Ecotoxicology and Environmental Safety 229 (2022) 113063

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Hydrolytic transformation mechanism of tetracycline antibiotics: Reaction

kinetics, products identification and determination in WWTPs
Shao-Fen Zhong a, b, c, Bin Yang a, b, *, Qian Xiong a, b, Wen-Wen Cai a, b, Zheng-Gang Lan a, b,
Guang-Guo Ying a, b, *
a
SCNU Environmental Research Institute, Guangdong Provincial Key Laboratory of Chemical Pollution and Environmental Safety & MOE Key Laboratory of Theoretical
Chemistry of Environment, South China Normal University, Guangzhou 510006, China
b
School of Environment, South China Normal University, Guangzhou 510006, China
c
School of Chemical Engineering and Energy Technology, Dongguan University of Technology, Dongguan 523808, China

A R T I C L E I N F O A B S T R A C T

Edited by Professor Bing Yan Antibiotic residues and antibiotic resistance have been widely reported in aquatic environments. Hydrolysis of
antibiotics is one of the important environmental processes. Here we investigated the hydrolytic transformation
Keywords: of four tetracycline antibiotics i.e. tetracycline (TC), chlortetracycline (CTC), oxytetracycline (OTC) and doxy­
Tetracyclines cycline (DC) under different environmental conditions, and determined their parents and transformation prod­
Antibiotics
ucts in the wastewater treatment plants (WWTPs). The results showed that the hydrolysis of the four tetracyclines
Hydrolysis
followed first-order reaction kinetics, and the acid-catalyzed hydrolysis rates were significantly lower than the
Transformation products
Wastewater treatment plants base-catalyzed and neutral pH hydrolysis rates. The effect of temperature on tetracycline hydrolysis was quan­
Mass spectrometry tified by Arrhenius equation, with Ea values ranged from 42.0 kJ mol− 1 to 77.0 kJ mol− 1 at pH 7.0. In total, nine,
six, eight and nine transformation products at three different pH conditions were identified for TC, CTC, OTC and
DC, respectively. The main hydrolysis pathways involved the epimerization/isomerization, and dehydration.
According to the mass balance analysis, 4-epi-tetracycline and iso-chlortetracycline were the main hydrolytic
products for TC and CTC, respectively. The 2 tetracyclines and 4 hydrolysis products were found in the sludge
samples in two WWTPs, with concentrations from 15.8 ng/g to 1418 ng/g. Preliminary toxicity evaluation for the
tetracyclines and their hydrolysis products showed that some hydrolysis products had higher predicted toxicity
than their parent compounds. These results suggest that the hydrolysis products of tetracycline antibiotics should
also be included in environmental monitoring and risk assessment.

1. Introduction hydrolysis (Kim et al., 2004; Li et al., 2008; Mitchell et al., 2014; Yi et al.,
2016). For example, it is difficult to detect penicillins in the aquatic
The large-scale consumption of antibiotics in human and animals can environment as they have been hydrolyzed, especially in alkaline
lead to presence of antibiotic residues in the environment, and also sewage (Christian et al., 2003). Hydrolysis of β-lactam antibiotics
develop antibiotic resistance, which has become a global concern ampicillin, cefalotin and cefoxitin are more likely to occur in alkaline
(Reardon, 2014; Wang et al., 2018). After released in the environment, systems (Mitchell et al., 2014). The enhanced hydrolysis technology has
these antibiotic residues may expose to various environmental processes been developed as a novel approach to treat high-concentration anti­
such as hydrolysis, photolysis and biodegradation. Hydrolysis has been biotic wastewater (Yi et al., 2016). Thus, hydrolysis of antibiotics is
known to be one of the common transformation pathways for various important to understanding their migration and transformation
antibiotics containing one or more easily hydrolyzable functional behavior in aquatic environments.
groups, e.g., ester, amide, imide and halogen (Biosic et al., 2017; Tetracycline antibiotics are extensively used for human therapy,
Kummerer, 2009; Zhang et al., 2015). Many antibiotics have been re­ animal husbandry and aquaculture because of their high efficiency, low
ported to have poor stability in aqueous solutions, and are susceptible to cost and broad-spectrum. Their yield and usage have been ranked the

* Corresponding authors at: SCNU Environmental Research Institute, Guangdong Provincial Key Laboratory of Chemical Pollution and Environmental Safety &
MOE Key Laboratory of Theoretical Chemistry of Environment, South China Normal University, Guangzhou 510006, China.
E-mail addresses: bin.yang@m.scnu.edu.cn (B. Yang), guangguo.ying@m.scnu.edu.cn (G.-G. Ying).

https://doi.org/10.1016/j.ecoenv.2021.113063
Received 5 October 2021; Received in revised form 22 November 2021; Accepted 4 December 2021
Available online 8 December 2021
0147-6513/© 2021 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
S.-F. Zhong et al. Ecotoxicology and Environmental Safety 229 (2022) 113063

first in China (Pan et al., 2015), and the second in the world (Gao et al., performance liquid chromatography-triple quadrupole mass spectrom­
2012). Tetracycline antibiotics mainly include tetracycline, chlortetra­ etry (UPLC-MS/MS). The USEPA ECOSAR software and gene recombi­
cycline, oxytetracycline and doxycycline, with very similar chemical nant luminescent bacteria were used to evaluate the toxicity of the
structures. After use, only a small part of tetracyclines can be metabo­ tetracyclines and their transformation products. The results can facili­
lized or absorbed by the body, and a high percentage (50–90%) of the tate a deep understanding of the fate of tetracycline antibiotics and their
tetracyclines can be excreted via urine and feces, subsequently found transformation products, as well as their long-term impacts on the
their way into the environment from wastewater discharge, manure environment.
disposal, aquaculture and animal grazing (Kummerer, 2009; Oberoi
et al., 2019). Currently, the overuse of antibiotics causes tetracyclines to 2. Materials and methods
be frequently detected at a concentration range of ng L− 1 to mg L− 1 in
surface water, groundwater and WWTPs’ wastewater (Chen et al., 2017; 2.1. Chemicals and materials
Dai et al., 2019; Zhou et al., 2013), which make it meaningful to study
the hydrolysis of tetracycline antibiotics in the real environment. Analytical grade hydrochloride salts of four tetracycline antibiotics
Tetracycline antibiotics have high water solubility and low log Kow, as (tetracycline (TC, 98%), oxytetracycline (OTC, 97%), chlortetracycline
shown in Text S1 and Table S1, Supplementary Information. Hydrolysis (CTC, 93%) and doxycycline (DC, 93%)) and seven hydrolysis products
has been found to be a very significant degradation process for tetra­ 4-epi-tetracycline (ETC), anhydrotetracycline (ATC), 4-epi-anhydrote­
cyclines with the half-lives ranging from less than 6 h up to 9.7 weeks tracycline (EATC), 4-epi-chlortetracycline (ECTC), iso-
(Loftin et al., 2008). According to recommendations of the European chlortetracycline (iso-CTC), 4-epi-oxytetracycline (EOTC) and 4-epi-
Medicines Agency (EMEA) and the US Food and Drug Administration doxycycline (EDC)) were purchased from Dr. Ehrenstorfer GmbH
(FDA), hydrolysis of pharmaceuticals in water should be implemented (Augsburg, Germany). Internal standard thiabendazole-D4 (TBD-D4)
by following OECD Test No. 111 Guideline (EPA, 2008; OECD, 2004). was obtained from Toronto Research Chemicals (North York, Canada).
Although some hydrolytic degradation data of tetracyclines are avail­ Buffers and all other reagents used in the experiments were of analytical
able in the literature, the test procedures did not all follow OECD Test grade. Milli-Q water (≥ 18.2 MΩ cm) was used to prepare all aqueous
No. 111 Guideline, which makes it difficult to compare the hydrolysis solutions. The HPLC grade acetonitrile and methanol were acquired
results (Kang et al., 2012; Loftin et al., 2008; Søeborg et al., 2004; Xuan from Merck (Darmstadt, Germany). Formic acid (HPLC grade, purity
et al., 2009; Yi et al., 2016). 98%) was supplied by ANPEL Laboratory Technologies (Shanghai,
Due to the high detection rate of tetracyclines in the natural envi­ China).
ronment, the monitoring and the potential risks of their transformation
products should not be ignored. It has been reported that transformation 2.2. Hydrolysis experiments of tetracyclines
products of tetracyclines are highly soluble and stable in receiving
water, and more active and/or toxic than their parents (Topal and Topal, Hydrolysis experiments of tetracyclines were conducted according to
2015). It was reported that tetracycline was prone to epimerization and the OECD Test No. 111 Guideline. The experiment was carried out in
dehydration, and transformed into hydrolysis products (Yang et al., term Amber-closed glass vials under dark conditions to avoid photo­
2020). Some hydrolytic products of tetracyclines have been identified as degradation. All glass containers, reagents and buffer solutions used in
4-epi-tetracycline, anhydrotetracycline, 4-epi-chlortetracycline and hydrolysis experiments were sterilized to prevent the effects of any
4-epi-oxytetracycline (Halling-Sørensen et al., 2002; Halling-Sørensen. biodegradation. The stock solutions of the above tetracyclines were
et al., 2003; Søeborg et al., 2004). The analysis of transformation dissolved in methanol and stored in − 20 ◦ C. Initial concentration of 1
products is quite challenging due to the lack of reliable standards mg/L of TC, OTC, CTC and DC used in the experiments were prepared by
(Escher and Fenner, 2011; Wang et al., 2020). With the improvement of dilution of single standard antibiotic stock solution with appropriate
instrument resolution, high-resolution mass spectrum has been intro­ phosphate buffer solution (Milli-Q water), and the final organic solvent
duced to identify known and yet unknown transformation products, content was negligible (lower than 1%). The buffer solution was pre­
making it possible to better understand the transformation mechanism pared by mixing K2HPO4 and KHPO4 at 10 mM. Small volume of 0.1 M
of tetracyclines. Besides, a previous study had reported high concen­ sodium hydroxide or 0.1 M phosphate acid was used for achieving pH
trations of oxytetracycline (238–1680 ng/g) and tetracycline 4.0, 7.0 and 9.0 with a variation of below 0.1 unit. All experiments were
(117–1650 ng/g) were found in the sludge samples in two WWTPs in performed in triplicate.
South China (Zhou et al., 2013), but their transformation products have
been rarely detected or not monitored in WWTPs. Tetracycline and 3 2.2.1. Tier 1
degradation products (4-epitetracycline, 4-epi-anhydrotetracycline, and The preliminary test for hydrolytic stability of TCs was carried out at
anhydrotetracycline) were reported at concentrations of micrograms per 50 ◦ C for 5 days. According to OECD 111 Guideline, only when the
liter and micrograms per kilogram in municipal wastewater treatment hydrolysis is more than 10% within this time, the tested substance can
plants (Topal et al., 2016; Topal and Topal, 2015). The concentrations be considered hydrolytically unstable, and usually further testing is
and toxicity of transformation products can be higher than those of their required.
parent compounds in some cases, warranting potential risk assessment
of transformation products (Wang et al., 2020). 2.2.2. Tier2
This study focused on the hydrolytic behavior of the four tetracycline The further test was performed at three different pH-values (4.0, 7.0,
antibiotics, including tetracycline, chlortetracycline, oxytetracycline 9.0) and at three different temperatures (10 ◦ C, 25 ◦ C, 50 ◦ C) to deter­
and doxycycline. The hydrolysis rate constants and half-lives were ob­ mine the kinetic behavior of unstable substances, defined by the pre­
tained under ambient temperature and pH according to OECD Test No. liminary test above.
111 Guideline. The Ultra performance liquid chromatography-
quadrupole time-of-flight mass spectrometry (UPLC-QTOF-MS) was 2.2.3. Tier3
employed to identify the hydrolysis products of tetracyclines at different Before and after hydrolysis experiments, the residual concentrations
acidic versus basic conditions, and the possible hydrolysis trans­ of tetracyclines were analyzed using high-performance liquid chroma­
formation pathways were tentatively proposed. In addition, the tetra­ tography (HPLC). The detailed HPLC method was given in Table S2. The
cyclines and their hydrolysis products, including knowns and unknowns, formation of hydrolysis products of tetracyclines and their structure
were monitored in two WWTPs by suspect screening of UPLC-QTOF-MS, elucidation were investigated by UPLC-QTOF-MS. A higher initial con­
and further quantified with authentic standards by using Ultra centration of tetracyclines (5 mg/L) was performed in the experiments

2
S.-F. Zhong et al. Ecotoxicology and Environmental Safety 229 (2022) 113063

in order to facilitate the detection and identification of their trans­ 2.4.2. Quantitative analysis using UPLC-MS/MS
formation products (Yang et al., 2018). The 2 tetracyclines and 4 hydrolysis products were found in the
above WWTPs samples. Among the 6 compounds, only 5 authentic
2.3. Instrumental analysis of hydrolysis products standards can be purchased, and therefore determined by UPLC-MS/MS
for confirmation and further quantification. The quantification of these
Hydrolysis products of tetracyclines were identified by using an 5 targets (TC, ETC, OTC, EOTC and EDC) in wastewater and sludge
Agilent 1290 Infinity LC system outfitted with an Agilent 6545 samples were conducted by ultra-performance liquid chromatography
quadrupole-time-of-flight mass spectrometer (UPLC-QTOF-MS). The coupled to Xevo TQ-S triple quadrupole mass spectrometer (Waters Co.,
separation was carried out on an Agilent Eclipse Plus C18 column (2.1 × Milford, MA, USA) with an electrospray ionization source in multiple-
150 mm, 1.8 µm). The injection volume was 5 μL. Two mobile phases A: reaction monitoring (MRM) mode (Jia et al., 2009; Topal et al., 2016;
0.1% formic acid in water and B: methanol were used at a flow rate of Topal and Topal, 2015). The detailed instrument information was pre­
0.4 mL min− 1. The gradient method started with 10% B and increased to sented in Text S5, and the ion transitions and retention times for these 5
30% B in 2 min, and then increased to 50% B at 10 min, then increased to confirmed compounds in UPLC-MS/MS were given in Table S3. The
90% B at 15 min which was held for 5 min (until 20 min), reverted to quality assurance (QA) and quality control (QC) during analysis were
10% B at 20.1 min and re-equilibrated for 4.9 min (from 20.1 to 25 min) described in Text S6. Detail recoveries, method detection limits (MDLs)
at 10% B. and method quantitative limits (MQLs) of 5 confirmed hydrolysis
The UPLC-QTOF-MS was operated with Dual AJS source in positive products and their parents were listed in Table S4. The range of MDLs
and negative electrospray ionization (ESI) mode, under the following and MQLs for sludge was 0.05–1.86 ng/L and 0.18–6.20 ng/L, respec­
conditions: sheath gas flow rate, 9 L min− 1; sheath gas temperature, tively. The average spiked recoveries of the target analytes in the sludge
350 ◦ C; nebulizer pressure, 40 psi; fragment voltage, 175 V; skimmer ranged from 80.8% to 129%, with a relative standard error (RSD) less
voltage, 65 V; eight-stage bar voltage, 750 V; capillary voltage, 3.5 kV. than 18.2%.
The UPLC-QTOF-MS accurate mass spectra was recorded across the
range 100–1700 m/z. First, samples were screened by MS1 scan mode at 2.5. Toxicity evaluation
a rate of 2.5 spectra/s and then analyzed by using auto-MS/MS spectra
for confirmation with 4 spectra/s. All the analysis were collected at three The toxicity values of the tetracyclines and their transformation
collision energy values of 10 V, 20 V and 40 V. products for fish, daphnia and green algae were predicted by USEPA
The MS and MS/MS data were acquired and processed using a ECOSAR (v1.11) in USEPA EPI Suite™ software. The acute toxicity of
sample-control comparison workflow for non-targeted screening by tetracycline hydrolysis solutions, as well as 4 tetracylines and 7 trans­
searching over sample chromatograms of zero-time, blank and control formation products were evaluated using the gene recombinant lumi­
samples with hydrolysis samples at different reaction times. Data nescent bacteria (E. coli HB101 pUCD607) according to the protocol
acquisition and analysis were performed using Agilent MassHunter described by Fang et al. (2020).
analysis software. Details of transformation product identification were
presented in Text S2. The standard curves of seven authentic reference
3. Results and discussion
standards of tetracyclines degradation products were prepared and
tested by UPLC-QTOF-MS with the above method. In the mass balance
3.1. Kinetics of hydrolytic degradation of tetracyclines
analysis, the concentrations of hydrolysis products were determined by
the external standard method.
The obtained results of preliminary tests (Tier 1) for determining the
hydrolytic stabilities of TC, OTC, CTC and DC at 50 ◦ C for 5 days,
2.4. Screening of environmental samples
revealed the hydrolytic instability of these 4 tetracyclines (Fig. S1), then
the higher tests (Tier 2) were conducted.
2.4.1. Qualitative analysis using UPLC-QTOF-MS
The hydrolysis of tetracyclines at the temperature of 25 ◦ C and pH
Wastewater and sludge samples were collected from two wastewater
4.0, 7.0 and 9.0 obeyed first-order reaction kinetics (R2 > 0.98)
treatment plants (denoted as WWTP1 and WWTP2) in Guangzhou for
(Fig. S2), which proved the hydrolysis should be a unimolecular reac­
screening analysis of tetracyclines and their hydrolysis products with
tion. The hydrolysis rate constants (kh) and half-lives (t1/2) at 25 ◦ C and
UPLC-QTOF-MS. The extraction of wastewater and sludge samples was
pH 4.0, 7.0 and 9.0 were listed in Table 1. The half-lives at pH 7.0 and
described in Text S3. An accurate mass database and a collision-induced-
25 ◦ C for TC, OTC, CTC and DC were 112 h, 66.0 h, 20.6 h and 107 h,
dissociation (CID) accurate mass spectral library of tetracyclines and
their hydrolysis products were established for actual sample screening.
Table 1
Mass spectrum data were obtained by standard solutions and above
Hydrolysis rate constants and half-lives of tetracyclines as a function of pH.
identified hydrolysis products containing the fragmentation ions and
structures. The database setup process was shown in Text S4. Data Compound Parameter pH

acquisition was performed in two steps for each actual sample (Bletsou 4.0 7.0 9.0
et al., 2015; Pang et al., 2018; Z. Wang et al., 2015). The first level of Tetracycline − 1
kh (h ) 1.8( ± 0.05) 6.2( ± 0.3)× 3.1( ± 0.1) ×
accurate mass-to-charge ratio information of the analyzed sample was × 10− 3 10− 3 10− 3
collected for the first time. By using "Find Company by Formula", the t1/2 (h) 3.85( ± 1.12( ± 0.06) 2.24( ± 0.07)
original data collected under full scanning mode of UPLC-QTOF-MS 0.11) × 102 × 102 × 102
Oxytetracycline kh (h− 1) 4.3( ± 0.2) 1.05( ± 0.05) 5.1( ± 0.2) ×
were analyzed automatically. According to the precise mass charge × 10− 3 × 10− 2 10− 3
ratio collected by QTOF-MS, the possible molecular formula of each t1/2 (h) 1.61( ± 6.6( ± 0.36) 1.36( ± 0.04)
extraction chromatography was given. For the second step, the Auto 0.07) × 102 × 101 × 102
MS/MS mode of QTOF-MS was used to collect the fragment ion infor­ Chlortetracycline kh (h− 1) 9.3( ± 0.3) 3.37( ± 0.2) 5.27( ± 0.24)
× 10− 3 × 10− 2 × 10− 1
mation, which was extracted by the "Find by Auto MS/MS" algorithm.
t1/2 (h) 7.45( ± 2.06( ± 0.09) 1.30 ± 0.061
Finally, the collected mass spectrum data information of the samples 0.21) × 101 × 101
was compared with the self-built UPLC-QTOF-MS database. If the Doxycycline kh (h− 1) 5.0( ± 0.3) 6.5( ± 0.4) × 9.9( ± 0.3) ×
QTOF-MS and MS/MS matching score of a certain compound were × 10− 3 10− 3 10− 3
higher than 70 points, the result can be defined as a positive result (Z. t1/2 (h) 1.39( ± 1.07( ± 0.05) 7.0( ± 0.26)
0.07) × 102 × 102 × 101
Wang et al., 2015).

3
S.-F. Zhong et al. Ecotoxicology and Environmental Safety 229 (2022) 113063

Table 2
Hydrolysis rate constants and half-lives of tetracyclines as a function of temperature.
Compound Parameter Temperature Ea (kJ mol− 1) A (h− 1) Q10

10 C
◦
25 C
◦
50 C
◦

Tetracycline kh (h− 1) 2.5( ± 0.1) × 10− 3 6.2( ± 0.2) × 10− 3 5.5( ± 0.28) × 10− 2 69.8 1.07 × 1010 2.39
t1/2 (h) 2.77( ± 0.11) × 102 1.12( ± 0.03) × 102 1.26( ± 0.04) × 101
Oxytetracycline kh (h− 1) 4.4( ± 0.3) × 10− 3 1.05( ± 0.05) × 10− 2 1.15( ± 0.06) × 10− 1 76.8 3.09 × 1011 2.61
t1/2 (h) 1.58( ± 0.11) × 102 6.6( ± 0.3) × 102 6.0 ± 0.344
Chlortetracycline kh (h− 1) 5.5( ± 0.3) × 10− 3 3.37( ± 0.1) × 10− 2 3.77( ± 0.15) × 10− 1 77.4 1.22 × 1012 2.63
t1/2 (h) 1.26( ± 0.06) × 102 2.06( ± 0.06) × 101 1.8 ± 0.0714
Doxycycline kh (h− 1) 3.1( ± 0.2) × 10− 3 6.5( ± 0.4)× 10− 3 2.46( ± 0.08) × 10− 2 42.6 1.91 × 10 5
1.70
t1/2 (h) 2.24( ± 0.1) × 102 1.07( ± 0.05) × 102 2.82( ± 0.09) × 101

Ea: Experimental activation energy; A: Pre-exponential factor; Q10: The change of hydrolysis rate with a temperature change of 10 ◦ C.
The calculation method of Ea, A and Q10 is given in Text S5.

respectively. In general, the trend of hydrolysis rate was hydrolysis rate factor for a change in 10 ◦ C (Q10) were similar to those
CTC>OTC>TC≈DC, and hydrolysis rates for DC and TC were roughly obtained for hydrolysis of antibiotics (Biosic et al., 2017; Loftin et al.,
equivalent under normal environmental conditions. These results were 2008; Mitchell et al., 2014; Yi et al., 2016). The range of Ea for the 4
comparable to the previous study that reported half-lives were 116 h, tetracyclines at pH 7.0 was 42–77 kJ mol− 1. The CTC and DC had the
41.0 h, 16.0 h at 22 ◦ C around neutral pH for TC, OTC and CTC, highest and lowest Ea, respectively. Thus, the hydrolysis of CTC was
respectively (Loftin et al., 2008). The difference in the hydrolysis rate more sensitive to temperature. Based on the hydrolysis rate constants at
among the 4 tetracyclines should be linked to their chemical structures. 25 ◦ C and 50 ◦ C, the Q10 was calculated to be 2.39, 2.61, 2.63 and 1.70
The tetracycline antibiotics are a class of tetrahydrobenzene derivatives for TC, OTC, CTC and DC, respectively. This increasing rate with tem­
with a dibasic tetraphenyl base structure (Fig. S3) (Dai et al., 2019). The perature for tetracyclines was the same as many other organic com­
substituents of R1, R2 and R3 positions may directly alter the acidity or pounds with the average rate increase of 2.5-fold for temperature rising
basicity of the reactive site on the molecule, induce steric hindrance, by 10 ◦ C (Mitchell et al., 2014). Similar to the variation trend of Q10 for 4
enhance/reduce the cross-conjugated systems existing in tetracycline tetracycline antibiotics, CTC obtained the highest value of the A value,
antibiotics (Loftin et al., 2008; Ji et al., 2016). Thus, the substituents of indicating the maximum number of effective collisions among activated
tetracyclines can affect their hydrolytic stability. The hydrolysis rate of molecules (Biosic et al., 2017). It is known that the reaction rates and
CTC was the fastest among the four tetracyclines under all pH condi­ activation energies are expected to be affected due to the occurrence of
tions, which may be explained by the steric repulsion effect of the different transition states, and therefore the main degradation pathways
chlorine substituent at R3 position, and the enhanced cross-conjugated of tetracyclines are different under acidic versus basic conditions (Hal­
systems. ling-Sørensen et al., 2002).
The tetracycline antibiotics are amphoteric organic compounds
containing two acidic functional groups (tricarbonyl amide and phenolic 3.3. Products identification and proposed transformation pathways
diketone) and one basic functional group (dimethylamine). Solution pH
is generally considered as one of the most important parameters since The reaction solutions for hydrolysis of tetracyclines were analyzed
different species will dominate in aqueous solutions as a function of the by using UPLC-QTOF-MS in both positive and negative electrospray
pH (Chen et al., 2017). Based on obtained results, acidic conditions ionization modes. Based on the accurate mass-charge ratio (m/z), the
produced longer hydrolysis half-lives for tetracyclines than other con­ molecular formulas were proposed for the potential transformation
ditions. Neutral pH solution was found to be the most favorable for the products of the tetracyclines. Their chemical structures were then
hydrolysis of TC and OTC, which was similar to previous reports (Kang analyzed by the characteristic fragment ions (Xiong et al., 2020). Table 3
et al., 2012; Xuan et al., 2009). It is important to note that since the pH of shows the accurate masses of the protonated molecules, errors, the
natural water, soil and fresh animal feces are nearly neutral, TC and OTC proposed formulas and ring double bond equivalent (RDBE) for tetra­
contamination in the environment could be reduced due to the rapid cyclines and their potential transformation products. The obtained
hydrolysis under neutral pH conditions. On the contrary, alkaline so­ MS/MS fragments and proposed structures of tetracyclines and their
lution accelerated the hydrolytic degradation of CTC and DC, with a main transformation products were summarized in Tables S5-S8. The
shorter half-life of hydrolysis. This was consistent with β-lactam anti­ extracted ion chromatograms and the high-resolution mass spectra for
biotics (Mitchell et al., 2014), florfenicol (Mitchell et al., 2013) and TC, CTC, OTC and DC were presented in Fig. S5 –S6, Fig. S7 –S8,
nitrofurantoin (Biosic et al., 2017). It should be noted that CTC was the Fig. S9–S10 and Fig. S11–S12, respectively. The evolution profiles of
most unstable of the studied tetracyclines, with merely the half-live of tetracyclines and their transformation products during hydrolysis reac­
1.3 h at pH 9.0. Thus, the hydrolysis effect should be considered in any tion process with different pH values were illustrated in Fig. S13-S16.
form of degradation of CTC solutions. Based on these results, the transformation pathways for TC, CTC, OTC
and DC were proposed in Figs. 1–4, respectively.
3.2. Effects of temperature on tetracyclines hydrolysis Acidic and alkaline pH values are associated with the use of tetra­
cyclines in human and veterinary therapy, as acidic pH occurs in the
The hydrolysis of tetracyclines could be significantly enhanced by stomach of humans and animals, and alkaline pH in fresh manure before
elevating temperature. As shown in Table 2 and Fig. S4, the hydrolysis spreading into agricultural fields. In addition, not only hydrolysis rate
half-lives at pH 7.0 for TC, OTC, CTC and DC were 12.6 h, 6.0 h, 1.8 h, but also the formed products are strongly influenced by the pH condi­
28.2 h at 50 ◦ C and 277 h, 158 h, 126 h, 224 h at 10 ◦ C, respectively. The tions. Thus, the anticipated hydrolytic products of tetracyclines were
hydrolysis rate constants were approximately 10 times higher at 50 ◦ C investigated at pH 4.0, 7.0 and 9.0, as shown in Table S9. On the whole,
than those at 25 ◦ C, which indicated hydrolytic stability of tetracyclines the transformation pathways of tetracyclines include epimerization/
was highly temperature dependent. isomerization, dehydration, decarbonylation, hydroxylation and
Temperature effects on hydrolysis of tetracyclines were quantified deamidation.
using the Arrhenius equation, as shown in Text S7. The obtained values
of activation energy (Ea, J mol− 1), frequency factor (A, h− 1) and average

4
S.-F. Zhong et al. Ecotoxicology and Environmental Safety 229 (2022) 113063

Table 3
UPLC-QTOF-MS data for tetracycline antibiotics and their hydrolysis products.
Compounds Retention time Experimental m/z Proposed formula Error MS/MS fragments Probable compounds
(min) [M+H]+ [M+H]+ (ppm)

Tetracycline 2.96 445.1610 C22H25N2O8 0.93 427.1502\410.1257

\154.0500
P445-2 2.54 445.1610 C22H25N2O8 0.91 427.1541\410.1239 4-epi-tetracyclinea(ETC)
\154.0500
P427-1 7.49 427.1505 C22H23N2O7 0.13 410.1231\321.0805 anhydrotetracyclinea(ATC)
\154.0493
P427-2 6.24 427.1505 C22H23N2O7 0.19 410.1238\321.0753 4-epi-
\154.0505 anhydrotetracyclinea(EATC)
P417-1 3.25 417.1645 C21H25N2O7 1.13 400.1378\382.1281
\354.1346
P417-2 2.64 417.1655 C21H25N2O7 0.96 400.1381\382.1276
\354.1287
P461-1 2.29 461.1560 C22H25N2O9 0.2 444.1304\426.1203
\398.1205
P461-2 1.93 461.1560 C22H25N2O9 1.04 444.1288\426.1191
\398.1196
P399 5.66 399.1555 C21H23N2O6 1.17 382.1264\337.2978
\201.0912
P402 2.70 402.1552 C21H24NO7 0.45 338.1421\385.1454
\175.0732
Chlortetracycline 4.41 479.1221 C22H24ClN2O8 1.03 462.0954\444.0837
\154.0497
P479-1 3.76 479.1215 C22H24ClN2O8 0.54 462.0954\444.0842 4-epi-chlortetracyclinea(ECTC)
\154.0498
P479-2 3.34 479.1215 C22H24ClN2O8 0.5 462.0943\196.9983 iso-chlortetracyclinea(iso-CTC)
P479-3 2.92 479.1215 C22H24ClN2O8 0.14 462.0952\196.9962 4-epi-iso-chlortetracycline
P495 3.66 495.1162 C22H24ClN2O9 0.36 478.0894\462.0953
P451-1 4.11 451.1245 C21H24ClN2O7 4.45 434.0982\416.0871
P451-2 3.45 451.1269 C21H24ClN2O7 0.96 434.0996\416.0935
Oxytetracycline 3.18 461.1554 C22H25N2O9 0.13 444.1330\426.1183
P461-1 2.83 461.1553 C22H25N2O9 0.3 444.1244\426.1178 4-epi-Oxytetracyclinea(EOTC)
P461-2 2.61 461.1555 C22H25N2O9 0.13 426.1181\444.1299 iso-Oxytetracycline
P461-3 2.27 461.1554 C22H25N2O9 0.31 444.1291\426.1181 4-epi-iso-Oxytetracycline
P443 4.09 443.1441 C22H23N2O8 0.66 426.1172\408.1091 α-apo-Oxytetracycline
P479 3.37 479.1642 C22H27N2O10 1.88 462.1383\226.0689
P433-1 3.03 433.1605 C21H25N2O8 0.04 416.1337\398.1242
P433-2 1.72 433.1603 C21H25N2O8 0.4 416.1345\398.1215
P418 4.32 418.1492 C21H24NO8 1.74 400.1402\198.0765
\102.0552
Doxycycline 6.02 445.1605 C22H25N2O8 1.08 429.1370\428.1337
\154.0494
DP445-1 3.03 445.1605 C22H25N2O8 0.26 429.1368\428.1335 4-epi-Doxycyclinea(EDC)
DP445-2 3.37 445.1663 C22H25N2O8 0.65 428.1323\429.1366 6-epi-Doxycycline
DP445-3 4.33 445.1663 C21H22N2O8 4.07 168.0636\213.0863
DP417-1 4.86 417.1659 C21H25N2O7 0.01 400.1382\293.0799
DP417-2 3.73 417.1655 C21H25N2O7 0.07 400.1385\293.0820
DP417-3 3.98 417.1653 C21H25N2O7 0.39 400.1390\293.0802
DP402 3.20 402.1554 C21H24NO7 1.68 187.0757\107.0130
\58.0648
DP463-1 5.03 463.1712 C22H27N2O9 0.43 402.1554\170.0454
DP463-2 4.34 463.1714 C22H27N2O9 0.35 445.1603\213.0877
a
The hydrolysis products have been confirmed by standard samples.

3.3.1. Epimerization/isomerization dimethyllamino group (Halling-Sørensen et al., 2002). This was

The epimerization has been reported to be an important initial consistent with the area-time trends of tetracyclines and their trans­
transformation process for tetracyclines, which forms the corresponding formation products (Fig. S13-S14). The 4-epi-chlortetracycline (ECTC)
4-epimers by epimerization at position C-4 (Halling-Sørensen et al., and 4-epi-tetracycline (ETC) were the primary hydrolysis products at the
2002). Thus, the P445-2 (4-epi-tetracycline), P479-1 (4-epi-chlorte­ beginning, while 4-epi-oxytetracycline (EOTC) and 4-epi-doxycycline
tracycline), P461-1 (4-epi-oxytetracycline) and P445-1 (4-epi-doxycy­ (EDC) were not particularly obvious. As shown in Table S9, four epi­
cline) were formed during hydrolysis of TC, CTC, OTC and DC, mers were found clearly at pH 4.0, and 7.0. The abundance of EDC and
respectively. Despite that these products have the same fragment ions as ETC was very low at pH 9.0, indicating that weak acidic and neutral pH
their parent compounds, they could be identified due to the different media favor epimerization of tetracyclines. What’s interesting is that
retention times and fragment ion abundances. In addition, they were ETC was the main transformation product at pH 9.0 despite having a low
further verified by comparing retention time, and mass-to-charge ratios abundance.
(m/z) as well as MS/MS spectra with the authentic reference standards. The epimerization is a reversible reaction, and the isomer and its
The epimerization was the initial dominant transformation process parent compound will reach equilibrium. However, the continuous
for CTC and TC (Loftin et al., 2008). CTC was more likely to have epi­ decline of the parent compound indicated having other transformation
merization at C-4 than TC due to the steric repulsion of C-7 chlorine pathways. The extracted ion chromatograms of CTC (m/z 479.1215) and
atom. However, OTC and DC were more difficult epimerization than TC OTC (m/z 461.1555) were shown in Fig. S7 and Fig. S9. In addition to
since the hydroxyl group at the position C-5 may bind with the the epimers mentioned above, there were also ion peaks with different

5
S.-F. Zhong et al. Ecotoxicology and Environmental Safety 229 (2022) 113063

Fig. 1. Proposed transformation pathways of tetracycline hydrolysis.

Fig. 2. Proposed transformation pathways of chlortetracycline hydrolysis.

retention times and similar fragments, for example, P479-2 (479.1215) in the whole process especially in an alkaline environment. It can be
and P461-2 (461.1555). A previous study found that the hydroxyl group seen from Fig. S14, the abundance of P479-2 showed an increasing trend
at tetracyclines’ C6 may attack the C11 carbonyl group to isomerize the with the reaction time.
respective iso-tetracyclines (Chen and Huang, 2011). Thus, we consid­
ered that they may be iso-chlortetracycline and iso-oxytetracycline, 3.3.2. Dehydration
which can be further epimerized to P479–3 (4-epi-iso-­ The P427-1 (m/z 427.1505) has 18 units difference from its parent
chlortetracycline) and P461–3 (4-epi-iso-oxytetracycline). The compound TC, which is exactly the mass of a water molecule. Thus, it
iso-chlortetracycline had been confirmed by the authentic standard. was suspected to be a tetracycline dehydration product. Liu et al.
Some studies indicated that CTC was particularly labile and prone to (2016b) proposed that the tautomerization of C11-C12 keto-enol can
form iso-CTC, which is a reversible first-order reaction (Diirckheimer, stabilize the second aromatic ring, therefore the most likely reaction site
1975; Halling-Sørensen et al., 2002). In our study, we also found P479-2 for dehydration was at C6. Although the C6 hydroxyl group is neither
(iso-chlortetracycline) was the main hydrolyzate of CTC at the three pH the most basic site in the molecule nor the thermodynamically favored

6
S.-F. Zhong et al. Ecotoxicology and Environmental Safety 229 (2022) 113063

Fig. 3. Proposed transformation pathways of oxytetracycline hydrolysis.

Fig. 4. Proposed transformation pathways of doxycycline hydrolysis.

site of protonation, the proton may migrate from a more basic amine site produce iso-chlortetracycline rather than anhydrochlortetracycline.
to the C6 hydroxyl group, thereby promoting the irreversible dehydra­ Fig. S13 showed that the abundance of P427-1 increased rapidly in the
tion reaction (Vartanian et al., 1998). As shown in Path 2 of Fig. 1, the first 10 days and then slowly decreased, reaching the lowest level on the
tertiary hydroxyl group at position C6 falls off and forms a double bond 40th day, while P399 (m/z = 399.1555) reached a top level on the 40th
at position 5a-6. Due to the influence of this new double bond, the day. The P399 was proposed to be the decarbonylation product of
double bonds on C11, C11a and C12 are transferred, leading to a stable P427-1 by losing 28 Da.
aromatic ring (Vartanian et al., 1998). This reaction is more likely to
occur under acidic conditions. Theoretically, OTC can also be hydro­ 3.3.3. Decarbonylation
lyzed to form anhydro-oxytetracycline (AOTC). Nonetheless, AOTC is The P417-1, P451-1, P433-1 and DP417-1 were observed to have 28
quite unstable in aqueous solution due to the hydroxyl group of C-5. The mass units differ from their parent compounds, i.e. a loss of CO from the
scission of ring B will produce isomeric phthalides α-apo-oxytetracy­ TC, CTC, OTC and DC. The CO could be formed through the rapid
cline. The P443 (α-apo-oxytetracycline, m/z = 443.1441) has been dissociation of CH3CO, which was produced by the C-C bond dissocia­
observed in acidic and neutral solutions, but the intermediate product of tion of acetylacetone (Choudhury and Lin, 1990). Tetracycline struc­
AOTC cannot be captured. In the case of the C6 hydroxyl group of tures have two enolic acetylacetone moieties in ring A and ring C. It was
chlortetracycline, as previously mentioned, it may be more inclined to reported that the decarbonylation is regularly accompanied by

7
S.-F. Zhong et al. Ecotoxicology and Environmental Safety 229 (2022) 113063

a-cleavage (Wagner, 1976), thus the pathway was more likely to occur at reaction time, indicating that 4-epi-oxytetracycline was only a little part
the first enolic acetylacetone moieties of the ring A. Furthermore, Liu of the hydrolysis products.
et al. (2016a) reported that the energy of C1-C12a (sp3 hybridization) is
lower than that of C1-C2 (sp2 hybridization). The tetracyclines are more 3.5. Screening and quantification of tetracyclines and their hydrolysis
likely to go through a cleavage at C1-C12a bond, generating a diradical products in the WWTPs
intermediate. Another diradical compound was produced by the loss of
CO. These two diradical compounds eventually form a decarbonylation 3.5.1. Screening of tetracyclines and their hydrolysis products
byproduct through ring closure (Liu et al., 2016a). The decarbonylation The screening results of tetracyclines and their hydrolysis products in
products of epimers P417–2, P451-2, P433-2 and DP417–2 were also two WWTPs were given in Table S10. The tetracyclines and their hy­
observed and showed very similar product ions spectra with P417–1, drolysis products could not be detected in the influents and effluents, but
P451–1, P433–1 and DP417–1, respectively. 6 target substances were detected in the sludge samples. The parent
compounds tetracycline and oxytetracycline, and the hydrolysis prod­
3.3.4. Hydroxylation ucts P445-2 (4-epi-tetracycline), P461-1 (4-epi-oxytetracycline),
The P461-1, P495, P479 and DP463-1were monohydroxylation P445–1(4-epi-doxycycline) and DP445-3 have high detection frequency
byproducts for TC, CTC, OTC and DC, respectively. Compared with in 8 sludge samples. The chromatogram, mass spectrum and fragment
another available double-bond C2–C3, the C11a–C12 double-bond of ions of the screening results matched with the database spectrum, as
tetracyclines was much more susceptible for free radical to attack shown in Fig. S19-S20. This result confirmed that tetracyclines and their
through electron-withdrawing substituent, and there are adjacent hydrolysis products were mainly adsorbed into sludge, as adsorption
electron-donating functional groups, making C11a-C12 the most was the main way to remove tetracycline antibiotics from the aqueous
possible hydration position (Wang et al., 2011; Chen et al., 2017). The phase in WWTPs (Zhou et al., 2012). It has also been reported that
main fragments correspond to the fragmentation of the hydroxylation tetracycline antibiotics and their degradation products were easily
product through loss of NH3, which was consistent with previous studies adsorbed on clay, organic matter and metal oxides through cation ex­
(Wang et al., 2011; Khan et al., 2010). For path 4 in Figs. 1–4, they might change, surface/cation complexation and bridging hydrophobic distri­
be formed by cycloaddition towards the C11a–C12 double-bond and a bution (Kong et al., 2012; Zhou et al., 2013). The composition profiles of
rearrangement with the hydroxyl at C12 position. Surprisingly, the OTC detected tetracyclines and their hydrolysis products in sludge samples
and DC do not rearrange, it may be associated with the hydroxyl groups were shown in Fig. S21. Based on the percentage of peak area, the
on C5, which limits the rearrangement of hydroxyl group at C12. It is proportion of parent compounds tetracycline and oxytetracycline
noted that all these hydroxylation transformation products can be ranged from 7.46–15.8% and 28.8–41.9%, respectively, while 4-epimer
detected under neutral conditions in this study, which may be related to had a higher amount in sludge samples. The 4-epi-tetracycline had the
pKa2 in the tetracyclines structure. highest proportion of approximately 55% in aerobic tank sludge
(AERS-WWTP2), 4-epi-oxytetracycline had the second-highest propor­
3.3.5. Deamidation tion (42%) in return sludge (RS-WWTP2), and 4-epi-doxycycline can be
The three products P402 (m/z 402.1552), P418 (m/z 418.1492) and screened in all sludge samples. These suggested that the 4-epimer
DP402 (m/z 402.1554) were formed at pH 9.0 with low abundance. constituted a large proportion of hydrolysis products for tetracyclines.
They have 43 mass units’ difference with their parents. They showed an The DP445–3 was frequently screened in the WWTP2 sample, and this
increasing trend in the time course. Based on the exact mass and pro­ was the first time to report the detection of hydrolysis products of
duction spectrum, they were tentatively identified as deamidation doxycycline in WWTPs. As the hydrolysis products had a higher peak
products via the cleavage of the amide bond (43 Da, -CONH2). In the area than their parents, indicating the importance of hydrolysis products
study by Jeong et al. (2010), the deamidation products for tetracyclines in the environmental fate of tetracyclines.
were generated by e−aq1 reaction with the amide group at the position of
ring A. They proposed a scheme that an ion radical can be formed with 3.5.2. Quantification of tetracyclines and their hydrolysis products
the addition of the e−aq1 in the keto-tautomer, and then followed by Among the above 6 tentatively screened tetracyclines and their hy­
drolysis products, 5 of them were further confirmed with available
elimination of formamide radical and enol product. These hydrolysis
authentic standards, and their concentrations were quantificationally
products had rarely reported before, since deamidation is only possible
analyzed using UPLC-MS/MS in these two WWTPs, as shown in
under acidic or alkaline conditions. Besides, high-resolution mass
Table S11. The concentration of tetracyclines and their hydrolysis
spectrometry improves the sensitivity of the detection of low-abundance
products ranged from 21.2 ng/g to 1418 ng/g and 15.8 ng/g to
deamidation substances.
318.3 ng/g in WWTP1 and WWTP2, respectively. By comparing the
concentration of quantified hydrolysis product with their parents of 2
3.4. Mass balance analysis WWTPs (Fig. S22.), 50% of hydrolysis product had an average concen­
tration higher than their parents, mainly 4-epi-tetracycline. Especially,
The mass balance analysis for the hydrolysis of tetracyclines was 4-epi-tetracycline had a concentration ratio of Transformation Product/
carried out in 5 mg/L initial concentration at pH 7.0 phosphate buffer Parent > 4 in dehydrated sludge and activated sludge samples from
solution, as shown in Fig. S17. Only products P445-1, P427-1, P427-2, WWTP1. According to the above hydrolysis mass balance analysis, the 4-
P479-1, P479-2, P461-1 and DP445-1 could be quantified by authentic epi-tetracycline (P445-2) was formed up to a ratio of 0.5 relative to its
reference standards, and other TPs cannot be established. Therefore, the parent TC concentration, indicating that 4-epi-tetracycline may be
total mass balance values decreased with the increasing hydrolysis time. produced by other pathways besides the hydrolysis reaction. In any case,
The main hydrolysis product for CTC was P479-2 with a yield of 60–80% such a high concentration of hydrolysis products confirmed the neces­
throughout the hydrolysis process. The CTC (5 mg/L) was almost sity to include them in the toxicity assessment.
completely converted into P479-2 (3.9 mg/L) on day 4. The mass bal­
ance experiment for TC showed that the total mass of TC and its hy­ 3.6. Toxicity evaluation
drolysis products P427-1, P427-2 and P445-2 accounted for more than
94% during 6 days of hydrolysis. The P445-2 was formed up to a ratio of The toxicity of TC, CTC, OTC, DC and their hydrolysis products for
0.5 relative to its parent TC concentration, indicating P445-2 was the fish, daphnia and green algae were predicted by USEPA ECOSAR
primary hydrolysis product for TC. The mass balance values of OTC and (v1.11). As shown in Table S12, Daphnia was more sensitive to tetra­
its available transformation products decreased significantly with the cyclines, both acute and chronic toxicity values were much lower than

8
S.-F. Zhong et al. Ecotoxicology and Environmental Safety 229 (2022) 113063

those of fish and green algae. All epimers had the same toxicity pre­ CRediT authorship contribution statement
diction as to their parent compounds, while the isomer had a lower acute
and chronic concentration value, especially for iso-chlortetracycline. Shao-Fen Zhong: Investigation, Data curation, Writing – original
The toxicity ratio was calculated by dividing the LC50 of parent com­ draft. Bin Yang: Laboratory Supervision, Writing – review & editing.
pound and its hydrolysis products (Fig. S23A). The 16 hydrolysis Qian Xiong: Investigation. Wen-Wen Cai: Data curation. Zheng-Gang
products had a toxicity ratio of more than 1, meaning they had higher Lan: Data analysis. Guang-Guo Ying: Conceptualization, Supervision,
predicted toxicity than their parent compounds. For example, the P443 Writing – review & editing, Funding acquisition.
was identified as a hydrolysis product formed by breaking ring B of the
unstable dehydration intermediate of oxytetracycline in our study, with
Declaration of Competing Interest
the highest toxicity ratio of 117. However no study can be found on its
occurrence and toxicity in the environment. Consistent with the previ­
The authors declare that they have no known competing financial
ous studies (Kühne et al., 2001), the dehydrated products such as
interests or personal relationships that could have appeared to influence
anhydrotetracycline (P427-1) have higher predicted toxicity (toxicity
the work reported in this paper.
ratio = 16) than that of tetracycline. That is probably due to the opening
of one ring to give iso-tetracycline, or aromatization of additional rings,
Acknowledgements
resulting in the alteration of basic skeleton of the tetracycline molecule.
In addition, we should consider the combined toxic effects of the
The authors would like to acknowledge the financial support from
mixture. Although the actual environmental concentration of trans­
the National Natural Science Foundation of China (U1701242,
formation product is lower, the mixture may also cause toxic effects.
42030703 and 41877358), and National Key Research Program from the
The computational toxicology prediction for tetracyclines and their
Ministry of Science and Technology of China (2020YFC1806900). Dr.
hydrolysis products is a preliminary toxicity evaluation, which needs to
Bin Yang acknowledges the Pearl River Talent Plan of Guangdong
be further verified and confirmed by aquatic toxicity tests. Thus, the
Province, China (2017GC010244) and Scientific Research and Tech­
gene recombinant luminescent bacteria (E.coli HB101 pUCD607) was
nology Development Program of Guangxi, China (2018AB36018).
employed to assess the toxicity change during the tetracyclines hydro­
lysis process, and the toxicity of 4 tetracyclines and their 7 trans­
formation products with available standards. As illustrated in Fig. S23B, Appendix A. Supporting information
the hydrolysis process could reduce the toxicity of CTC by decreasing
24% of the inhibition ratio. This suggested that CTC was eventually Supplementary data associated with this article can be found in the
transformed into products less toxic for E. coli HB101 pUCD607. On the online version at doi:10.1016/j.ecoenv.2021.113063.
other hand, the bioluminescence inhibition ratio of TC, OTC and DC
increased after 120 h hydrolysis. This result may be due to the higher References
toxicity ratio of the hydrolysis products P443 and P427–1 etc. The EC50
Biosic, M., Skoric, I., Beganovic, J., Babic, S., 2017. Nitrofurantoin hydrolytic
values was calculated by using the dose-response fitting tool in Graph­ degradation in the environment. Chemosphere 186, 660–668.
Pad Prism 5 software and provided in Table S13. The EC50 values ob­ Bletsou, A.A., Jeon, J., Hollender, J., Archontaki, E., Thomaidis, N.S., 2015. Targeted and
tained for TC, OTC, CTC and DC were 107 mg/L, 707 mg/L, 173 mg/L non-targeted liquid chromatography-mass spectrometric workflows for
identification of transformation products of emerging pollutants in the aquatic
and 447 mg/L, respectively. Among the 7 transformation products, 5 environment. TrAC-Trend Anal. Chem. 66, 32–44.
transformation products had lower EC50 values than their parents, Chen, W.R., Huang, C.H., 2011. Transformation kinetics and pathways of tetracycline
indicating increased acute toxicity to the luminescent bacteria from the antibiotics with manganese oxide. Environ. Pollut. 159, 1092–1100.
Chen, Y.Y., Ma, Y.L., Yang, J., Wang, L.Q., Lv, J.M., Ren, C.J., 2017. Aqueous tetracycline
transformation products. The ATC had the strongest toxicity toward the
degradation by H2O2 alone: removal and transformation pathway. Chem. Eng. J.
luminescent bacteria, with an EC50 of 11.4 mg/L, which was consistent 307, 15–23.
with the previous study by algal cells (Xu et al., 2019). Consequently, Choudhury, T.K., Lin, M.C., 1990. Homogeneous pyrolysis of acetylacetone at high
temperatures in shock waves. Int. J. Kinet. 20, 491–504.
more attention should be paid to the environmental risk of the trans­
Christian, T., Schneider, R.J., Färber, H.A., Skutlarek, D., Meyer, M.T., Goldbach, H.E.,
formation products of tetracyclines. 2003. Determination of antibiotic residues in manure, soil, and surface waters. Acta
Hydrochim. Hydrobiol. 31, 36–44.
4. Conclusions Dai, Y., Liu, M., Li, J., Yang, S., Sun, Y., Sun, Q., Wang, W., Lu, L., Zhang, K., Xu, J.,
Zheng, W., Hu, Z., Yang, Y., Gao, Y., Liu, Z., 2019. A review on pollution situation
and treatment methods of tetracycline in groundwater. Sep. Sci. Technol. 55,
The investigated four tetracyclines (TC, CTC, OTC and DC) were 1005–1021.
found to be more stable in acidic solution than that of the neutral and Diirckheimer, W., 1975. Tetracyclines: chemistry, biochemistry, and structure-activity
relations. Angew. Chem. Int. Ed. 14, 721–734.
alkaline conditions. High temperature favored tetracyclines hydrolysis EPA, 2008. Fate, Transport and transformation test guidelines OPPTS 835.2120.
at pH 7.0 solutions. A total of 9, 6, 8 and 9 hydrolysis products were Escher, B.I., Fenner, K., 2011. Recent advances in environmental risk assessment of
detected for TC, CTC, OTC and DC, respectively. They can be formed by transformation products. Environ. Sci. Technol. 45, 3835–3847.
Fang, G.Z., Hu, L.X., Huang, G.Y., Zhao, J.L., Ying, G.G., 2020. The application of gene
epimerization/isomerization, dehydration, decarbonylation, hydroxyl­ recombinant luminescent bacteria to environmental sample toxicity test. J. South
ation and deamidation. The epimerization/isomerization and dehydra­ China Norm. Univ. Nat. Sci. Ed. 52, 60–67 (In Chinese).
tion were the main hydrolysis pathways. The computational toxicology Gao, P., Munir, M., Xagoraraki, I., 2012. Correlation of tetracycline and sulfonamide
antibiotics with corresponding resistance genes and resistant bacteria in a
results showed that 16 hydrolysis products had higher toxicity than their
conventional municipal wastewater treatment plant. Sci. Total Environ. 421–422,
parent compounds, with toxicity ratio ranging from 1.7 to 117. The 173–183.
hydrolysis products increased the bioluminescence toxicity of TC, OTC Halling-Sørensen, B., Sengeløv, G., Tjørnelund, J., 2002. Toxicity of tetracyclines and
tetracycline degradation products to environmentally relevant bacteria, including
and DC solutions. Screening of two WWTPs showed the presence of 2
selected tetracycline-resistant bacteria. Arch. Environ. Contam. Toxicol. 42,
tetracyclines and 4 hydrolysis products in the sludge samples, with 263–271.
concentrations ranging from 15.8 ng/g to 1418 ng/g. The 4-epi-tetracy­ Halling-Sørensen, B., Lykkeberg, A., Ingerslev, F., Blackwell, P., Tjørnelund, J., 2003.
cline had an average concentration higher than its parent. Future Characterisation of the abiotic degradation pathways of oxytetracyclines in soil
interstitial water using LC–MS–MS. Chemosphere 50, 1331–1342.
monitoring and risk assessment studies on the environmental fate of Jeong, J., Song, W., Cooper, W.J., Jung, J., Greaves, J., 2010. Degradation of tetracycline
tetracyclines should also be taken into account the potential presence of antibiotics: mechanisms and kinetic studies for advanced oxidation/reduction
transformation products resulting from the hydrolysis process. processes. Chemosphere 78 (5), 533–540.
Ji, Y., Shi, Y., Dong, W., Wen, X., Jiang, M., Lu, J., 2016. Thermo-activated persulfate
oxidation system for tetracycline antibiotics degradation in aqueous solution. Chem.
Eng. J. 298, 225–233.

9
S.-F. Zhong et al. Ecotoxicology and Environmental Safety 229 (2022) 113063

Jia, A., Xiao, Y., Hu, J., Asami, M., Kunikane, S., 2009. Simultaneous determination of Søeborg, T., Ingerslev, F., Halling-Sørensen, B., 2004. Chemical stability of
tetracyclines and their degradation products in environmental waters by liquid chlortetracycline and chlortetracycline degradation products and epimers in soil
chromatography-electrospray tandem mass spectrometry. J. Chromatogr. A 1216, interstitial water. Chemosphere 57, 1515–1524.
4655–4662. Topal, M., Topal, E.I.A., 2015. Occurrence and fate of tetracycline and degradation
Kang, H.J., Lim, M.Y., Kwon, J.H., 2012. Effects of adsorption onto silica sand particles productsin municipal biological wastewater treatment plantand transport of them in
on the hydrolysis of tetracycline antibiotics. J. Environ. Monit. 14, 1853–1859. surface water. Environ. Monit. Assess. 187, 1–9.
Khan, M.H., Bae, H., Jung, J.Y., 2010. Tetracycline degradation by ozonation in the Topal, M., Şenel, G.U., Öbek, E., Topal, E.I.A., 2016. Investigation of relationships
aqueous phase: proposed degradation intermediates and pathway. J. Hazard. Mater. between removals of tetracycline and degradation products and physicochemical
181, 659–665. parameters in municipal wastewater treatment plant. J. Environ. Manag. 173, 1–9.
Kim, Y.H., Heinze, T.M., Beger, R., Pothuluri, J.V., Cerniglia, C.E., 2004. A kinetic study Vartanian, V.H., Goolsby, B., Brodbelt, J.S., 1998. Identification of tetracycline
on the degradation of erythromycin A in aqueous solution. Int. J. Pharm. 271, 63–76. antibiotics by electrospray ionization in a quadrupole ion trap. J. Am. Soc. Mass
Kong, W., Li, C., Dolhi, J.M., Li, S., He, J., Qiao, M., 2012. Characteristics of Spectr. 9, 1089–1098.
oxytetracycline sorption and potential bioavailability in soils with various physical- Wagner, P.J., 1976. Chemistry of excited triplet organic carbonyl compounds. Triplet
chemical properties. Chemosphere 87, 542–548. States III 1–52.
Kühne, M., Hamscher, G., Körner, U., Schedl, D., Wenzel, S., 2001. Formation of Wang, J., Zhi, D., Zhou, H., He, X., Zhang, D., 2018. Evaluating tetracycline degradation
anhydrotetracycline during a high-temperature treatment of animal-derived feed pathway and intermediate toxicity during the electrochemical oxidation over a Ti/
contaminated with tetracycline. Food Chem. 75, 423–429. Ti4O7 anode. Water Res. 137, 324–334.
Kummerer, K., 2009. Antibiotics in the aquatic environment - a review - part I. Wang, X., Yu, N., Yang, J., Jin, L., Guo, H., Shi, W., Zhang, X., Yang, L., Yu, H., Wei, S.,
Chemosphere 75, 417–434. 2020. Suspect and non-target screening of pesticides and pharmaceuticals
Li, D., Yang, M., Hu, J., Zhang, Y., Chang, H., Jin, F., 2008. Determination of penicillin G transformation products in wastewater using QTOF-MS. Environ. Int. 137, 105599.
and its degradation products in a penicillin production wastewater treatment plant Wang, Y., Zhang, H., Zhang, J., Lu, C., Huang, Q., Wu, J., Liu, F., 2011. Degradation of
and the receiving river. Water Res. 42, 307–317. tetracycline in aqueous media by ozonation in an internal loop-lift reactor.
Liu, Y., He, X., Fu, Y., Dionysiou, D.D., 2016a. Degradation kinetics and mechanism of J. Hazard. Mater. 192, 35–43.
oxytetracycline by hydroxyl radical-based advanced oxidation processes. Chem. Eng. Wang, Z., Chang, Q., Kang, J., Cao, Y., Ge, N., Fan, C., Pang, G.F., 2015b. Screening and
J. 284, 1317–1327. identification strategy for 317 pesticides in fruits and vegetables by liquid
Liu, Y., He, X., Fu, Y., Dionysiou, D.D., 2016b. Kinetics and mechanism investigation on chromatography-quadrupole time-of-flight high resolution mass spectrometry. Anal.
the destruction of oxytetracycline by UV-254nm activation of persulfate. J. Hazard. Methods 7, 6385–6402.
Mater. 305, 229–239. Xiong, Q., Liu, Y.S., Hu, L.X., Shi, Z.Q., Cai, W.W., He, L.Y., Ying, G.G., 2020. Co-
Loftin, K.A., Adams, C.D., Meyer, M.T., Surampalli, R., 2008. Effects of ionic strength, metabolism of sulfamethoxazole by a freshwater microalga Chlorella pyrenoidosa.
temperature, and pH on degradation of selected antibiotics. J. Environ. Qual. 37, Water Res. 175, 115656.
378–386. Xu, D., Xiao, Y., Pan, H., Mei, Y., 2019. Toxic effects of tetracycline and its degradation
Mitchell, S.M., Ullman, J.L., Teel, A.L., Watts, R.J., Frear, C., 2013. The effects of the products on freshwater green algae. Ecotoxicol. Environ. Saf. 174, 43–47.
antibiotics ampicillin, florfenicol, sulfamethazine, and tylosin on biogas production Xuan, R., Arisi, L., Wang, Q., Yates, S.R., Biswas, K.C., 2009. Hydrolysis and photolysis of
and their degradation efficiency during anaerobic digestion. Bioresour. Technol. oxytetracycline in aqueous solution. J. Environ. Sci. Health B 45, 73–81.
149, 244–252. Yang, B., Xu, C., Kookana, R.S., Williams, M., Du, J., Ying, G., Gu, F., 2018. Aqueous
Mitchell, S.M., Ullman, J.L., Teel, A.L., Watts, R.J., 2014. pH and temperature effects on chlorination of benzodiazepines diazepam and oxazepam: kinetics, transformation
the hydrolysis of three beta-lactam antibiotics: ampicillin, cefalotin and cefoxitin. products and reaction pathways. Chem. Eng. J. 354, 1100–1109.
Sci. Total. Environ. 466–467, 547–555. Yang, S., Liu, Y., Shen, C., Li, F., Yang, B., Huang, M., Yang, M., Wang, Z., Sand, W.,
Oberoi, A.S., Jia, Y., Zhang, H., Khanal, S.K., Lu, H., 2019. Insights into the fate and 2020. . Rapid decontamination of tetracycline hydrolysis product using
removal of antibiotics in engineered biological treatment systems: a critical review. electrochemical CNT filter: mechanism, impacting factors and pathways.
Environ. Sci. Technol. 53, 7234–7264. Chemosphere 244, 125525.
OECD, 2004. OECD guidelines for the testing of chemicals, Test No. 111: Hydrolysis as a Yi, Q., Gao, Y., Zhang, H., Zhang, H., Zhang, Y., Yang, M., 2016. Establishment of a
Function of pH. 〈https://www.oecd-ilibrary.org/environment/test-no-111-hydr pretreatment method for tetracycline production wastewater using enhanced
olysis-as-a-function-of-ph_9789264069701-en〉. hydrolysis. Chem. Eng. J. 300, 139–145.
Pan, S.F., Zhu, M.P., Chen, J.P., Yuan, Z.H., Zhong, L.B., Zheng, Y.M., 2015. Separation Zhang, H., Xie, H., Chen, J., Zhang, S., 2015. Prediction of hydrolysis pathways and
of tetracycline from wastewater using forward osmosis process with thin film kinetics for antibiotics under environmental pH conditions: a quantum chemical
composite membrane-Implications for antibiotics recovery. Sep. Purif. Technol. 153, study on cephradine. Environ. Sci. Technol. 49, 1552–1558.
76–83. Zhou, L.J., Ying, G.G., Liu, S., Zhao, J.L., Chen, F., Zhang, R.Q., Peng, F.Q., Zhang, Q.Q.,
Pang, G.F., Fan, C.L., Chang, Q.Y., Li, J.X., Kang, J., Lu, M.L., 2018. Screening of 485 2012. Simultaneous determination of human and veterinary antibiotics in various
pesticide residues in fruits and vegetables by liquid chromatography-quadrupole- environmental matrices by rapid resolution liquid chromatography-electrospray
time-of-flight mass spectrometry based on TOF accurate mass database and QTOF ionization tandem mass spectrometry. J. Chromatogr. A 1244, 123–138.
spectrum library. J. AOAC Int. 101, 1156–1182. Zhou, L.J., Ying, G.G., Liu, S., Zhao, J.L., Yang, B., Chen, Z.F., Lai, H.J., 2013. Occurrence
Reardon, S., 2014. Antibiotic resistancesweeping developing world. Nature 509, and fate of eleven classes of antibiotics in two typical wastewater treatment plants in
141–142. South China. Sci. Total Environ. 452, 365–376.

10