Photochemical Transformation of Aminoglycoside Antibiotics in

Feb 17, 2016 - A control experiment of centrifuging 5.2 mg of C L–1 SRNOM without aminoglycosides was conducted, showing that most of SRNOM was reta...
0 downloads 10 Views 800KB Size
Subscriber access provided by - Access paid by the | UCSB Libraries

Article

Photochemical transformation of aminoglycoside antibiotics in simulated natural waters Rui Li, Cen Zhao, Bo Yao, Dan Li, Shuwen Yan, Kevin E. O'Shea, and Weihua Song Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b05234 • Publication Date (Web): 17 Feb 2016 Downloaded from http://pubs.acs.org on February 18, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 28

Environmental Science & Technology

1 2 3

Photochemical transformation of aminoglycoside antibiotics in

4

simulated natural waters

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

Rui Li1, Cen Zhao2, Bo Yao1, Dan Li1, Shuwen Yan1, Kevin E. O’Shea2, and Weihua Song1,*

1

Department of Environmental Science & Engineering, Fudan University, Shanghai, 200433, China

2

Department of Chemistry & Biochemistry, Florida International University, Miami, FL, 33199, United States

*corresponding author: email: [email protected] Tel: (+86)15821951698

Prepared for Environ. Sci. & Technol.

32

1

ACS Paragon Plus Environment

Environmental Science & Technology

33 34

Abstract Aminoglycoside antibiotics are widely used in human therapy and veterinary medicine. We

35

report herein a detailed study on natural organic matter (NOM) photosensitized degradation of

36

aminoglycosides in aqueous media under simulated solar irradiation. It appears that the direct

37

reaction of 3NOM* with aminoglycosides is minor. The contributions of reactive oxygen species

38

(ROS) in the bulk solutions, are also unimportant as found by an assessment based on steady state

39

concentrations and bimolecular reaction rate constants in a homogeneous reaction model. The

40

inhibition of the photodegradation by isopropamide is rationalized through competitive sorption with

41

aminoglycosides on the NOM surface, while the addition of isopropanol negligibly affects

42

degradation because it quenches HO• in the bulk solution, but not HO• localized on the NOM surface

43

where aminoglycosides reside. Therefore sorption-enhanced photo transformation mechanism is

44

proposed. The sorption of aminoglycosides on NOM follows dual mode model involving Langmuir-

45

and linear-isotherms. The steady state concentration of HO• on the surface of NOM is calculated as

46

10-14 M, 2 orders of magnitude higher than in the bulk solution. This fundamental information is

47

important in the assessment of the fate and transport of aminoglycosides in aqueous environments.

48

2

ACS Paragon Plus Environment

Page 2 of 28

Page 3 of 28

49 50

Environmental Science & Technology

Introduction The occurrence, transformation, and risk of antibiotics in the aquatic ecosystems are emerging

51

environmental issues, and draw great attention from scientists, engineers and public.1-3 Among

52

antibiotics, aminoglycosides are an important group that consists of several aminosugar moieties

53

linked to an aminocyclitol component in their molecule structures. They can bind to the 16S rRNA in

54

the small ribosomal subunit of bacteria, causing serious interference in the translation process and

55

leading to bacterial death.4-6 Consequently, they exhibit effective antibacterial activity with the

56

treatment of both gram-positive and gram-negative infections.7, 8 Aminoglycoside antibiotics have

57

been widely used in human therapy against diseases and as veterinary medicines to promote growth

58

and prevent infectious diseases.9, 10 Although the possible interactions of these pollutants with living

59

organisms in the environment are not well documented, it is highly likely that the aminoglycosides

60

exert toxic effects on algae and invertebrates. More important, they can favor the development of

61

multi-resistant strains in microorganism.11-14

62

A limited number of studies have reported on the detection and transformation of

63

aminoglycosides in the aquatic environments, probably due to lack of suitable analytical methods.

64

Hu et al. employed the 9-fluorenylmethyl chloroformate (FMOC-Cl) derivatization method to

65

identify 98.2 ± 10.0 ppb aminoglycosides in the waste effluent from a pharmaceutical manufactory.15

66

In the food chemistry, aminoglycosides residues have been widely reported in animal derived food

67

such as milk, meat, egg and so on.16, 17 New analytical methods are needed to identify trace amounts

68

of aminoglycosides in the aquatic environments.

69

Previous studies have demonstrated that urban sewage treatment systems may not completely

70

remove antibiotics.18, 19 When antibiotics are utilized in livestock, they may also enter natural waters

71

directly or through contaminated manure utilization. A variety of treatment methods have been

72

attempted to remove these contaminants, including membrane treatment, ozonation process, and

73

adsorption technologies. Photodegradation has also been explored as a feasible method to treat 3

ACS Paragon Plus Environment

Environmental Science & Technology

74

pharmaceutical contaminants in natural and engineered systems.20-22 The direct photodegradation of

75

contaminants requires the overlap of absorption spectra within the irradiation spectrum.23, 24 The

76

indirect photodegradation can be promoted via photosensitizers. Natural organic matter (NOM) can

77

function as an effective photosensitizer by absorbing sunlight and reach excited states. Then the

78

excited states of NOM (3NOM*) can react with dissolved oxygen and form a series of reactive

79

oxygen species (ROS), such as singlet oxygen (1O2), superoxide radical (O2•-), hydroxyl radical (HO•)

80

and so on.25-27 Organic contaminants can subsequently be degraded by these ROS or react by transfer

81

of electrons or energy with excited states of photosensitizers.28-30 Previous studies have showed that

82

organic compounds can be adsorbed onto NOM, seriously affecting the rate of contaminant

83

degradation, biological uptake ability, evaporation property and transformation in sediments.31

84

Consequently, the photosensitized degradation process may be altered due to the interaction with

85

NOM. Previous studies32, 33 revealed sorption-enhanced indirect photo-transformation of cationic

86

histidine and histamine in NOM enriched solutions. The microheterogeneous distribution of 1O2 in

87

irradiated NOM solution leads to overall enhanced reaction rates.34 To our best knowledge, we report

88

herein the first detailed study focused on the microheterogeneous distribution of HO• on the NOM

89

surface.

90

A series of aminoglycosides, streptomycin, kanamycin, gentamycin, tobramycin and amikacin

91

sharing common organic functional groups, are selected as target compounds in this study.

92

Streptomycin is the first aminoglycoside antibiotic. Discovered in 1940s, it is still used due to low

93

cost and reliable activity.6, 35 All remaining aminoglycoside antibiotics are widely used in human and

94

veterinary medicine.

95

In this study, a liquid chromatography tandem mass spectrometry (LC-MS/MS) method

96

without derivatization was developed to detect the aminoglycosides. A series of experiments were

97

conducted to explore the roles of ROS in the photosensitized degradation. Our results suggested

98

aminoglycoside antibiotics are adsorbed on NOM and isopropamide was employed to confirm the 4

ACS Paragon Plus Environment

Page 4 of 28

Page 5 of 28

99

Environmental Science & Technology

role of sorption. Our results indicate that the reaction of aminoglycosides with HO• on the surface of

100

NOM, defined as micro-heterogeneous reaction, is critical in photodegradation. In addition the

101

cytotoxicity of the products was assessed and demonstrated that the NOM induced photodegradation

102

process can effectively remove the toxicity of aminoglycosides.

103

Materials and methods

104

Chemicals. Streptomycin (≥ 95%), kanamycin (≥ 94%), tobramycin (≥ 94%) were purchased

105

from TCI (Tokyo Chemical Industry), gentamycin and amikacin (≥ 98%) were purchased from BBI

106

(Bio Basic Inc.), structures are shown in Scheme 1. Furfuryl alcohol (FFA, ≥ 99%),

107

furan-2-carbaldehyde (FAD, ≥ 99%), Rose Bengal (RB), terephthalic acid (TA),

108

3-methoxyacetophenone, trimethylphenol, isopropanol (IPA), isopropamide, and formic acid were

109

purchased from Sigma-Aldrich. 2-hydroxyl terephthalic acid (2HTA) was synthesized using a

110

literature method.36 Deuterium oxide (D2O, 99.9%) was obtained from Cambridge Isotope

111

Laboratories. Methanol and Acetonitrile (J.T. Baker) were of HPLC grade. Compressed nitrogen and

112

oxygen were purchased from Fudan Spring Inc. Suwannee river natural organic matter (SRNOM)

113

was obtained from the International Humic Substances Society (IHSS). All the prepared solutions

114

contained 2.0 mM phosphate buffer and were adjusted to pH 7.0 using HCl or NaOH.

115

(Insert Scheme 1)

116

LC-MS/MS method for aminoglycosides. The concentrations of aminoglycosides were

117

determined by LC-triple Quadrupole MS/MS with electrospray ionization source (Agilent

118

1290-6430). A sample volume of 15.0 µL was injected into the column (Shiseido, Capcell Pak ST,

119

2.0×150 mm) thermostated at 40 oC. The mobile phase was 92% H2O and 8% acetonitrile, both

120

containing 0.2% formic acid. The flow rate was 0.3 mL min-1. Mass spectra were obtained in positive

121

ion mode and MS parameters were optimized as follows: nebulizer nitrogen gas flow rate (11 L min-1)

122

and pressure (35 psi), capillary voltage (4000 V). The details regarding precursor ions and product

123

ions are shown in Table S1, Supporting Information (SI). 5

ACS Paragon Plus Environment

Environmental Science & Technology

124

Page 6 of 28

Photolysis experiments. To study the direct photolysis, aminoglycoside solutions were

125

prepared in ultrapure water with 2.0 mM phosphate buffer (pH = 7.0), and irradiated in a solar

126

simulator (Suntest XLS+) with a 1700W Xenon lamp equipped with 290 nm cut off filter. The

127

absolute irradiance spectrum of the solar simulator was recorded using a spectrometer (USB-4000,

128

Ocean Optics Inc.), as illustrated in Figure S1, SI. A temperature control unit (Suncool®) fixed the

129

temperature at 25 °C. To study the indirect photolysis, aminoglycosides (3.0 µM) containing 5.2 mg

130

C L-1 SRNOM were exposed to the solar simulator as above. The TOC contents of the solutions were

131

acquired using a TOC analyzer (Shimadzu, TOC − CPH/CN). Experiments were performed in D2O

132

to explore the role of 1O2. The isopropanol or isopropamide were employed as HO• scavengers. The

133

effect of the triplet excited state of NOM was studied in the solutions purged with nitrogen, or

134

oxygen gas.

135

The bimolecular reaction rate constants of aminoglycosides with HO•. Bimolecular

136

reaction rate constants of HO• with five aminoglycosides were determined using electron pulse

137

radiolysis. This task was performed at the Notre Dame Radiation Laboratory in US with the 8-MeV

138

Titan Beta model TBS-8/16-1S linear accelerator, which has been described elsewhere.37, 38

139

Dosimetry was performed using N2O-saturated, 1.00 × 10-2 M KSCN solutions at λ= 472 nm, with

140

average doses of 3-5 Gy per 2-3 ns pulse. All experimental data were determined by averaging 8-10

141

replicate pulses using the continuous flow mode of the instrument.

142

Since the reaction intermediates of HO• with aminoglycosides do not show UV-vis absorption,

143

the intermediate buildup method could not be applied for rate constant measurement. Therefore the

144

HO• reaction rate constants with aminoglycosides were determined using SCN- competition kinetics

145

based on the monitoring of (SCN)2•- absorption. Eqs 1 and 2 show the respective reactions of

146

aminoglycosides and SCN- with HO•.

147

HO•

+

aminoglycosides

148

HO•

+

SCN-(+ SCN-)



→ 



H2 O

+ intermediate

(1)

HO-

+ (SCN)2•-

(2)

6

ACS Paragon Plus Environment

Page 7 of 28

Environmental Science & Technology

149

This competition can be analyzed to give the expression:

150

[(SCN) •2- ]0 k [Aminoglycosides ] = 1+ 1 •[(SCN) 2 ] k 2 [SCN]

(3)

151

Where [(SCN)2•-]0 is the absorbance of the transient at 472 nm (A472nm) when only SCN- is present.

152

Figure 1 shows that the absorption intensity of (SCN)2•- was gradually reduced as aminoglycosides

153

concentration increased, implying competitive involvement of aminoglycosides in HO• reaction with

154

SCN-. A plot of eq3 ([(SCN) 2•-]0/[(SCN) 2•-] vs. [Aminoglycosides]/[SCN-]) shows a linear

155

correlation with a slope of k1/k2 (Figure 1, Insert). Using k2 (HO• + SCN-) = 1.16 × 1010 M-1 s-1, the rate

156

constant for the HO• reaction with aminoglycosides was calculated as k1. The HO• reaction rate

157

constants of all five aminoglycosides are in the range of 109 M-1 s-1 and the details are summarized in

158

Table S2, SI. These hydroxyl radical reaction rate constants also provide fundamental information

159

necessary to apply advanced oxidation processes (AOPs) to the treatment of aqueous aminoglycoside

160

wastes. (Insert Figure 1)

161 162

The bimolecular reaction rate constants of aminoglycosides with 1O2. Irradiated solutions

163

of RB in distilled water generate 1O2.29 To determine the 1O2 bimolecular reaction rate constants,

164

solutions containing 3.0 µM of the aminoglycosides, 1.6 mM FAD and 0.10 mM RB in pH 7.0

165

phosphate buffer were irradiated in a solar simulator. Aliquots were removed at time intervals for

166

aminoglycosides and FAD analysis using LC-MS/MS and LC-DAD respectively, the details are

167

shown in Text S1, SI. The 1O2 reaction rate constants of all five aminoglycosides are reported in

168

Table S2 of SI.

169

The steady state concentrations of ROS under simulated solar irradiation. To study the

170

steady-state concentrations of HO• and 1O2 in bulk solutions, TA and FFA were employed as

171

chemical probes. For HO•, varied concentrations of TA (3.0 to 200 µM) were employed to trap the

172

radical and produce 2HTA, which was measured by HPLC-fluorescence (λexcitation= 315 nm, λemission=

173

425 nm). 39 The formation yield of 2HTA was estimated as 0.28 based on the literature.40 The 7

ACS Paragon Plus Environment

Environmental Science & Technology

174

experimental details and calculations for steady-state concentration of HO• are presented in the Text

175

S2 of SI. For 1O2, varied concentrations of FFA were engaged to trap 1O2. The loss of FFA was

176

measured by HPLC-UV.41 The details are shown in Text S2, SI.

177

Sorption of Aminoglycosides on the NOM. The sorption experiments were performed using

178

Molecular weight cut off (MWCO) filter method.42 Aliquots of 4.0 mL aminoglycosides solutions

179

with 5.2 mg C L-1 SRNOM were transferred to centrifuge tubes fitted with a 3,000 MWCO filter

180

(Millipore Inc.), then centrifuged at 3500 g for 5 mins (Beckman Coulter, Avanti J-26 XPI). Only a

181

part of the solutions passed through the MWCO filter membrane. The control experiment of

182

centrifuging 5.2 mg C L-1 SRNOM without aminoglycosides was conducted, showing that most of

183

SRNOM was retained by the MWCO filter, based on the measurement of UV-vis spectra (Agilent,

184

Cary 60) that are shown in Figure S2 of SI. Hence, the MWCO filter is a suitable method to explore

185

the adsorption behavior of aminoglycosides on the NOM. Both aliquots of supernatants (sorbed) and

186

the effluents (free) were collected for direct LC-MS/MS test. No NOM matrix effect has been

187

observed during the analysis.

188

Cytotoxicity assay. The toxicity of streptomycin and aliquots of the irradiated solutions were

189

assayed through the inhibition ratios of bioluminescence of the marine bacterium Vibrio fischeri. The

190

5 mg freeze dried bacteria were re-suspended in 5 mL of supplemented seawater complete medium

191

(SSWC medium, Peptone from casein [5% (w/v)], Yeast extract [0.5% (w/v)], Glycerol [0.3% (v/v)],

192

NaCl [3% (w/v)], NaH2PO4 [44.2 mM], K2HPO4 [12.1 mM], MgSO4⋅7 H2O [0.8 mM], (NH4)2HPO4

193

[3.8 mM]; pH = 7) and incubated at 20 oC for 24 hrs at 150 rpm. Then the bacterial suspensions were

194

diluted by SSWC medium until the optical density OD600 reached 0.1 and used in the following

195

experiments. The tests were performed on 96-well plates. The standard solutions of streptomycin

196

were in the range of 7 nM to 15 µM. The initial concentration of streptomycin (3.0 µM) in solutions

197

subjected to irradiation was outside the sensitive region of the calibration curve and hence these

198

solutions were diluted 3 times accordingly before running the assays. The samples and controls were 8

ACS Paragon Plus Environment

Page 8 of 28

Page 9 of 28

Environmental Science & Technology

199

loaded on the 96-well plates with 100 µL in each well and three duplicates were carried out for each

200

sample. Then 100 µL bacterial suspensions were added to each well of the plates, these mixed

201

solutions were incubated at 20 oC and at 150 rpm for 12 hrs. Subsequently, bioluminescence was

202

measured with a plate reader (BioTek, Synergy HT), and the inhibition ratios of bioluminescence

203

were calculated for each sample compared with the bioluminescence of a blank control to assess the

204

toxicity of aminoglycosides and their decomposition products. Meanwhile the residual concentration

205

of streptomycin was also measured by LC-MS/MS.

206

Results and Discussions The roles of ROS in the photodegradation of aminoglycosides. A series of experiments

207 208

were conducted to explore the photochemical transformation of aminoglycoside antibiotics under

209

simulated solar irradiation. The control experiments showed that the direct photodegradation of

210

aminoglycosides at pH 7.0 in ultrapure water was negligible, as illustrated in Figure 2. It was due to

211

the lack of UV-Vis absorption of aminoglycosides within the solar irradiation spectrum.43, 44 The

212

photodegradation of aminoglycosides in the presence of SRNOM was observed, which suggested

213

that photosensitized degradation is critical in the environmental photo-transformation of

214

aminoglycosides. The indirect photodegradation followed pseudo first order kinetics, and the rate

215

constants of kanamycin and streptomycin were 0.226 and 0.221 hr-1 respectively.

216

(Insert Figure 2)

217

In order to probe the roles of 3NOM* and ROS in the photolytic process, a series of ROS

218

inhibited/enhanced studies were conducted. To distinguish the influence of 3NOM*, O2 and N2

219

saturated experiments were performed and compared to air saturated condition. Since O2 is known to

220

be a triplet excited state quencher and yield 1O2, in the absence of O2 (under N2 purge) the role of

221

3

222

aminoglycosides degradation in different gas saturated solutions followed the order: oxygen

223

saturated > air saturated > nitrogen saturated, indicating that the direct reaction of aminoglycosides

NOM* would be enhanced while the role of 1O2 would be minimized.45 The rates of

9

ACS Paragon Plus Environment

Environmental Science & Technology

224

with 3NOM* is of minor importance. Furthermore 3-methoxyacetophenone, a model compound, is

225

applied to simulate 3NOM*.46 Trimethylphenol is used as a reference compound25 to measure the

226

reactivity of 3NOM* with aminoglycosides. As shown in Figure S3 of SI, the photosensitized

227

reaction rates of 3-methoxyacetophenone with trimethylphenol is 0.0905 ± 0.0018 min-1, which is

228

60-fold faster than the reaction with aminoglycosides (0.0014 ± 0.0008 min-1). Considering that the

229

reaction rate constant of trimethylphenol with typical 3NOM* is reported as 1.8 × 109 M-1 s-1,47 we

230

estimated that the reaction rate constants of aminoglycosides with 3NOM* are less than 3 × 107 M-1

231

s-1. Thus as, it further suggests that the direct reaction of 3NOM* plays a minor role in the

232

photosensitized degradation of aminoglycosides.

233

To explore the roles of 1O2 in the degradation process, SRNOM was dissolved in D2O instead of

234

H2O. The contribution of 1O2 mediated processes is enhanced in D2O, because 1O2 life time is much

235

longer in D2O compared to H2O.28, 29, 48 The deuterium solvent isotope effects (kD/kH) for kanamycin

236

and streptomycin photodegradation are 1.14 and 1.11 respectively, suggesting that 1O2 plays a minor

237

role to remove aminoglycosides under simulated solar irradiation.

238

To explore the role of HO•, a powerful HO• scavenger, isopropanol (52mM), was added to the

239

aminoglycosides/SRNOM solutions. The degradation rates of kanamycin and streptomycin within

240

isopropanol aqueous solutions were 0.197 and 0.206 hr-1 respectively. Compared with the

241

degradation rate of aminoglycosides in SRNOM, only a slight inhibition of photochemical

242

degradation rate was observed, as illustrated in Figure 2. The presence of isopropanol, a HO•

243

scavenger, does not inhibit the photochemical processes. Carbonate radical (CO3•-) can also be an

244

important ROS presented in the sunlit surface water.49, 50 It can be generated through

245

bicarbonate/carbonate trapping of HO• and electron transfer with 3NOM*. As shown in Figure S4 of

246

SI, high concentration of bicarbonate has little impact on the photodegradation rate of

247

aminoglycosides. The results confirm that bicarbonate cannot apparently inhibit or enhance the

248

NOM induced photochemical degradation of aminoglycosides. 10

ACS Paragon Plus Environment

Page 10 of 28

Page 11 of 28

Environmental Science & Technology

To further quantitatively investigate the role of ROS in the photosensitized degradations, the

249 250

steady-state concentrations of 1O2 and HO• in bulk solutions were determined using the chemical

251

probes FFA and TA respectively. As demonstrated in Test S2 of SI, the steady-state concentration of

252

1

253

concentrations and the bimolecular reaction rate constants reported in Table S2, the contributions of

254

both 1O2 and HO• in bulk solutions were less than 7% of total observed degradation. These results

255

were consistent with ROS scavenger/enhanced experiments. Through the experimental results above,

256

the dominating factor in the photodegradation of aminoglycosides remains unclear. We propose that

257

a homogenous description of NOM photochemical process is insufficient to understand the NOM

258

induced photo-transformation of aminoglycosides. With this in mind, we reflect on previous studies

259

demonstrating that 1O2 levels were elevated in the core of NOM and only a small fraction of 1O2

260

could diffuse into the bulk solutions.34 Herein we hypothesized that other ROS (i.e., 3NOM* and HO•)

261

could also have microheterogeneous distributions. NOM-sorbed aminoglycosides could experience

262

enhanced rates of photodegradation due to being exposed to high localized ROS concentrations at or

263

near the surface of NOM. To accurately predict transformation in this system, the sorption of

264

aminoglycosides to NOM was investigated as described below.

265

O2 was (4.46 ± 0.11) × 10-13 M, and HO• was (8.59 ± 0.09) × 10-16 M. Based on the steady state

The micro-heterogeneous reaction in photosensitized degradation. Previous reports suggest

266

that NOM is effective to sorb hydrophobic pollutants through van der Waals interactions.51, 52 This

267

has a profound influence on the transportability, bioavailability, toxicity, and ultimate fate of organic

268

pollutants in natural waters. The interaction of NOM with aminoglycosides could also have a

269

pronounced influence on their photo-transformation. At neutral pH, NOM is negatively charged due

270

to the deprotonation of carboxylic and phenolic moieties.48, 53, 54 Aminoglycosides possess ionizable

271

amino groups, which exist as positively charged forms.55, 56 For example, the two pKa values of

272

kanamycin are 9.52 and 12.94, and for streptomycin they are 11.51 and 13.40 respectively. All

273

aminoglycosides possess positive charges at neutral pH owing to the presence of amino groups.57-59 11

ACS Paragon Plus Environment

Environmental Science & Technology

274

The electrostatic attraction between the positively charged aminoglycosides and negatively charged

275

NOM could be a major driving force in the strong interactions, which is generally considered to be

276

significantly stronger than the van der Waals interactions often associated with the sorption of

277

organic compounds on NOM. To disturb the strong electrostatic interactions (associations) of

278

cationic aminoglycosides with anionic NOM, isopropamide was added as a cationic HO• scavenger,

279

and compared with isopropanol, a neutral HO• scavenger. Isopropamide and isopropanol share the

280

same isopropyl group, both have similar bimolecular reaction rate constants with HO• and negligible

281

rate constants with other ROS.60-62 However, isopropamide has a positive charge at neutral pH and

282

thus can compete for aminoglycosides on the negatively charged surface of NOM, as showed in

283

Figure S5 of SI. Figure 3 reports that the photodegradation rates of kanamycin and streptomycin

284

were significantly decreased with the addition of varied concentrations of isopropamide, while the

285

addition of isopropanol had no effect. The sorption of aminoglycosides on NOM would thus be

286

critical and generation of HO• at the surface of NOM would play a key role in indirect

287

photodegradation processes. (Insert Figure 3)

288 289

The relationships of photodegradation rates and sorption behaviors among varied

290

aminoglycosides. Based on the above experiments, it is reasonable to envision that the sorption of

291

aminoglycosides on NOM could lead to a faster micro-heterogeneous reaction. Thus, a quantitative

292

investigation of the sorption of various aminoglycosides on NOM was necessary. Unlike soil

293

(sediment)-water systems, the determination of sorption between dissolved matters is challenging.

294

Several research groups have developed dialysis or size exclusion chromatography (SEC) to

295

investigate the sorption of inorganic and hydrophobic pollutants onto natural colloids in aquatic

296

environments.48, 52, 63 With this in mind, MWCO filters were carefully chosen for the sorption studies.

297

This method separates compounds based on the change of apparent molecular size and the details are

298

described in previous studies.54 12

ACS Paragon Plus Environment

Page 12 of 28

Page 13 of 28

Environmental Science & Technology

299

Our results showed that the sorption was described by a combination model involving

300

Langmuir- and linear- equations for each aminoglycoside (Figure S6 of SI). This dual-mode model

301

involves site-limited adsorption (Langmuir isotherm) and linear absorption (linear isotherm with

302

partition coefficient, Kip), and the resultant combined equation is:

303

 =

,     

(4)

+  

304

As shown in Table 1, the sorption parameters indicated that the Langmuir isotherm portion could be

305

defined as the electrostatic attraction between aminoglycosides and NOM, while the linear isotherm

306

portion could be driven by van der Waals interactions.48, 64 It is apparent that the adsorption

307

(Langmuir isotherm) is significantly stronger than the absorption (linear isotherm). NOM surface had

308

different maximum adsorption capacities as summarized in Table 1. In other words, the NOM

309

adsorption abilities for different aminoglycosides were varied. For streptomycin, the maximum

310

adsorbed concentration on SRNOM was about 580 µmol g-1 C, nearly 50% lower than for other

311

aminoglycosides. Towards other aminoglycosides, the adsorption capacities of NOM were also

312

slightly different. The different adsorption capacities among various aminoglycosides are probably

313

due to the different charge conditions and three-dimensional conformation. Toward streptomycin, the

314

adsorption behavior was quite weak, possibly because the amidine groups of streptomycin hold

315

positive charges more unstable than the general amino groups of other aminoglycosides.65

316 317

(Insert Table 1) The experiments conducted subsequently, confirmed the positive relationship between sorption

318

and photodegradation for all five kinds of aminoglycosides. The photodegradation rate constants

319

using different initial concentrations of each aminoglycoside (from 0.3 µM to 24 µM) with 5.2 mg C

320

L-1 SRNOM irradiated in the same conditions were tested by LC-MS/MS, and the sorption ratios

321

were calculated by the dual-mode model simultaneously. Figure 4a and Figure S7 of SI show that all

322

tested aminoglycosides behave in such a way as to imply that stronger sorption leads to faster

323

degradation rates. While different aminoglycosides present different photodegradation rate constants 13

ACS Paragon Plus Environment

Environmental Science & Technology

324 325

for the same sorption ratios, this is likely due to diverse HO• reaction rate constants. (Insert Figure 4)

326

To get further insight, we hypothesized that the efficiency of photodegradation might be

327

connected with both the sorption ratios and HO• reaction rate constants. Overall a linear relationship

328

between the photodegradation rates and sorption ratios × HO• rate constants was observed, as

329

demonstrated in Figure 4b. It further proved the fact that the aminoglycosides share a similar

330

degradation mechanism, controlled by HO• from heterogeneous reaction processes.

331

Cytotoxicity assessment. In general the photodegradation leads to complex mixtures of

332

products in low overall yields. It is a daunting mission to isolate the products and assess their

333

individual biological activities. Therefore we applied the luminescent bacteria Vibrio fischeri, to

334

assess the cytotoxicity of the treated solutions at various irradiation times. Since all aminoglycosides

335

share a similar toxicology mechanism, we only employed streptomycin as a model aminoglycoside

336

for cytotoxicity assessment. With Four Parameter Logistic Equation fitting, a calibration curve for

337

the bacteria inhibition as a function of the concentration of streptomycin was constructed as

338

illustrated in Figure 5a. The inhibition curve of streptomycin showed an IC50 of 0.57 µM. Through

339

the standard curve, the residual cytotoxicity of the solutions could be converted to the

340

aminoglycoside concentrations. As revealed in Figure 5b, cell toxicity in the mixture decreased as the

341

irradiation time was increased. LC-MS/MS determination of the concentration of streptomycin

342

indicates that the observed biological activity of the treated samples parallels the concentration of

343

streptomycin. It indicated that the streptomycin residual was responsible for most of the cytotoxicity.

344

In the other words, it was implied that streptomycin products are not formed to a significant extent

345

and/or do not exhibit important toxicity under our experimental conditions. It should be noted that

346

we only performed the cytotoxicity test on streptomycin using the luminescent bacteria. The

347

potential ecotoxicological effects of photoproducts of aminoglycosides need be addressed to better

348

understand the environmental impacts of aminoglycosides. 14

ACS Paragon Plus Environment

Page 14 of 28

Page 15 of 28

Environmental Science & Technology

(Insert Figure 5)

349 350 351

Environmental Significance In this study, the indirect photodegradation mechanisms of aminoglycosides have been

352

investigated. Our results suggest that the photodegradation rates of aminoglycosides are linearly

353

related with their sorption ratios on NOM. The micro-heterogeneous reaction is dominating for the

354

photodegradation in the low concentration range, and HO• on the surface of NOM ([HO•]surface) plays

355

a major role. Therefore [HO•]surface could be calculated using the formula:

356

357 358 359

 ! "

Substrate

(5)

= −$%&,' aminoglycoside!HO• !789'

(6)



Considering the [HO•]surface keeps constant during the reaction, :;?@AB>CB>C