See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/23556244

Chemical fate and changes in mutagenic activity of antibiotics nitrofurazone

and furazolidone during aqueous chlorination

Article in The Journal of Toxicological Sciences · January 2009

DOI: 10.2131/jts.33.621 · Source: PubMed

CITATIONS READS

10 172

5 authors, including:

Tsuyoshi Kawakami
National Institute of Health Sciences
81 PUBLICATIONS 450 CITATIONS

SEE PROFILE

All content following this page was uploaded by Tsuyoshi Kawakami on 21 December 2015.

The user has requested enhancement of the downloaded file.

 The Journal of Toxicological Sciences (J. Toxicol. Sci.) 621
Vol.33, No.5, 621-629, 2008

Original Article

Chemical fate and changes in mutagenic activity of

antibiotics nitrofurazone and furazolidone during aqueous
chlorination
Hitomi Nakamura1, Tsuyoshi Kawakami1,3, Tatsuhiro Niino2, Yasuo Takahashi1
and Sukeo Onodera1

Faculty of Pharmaceutical Sciences, Tokyo University of Science, 2641 Yamazaki, Noda, Chiba 278-8510, Japan
1

2Mitsubishi Chemical Safety Institute Ltd., 1000 Kamoshida-cho, Aoba-ku, Yokohama 227-0033, Japan

3Present address: Division of Medical Devices, National Institute of Health Sciences, 1-18-1 Kamiyoga, Setagaya-ku,

Tokyo 158-8501, Japan

(Received August 26, 2008; Accepted October 14, 2008)

ABSTRACT — Reactions of nitrofuran antibiotics (nitrofurazone (NFZ) and frazolidone (FZD)) with
hypochlorite in aqueous solution were investigated under the conditions that simulate wastewater disin-
fection. The chlorination byproducts were determined by high performance liquid chromatography. At
WKHOHYHOVRIȝ01)=UHDFWHGUDSLGO\ZLWKIUHHFKORULQHLQQHXWUDOS+  ZKLOHWKH)='K\SRFKOR-
rite reaction was reasonably slow under the same pH. Nevertheless, the strong mutagenic parents dis-
appeared completely after the hypochlorite reactions, and the chlorination byproducts were observed to
exert a weak mutagenic effect on Salmonella typhimurium TA100 without S9-mix. The extent of the reac-
tions depended on the chlorine dose, solution pH and compound structures.

Key words:1LWURIXUD]RQH)XUD]ROLGRQH:DWHUFKORULQDWLRQ0XWDJHQLFLW\

INTRODUCTION ed to be found in aquatic environments (Samuelsen et

The focus of environmental research has been extended
from traditional pollutants such as polychlorinated biphe-
nyls, polycyclic aromatic hydrocarbons, and pesticides, to
emerging pollutants such as pharmaceuticals and person-
al care products (PPCPs), some of which may be carci-
nogenic, mutagenic, or reproductively toxic (Tixier et al.,
2003; Ashton et al., 2004; Richardson and Ternes, 2005).
Pharmaceutical substances are used in human and vet-
erinary medicine and can enter the aquatic environment
following their manufacture, use, or ingestion/excretion.
The rapid increase in use of pharmaceutical products is
a new environmental problem and has received attracted
worldwide (Lin et al., 2005; Petrovic et al., 2005; Van De
Steene et al., 2006).
Recent studies have illustrated the omnipresence of
QXPHURXVDQWLEDFWHULDOVLQPXQLFLSDOZDVWHZDWHUHIÀXHQWV
and affected surface waters (Kolpin et al., 2001; Kim and
Carlson, 2007). Based on the production and use, 1 class
of compounds, i.e., the nitrofuran antibiotics and their
structurally related transformation products are expect-
al., 1991; Lunestad et al., 1995; Pothuluri et al., 1998;
Hirsch et al +RZHYHUQRQHRIWKHPDMRU¿HOG
studies conducted thus far have included the parent com-
pounds in their analysis suite. It has been estimated that
less than 15% of the PPCPs predicted to be found in the
environment are actually sought analytically (Ternes et
al., 2004).
Structure of nitrofuran antibiotics are shown in Table 1.
These antibiotics are used in the treatment of genitouri-
nary, gastrointestinal, and dermatological infections in
both humans and animals (Paul and Paul, 1964). One
nitrofuran antibiotic, furazolidone (FZD), is used in the
WUHDWPHQWRIEDFWHULDOGLVHDVHVLQ¿VKIDUPV 6DPXHOVHQ
et al., 1991; Lunestad et al., 1995; Pothuluri et al., 1998).
Nitrofurazone (NFZ) is used in the treatment of bacteri-
DOGLVHDVHVLQDODUJHQXPEHURI¿VKHV)XUWKHUPRUHLQ
humans, nitrofurantoin is used against a wide variety of
gram-positive and gram-negative bacteria; it is prima-
rily used as an oral treatment for urinary tract infections
(Guay, 2001).
Use of nitrofurans has already been banned in the

Correspondence: Sukeo Onodera (E-mail: onodera@rs.noda.tus.ac.jp)

Vol. 33 No. 5
622

H. Nakamura et al.

Table 1. Nitrofuran antibiotics and related compounds used in aqueous chlorination studies.
Chemicals CAS No. Formulae 0ROHFXODU:HLJKW Structure
H
NO2 O N NH
Nitrofurazone (NFZ) 59-87-0 C6H6N4O4 198.14 N
O

NO2 O N O
Furazolidone (FZD) 67-45-8 C8 H 7 N 3 O 5 225.16 N
O
O
5-Nitro-2-furaldehyde (NFA) 698-63-5 C5H3NO4 141.09 NO2 O CH

O
NO2 O C OH
5-Nitro-2-furoic acid (NFOA) 645-12-5 C5H3NO5 157.08

NH2
3-Amino-2-oxazolidinone (AOZ) 80-65-9 C3 H 6 N 2 O 2 122.55 N O

O
6HPLFDUED]LGH 6(0 563-41-7 CH5N3O 75.08 H2N
N NH2
H

European Union (EU) since 1993 due to doubts regard-
ing the safety of the protein-bound residues of these drugs
in edible products (Tatsumi et al0F&DOOD
Van Koten-Vermeulen et al., 1993). FZD used in food
production has also been completely prohibited world-
wide since 1995. Nevertheless, these antibiotics are still
used in some countries. In fact, NFZ is currently used for
the treatment of bacterial diseases in a large number of
¿VKLQ-DSDQ7KHVHFRPSRXQGVDUHWKXVLQWURGXFHGLQWR
the water systems. Therefore, chlorination is extensive-
ly used in wastewater treatment in order to disinfect and
GHRGRUL]HHIÀXHQWVSULRUWRGLVFKDUJHSDUWLFXODUO\ZKHQ
the water may subsequently be used for recreational pur-
poses or as a source of potable water.
In order to provide further insight into the possible role
of organic compounds in the formation of chlorine-sub-
stituted compounds and that of chlorine-induced muta-
gens, we continued our study of the aqueous chlorination
chemistry of organic compounds (Onodera et al., 2008;
Tomiyama et al., 2008). The objectives of the present
study were (1) to quantify the reaction kinetics of NFZ
(and FZD) with free available chlorine at various pH
values, (2) to identify probable transformation prod-
ucts under conditions relevant to chlorine-based munici-
pal wastewater disinfection processes, and (3) to evaluate
mutagenic activities of both the reaction products and the
parent compounds.
MATERIALS AND METHODS
Chemicals
NFZ and FZD of analytical reagent grade (Kanto
Chemicals, Tokyo, Japan) were used for the food residue
analysis. 5-Nitrofuraldehyde (NFA), 5-nitro-2-furoic acid
(NFOA), 3-amino-2-oxazolidinone (AOZ), and semicar-
ED]LGH 6(0 K\GURFKORULGHZKLFKDUHH[SHFWHGWREH
formed during the chlorination of the parent compounds
with free available chlorine, were obtained from Tokyo
Chemical Industry Co., Ltd. (Tokyo, Japan), except for
AOZ which was obtained from Sigma-Aldrich (St. Lou-
LV0286$ 6WDQGDUGVROXWLRQVRIWKHVHFRPSRXQGV
as well as their mixtures were prepared by dissolving
the compounds in methanol and diluting subsequently.
Organic solvents (methanol, diethyl ether and ethyl ace-

Vol. 33 No. 5

0XWDJHQLFDFWLYLW\RIQLWURIXUDQVGXULQJDTXHRXVFKORULQDWLRQ
tate) of analytical reagent grade (Wako Pure Chemical
Industry Co., Ltd, Osaka, Japan) were used for the pesti-
cide residue analysis. Hypochlorite solution was prepared
by diluting sodium hypochlorite solution (ca. 5% availa-
ble chlorine; Kanto Chemicals) in phosphate buffer solu-
tion. The hypochlorite concentrations were determined by
iodometric titration.
Aqueous chlorination of nitrofuran antibiotics
A mixture of 100 ml of hypochlorite solution and each
nitrofuran antibiotic dissolved in 1 ml of methanol was
shaken in a brown separatory funnel to avoid its pho-
to-degradations. After the desired reaction time, residu-
al chlorine was removed by the addition of an equivalent
YROXPHRIVRGLXPVXO¿WHVROXWLRQ7KHVHZDWHUVDPSOHV
ȝO ZHUHGLUHFWO\DQDO\]HGE\KLJKSHUIRUPDQFHOLT-
uid chromatography (HPLC). The reaction mixture was
WKHQDFLGLILHGWRS+ZLWK0K\GURFKORULFDFLG
and 20 g of sodium chloride was added before extract-
ing 3 volumes (20 ml) of diethyl ether/ethyl acetate (3:7).
The solvents were dried over anhydrous sodium sulfate,
and 1 ml of methanol was added to prevent the evapora-
tion of the reaction products during concentration under
vacuum at 40°C for obtaining suitable volumes for muta-
genicity tests.
Analytical methods
The reaction products of each nitrofuran antibiotic with
hypochlorite in aqueous solution were analyzed using a
Shimadzu LC-6AD liquid chromatograph (Shimadzu,
Kyoto, Japan) with UV detection (Shimadzu SPD-20AV).
7KH2'60LJKW\VLO 53 DQDO\WLFDOFROXPQ FP
[PP,' ZDVSDFNHGZLWKȝPSDUWLFOHVL]HSDFN-
LQJPDWHULDOIRU+3/&DQGWKH2'60LJKW\VLOJXDUG
FROXPQVZHUHSDFNHGZLWKZLWKȝPSDUWLFOHVL]HSHO-
licular material (Kanto Chemicals). The conditions were
as follows: column temperature, 40°C; mobile phase,
PHWKDQROP0SKRVSKDWHEXIIHU YY ÀRZ
rate, 1 ml/min; and detection wavelengths, 265 and 365
nm. A Shimadzu Chromatopac C-R6A data processor was
used to determine the retention times and the peak areas
on the chromatograms. The residual amounts of the nitro-
furan antibiotics and their degradation products were esti-
mated from each peak area, compared with their calibra-
tion curves.
Individual nitrofuran antibiotics and their reaction
products were identified by comparing of their reten-
tion times with those of the authentic standards. In addi-
tion, absorption spectra of buffered solutions containing
10 mg/l of the antibiotic reagents were examined before
and after chlorination with hypochlorite at wavelengths
623
between 190 and 500 nm by using a Shimadzu UV mini
1240 spectrophotometer (Shimadzu, Kyoto, Japan).
Mutagenicity tests
The mutagenicity of the samples was tested accord-
LQJWRWKHPHWKRGRI0DURQDQG$PHV  ZLWKPLQRU
PRGL¿FDWLRQV%HFDXVHPRVWRIWKHPXWDJHQVIRUPHGLQ
the reactions of the organic compounds with chlorine in
water have been shown to be positive in Salmonella typh-
imurium strain TA100 without S9 mix (Onodera et al.,
1998), this strain was employed throughout the experi-
ments. The samples were dissolved in dimethylsulfoxide
'062 DQGSUHLQFXEDWHGZLWKDWHVWVWUDLQDWƒ&IRU
30 min with a phosphate buffer. After addition of the test
samples, the plates were incubated at 37°C for 2 days.
The assay was performed in triplicate for each sample and
the mutagenic activity expressed as the mean of the rever-
tants. The mutagenic activity of the antibiotics, their deg-
radation derivatives, and the chlorination products were
estimated by least-squares regression analysis of the ini-
tial slopes of the dose-response curves.
RESULTS AND DISCUSSION
Kinetics of the hypochlorite reactions
In a preliminary examination of the reactions of the
nitrofuran antibiotics (NFD and FZD) with hypochlorite
in a dilute aqueous solution at pH 7.0, the decrease in the
concentration of free available chlorine (HOCl) during
contact with each of these compounds was measured by
XVLQJD0RGHO5&5HVLFORFRUGHU %LRQLFV,QVWUXPHQW
Co., Tokyo, Japan) at 20°C. The decrease in the concen-
trations of these antibiotics was also determined under the
same conditions by HPLC analyses of the reaction mix-
tures. Furthermore, the individual reactions of the antibi-
otics with hypochlorite in water at pH 5.0, 7.0, and 9.0
were measured by iodometric titration.
Fig. 1 shows the time courses of HOCl consumption
E\WKHQLWURIXUDQDQWLELRWLFV ȝPROO DQGWKHUHVLG-
ual concentrations of the parent compounds in buffered
solution at pH 7.0. As shown in Fig. 1, HOCl was large-
O\FRQVXPHGE\1)'GXULQJWKH¿UVWPLQIROORZHGE\
considerably slower secondary reactions. Approximately
RIWKHDFWLYHFKORULQHZDVZLWKLQWKH¿UVWPLQRI
the 1 hr with NFZ (Fig. 1A). In addition, NFZ in buffered
solution at pH 7.0 disappeared rapidly within 1 min after
treatment with chlorine under the same condition (Fig.
1B).
In contrast to the NFZ-hypochlorite reaction, a rapid
but small consumption of HOCl by FZD occurred dur-
ing the first 10 min, followed by very slow secondary
Vol. 33 No. 5
624

H. Nakamura et al.

㪈㪇
㪥㪝㪱
㪝㪱㪛
㪏

㪇
㪇 㪊㪇 㪍㪇 㪐㪇 㪈㪉㪇
㪫㫀㫄㪼㩿㫄㫀㫅㪀

Fig. 1. Time courses of chlorine consumptions by the nitrofuran antibiotics (A) and the residual amounts of NFZ and FZD (B) in
EXIIHUHGZDWHUWKDWRIS+DQGURRPWHPSHUDWXUH,QLWLDOFRQFHQWUDWLRQRIWKHQLWURIXUDQFRPSRXQGV ȝ0DQGWKDWRI
K\SRFKORULWH ȝ0 PJO 

reactions (Fig. 1A). In addition, HPLC analyses showed in chlorinated water at pH 5.0, 7.0, and 9.0. As a general
that FZD was extremely stable in buffered solution at trend, the rate of chlorine consumption by both the anti-
pH 7.0, despite the presence of excess of chlorine (Fig. biotics, particularly FZD, decreased with an increase in
1B). The chlorinated FZD derivatives formed in the solu- WKHVROXWLRQS+GXULQJWKH¿UVWKURIWKHK\SRFKORULWH
tion may be converted back to the parent compound by reactions. In contrast with the chlorine consumption rate,
the addition of strong nucleophiles, such as thiosulfate or the chlorine demands of both the antibiotics in alkaline
ELVXO¿WH %HGQHUDQG0DF&UHKDQ 7KHGLIIHUHQFH solutions after 24 hr were found to be larger than those
in the reactivity of NFZ and FZD toward chlorine can be observed under acidic and neutral conditions. This can be
explained in terms of the chemical structures of their side explained in terms of the concentrations of undissociated
chains (Table 1). hypochlorous acid (pKa = 7.5) in the chlorinated water.
Table 2 summaries the chlorine consumptions by the
2 individual antibiotics in the presence of buffer reagent Chlorination byproducts and reaction pathways
A previous study has reported the major product of
nitrofurans photolysis in water to be NFA and the product
Table 2. Chlorine (Cl) demands of NFZ and FZD in a of further oxidation of NFA to be 5-hydroxymethylene-2-
 ȝPROO VROXWLRQ DIWHU UHDFWLRQ IRU  DQG  + IXUDQRQH +0)  %XVNHUDQG%HLMHUVEHUJHQ
hr, respectively, with excess of hypochlorite at its chemical structure shown in Fig. 2 (c)). It has also been
20°C reported that a geometrical isomerization about the C = N
double bond forms anti-nitrofurans (Padwa and Albrecht,
Chemicals Cl demand (mol/mol)
Solution pH 1974; Quilliam et al., 1987; Edhlund et al., 2006), and
tested after 1 hr after 24 hr aqueous photolysis of all nitrofuran antibiotics produces 5-
5 7.23 11.29 nitrofuraldehyde azine (NFAA) (Quilliam et al., 1987; its
NFZ 7 7.72 11.09 chemical structure shown in Fig. 2 (e)). Although several
9 7.16 12.47 reports have been published on the photodecomposition
of nitrofuran antibiotics in water (Padwa and Albrecht,
5 2.36 3.15
1974; Busker and Beijersbergen, 1987; Quilliam et
FZD 7 2.14 3.15 al., 1987; Edhlund et al., 2006), there have been only a
9 1.58 3.83 few reports on the chemical fate of these compounds in

Vol. 33 No. 5
 625

0XWDJHQLFDFWLYLW\RIQLWURIXUDQVGXULQJDTXHRXVFKORULQDWLRQ

water during exposure to hypochlorite (Dodd and Huang,
2007).
Fig. 3 shows typical HPLC chromatograms of the reac-
WLRQSURGXFWVLQDTXHRXV1)=DQG)='VROXWLRQV ȝ0 
after treatment with chlorine at 4 eq. HOCl per mol of
NFZ solution and 100 eq. HOCl per mol of FZD solu-
tion and at pH 7.0 for 1 hr in dark. As shown in Fig. 3,
several reaction products could be seen on the chromato-
gram obtained from the chlorinated NFD and FZD solu-
tions. Some of the peaks appearing on the chromatograms
ZHUHFRQ¿UPHGRUGHWHUPLQHGDVWKHFKORULQDWLRQE\SURG-
ucts by comparing their retention times with those of the
authentic standards.
A geometrical isomer of syn-NFZ (the isomer, indicat-
ed in Fig. 3A and its chemical structure shown in Fig. 2
(a)) that is considered to be formed by the conversion of
V\QWRDQWL1)= )LJ E ZDVFRQ¿UPHGWREHSUHVHQW
in the chlorine-treated NFZ solution even under dark con-
to be NFOA by comparing the UV-vis absorbance spec-
tra and HPLC retention time of the product with those of
DQDXWKHQWLFVWDQGDUG$2=DQG6(0ZHUHQRWGHWHFWHG
in the chlorinated NFZ and FZD solutions because the
UV-absorbance intensities of these compounds were not
enough for detection.
A number of polar products with short retention times
were observed in the reverse-phase HPLC chromatogram
obtained from the chlorine-treated NFZ and FZD solu-
tions, particularly under an alkaline condition. Some of
these products were considered to be 5-hydroxy-2-furo-
LFDFLG +)$)LJ G RU+0)VLQFHWKHVHSURGXFWV
could be detected by the UV detector at 265 nm but not
at 365 nm. Furthermore, 2 reaction products with rather
long retention times than those of the parent compounds

㪥㪝㪱㩷㫀㫊㫆㫄㪼㫉
㪥㪝㪱
ditions. This observation is in agreement with that of pre- 㪘
㪩㪼㫊㫇㫆㫅㫊㪼㩷㩿㱗㪔㪉㪍㪌㫅㫄㪀㩷

vious studies (Padwa and Albrecht, 1974; Quilliam et al.,

1987; Edhlund et al., 2006) that demonstrated the anti-
nitrofurans to be present in nitrofuran photolysis solution.

㪬㫅㫂㫅㫆㫎㫅㩷㪉
However, the presence of a compound corresponding to
DQWL)='FRXOGQRWEHFRQ¿UPHGLQWKHFKORULQDWHG)='
㪬㫅㫂㫅㫆㫎㫅㩷㪈

solutions under the same conditions (Fig. 3B). 㪬㫅㫂㫅㫆㫎㫅㩷㪊

㪥㪝㪘
㪥㪝㪦㪘

NFA (Fig. 3A) that is considered to be formed by the

hydrolysis of the azomethine (C = N) bond was found to
be an intermediate product in the chlorinated NFZ solu-
WLRQ,WZDVFRQ¿UPHGE\FRPSDULQJWKH89YLVDEVRUE- 㪋 㪏 㪈㪉 㪈㪍
ance spectra and HPLC retention time of the product with 㪩㪼㫋㪼㫅㫋㫀㫆㫅㩷㫋㫀㫄㪼㩷㩿㫄㫀㫅㪀
those of an authentic standard. However, NFA was not
observed in the chromatogram obtained from the chlo-
㪬㫅㫂㫅㫆㫎㫅㩷㪋

㪝㪱㪛

rinated FZD solutions (Fig. 3B), because of the overlap- 㪙

㪩㪼㫊㫇㫆㫅㫊㪼㩷㩿㱗㪔㪉㪍㪌㫅㫄㪀㩷

ping of both peaks of NFA. The compound correspond-

ing to the second peak in Fig. 3A and B was identified

㩿㪸㪀 㩿㪹㪀 NH2

H H
㪬㫅㫂㫅㫆㫎㫅㩷㪌

H HH C O
O
㪬㫅㫂㫅㫆㫎㫅㩷㪎

N
㪬㫅㫂㫅㫆㫎㫅㩷㪍
㪥㪝㪦㪘

N C NH N
NO2 O C N 2 NO2 O C
H H H
㩿㪺㪀 㩿㪻㪀 O
H 㪋 㪏 㪈㪉 㪈㪍
O C HO C OH 㪩㪼㫋㪼㫅㫋㫀㫆㫅㩷㫋㫀㫄㪼㩷㩿㫄㫀㫅㪀
O OH O

Fig. 3. HPLC traces of the products of reaction of nitrofuran

㩿㪼㪀 antibiotics (NFZ and FZD) with hypochlorite in buff-
N N ered aqueous solution at pH 7 and at room temperature
O2N O C C O NO2
for 1 hr. Aqueous NFZ solution was treated with 4 eq.
HOCl per mol of the compound, while aqueous FZD
Fig. 2. Chemical structures of syn-NFZ (a), anti-NFZ (b), was treated with 100 eq. HOCl per mol of the com-
+0) F +)$ G DQG1)$$ H  pound.

Vol. 33 No. 5
626

H. Nakamura et al.

were observed in the chromatograms (Fig. 3). These prod- behavior than that of the parent compounds.
ucts are considered to be chloramines or chlorine-substi- Fig. 4 shows the results of HPLC determinations of the
tuted products of NFZ (Dodd and Huang, 2007) and FZD, chlorinated NFZ and FZD solutions at various pH values
because these compounds could be detected by the UV for 1 hr. Because the reactivity of NFZ toward hypochlo-
detector at 365 nm and exhibited comparatively non-polar rite in water was very high and that of FZD toward chlo-

㪉㪇 㪥㪝㪱 㪈㪇 㪉㪇 㪈㪇
㪥㪝㪱㩷㫀㫊㫆㫄㪼㫉㩷
㪝㪱㪛㩷
㪬㫅㫂㫅㫆㫎㫅㩷㪉
㪬㫅㫂㫅㫆㫎㫅㩷㪈㩷 㪏 㪬㫅㫂㫅㫆㫎㫅㩷㪎 㪏
㪈㪌 㪈㪌
㪥㪝㪦㪘㩷 㪬㫅㫂㫅㫆㫎㫅㩷㪋
㪥㪝㪘 㪍 㪥㪝㪦㪘㩷 㪍
㫇㪟㪌 㫇㪟㪌
㪈㪇 㪈㪇 㪬㫅㫂㫅㫆㫎㫅㩷㪌
㪋 㪋

㪌 㪌
㪉 㪉
㪩㪼㫊㫀㪻㫌㪸㫃㩷㪺㫆㫅㪺㪼㫅㫋㫉㪸㫋㫀㫆㫅㫊㩷㩿㫄㪾㪆㪣㪀㩷㫆㪽㩷㪥㪝㪱㪃㩷㪥㪝㪱㩷㫀㫊㫆㫄㪼㫉㪃㩷㪸㫅㪻㩷㪬㫅㫂㫅㫆㫎㫅㩷㪉

㪩㪼㫊㫀㪻㫌㪸㫃㩷㪺㫆㫅㪺㪼㫅㫋㫉㪸㫋㫀㫆㫅㩷㩿㫄㪾㪆㪣㪀㩷㫆㪽㩷㪥㪝㪦㪘㪃㩷㪬㫅㫂㫅㫆㫎㫅㩷㪋㪃㩷㪸㫅㪻㩷㪬㫅㫂㫅㫆㫎㫅㩷㪌㩷
㪩㪼㫊㫀㪻㫌㪸㫃㩷㪺㫆㫅㪺㪼㫅㫋㫉㪸㫋㫀㫆㫅㫊㩷㩿㫄㪾㪆㪣㪀㩷㫆㪽㩷㪥㪝㪘㪃㩷㪥㪝㪦㪘㪃㩷㪸㫅㪻㩷㪬㫅㫂㫅㫆㫎㫅㩷㪈

㪇 㪇 㪇 㪇
㪩㪼㫊㫀㪻㫌㪸㫃㩷㪺㫆㫅㪺㪼㫅㫋㫉㪸㫋㫀㫆㫅㫊㩷㩿㫄㪾㪆㪣㪀㩷㫆㪽㩷㪝㪱㪛㩷㪸㫅㪻㩷㪬㫅㫂㫅㫆㫎㫅㩷㪎㩷

㪇 㪌 㪈㪇 㪈㪌 㪉㪇 㪇 㪌㪇 㪈㪇㪇 㪈㪌㪇 㪉㪇㪇

㪉㪇 㪈㪇 㪉㪇 㪈㪇
㫇㪟㪎 㫇㪟㪎
㪏 㪏
㪈㪌 㪈㪌

㪍 㪍
㪈㪇 㪈㪇
㪋 㪋
㪌 㪌
㪉 㪉

㪇 㪇 㪇 㪇
㪇 㪌 㪈㪇 㪈㪌 㪉㪇 㪇 㪌㪇 㪈㪇㪇 㪈㪌㪇 㪉㪇㪇

㪉㪇 㪈㪇 㪉㪇 㪈㪇
㫇㪟㪐 㫇㪟㪐
㪏 㪏
㪈㪌 㪈㪌

㪍 㪍
㪈㪇 㪈㪇
㪋 㪋

㪌 㪌
㪉 㪉

㪇 㪇 㪇 㪇
㪇 㪌 㪈㪇 㪈㪌 㪉㪇 㪇 㪌㪇 㪈㪇㪇 㪈㪌㪇 㪉㪇㪇
㪤㫆㫃㫆㫉㩷㫉㪸㫋㫀㫆㩷㫆㪽㩷㪟㪦㪚㫃㩷㫇㪼㫉㩷㪺㫆㫄㫇㫆㫌㫅㪻 㪤㫆㫃㫆㫉㩷㫉㪸㫋㫀㫆㩷㫆㪽㩷㪟㪦㪚㫃㩷㫇㪼㫉㩷㪺㫆㫄㫇㫆㫌㫅㪻

Fig. 4. Residual amounts of chlorination byproducts of nitrofuran antibiotics (NFZ and FZD) in chlorinated aqueous solutions at
room temperature after 1 hr and at various molar ratios of HOCl per mol of compound. Initial concentration of the nitrofuran
antibiotics = 20 mg/l. The residual amount of the unknown product 1, 4, and 5 were derived from the HPLC peak areas,
relative to the peak area of NFOA. The residual amounts of unknown products 2 and 7 were derived from the HPLC peak
area, relative to the peak areas of the parent compound, and those of the other products, from each standard compound.

Vol. 33 No. 5

0XWDJHQLFDFWLYLW\RIQLWURIXUDQVGXULQJDTXHRXVFKORULQDWLRQ
rine was very low (Fig. 1), these compounds were treated
with chlorine at various molar ratios of HOCl per com-
pound. The concentration of the parent NFZ in water
decreased with an increase in the molar ratio of hypochlo-
rite to the compound (Fig. 4, left). NFZ was complete-
ly decomposed at the molar ratio of over 10 in the chlo-
rinated water. However, FZD remained in the chlorinated
water even at the molar ratio of 200 for the 1 hr reac-
tion (Fig. 4, right). Complete disappearance of FZD was
observed after treatment with hypochlorite at the molar
ratio of 200 for 48 hr in alkaline solution.
A high concentration anti-NFZ was found in aque-
ous NFZ solution on treatment with chlorine at a mod-
erate molar ratio of HOCl to NFZ (Fig. 4, left). At high
molar ratios of HOCl to NFZ, NFZ disappeared com-
pletely in the presence of excess active chlorine under the
experimental conditions. NFA, NFOA, and chloramines
(or chlorine-substituted products) of NFZ generated were
comparatively stable in acidic and neutral solutions. How-
ever, these products were degradable in the alkaline con-
dition (Fig. 4, left). Direct formation of NFOA was not
observed during further chlorination of NFA in buffered
water at pH 7.0. Unknown 1, which was polar products,
was detected at comparatively high concentrations in the
NFZ-hypochlorite reactions at pH 9.0 (Fig. 4, left).
On the other hand, it was found that the reaction of
FZD with hypochlorite in water proceeds under alkaline
conditions rather than in neutral and acidic conditions
(Fig. 4, right). Samuelsen et al. (1991) demonstrated that
FZD present in the sediment of a Norwegian salmon farm
was actively metabolized by the microorganisms in the
sediment, and the half-life of the parent compound at 4°C
was calculated to be 18 hr. In addition, Lunestad et al.
(1995) reported that FZD in seawater was degraded and
lost its antibacterial activity after 21 days of daylight radi-
ation at sea level (50 mg/l) and at a depth of 1 m from the
sea surface. However, FZD was stable in the chlorinat-
ed water (Fig. 4, right) even at high chlorine doses, indi-
FDWLQJWKHLQWURGXFWLRQRIZDVWHZDWHUHIÀXHQWVFRQWDLQLQJ
residual FZD into the receiving waters. Unknown 4 and
5, which were polar products, were detected at compara-
tively high concentrations in the FZD-hypochlorite reac-
tions at pH 9.0 (Fig. 4, right).
Hypochlorous acid plays multiple roles in the degrada-
tion of a variety of organic compounds, especially when
it is in equilibrium with hypochlorite at a pH close to its
pKa (7.5). In the reactions of the nitrofuran antibiotics
with chlorine in aqueous solution, the target compounds
rapidly underwent an anti-/syn isomerization to form anti-
nitrofurans in the initial step of the reactions. Subsequent-
ly, chlorination of most of the active sites on the target
627
compounds produced chlorinated nitrofurans. These reac-
tions were followed by the hydrolysis of azomethine (C
= N) bond to form NFA and the side chain moieties. The
residual amounts of their chlorination products in water
were found to be strongly dependent on the solution pH.
Mutagenicity tests
FZD has been found to exert mutagenic effects on
EscherichiaFROL:3 0F&DOODDQG9RXWVLQRV
Quilliam et al., 1987) and S. typhimurium TA100 (Wang
et al., 1975; Tatsumi et al0F&DOOD ,QWKLV
study, Ames assay was reperformed for FZD, NFZ, NFA,
NFOA and sodium azide (NaN3) by using the S. typhimu-
rium strain TA100 to compare the mutagenic response of
each compound. Table 3 summarizes the results of the
Ames assay conducted for these compounds by using the
TA100 strain without rat liver homogenate (S9 mix). The
strongest mutagenic activity was observed for the FZD
standard (net revertants = c.a. 2,588/nmol); its activi-
ty was approximately 13.5, 980, 18,500, and 15.3 times
higher than the activities observed for the NFZ, NFA,
NFOA, and NaN3 standards, respectively. Enzymatic acti-
vation with S9 mix reduced slightly both toxic and muta-
genic effects of nitrofuran antibiotics (NFZ and FZD) and
their degradation compounds (NFA and NFOA) to the
tester strain.
Fig. 5 illustrates the mutagenic responses obtained
by exposing the tester strain TA100 without S9 mix to
ethyl acetate extracts from the chlorinated aqueous NFZ
and FZD solutions (20 mg/l). Aqueous NFZ solution (pH
7.0) was treated with hypochlorite at 20 eq. HOCl/com-
pound for 1 hr, whereas aqueous FZD solution (pH 9.0)
was treated at 200 eq. HOCl/compound for 48 hr. Under
these conditions, a complete decomposition of the par-
ent compounds NFZ and FZD was observed in the chlo-
rinated NFZ and FZD solutions. No mutagenic response
was detected for the chlorinated FZD solution, whereas
the chlorination byproduct(s) of NFZ exerted mutagenic
effects on the TA100 strain without S9 mix.
The mutagenic response of the ethyl acetate extract of
the NFZ-hypochlorite reaction could not be explained by
the presence of NFA and NFOA in the extract (Fig. 4),
because their lower mutagenic activities as compared with
those of the parent compounds (Table 3). The mutagenic
activity of the extract might be due to the presence of an
unknown compound (compound 2 in Fig. 3), because this
compound was considered to be a chlorine-substituted
NFZ. Therefore, further investigation will be conducted
to determine the exact structure and mutagenic response
of this unknown compound after separating it from the
reaction mixture of NFZ and hypochlorite in water.
Vol. 33 No. 5
628

H. Nakamura et al.

Table 3.0XWDJHQLFHIIHFWVRIQLWURIXUD]RQHIXUD]ROLGRQHDQGVRGLXPD]LGHVWDQGDUGVRQ6typhimurium TA100 strain

without S9 mix.
Revertantsa)
'RVH ȝJ SODWH FZD NFZ NFA NFOA NaN3
0.1 1200 ± 390 146 ± 27 - b) - -
0.25 1567 ± 192 - - - 714 ± 230
0.5 537 ± 179 546 ± 172 - - 1340 ± 331
1 92 ± 41 948 ± 258 - - 2225 ± 553
2.5 Toxc) - 93 ± 12 56 ± 8 -
5 - 2210 ± 251 122 ± 6 66 ± 9 -
10 - 776 ± 106 231 ± 26 77 ± 14 -
25 - Tox 163 ± 27 109 ± 10 -
50 - - 88 ± 21 176 ± 11 -
100 - - Tox 278 ± 18 -
Net revertants / nmol 2588 192 2.64 0.14 169.6
6SRQWDQHRXV '062 61 ± 5
a)Average values and standard deviations of 3 plates per dose in each test
b) Not determined
C) Toxic effect toward tester strain

㪋㪇 㪥㪝㪱 㪋㪇 㪝㪱㪛
㩿㪘㪀 㩿㪙㪀
㫉㪼㫍㪼㫉㫋㪸㫅㫋㫊㪆㩷㫊㫇㫆㫅㫋㪸㫅㪼㫆㫌㫊㩷㫉㪼㫍㪼㫉㫋㪸㫅㫋㫊㪀㩷㩷

㪚㪿㫃㫆㫉㫀㫅㪸㫋㫀㫆㫅 㪚㪿㫃㫆㫉㫀㫅㪸㫋㫀㫆㫅
㪹㫐㫇㫉㫆㪻㫌㪺㫋㫊
㪤㫌㫋㪸㪾㪼㫅㫀㪺㫀㫋㫐㩷㫉㪸㫋㫀㫆㩷㩿㫅㫌㫄㪹㪼㫉㩷㫆㪽㩷

㪹㫐㫇㫉㫆㪻㫌㪺㫋㫊
㪊㪇 㪊㪇

㪉㪇 㪉㪇

㪈㪇 㪈㪇

㪇 㪇
㪇 㪉 㪋 㪍 㪏 㪈㪇 㪇 㪉 㪋 㪍 㪏 㪈㪇
㪛㫆㫊㪼㩷㩿㱘㪾㪃㩷㪸㫊㩷㫊㫋㪸㫉㫋㫀㫅㪾㩷㫄㪸㫋㪼㫉㫀㪸㫃㪀㩷㫇㪼㫉㩷㫇㫃㪸㫋㪼 㪛㫆㫊㪼㩷㩿㱘㪾㪃㩷㪸㫊㩷㫊㫋㪸㫉㫋㫀㫅㪾㩷㫄㪸㫋㪼㫉㫀㪸㫃㪀㩷㫇㪼㫉㩷㫇㫃㪸㫋㪼

Fig. 5. 0XWDJHQLFUHVSRQVHVRIQLWURIXUDQDQWLELRWLFV1)' $ DQG)=' % DQGWKHLUFKORULQDWLRQE\SURGXFWVRQ6typhimurium

TA100 without S9 mix. Chlorination by products were obtained by reaction in 20 eq. HOCl/NFZ solution at room tempera-
ture for 1 hr and 200 eq HOCl/FZD solution at pH 9 for 48 hr.

Aqueous chlorine is used to disinfect sewage water with hypochlorite in water (Fig. 3). The production and
and raw water that is then supplied for human consump- residual amounts of these compounds greatly depend-
tion. In this study nitrofuran antibiotic NFZ was shown ed on the molar ratios of HOCl to the compound and the
to produce many decomposition products after reaction reaction pH, with higher concentrations at moderate chlo-

Vol. 33 No. 5

0XWDJHQLFDFWLYLW\RIQLWURIXUDQVGXULQJDTXHRXVFKORULQDWLRQ
rine doses and under acidic and neutral conditions. Acidic
or alkaline chlorinated waters are generally neutralized or
diluted in the receiving water at water treatment stations.
Therefore, most of the residual products are destroyed by
this process and public health hazard may be substantial-
ly decreased.
ACKNOWLEDGMENTS
The authors thank all the members of the Environmen-
tal Science Laboratory, Faculty of Pharmaceutical Sci-
ences, Tokyo University of Science, for their cooperation
during the course of this study.
REFERENCES
$VKWRQ'+LOWRQ0DQG7KRPDV.9  ,QYHVWLJDWLQJWKH
environmental transport of human pharmaceuticals to streams in
the United Kingdom. Sci. Total Environ., 333, 167-184.
%XVNHU5:DQG%HLMHUVEHUJHQYRQ+HQHJRXZHQ*0-  
The photolysis of 5-nitrofurfural in aqueous solutions: Nucle-
ophilic substitution of the nitro group. Photochem. Photobiol.,
45, 331-335.
%HGQHU0DQG0DF&UHKDQ:$  5HDFWLRQVRIWKHDPLQH
containing drugs fluoxetine and metoprolol during chlorina-
tion and dechlorination processes used in wastewater treatment.
Chemosphere, 65, 2130-2137.
'RGG0&DQG+XDQJ&+  $TXHRXVFKORULQDWLRQRIWKH
antibacterial agent trimethoprim: Reaction kinetics and path-
ways. Water Res., 41, 647-655.
(GKOXQG%/$UQROG:$DQG0F1HLOO.  $TXDWLFSKR-
tochemistry of nitrofuran antibiotics. Environ. Sci. Technol., 40,
5422-5427.
Guay, D.R. (2001): An update on the role of nitrofurans in the man-
agement of urinary tract infections. Drugs, 61, 353-364.
Hirsch, R., Ternes, T., Haberer, K. and Kratz, K.-L. (1999): Occur-
rence of antibiotics in the aquatic environment. Sci. Total Envi-
ron., 225, 109-118.
Kim, S.C. and Carlson, K. (2007): Temporal and spatial trends in
the occurrence of human and veterinary antibiotics in aqueous
and river sediment matrices. Environ. Sci. Technol., 41, 50-57.
.ROSLQ':)XUORQJ(70H\HU077KXYRQDQ(0=DXJJ
S.D., Barder, L.B. and Buxton, H.P. (2001): Pharmaceuticals,
hormones, and other organic wastewater contaminants in U.S.
streams, 1999-2000 A national reconnaissance. Environ. Sci.
Technol., 36, 1202-1211.
Lin, W.C., Chen, H.C. and Ding, W.H. (2005): Determination of
pharmaceutical residues in waters by solid-phase extraction and
large-volume on-line derivatization with gas chromatography-
mass spectrometry. J. Chromatogr. A, 1065, 279-285.
Lunestad, B.T., Samuelsen, O.B., Fjelde, S. and Ervik, A. (1995):
Photostability of eight antibacterial agents in seawater. Aquacul-
ture, 134, 217-225.
0F&DOOD'5DQG9RXWVLQRV'  2QWKHPXWDJHQLFLW\RI
QLWURIXUDQV0XWDW5HV26, 3-16.
0F&DOOD '5   0XWDJHQLFLW\ RI QLWURIXUDQ GHULYDWLYHV
5HYLHZ(QYLURQ0XWDJHQ5, 745-765.
0DURQ'0DQG$PHV%1  5HYLVHGPHWKRGVIRUWKHSal-
View publication stats
629
monella PXWDJHQLFLW\WHVW0XWDW5HV113, 173-215.
Onodera, S., Yoshimatsu, K., Saitoh, H. and Uchida, A. (1998):
Behavior of mutagenic formation from phenolic compounds in
water disinfection with chlorine and their mutagenic potential
formation. Jpn. J. Toxicol. Environ. Health, 44, 289-299.
Onodera, S., Takahashi, T., Takemoto, S. and Oh-I, T. (2008): Char-
acterization of polyhalogenated 4-methylphenol dimers (Br/Cl-
predioxins) formed during aqueous chlorination of 4-methylphe-
nol solution in the presence of bromide ion. J. Health Sci., 54,
423-431.
Padwa, A. and Albrecht, F. (1974): Photochemical syn-anti isomer-
ization about the carbon-nitrogen double bond. J. Am. Chem.
Soc., 96, 4849-4857.
3DXO+(DQG3DXO0)  7KHQLWURIXUDQVFKHPRWKHUDSHX-
tic properties. In Experimental Chemotherapy, Vol. 2, (Schnitzer,
R.A. and Hawking, F., eds) pp. 307-370, Academic Press, New
York.
3HWURYLF0+HUQDQGR0''LD]&UX]06DQG%DUFHOR'
(2005): Liquid chromatography-tandem mass spectrometry for
the analysis of pharmaceutical residues in environmental sam-
ples: A review. J. Chromatogr. A, 1067, 1-14.
3RWKXOXUL -9 1DZD] 06 .KDQ$$ DQG &HUQLJOLD &(
(1998): Antimicrobial use in aquaculture: Environmental fate
and potential for transfer of bacterial resistance genes. Recent
5HV'HY0LFURELRO2, 351-372.
4XLOOLDP0$0F&DUU\%(+RR.+0F&DOOD'5DQG
9DLWHNXQDV6  ,GHQWL¿FDWLRQRIWKHSKRWRO\VLVSURGXFWV
of nitrofurazone irradiated with laboratory illumination. Can. J.
Chem., 65, 1128-1132.
Richardson, S.D. and Ternes, T.A. (2005): Water analysis: Emerging
contaminants and current issues. Anal. Chem., 77, 3807-3838.
Samuelsen, O.B., Solheim, E. and Lunestad, B.T. (1991): Fate and
microbiological effects of furazolidone in a marine aquaculture
sediment. Sci. Total Environ., 108, 275-283.
Tatsumi, K., Ou, T., Yamada, H., Yoshimura, H., Koga, H. and
+RULXFKL7  ,VRODWLRQDQGLGHQWL¿FDWLRQRIWKHPHWDER-
lite of N-(5-nitro-2-furfurylidene)-3-amino-2-oxazolidone (fura-
zolidone). J. Pharmacobio. dyn., 1, 256-261.
Ternes, T.A., Joss, A. and Siegris, H. (2004): Scrutinizing pharma-
ceuticals and personal care products in wastewater treatment.
Environ. Sci. Technol., 38, 392-399.
7L[LHU&6LQJHU+32HOOHUV6DQG0XOOHU65  2FFXU-
UHQFHDQGIDWHRIFDUEDPD]HSLQHFOR¿EULFDFLGGLFORIHQDFLEX-
profen, ketoprofen, and naproxen in surface waters. Environ.
Sci. Technol., 37, 1061-1068.
Tomiyama, S., Takahashi, Y., Yaguchi, K. and Onodera, S. (2008)
Chlorination byproducts of epoxy resin hardener and mutagenic
assay of their products. J. Health Sci., 54, 17-22.
9DQ'H6WHHQH-&0RUWLHU.$DQG/DPEHUW:(  
Tackling matrix effects during development of a liquid chro-
matographic-electrospray ionisation tandem mass spectrometric
analysis of nine basic pharmaceuticals in aqueous environmental
samples. J. Chromatogr. A, 1123, 71-81.
9DQ .RWHQ9HUPHXOHQ -(0 :RXWHUV 0)$ DQG 9DQ
Leeuwen, F.X.R. (1993): Toxicological evaluation of certain vet-
erinary drug residues in food. Report of the 40th meeting of the
joint FAO/WHO expert committee on food additives (JEC-FA),
World Health Organization, Geneva pp.85-123.
:DQJ&<0XUDRND.DQG%U\DQ*7  0XWDJHQLFLW\RI
nitrofurans, nitrothiophenes, nitropyrroles. Nitroimidazole, ami-
nothiophenes, and aminothiazoles in Salmonella typhimurium.
Cancer Res., 35, 3611-3617.
Vol. 33 No. 5