Identification of methylglyoxal as a major mutagen in wood and bamboo pyroligneous acids

Abstract

To identify the major mutagen in pyroligneous acid (PA), 10 wood and 10 bamboo pyroligneous acids were examined using the Ames test in Salmonella typhimurium strains TA100 and TA98. Subsequently, the mutagenic dicarbonyl compounds (DCs), glyoxal, methylglyoxal (MG), and diacetyl in PA were quantified using high-performance liquid chromatography, and the mutagenic contribution ratios for each DC were calculated relative to the mutagenicity of PA. Eighteen samples were positive for mutagens and showed the strongest mutagenicity in TA100 in the absence of S9 mix. MG had the highest mutagenic contribution ratio, and its presence was strongly correlated with the specific mutagenicity of PA. These data indicate that MG is the major mutagen in PA.

Graphical abstract

Pyroligneous acids showed the mutagenicity by the Ames test, and methylglyoxal was identified as a major mutagen.

Key words:

• pyroligneous acid
• mutagen
• Ames test
• dicarbonyl compound
• methylglyoxal

Pyroligneous acid (PA) is obtained by cooling the smoke that is generated during the carbonization of wood and is a predominant by-product of charcoal production in Japan. It is known that PA has bactericidal,¹⁾ antiviral,²⁾ antifungal,³⁾ and odor masking⁴⁾ activities because it contains multiple compounds, such as organic acids, phenols, and carbonyl compounds.^5,6) Accordingly, PA is sold at gardening stores as an agricultural antiseptic and soil improvement agent and is sold at drugstores as a bath agent. In addition, PA is used as a food additive to give a smoked flavor to foodstuffs.^7,8) However, several studies show that PA was mutagenic in the Ames tests using Salmonella typhimurium strains TA100 and TA98.⁹⁾ Furthermore, these studies have produced diverse results, and the primary mutagen in PA has not been sufficiently identified.^10–13) Nonetheless, because wood smoke has been shown to be mutagenic,¹⁴⁾ it can be assumed that the major mutagen in PA is formed during carbonization.

Similar to PA, roasted coffee beans are produced by a high-temperature heating process and contain mainly carbohydrate as raw materials. It had been reported that the mutagens of coffee beans were produced during the roasting process.¹⁵⁾ Roasted coffee beans showed mutagenicity in the TA100 in the absence of S9 mix,¹⁶⁾ and methylglyoxal (MG) was shown to be the primary mutagen.¹⁷⁾ MG is one of the several dicarbonyl compounds (DCs) that are formed during decomposition of carbohydrates.^18,19) DCs have demonstrated mutagenicity in TA100²⁰⁾ and are likely mutagenic components of PA because the D-glucose polymer cellulose is a major constituent of wood.

Most studies of PA mutagenicity in S. typhimurium have been conducted on smoke condensates from wood. In Japan, both wood pyroligneous acid (WPA) and bamboo pyroligneous acid (BPA) are commercially available and have been sold in equal proportions. In this study, the mutagenicity of 10 WPAs and 10 BPAs were examined using the Ames test in TA100 or TA98, and furthermore, MG, glyoxal (G), and diacetyl (DA) in WPAs and BPAs were quantified using high-performance liquid chromatography (HPLC). The involvement of each DC in the mutagenicity of PA was then investigated, and the major PA mutagen was determined.

Materials and methods

PA samples

WPAs and BPAs were purchased as liquids from gardening or drug stores, and WPA and BPA samples were denoted W-1–W-10 and B-1–B-10, respectively.

Reagents

G and MG were purchased from Sigma-Aldrich Corp. (St. Louis, MO, USA). DA, quinoxaline, 2-methyl-quinoxaline, 2,3-dimethyl-quinoxaline, and 1,2-phenylenediamine were purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). HPLC grade methanol and formic acid were purchased from Kanto Chemical Co., Inc. (Tokyo, Japan). S9 was purchased from Kikkoman Biochemifa Company (Tokyo, Japan).

Instrument

The C18 cartridge Sep-Pak Plus was purchased from Waters Corporation (Milford, MA, USA), and the automatic colony counter MCL-2000 was purchased from SK-Electronics CO. Ltd. (Kyoto, Japan). The HPLC system (Shimadzu Corporation, Kyoto, Japan) was equipped with a system controller (CBM-20A), a pump (LC-20AD), a diode array detector (SPD-M20A), an auto sampler (SIL-20AC), a column oven (CTO-20AC), and an Inertsil ODS-4 column (5 μm particle size, 4.6 mm i.d. × 250 mm, GL Sciences Inc., Tokyo, Japan).

Mutagenicity test

The Ames test was performed according to a previously described preincubation method²¹⁾ using S. typhimurium strains TA100 and TA98 in the absence and presence of S9 mix. Both strains were kindly provided by Dr Bruce N. Ames, University of California, Berkeley, California. The S9 mix contained 50 μL of S9, which was prepared from the livers of male Sprague Dawley rats by treatment with phenobarbital and 5,6-benzoflavone in a total volume of 500 μL. Revertants were counted using a colony counter after incubation for 48 h at 37 °C, and mutagenicity tests were performed in more than 2 plates. Distilled water was used as a negative control, and 2-(2-furyl)-3-(5-nitro-2-furyl)acrylamide or 2-aminoanthracene was used as positive controls. PA mutagenicity was recorded as positive when PA samples induced more than twice the number of negative control revertants. Specific mutagenicity values of PA represented net revertants per 1 μL of sample and were calculated using dose–revertant response curves.

Determination of DCs in PA

G, MG, and DA in PA were derivatized to quinoxaline, 2-methylquinoxaline, and 2.3-dimethylquinoxaline, respectively, using 1,2-phenylenediamine in a modified quinoxaline derivatization method.²²⁾ Briefly, PA was diluted 10-fold with 0.2 M phosphate buffer solution (pH 8.0) and 2 mL of diluted PA was added to 2 mL of 0.3 M 1,2-phenylenediamine solution containing 0.2 M phosphate buffer (pH 8.0) and was allowed to react for 20 min at 37 °C. After cooling to room temperature, distilled water was added to a final volume of 10 mL. Subsequently, 5 mL aliquots were loaded onto a Sep-Pak Plus C18 cartridge that had been conditioned with 5 mL of methanol and 5 mL of distilled water and then washed with 5 ml of distilled water. Subsequently, the cartridge was eluted with 10 mL of 50% (v/v) methanol, and eluates were analyzed using HPLC with a mobile phase of 50%(v/v) methanol containing 0.1% (v/v) formic acid, a flow rate of 1.0 mL/min, injection volume of 10 μL, column temperature of 40 °C, and detection wavelength of 315 nm. A recovery test was performed 3 times using B-1, and G, MG, and DA were added at 4.0, 2.4, and 0.4 μmol/mL, respectively.

Calculation of mutation incidence

The effect on the mutation expression by the cytotoxicity of PA was examined using the Ames tests. Tests of mutagenicity were simultaneously performed in PA with (A) and without (B) added 0.3 μmol MG for each dose using TA100 in the absence of S9 mix. PA was prepared at doses of 20, 40, 60, and 80 μL, and 0.3 μmol MG was added to the same doses in parallel. As a control, 0.3 μmol MG (C) was used, and all sample solutions were adjusted to 100 μL. The recovery rate of revertants by added MG at each dose was calculated using the following equation and was defined as the mutation incidence of PA:

A regression line of mutation incidence of PAs was obtained using the values calculated in each dose from the above equation.

Calculation of mutagenic contribution ratios of DCs in PA

Mutagenic contribution ratios were expressed as percentages of mutagenicity from individual DCs relative to the total mutagenicity of PA and were calculated for the PA dose that showed twice the number of negative control revertants. The number of the net revertants induced by DC contents of PA (D) and the number of the net revertants of PA (E_c) were calculated using dose-response curves from Ames tests. E_c values were calculated by correcting the number of the net revertants of PA (E_m) with mutation incidence rate (F), and E_m values were the actual measured number of the net revertants of PA. Mutagenic contribution ratios were calculated under TA100 in the absence of S9 mix as follows:

Statistical analysis

Differences were identified using Student’s t-test and were considered significant when p < 0.05.

Results and discussion

Mutagenicity of PA

In the Ames tests of PA in TA100 and TA98 (Tables 1 and 2), 18 (9 WPA and 9 BPA) of 20 samples had positive mutagenicity. In TA100, 17 samples (8 WPA and 9 BPA) were positive in the absence of S9 mix, and 6 samples (2 WPA and 4 BPA) were positive in the presence of S9 mix. In TA98, 10 samples (4 WPA and 6 BPA) were positive in the absence of S9 mix, and only B-8 was positive in the presence of S9 mix. It was found that PA samples showed the strongest mutagenicity in TA100 in the absence of S9 mix because the values of specific mutagenicity in TA100 in the absence of S9 mix were the highest in the Ames test under various conditions. In particular, W-2 and B-10 were highly mutagenic in TA100 in the absence of S9 mix and led to the generations of more than 1000 revertants. In agreement with these observations, previous studies showed stronger mutagenicity of PA in TA100 than in TA98 but showed differing effects of metabolic activation.^10–12) In this study, the mutagenicity of PA was strongest in TA100 in the absence of S9 mix, although mutagenicity was reduced by metabolic activation in both TA100 and TA98. Similarly, Nakama et al. and Kim Oanh et al. reported that the mutagenicity of smoke generated by burning wood was strongest in TA100 in the absence of S9 mix.^11,14)

Table 1. Mutagenicity tests of pyroligneous acids in Salmonella typhimurium TA100.

Sample	–S9 mix								+S9 mix
	Revertants							Specific mutagenicity (revertants/μL)	Revertants							Specific mutagenicity (revertants/μL)
	Negative control^a	AF-2^b (0.01 μg)	Dose (μL)						Negative control	2AA^c (1.0 μg)	Dose (μL)
			5	10	25	50	100				5	10	25	50	100
W-1	111	411	154	148	275	410	94*	6.0	126	805	152	150	191	281	365*	3.0
W-2	111	411	169	207	597	1522	3426	31.4	126	805	141	149	149	150	209	－
W-3	111	411	126	117	151	222	378	2.5	126	805	129	126	144	146	228	－
W-4	124	605	121	131	175	215	88*	(1.8)^d	123	1196	126	147	156	182	266	1.4
W-5	124	605	132	143	157	207	292	1.7	123	1196	118	131	170	168	166	－
W-6	124	605	122	122	118	131	144	－^e	123	1196	125	125	130	148	150	－
W-7	124	605	128	139	167	218	399	2.5	123	1196	120	124	154	169	175	－
W-8	124	605	120	135	162	169	250	1.2	123	1196	121	130	132	157	163	－
W-9	124	605	136	145	170	287	100*	2.9	123	1196	133	168	166	170	201*	－
W-10	128	734	123	142	219	374	956	7.4	124	984	104	106	125	121	131	－
B-1	111	411	182	227	410	75*	0*	12.0	126	805	142	161	242	260*	90*	3.1
B-2	124	605	121	131	138	143	150	－	123	1196	117	111	108	121	112	－
B-3	111	605	107	106	107	152	325	1.8	126	805	131	133	142	154	200	－
B-4	124	605	126	143	157	178	254	1.3	123	1196	119	120	164	150	201	－
B-5	124	605	137	164	211	269	410	2.9	123	1196	135	160	158	191	251	1.3
B-6	124	605	141	176	221	286	100*	3.4	123	1196	129	140	151	210	170*	－
B-7	124	605	129	153	171	219	409	2.6	123	1196	123	153	161	188	281	1.5
B-8	111	411	109	141	195	466	724	6.1	126	805	140	149	171	216	391	2.4
B-9	124	605	127	147	168	305	84*	3.2	123	1196	119	140	180	198	196*	－
B-10	111	411	132	140	180	375	1031	8.1	126	805	124	121	113	140	169	－

Notes: Numbers of revertants for samples, AF-2 and 2AA are presented as the average of 2 plates, and the number of the negative control revertants is presented as the average of 5 plates. Specific mutagenicity values represent the number of the net revertants per 1 μL of sample, and were calculated from dose–revertant response curves.

^a Distilled water;

^b 2-(2-Furyl)-3-(5-nitro-2-furyl)acrylamide;

^c 2-Aminoanthracene;

^d Mutagenicity was pseudo-positive;

^e Mutagenicity was negative;

* Growth inhibition was observed.

Table 2. Mutagenicity tests of pyroligneous acids in Salmonella typhimurium TA98.

Sample	–S9 mix								+S9 mix
	Revertants							Specific mutagenicity (revertants/μL)	Revertants							Specific mutagenicity (revertants/μL)
	Negative control^a	AF-2^b	Dose (μL)						Negative control	2AA^c	Dose (μL)
		(0.1 μg)	5	10	25	50	100			(0.5 μg)	5	10	25	50	100
W-1	19	520	21	29	34	47	16*	0.6	26	227	31	31	26	35	35*	−
W-2	19	520	19	24	39	82	37	1.1	26	227	27	30	31	38	31	−
W-3	19	520	24	19	27	44	48	0.3	26	227	29	32	32	40	51	−
W-4	25	501	27	32	29	42	25*	−^d	30	327	32	31	42	46	43*	−
W-5	25	501	26	29	33	35	46	−	30	327	26	33	37	33	36	−
W-6	25	501	26	29	26	33	33	−	30	327	29	31	32	35	37	−
W-7	25	501	30	37	45	53	51	0.6	30	327	32	41	41	40	51	−
W-8	25	501	26	37	38	40	45	−	30	327	31	35	36	38	35	−
W-9	25	501	26	37	36	32	21*	−	30	327	29	49	34	35	31*	−
W-10	25	453	29	22	22	31	45	−	23	245	25	28	26	25	29	−
B-1	19	520	23	33	30	7*	0*	−	26	227	35	32	34	24*	8*	−
B-2	25	501	24	22	22	27	25	−	30	327	26	39	32	34	32	−
B-3	19	520	21	21	28	36	61	0.4	26	227	32	27	33	33	41	−
B-4	25	501	25	32	37	37	47	−	30	327	27	40	33	49	42	−
B-5	25	501	25	38	41	44	53	0.3	30	327	34	38	37	37	48	−
B-6	25	501	29	37	38	53	26*	0.6	30	327	33	42	45	41	28*	−
B-7	25	501	24	29	34	41	54	0.3	30	327	31	34	38	41	40	−
B-8	19	520	24	38	43	102	89*	1.5	26	227	28	30	40	52	60	0.4
B-9	25	501	27	36	33	47	25*	−	30	327	33	39	36	39	20*	−
B-10	19	520	24	22	24	37	58	0.4	26	227	25	30	23	25	37	−

Notes: Numbers of revertants for samples, AF-2 and 2AA are presented as the average of 2 plates, and the number of the negative control revertants is presented as the average of 5 plates. Specific mutagenicity values represent the number of the net revertants per 1 μL of sample, and were calculated from dose–revertant response curves.

^a Distilled water;

^b 2-(2-Furyl)-3-(5-nitro-2-furyl)acrylamide;

^c 2-Aminoanthracene;

^d Mutagenicity was negative;

* Growth inhibition was observed.

Determination of DCs in PA

DC contents of PA (Table 3) were determined with a detection limit of 0.03 μmol/mL and a quantitation limit of 0.1 μmol/mL. Mean recovery rates ± coefficients of variation of G, MG, and DA were 98.5 ± 1.4, 90.1 ± 2.1, and 96.1 ± 1.8%, respectively. HPLC chromatograms of W-1 and B-1 were shown in Fig. 1. The highest DC contents were observed in sample B-1 and were present at a total of 0.83 mg/mL. DC contents and composition ratios of the 3 DCs, G, MG, and DA in PA were quite different between samples. Some among the 20 samples, e.g., a pair of W-1 and B-1, W-9 and B-9, and W-10 and B-10, are produced by the same maker. With regard to W-1 and B-1, although the composition ratios of DCs were different, the total DC content was the highest value in the group of WPAs or BPAs. With regard to W-10 and B-10, DC contents were lower values or less than the quantification limit, although these samples showed relatively strong mutagenicity and the equivalent specific mutagenicity values in TA100 in the absence of S9 mix. With regard to W-9 and B-9, DC contents were comparatively high, and the composition ratios of DCs were similar. It was suggested that DC contents and the composition ratios of DCs were similar in WPA and BPA of the same maker. Therefore, differing DC contents and the composition ratios of DCs in the present 20 PA samples may be caused by different production heating temperature and time for each maker. In agreement, previous studies showed that the composition ratios of DCs in coffee beans and cookies fluctuated with changes in heating times,^23,24) and MG production during heating varied with temperatures and heating times.¹⁸⁾ Therefore, differing contents and composition ratios of DCs in the present PA samples likely reflect differences in heating temperatures and times during manufacture. However, no significant differences in DC contents were observed between WPA and BPA samples by statistical analysis. DCs are the decomposition products of carbohydrate combustion and are the major components of common WPAs and BPAs. Previous reports also show no large differences in composition of most volatile and odor components between WPA and BPA,^5,25) suggesting similar compositions of these PAs.

Table 3. Contents of dicarbonyl compounds in pyroligneous acids.

	Content (μmol/mL)			Total (mg/mL)
Sample	G^a	MG^b	DA^c
W-1	5.2	3.7	0.4	0.60
W-2	0.2	0.5	tr^d	0.05
W-3	0.9	1.0	0.1	0.13
W-4	0.1	1.3	0.3	0.13
W-5	0.6	0.6	tr	0.07
W-6	ND^e	ND	ND	−^f
W-7	tr	0.4	ND	0.03
W-8	0.2	0.5	tr	0.05
W-9	0.5	2.4	0.1	0.21
W-10	0.3	0.2	ND	0.03
B-1	5.4	6.5	0.6	0.83
B-2	tr	ND	tr	−
B-3	0.7	0.6	tr	0.08
B-4	0.7	0.5	tr	0.08
B-5	2.3	1.5	0.1	0.25
B-6	0.9	1.0	ND	0.12
B-7	0.4	0.7	0.1	0.08
B-8	1.1	1.2	0.3	0.18
B-9	0.6	2.6	0.2	0.24
B-10	tr	tr	ND	−

Note: Values are the average of 2 times measurement.

^a Glyoxal;

^b Methylglyoxal;

^c Diacetyl;

^d 0.03 μmol/mL ≤ tr < 0.1 μmol/mL;

^e ND < 0.03 μmol/mL;

^f Not calculated.

Fig. 1. HPLC chromatograms of W-1 and B-1 after derivatization.

Notes: Glyoxal, methylglyoxal, and diacetyl in W-1 and B-1 were quantified as quinoxaline (Q), 2-methylquinoxaline (2-MQ), and 2.3-dimethylquinoxaline (2.3-DMQ), respectively, by HPLC after quinoxaline derivatization.

Mutagenic contribution ratios of DCs from PA in TA100 in the absence of S9 mix

Most PA samples contained DCs and were mutagenic. Thus, in TA100 in the absence of S9 mix, the mutagenic contribution ratios of DCs were calculated as percentages of total PA mutagenicity. We calculated the mutagenic contribution ratios of DCs not only positive PA samples but also W-4, because W-4 was regarded as pseudo-positive, showing 1.5–2 times the number of negative control revertants. These analyses showed the mutagenic contribution ratios of MG in W-1, W-9, B-1, and B-9 exceed 100% (data are not shown), and these PA samples showed clear cytotoxicity in Ames tests (Table 1). The presence of bactericidal components in PA has been previously reported,^1,5–7) and the cytotoxicity of these has been shown in S. typhimurium.^10–12) The present MG mutagenic contribution ratios exceeding 100% may reflect the underestimation of the mutagenicity of PA due to PA cytotoxicity. Therefore, to examine the effects on mutation expression by various components including toxicity to test bacteria, we performed the Ames test of PAs with and without added MG. Specifically, the recovery rate of the number of net revertants derived from MG that added to PA was calculated, and the recovery rates of MG were regarded as the mutation incidence. In a previous report, MG was identified as the most mutagenic of the 3 DCs in TA100 in the absence of S9 mix (about 18 times the mutagenicity of G).²⁶⁾ Following similar observations (data are not shown), we used MG for recovery tests and confirmed that 0.3 μmol MG was optimum mutagenic dose but not cytotoxic. Under these conditions, the mutation incidence rates decreased with increasing PA doses, except in the case of the W-5 (Fig. 2). The mutation incidence rates of W-1, W-4, W-9, B-1, B-6, and B-9 particularly decrease compared with those of other samples, and these PA samples showed marked growth inhibition (Fig. 2, Table 1). Therefore, it was found that the mutation expression was decreased with increasing PA doses because components toxic to test bacteria increase with increasing PA doses. In addition, it was suggested that toxic components and amounts were different between samples because the mutation incidence showed a significantly different decreasing trend between samples. Therefore, to calculate the mutagenic contribution ratios, we determined to correct the number of the net revertants of PA by the mutation incidence rates with using a method for adjusting the decrease in the mutation expression level with the values of the mutation incidence rates. For the consistency of mutagenic contribution ratios of DCs, calculations were performed at doses that corresponded with 2-fold increases in the number of the revertants compared with the negative control, according to the threshold of positive PA mutagenicity. Mutagenic contribution ratios ranged 0.0–8.8% for G and 0.0–93.0% for MG, and they ranged 1.7–101.8% among all DCs (Table 4). With regard to DA, mutagenic contribution ratio was not calculated because DA showed negative mutagenicity due to the cytotoxicity in this study. DCs contributed more than 25% of the mutagenicity of 12 PA samples (60% of all samples), and MG accounted for the highest mutagenic contribution ratio in all PA samples. In particular, W-1, W-9, B-1, and B-9 contained MG at concentrations of more than 2.4 μmol/mL, and MG contributed more than 70% of the mutagenicity. These data suggest that MG is the major mutagen in PA. However, despite the strong mutagenicity of W-2, W-10, and B-10, MG contents were low and showed mutagenic contribution ratios of less than 1.7%. Correlations between specific mutagenicity and MG contents among all 20 samples and after the exclusion of W-2, W-10, and B-10 showed coefficient (r) values of 0.17 and 0.90, respectively. Consequently, with the exception of W-2, W-10, and B-10, MG was a major mutagen in TA100 in the absence of S9 mix. On the other hand, the mutagenic activities of W-2, W-10, and B-10 indicate the presence of other mutagens, and these are currently under investigation.

Fig. 2. Mutation incidence rates in methylglyoxal recovery tests in Salmonella typhimurium TA100 in the absence of S9 mix.

Notes: Methylglyoxal (MG) recovery tests were conducted for all mutagenic pyroligneous acids (PA) using the Ames test. PAs were prepared at doses of 20, 40, 60, and 80 µL, and simultaneous Ames tests were performed in PA with and without addition of 0.3 µmol MG; 0.3 µmol MG was used as a reference material. Recovery rates were calculated as percentages after subtracting the number of the net revertants of PA from the number of the net revertants of PA added 0.3 µmol MG and dividing by the number of the net revertants of 0.3 µmol MG. Hence, the recovery rates of MG represent mutation incidence rates of PAs.

Table 4. Mutagenic contribution ratios of dicarbonyl compounds to the mutagenicity of pyroligneous acids in Salmonella typhimurium TA100 in the absence of S9 mix.

	Mutagenic contribution ratio (%)			Total (%)
Sample	G^a	MG^b	DA^c
W-1	8.8	93.0	−^d	101.8
W-2	0.0	1.7	−	1.7
W-3	2.0	52.6	−	54.6
W-4	0.1	31.6	−	31.7
W-5	3.3	54.0	−	57.3
W-6	−	−	−	−
W-7	0.0	16.3	−	16.3
W-8	0.2	17.3	−	17.5
W-9	0.4	70.4	−	70.8
W-10	0.1	1.6	−	1.7
B-1	3.1	78.4	−	81.5
B-2	−	−	−	−
B-3	2.0	32.4	−	34.4
B-4	1.6	24.4	−	26.0
B-5	6.1	61.5	−	67.6
B-6	0.7	17.4	−	18.1
B-7	0.7	28.8	−	29.5
B-8	1.1	28.9	−	30.0
B-9	0.4	84.0	−	84.4
B-10	0.0	0.0	−	0.0

^a Glyoxal;

^b Methylglyoxal;

^c Diacetyl;

^d Not calculated.

In TA100 in the presence of S9 mix, MG showed the strongest mutagenicity among 3DCs, and showed the mutagenicity in a dose-dependent manner at higher level than 0.75 μmol by dose-response curve (data are not shown). B-1 contains the highest concentration of MG between samples and showed the mutagenicity positive; however, the amount of MG was in low level to induce the mutation. Asita et al.¹⁰⁾ reported that in the TA100 or TA98 in the presence of S9 mix, the contribution of the polycyclic aromatic hydrocarbons (PAHs) to the mutagenicity of the wood smoke condensates could not be determined because of the concentrations of the PAHs was much lower level for inducing the mutation. These results suggested that many more mutagens contained in PA might contribute to the mutagenicity of PA in the presence of S9 mix.

In this study, we showed that MG is the major mutagen of WPAs and BPAs in TA100 in the absence of S9 mix. However, although PA has been used in various applications such as a bath agent and food additive whereby it is in direct contact with the human body, it appeared that PA has a serious safety issue because of the mutagenicity. Therefore, it is important to explore and evaluate the mutagenic substance in PA to ensure the safety of PA. The mutagens of PA are produced in the manufacturing process; so when the production conditions of PA are examined in a way the generation of MG is controlled or the formed MG is removed, PA might become safe. In this study, no significant differences in DCs contents were observed between WPA and BPA samples; however, it appears that there is great variability among quality of PA because of differences in the materials and production method based on the maker. At the present day, with respect to the components of PA, standard values of PAHs including benzo[a]pyrene have been established by JECFA (Joint FAO/WHO Expert Committee on Food Additives). However, the risk of these mutagenic substances has not been managed in Japan. Therefore, the setting of standard values of harmful substances and mutagenic substances such as MG in PA is desired in future. The present data will contribute to further studies of the risks associated with multiple harmful compounds that are present in consumables and the environment.

Author Contribution

A.O., M.A. and H.N. conceived and designed the experiments, the results, and wrote the manuscript. A.O. and M.A. performed the experiments and analyzed the data.

Disclosure statement

No potential conflict of interest was reported by the authors.