that in DOK cells treated with alcohol-containing mouthwash there
were 296 genes uniquely regulated by brand A+ while in contrast 2534 genes were uniquely
regulated by brand B+ (Figure 3A). Additional analysis of these showed that there were 101
genes commonly regulated across both brands (Figure 3A). Similar analysis was performed
for the OKF6-TERT treated cells (Figure 3B) although the numbers of significantly regulated
genes were much lower with 58 and 37 genes unique to A+ and B+ respectively. There were
no genes commonly regulated by both brands of alcohol-containing mouthwash in OKF6-
TERT cells (Figure 3B).

Gene Ontology and Pathway Enrichment analysis of mouthwash induced genes

To explore the biological meaning of the regulated genes we performed statistical analysis
for enrichment of gene ontology (biological processes or molecular function) and biological
pathways (KEGG or Wikipathways databases) using the gene subsets generated by Venn
analysis (Figure 3). In particular, we focussed upon the genes which were differentially
regulated by the alcohol-containing mouthwashes but not the alcohol-free mouthwashes.
Our analysis of gene ontology in these datasets yielded statistically significant enriched
terms especially for experiments involving DOK cells, however, they broadly related to
transcriptional regulation (i.e. transcription factors including ARID5A, SP100, ASCL2 and
DLX4) and did not yield particular biological insight. Therefore, they are not presented here.
For pathway analysis of gene regulation specific to alcohol-containing mouthwashes A+ and
B+ there were 20 and 104 significantly enriched pathways respectively in the treated DOK
cells. The top 10 pathways for each mouthwash are shown in Table 3. As can be seen, two
pathways, viral carcinogenesis (hsa05203) and DNA damage response (WP710), were
common to the most highly ranked pathways for both mouthwash treatments. For analysis

This article is protected by copyright. All rights reserved.

 of the OKF6-TERT treatments there were few significant pathways or ontology terms found
with the exception of alcohol-containing mouthwash A+ where 4 pathways were enriched:
Accepted Article
TNF signalling (hsa04668), cytokine-cytokine receptor interaction (hsa04060), rheumatoid
arthritis (hsa05323) and Jak-STAT signalling (hsa04630). Pathway analysis of genes
exclusively regulated by the alcohol-free mouthwashes A- and B- did not yield any enriched
pathways in either cell line.

DISCUSSION

The principal aim of this study was to use cellular assays of genotoxicity and RNA-seq to
investigate the in vitro effects of exposure to alcohol-containing and alcohol-free
commercial mouthwashes upon oral keratinocytes. In addition, we also characterised the
different responses between normal and dysplastic cells. The literature in this area is largely
epidemiological and there is a deficiency of in vitro studies which we aimed to address. Our
initial investigation of the genotoxic effects of alcohol and mouthwash preparations showed
a genotoxic effect and were entirely consistent with this effect being exerted by alcohol.
Even a short term acute in vitro exposure resulted in evidence of DNA damage in both OKF6-
TERT and DOK cells as assessed by Comet assay. This is consistent with the in vitro evidence
for DNA damage by ethanol and that it can exert a genotoxic effect additional to that
initiated by its acetaldehyde metabolites 27.

Although it did appear that the magnitude of the DNA damage effect was slightly greater in
DOK cells, overall both cell lines exhibited broadly similar DNA damage effects across a
variety of treatments. In contrast there was a striking difference between OKF6-TERT and
DOK cells with respect to the transcriptomic response. We have previously demonstrated
that DNA damage repair mechanisms are more active in the normal derived OKF6-TERT cells
than in dysplastic derived DOK cells 28 and these findings are consistent with that previous
study. Whether this enhanced sensitivity of dysplastic cells is a feature of similar in vitro
models and can be generalised to in vivo clinical oral epithelial dysplasia requires further
investigation. It does raise the possibility that dysplastic tissue is more susceptible to

This article is protected by copyright. All rights reserved.

 genotoxic effects of ethanol and suggests an additional mechanism for the synergistic
effects of alcohol and tobacco on oral mucosal changes particularly in oral potentially
Accepted Article
malignant disorders (OPMD), and may explain the added risk of malignant transformation in
patients with OPMD who smoke and also use alcohol-containing mouthwashes on a regular
basis. It has been shown that mouthwash users who smoke are at greater risk of developing
OSCC than non-smoking users (OR = 9.15 for ≥ twice daily mouthwash use) 13,29. This
compares to an OR of 4.98 for former smokers, and 5.12 for current alcohol drinkers, which
is not significantly different from non-smokers (OR = 5.86). The combined effects of smoking
and chronic daily use of alcohol-containing mouthwash may result in continued mutagenic
events within an already sensitised field, promoting continued epithelial transformation. It is
possible to identify several groups of alcohol-containing mouthwash users who could
theoretically be at higher risk with chronic use of high alcohol-containing mouthwash.
Firstly, subjects who smoke and use alcohol-containing mouthwash are regularly exposed to
both tobacco carcinogens and ethanol, the synergy of which has been highlighted above.
Epidemiological studies have also shown that current and past smokers are more likely to
use mouthwash 11,12,29,30. Secondly, use of alcohol-containing mouthwash by patients with
oral epithelial dysplasia has the potential for concern, since continued exposure to ethanol
may act to facilitate progression towards malignancy. Patients with oral epithelial dysplasia
tend to be smokers, and the oral epithelium in these patients is already transformed,
placing them at heightened risk of further cellular and molecular damage should they
engage in chronic use of alcohol-containing mouthwash. It is also possible that the discovery
of an oral lesion by a patient may act as the motivating factor for mouthwash use, which
would place the patient at increased risk of further damage to an existent lesion.

Overall the transcriptomic responses across the experiments were diverse in magnitude
with distinct responses seen, not only between the cell lines, but also differentially between
the mouthwash brands. Given the differences in composition across the brands these
differences are not unexpected although the complex composition does make dissection of
the effects problematic. We did not see significant differential expression in comparisons of
the alcohol-containing and alcohol-free preparations within each brand. However, the
composition differences other than alcohol may have had a confounding effect.

This article is protected by copyright. All rights reserved.

 Unfortunately, for practical reasons it was not feasible to perform RNA-sequencing using all
of the preparations used in our preliminary DNA damage study. The focus of this analysis
Accepted Article
was therefore to investigate the biological effects at the global gene expression level and
thereby to elucidate biological meaning and distinguishing features through ontology and
pathway analysis.

Pathway analysis demonstrated that in DOK cells treated with the alcohol-containing
mouthwashes the specifically regulated genes were enriched for many pathways associated
with carcinogenesis and DNA damage response. Several pathways were found common to
the two mouthwash brands indicating a commonality of biological effects even if the
common gene numbers (101 by Venn analysis) were relatively low. Notably this included the
DNA damage response pathway which is consistent with the genotoxic effects seen in the
Comet assay. The absence of such a finding in the transcriptome of OKF6-TERT cells may
reflect the differences in DNA damage response in these cells as previously reported by us28.
The 3 significant pathways enriched in OKF6-TERT cells treated with Brand A+ all related to
inflammatory signalling and indicate distinct biological effects between the cell lines.
Despite the large number of genes regulated by the alcohol-free mouthwashes A- in DOK
cells and B- in OKF6-TERT cells, we did not identify enrichment of specific biological
pathways. Since the results of the cellular assays indicate an absence of genotoxicity further
investigation of the individual components of these mouthwashes would be required to
elucidate this effect.

The data we present here provides insight into the biological and gene regulatory
mechanisms underlying the effects of alcohol-containing mouthwash on oral epithelial cells.
Whether these in vitro findings are reflected by similar in vivo effects requires further
investigation.

This article is protected by copyright. All rights reserved.

 CONCLUSIONS
Accepted Article
Alcohol-containing mouthwashes are genotoxic in vitro to normal and dysplastic oral
keratinocytes and induce widespread changes in gene expression. Dysplastic cells are more
susceptible to the transcriptomic effects of alcohol-containing mouthwash.

ACKNOWLEDGEMENTS

The authors thank Stephen Rudd, Roxane Legaie and Anne Bernard from the Queensland
Facility for Advanced Bioinformatics (QFAB) for assistance with bioinformatics analysis and
Richard Allcock (LotteryWest State Biomedical Facility Genomics, Western Australia) for next
generation sequencing. This study was supported by grants from The Queensland
Government Smart Futures Co-investment Fund, Cancer Australia, Agilent Technologies, Life
Technologies and The Australian Dental Research Foundation awarded to CSF.

CONFLICT OF INTEREST

None to declare.

REFERENCES

1. Leemans CR, Braakhuis BJM, Brakenhoff RH. The molecular biology of head and neck cancer. Nat
Rev Cancer. 2011;11(1):9-22. doi:10.1038/nrc2982

2. Turati F, Garavello W, Tramacere I, et al. A meta-analysis of alcohol drinking and oral and
pharyngeal cancers: results from subgroup analyses. Alcohol Alcohol Oxf Oxfs. 2013;48(1):107-
118. doi:10.1093/alcalc/ags100

3. Seitz HK, Stickel F. Molecular mechanisms of alcohol-mediated carcinogenesis. Nat Rev Cancer.
2007;7(8):nrc2191. doi:10.1038/nrc2191

4. Väkeväinen S, Tillonen J, Agarwal DP, Srivastava N, Salaspuro M. High salivary acetaldehyde after
a moderate dose of alcohol in ALDH2-deficient subjects: strong evidence for the local
carcinogenic action of acetaldehyde. Alcohol Clin Exp Res. 2000;24(6):873-877.

5. Brooks PJ, Zakhari S. Acetaldehyde and the genome: beyond nuclear DNA adducts and
carcinogenesis. Environ Mol Mutagen. 2014;55(2):77-91. doi:10.1002/em.21824

This article is protected by copyright. All rights reserved.

 6. Druesne-Pecollo N, Tehard B, Mallet Y, et al. Alcohol and genetic polymorphisms: effect on risk
of alcohol-related cancer. Lancet Oncol. 2009;10(2):173-180. doi:10.1016/S1470-
Accepted Article 2045(09)70019-1

7. Linhart K, Bartsch H, Seitz HK. The role of reactive oxygen species (ROS) and cytochrome P-450
2E1 in the generation of carcinogenic etheno-DNA adducts. Redox Biol. 2014;3:56-62.
doi:10.1016/j.redox.2014.08.009

8. Howie NM, Trigkas TK, Cruchley AT, Wertz PW, Squier CA, Williams DM. Short-term exposure to
alcohol increases the permeability of human oral mucosa. Oral Dis. 2001;7(6):349-354.

9. Lachenmeier DW, Gumbel-Mako S, Sohnius E-M, Keck-Wilhelm A, Kratz E, Mildau G. Salivary

acetaldehyde increase due to alcohol-containing mouthwash use: a risk factor for oral cancer.
Int J Cancer. 2009;125(3):730-735. doi:10.1002/ijc.24381

10. Moazzez R, Thompson H, Palmer RM, Wilson RF, Proctor GB, Wade WG. Effect of rinsing with
ethanol-containing mouthrinses on the production of salivary acetaldehyde. Eur J Oral Sci.
2011;119(6):441-446. doi:10.1111/j.1600-0722.2011.00886.x

11. Wynder EL, Kabat G, Rosenberg S, Levenstein M. Oral cancer and mouthwash use. J Natl Cancer
Inst. 1983;70(2):255-260.

12. Kabat GC, Hebert JR, Wynder EL. Risk factors for oral cancer in women. Cancer Res.
1989;49(10):2803-2806.

13. Guha N, Boffetta P, Wünsch Filho V, et al. Oral health and risk of squamous cell carcinoma of the
head and neck and esophagus: results of two multicentric case-control studies. Am J Epidemiol.
2007;166(10):1159-1173. doi:10.1093/aje/kwm193

14. Divaris K, Olshan AF, Smith J, et al. Oral health and risk for head and neck squamous cell
carcinoma: the Carolina Head and Neck Cancer Study. Cancer Causes Control CCC.
2010;21(4):567-575. doi:10.1007/s10552-009-9486-9

15. Ahrens W, Pohlabeln H, Foraita R, et al. Oral health, dental care and mouthwash associated with
upper aerodigestive tract cancer risk in Europe: the ARCAGE study. Oral Oncol. 2014;50(6):616-
625. doi:10.1016/j.oraloncology.2014.03.001

16. Gandini S, Negri E, Boffetta P, La Vecchia C, Boyle P. Mouthwash and oral cancer risk
quantitative meta-analysis of epidemiologic studies. Ann Agric Environ Med AAEM.
2012;19(2):173-180.

17. Boffetta P, Hayes RB, Sartori S, et al. Mouthwash use and cancer of the head and neck: a pooled
analysis from the International Head and Neck Cancer Epidemiology Consortium. Eur J Cancer
Prev Off J Eur Cancer Prev Organ ECP. 2016;25(4):344-348. doi:10.1097/CEJ.0000000000000179

18. Zamora-Perez AL, Mariaud-Schmidt RP, Fuentes-Lerma MG, et al. Increased number of
micronuclei and nuclear anomalies in buccal mucosa cells from people exposed to alcohol-
containing mouthwash. Drug Chem Toxicol. 2013;36(2):255-260.
doi:10.3109/01480545.2012.737804

19. Rodrigues F, Lehmann M, do Amaral VS, Reguly ML, de Andrade HHR. Genotoxicity of three
mouthwash products, Cepacol, Periogard, and Plax, in the Drosophila wing-spot test. Environ
Mol Mutagen. 2007;48(8):644-649. doi:10.1002/em.20332

This article is protected by copyright. All rights reserved.

 20. Dalley AJ, AbdulMajeed AA, Upton Z, Farah CS. Organotypic culture of normal, dysplastic and
squamous cell carcinoma-derived oral cell lines reveals loss of spatial regulation of CD44 and
Accepted Article p75 NTR in malignancy. J Oral Pathol Med Off Publ Int Assoc Oral Pathol Am Acad Oral Pathol.
2013;42(1):37-46. doi:10.1111/j.1600-0714.2012.01170.x

21. Speit G, Hartmann A. The comet assay: a sensitive genotoxicity test for the detection of DNA
damage and repair. Methods Mol Biol Clifton NJ. 2006;314:275-286. doi:10.1385/1-59259-973-
7:275

22. Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform.
Bioinforma Oxf Engl. 2009;25(14):1754-1760. doi:10.1093/bioinformatics/btp324

23. DeLuca DS, Levin JZ, Sivachenko A, et al. RNA-SeQC: RNA-seq metrics for quality control and
process optimization. Bioinforma Oxf Engl. 2012;28(11):1530-1532.
doi:10.1093/bioinformatics/bts196

24. Anders S, Pyl PT, Huber W. HTSeq—a Python framework to work with high-throughput
sequencing data. Bioinformatics. 2015;31(2):166-169. doi:10.1093/bioinformatics/btu638

25. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq
data with DESeq2. Genome Biol. 2014;15(12):550. doi:10.1186/s13059-014-0550-8

26. Kuleshov MV, Jones MR, Rouillard AD, et al. Enrichr: a comprehensive gene set enrichment
analysis web server 2016 update. Nucleic Acids Res. 2016;44(W1):W90-97.
doi:10.1093/nar/gkw377

27. Kayani MA, Parry JM. The in vitro genotoxicity of ethanol and acetaldehyde. Toxicol Vitro Int J
Publ Assoc BIBRA. 2010;24(1):56-60. doi:10.1016/j.tiv.2009.09.003

28. Jessri M, Dalley AJ, Farah CS. Deficient double-strand break repair in oral squamous cell
carcinoma cell lines. J Oral Pathol Med. 2017;46(9):695-702. doi:10.1111/jop.12576

29. Winn DM, Blot WJ, McLaughlin JK, et al. Mouthwash use and oral conditions in the risk of oral
and pharyngeal cancer. Cancer Res. 1991;51(11):3044-3047.

30. Eliot MN, Michaud DS, Langevin SM, McClean MD, Kelsey KT. Periodontal disease and
mouthwash use are risk factors for head and neck squamous cell carcinoma. Cancer Causes
Control CCC. 2013;24(7):1315-1322. doi:10.1007/s10552-013-0209-x

This article is protected by copyright. All rights reserved.

 Figure legends
Accepted Article FIGURE 1 Representative comet images of OKF6-TERT cells showing (A) ~10% tail DNA
migration (B) ~30% tail DNA migration (C) ~60% tail DNA migration (D) ~95% tail DNA
migration. Original magnification x 400.

FIGURE 2 Percentage tail intensity in OKF6-TERT and DOK cells exposed to (A) varying
concentrations of ethanol, (B) Brand A mouthwash preparations, (C) Brand B mouthwash
preparations, (D) comparison of Brand A and Brand B alcohol containing mouthwash only.
Positive control (Pos.) is 0.3% H2O2. Figures are Tukey-style boxplots showing median and
outliers. Each bar represents a single experiment comprising 50-100 comets.

FIGURE 3 Venn analysis of differentially expressed genes in (A) DOK cells and (B) OKF6-TERT
cells.

This article is protected by copyright. All rights reserved.

 TABLE 1 Mouthwash treatment solutions prepared from Brands A and B
Accepted Article Code Description

A+ alcohol containing, original

A+W alcohol containing, alcohol evaporated, reconstituted with water

A+E alcohol containing, alcohol evaporated, reconstituted with alcohol

A- alcohol free, original

A-W alcohol free, alcohol evaporated, reconstituted with water

A-E alcohol free, alcohol evaporated, reconstituted with alcohol

B+ alcohol containing, original

B+W alcohol containing, alcohol evaporated, reconstituted with water

B+E alcohol containing, alcohol evaporated, reconstituted with alcohol

B- alcohol free, original

B-W alcohol free, alcohol evaporated, reconstituted with water

B-E alcohol free, alcohol evaporated, reconstituted with alcohol

N negative control, alcohol free, vehicle only

This article is protected by copyright. All rights reserved.

 TABLE 2 Overview of differential expression analysis
Accepted Article Comparison OKF6-TERT DOK

A+ vs N 91 (13/78) 2399 (1141/1258)

A- vs N 126 (55/71) 4484 (2115/2369)

A+ vs A- 1 (0/1) 1165 (734/431)

B+ vs N 209 (95/114) 2860 (1656/1204)

B- vs N 1573 (709/864) 394 (232/162)

B+ vs B- 4 (0/4) 6 (1/5)

data is total genes (upregulated/downregulated)

TABLE 3 Top 10 Enriched Pathways for alcohol mouthwash in DOK cells

Pathway Score* -log10P

A+ vs N unique geneset (296 genes)
Integrated Pancreatic Cancer Pathway WP2377 23.64 2.46
Viral carcinogenesis hsa05203 22.13 2.95
p53 Signaling pathway hsa04115 20.80 2.95
Retinoblastoma (RB) in Cancer WP2446 18.28 2.46
Androgen receptor Signaling pathway WP138 18.25 2.46
DNA Damage Response (only ATM dependent) WP710 17.79 2.05
Senescence and Autophagy in Cancer WP615 15.79 2.09
G1 to S cell cycle control WP45 13.50 1.91
TGF-beta Signaling Pathway WP366 12.27 1.80
Integrated Cancer Pathway WP1971 11.83 1.81
B+ vs N unique geneset (2534 genes)
Insulin Signaling WP481 30.77 4.90
Proteoglycans in cancer hsa05205 27.58 3.88
Alcoholism hsa05034 27.13 3.88
Chronic myeloid leukemia hsa05220 24.92 3.81
Viral carcinogenesis hsa05203 24.23 3.68
DNA Damage Response (only ATM dependent) WP710 21.86 3.44
p38 MAPK Signaling Pathway WP400 21.76 3.57
Primary Focal Segmental Glomerulosclerosis WP2572 21.47 3.60
AGE-RAGE Signaling pathway hsa04933 21.15 3.31
EGF/EGFR Signaling Pathway WP437 20.94 3.47
Results were filtered for Padj<0.05 and ranked by EnrichR Combined Score*
Pathways from the KEGG (hsa) or Wikipathways (WP) databases

This article is protected by copyright. All rights reserved.

Accepted Article

This article is protected by copyright. All rights reserved.

Accepted Article

This article is protected by copyright. All rights reserved.