Relationship between acetaldehyde concentration in mouth air and characteristics of microbiota of tongue dorsum in Japanese healthy adults: a cross-sectional study

Abstract Acetaldehyde, associated with consumption of alcoholic beverages, is known to be a carcinogen and to be related to the tongue dorsum. Objective The aim of this study was to investigate the relationship between acetaldehyde concentration in mouth air and bacterial characteristics on the tongue dorsum. Methodology Thirty-nine healthy volunteers participated in the study. Acetaldehyde concentrations in mouth air were evaluated by a high-sensitivity semiconductor gas sensor. A 16S rRNA gene sequencing technique was used to compare microbiomes between two groups, focusing on the six samples with the highest acetaldehyde concentrations (HG) and the six samples with lowest acetaldehyde concentrations (LG). Results Acetaldehyde concentration increased in correlation with the increase in bacterial count (p=0.048). The number of species observed in the oral microbiome of the HG was higher than that in the oral microbiome of the LG (p=0.011). The relative abundances of Gemella sanguinis, Veillonella parvula and Neisseria flavescens in the oral microbiome of the HG were higher than those in the oral microbiome of the LG (p<0.05). Conclusion Acetaldehyde concentration in mouth air was associated with bacterial count, diversity of microbiome, and relative abundance of G. sanguinis, V. parvula, and N. flavescens.


Introduction
Acetaldehyde, associated with consumption of alcoholic beverages, is known as a gaseous carcinogen leading to oral, esophageal, and gastrointestinal tract cancers. [1][2][3] The 2012 Monograph of the International Agency for Research on Cancer concluded that "alcohol consumption is carcinogenic to humans; ethanol in alcoholic beverages is carcinogenic to humans; and acetaldehyde associated with the consumption of alcoholic beverages is carcinogenic to humans". 4 Previous studies have shown that acetaldehyde causes point mutations in DNA and formation of DNA adducts, inducing sister chromatid exchange and gross chromosomal aberrations. [5][6][7] Acetaldehyde can be generated in the human oral cavity by microorganisms such as yeasts and bacteria.
For example, studies demonstrated that oral Candida species are capable of producing significant amounts of acetaldehyde from ethanol and glucose in vitro. 8 In addition, Neisseria and Streptococcus species in saliva samples have been investigated regarding their role in acetaldehyde production. 9 However, the relationship between microbiome and physiological acetaldehyde concentration in mouth air is uncertain. It is important to identify the microbial factors related to acetaldehyde accumulation in mouth air, because individuals who have oral microbiome with high ability of producing acetaldehyde could have a high risk of cancer by consumption of alcoholic beverages.
A previous study revealed that acetaldehyde concentration in mouth air is associated with the tongue coating, 10 which may be a source of local production of this compound by the oral microbiome.
Acetaldehyde concentration in healthy adults with a tongue-coating-status score of 3 (i.e., coating covering more than two-thirds of the tongue dorsum surface) was significantly higher than in healthy adults with a score of 0/1 (i.e., no coating or coating covering less than one-third of the tongue dorsum surface, respectively). 10 The tongue coating consists of bacteria, large quantities of desquamated epithelial cells, blood metabolites, different kinds of food remnants, and leucocytes derived from periodontal pockets. 11,12 However, little is known about which organisms of the tongue coating microbiome affect the acetaldehyde concentration in mouth air.
The hypothesis was that the acetaldehyde concentration in mouth air is related to the presence of certain organisms of the microbiome on the tongue. measurement. As acetaldehyde is highly volatile, air contamination in the oral cavity was avoided as much as possible. Notably, during collection of mouth air (using a syringe), each participant breathed through the nose. Immediately upon collection of the sample gas into the syringe, the sample was injected and parameters were measured. 10,13 An injection of the sample gas (5 mL) from the syringe into the detector initiated the measurement automatically.
Measurement was completed in eight minutes.
To assess the reproducibility of the sampling, the two-days experiment was set for calibration. Each measurement was performed in duplicate. In the intraand inter-assay, the error was below 5%.
To assess the reproducibility of the measurement, defined samples containing 100-10,000 ppb acetaldehyde were used for calibration. Each measurement was performed in duplicate. Both intra-and inter-assay coefficients of variation were below 5%.

Oral examination
Status of tongue coating was assessed according to its distribution area, with scoring as follows: 0, none visible; 1, less than one-third of the tongue dorsum surface covered; 2: less than two-thirds of the tongue dorsum surface covered; 3: more than two-thirds of the tongue dorsum surface covered. 14 All clinical procedures were performed by one of four trained dentists (A.Y., M. Y., T. M., R. Y., and D. E.). Scoring among the dentists was calibrated by confirming that these dentists showed good intra-and inter-examiner agreement for the examination, as evaluated by kappa statistics over 0.8. The results of measurement then were shown on a liquid crystal display. The sample was collected from the median area of the tongue dorsum using a swab (men-tip ® ; J.C.B. Industry Limited, Tokyo, Japan). 15 The collection pressure was about 21 g, and a 1-cm span was sampled by rubbing the swab back and forth three times. 16 To assess the reproducibility of the sampling, the measurement was performed in duplicate at the same time. In the intra-assay, the error was below 5%.

Detection of Candida species
Acetaldehyde can be generated in the human oral cavity by microorganisms such as Candida and bacteria 8 . We detected Candida and bacteria separately. We used CHROMagar Candida medium (CHROMagar Candida, Kanto Chemical Co., Inc., Tokyo, Japan) (pH 6.1) to detect Candida albicans, Candida tropicalis, and Candida krusei. The medium comprised (per liter) peptone (10 g), glucose (20 g), agar (15 g), chloramphenicol (0.5 g), and Chromogenic IX (2 g), and was prepared according to the manufacturer's instructions. All samples swabbed from the oral mucosa and tongue were plated on the medium and then permitted to grow for 48 hours at 37°C. The color and morphology of the resulting colonies were recorded, and the organism classification was based on comparison to photographic images, including green colonies for C. albicans, steel blue colonies for C. tropicalis, and rose-colored colonies for C. krusei. 17

Assessment of alcohol sensitivity
Alcohol sensitivity, which can reflect acetaldehyde production in the human body, 18 was assessed by inferring the acetaldehyde dehydrogenase (ALDH) genotype of each participant. The ethanol patch test (ASK Human Care, Co., Ltd., Tokyo, Japan) was used to infer the ALDH genotype of each participant. 19 Briefly, a plaster patch impregnated with ethanol was fixed on adhesive tape. The plaster patch was attached to the inner surface of the arm for 20 minutes and removed according to the manufacturer's procedure.
Patients whose patch area exhibited erythema after plaster removal were judged to be positive for reaction to alcohol and inferred to have the ALDH genotype (ALDH2*1/*2 or *2/*2). If negative, the participants were assigned to the ALDH genotype (ALDH2*1/*1).

Questionnaire
In addition to age, sex, and general condition, characterized as "never", "past", and "current". 20 Information regarding alcohol consumption was characterized as "never", "light" (less than five days per week), "moderate" (five or more days per week, less than twice a day), and "heavy" (five or more days per week, more than twice a day). 21 Sample collection for identification of bacteria on tongue dorsum LG) acetaldehyde concentrations in the mouth air. These samples were collected between 11:00 am and 12:00 pm. Each sample was collected from the median area of the tongue dorsum using a swab (men-tip ® ) that had been previously wetted by immersion in 5 mL of pure water and then rolled over the sampled surface using moderate pressure (21 g) and circular motion. 22 The 12 samples then were extracted (as described below) to permit investigation of the characteristics of the respective microbiomes.

Identification of bacteria on tongue dorsum
The characteristics of the oral microbiomes from the HG and LG were compared by focusing on the relative abundance of bacteria that had been previously used to analyze the diversity of the microbiome on the tongue. As noted above, we selected the bacteria that previously 9 had been shown to be capable of producing acetaldehyde. The Mann-Whitney U test was used to analyze the significance of differences in the relative abundances of bacteria between the microbiomes of the two groups.

Results
Thirty-nine subjects (12 males and 27 females; 20-30 years old) completed this study. Table 1 shows the characteristics of the study participants. More than half of the patients were female. There were no  Table 2).
There was a significant correlation between the bacterial count measured using the DEPIM method and acetaldehyde concentration in mouth air (        LG. Data were presented as median with error bars representing the 25th and 75th percentiles (n=6/group). These were significant differences in relative abundance of some species between HG (blue bar) and LG (red bar)

Discussion
In this study, the acetaldehyde concentration was positively associated with bacterial count on the tongue; diversity of the tongue microbiome; and relative abundance of G. sanguinis, V. parvula, and N.
flavescens, species known to produce acetaldehyde.
To the best of our knowledge, this is the first report sanguinis in the HG microbiomes were significantly higher than those in the LG microbiomes. In contrast, the relative abundances of P. histicola and S. parasanguinis in the HG microbiomes were significantly lower than those in the LG microbiomes. However, a recent study reported results that differed from those of the present study. In that study, participants with type-I oral microbial communities (defined as those consisting specifically of P. histicola and S. parasanguinis) showed higher acetaldehyde production than those with the type-II oral microbial communities (defined as those consisting specifically of N. flavescens and G. microbiomes of the HG were significantly higher than those in the microbiomes of the LG. A previous study reported that N. flavescens exhibited higher acetaldehyde production capacity than other bacteria that can generate acetaldehyde. 9 Neisseria species are associated with oral lichen planus, which is known to be a precancerous condition. 37 N. flavescens on the tongue may contribute to the incidence of oral cancer, through the higher production of acetaldehyde. The participants in this study may be representative of the general population. First, the median bacterial count was 1.15×10 7 CFUs/mL. This value was within the range (9.5×10 6 -2.9×10 7 CFUs/mL) defined (using the same device) in a previous study of Japanese perioperative patients. 38

Conclusions
This study revealed that acetaldehyde concentration in mouth air was positively associated with bacterial count; diversity of the microbiome; and the relative abundance of G. sanguinis, V. parvula, and N. flavescens, species that are known to be capable of producing acetaldehyde.