Chlorotyrosines versus Volatile Byproducts from Chlorine Disinfection

Subscriber access provided by UNIV OF ARIZONA

Remediation and Control Technologies

Chlorotyrosines versus Volatile Byproducts from Chlorine Disinfection During Washing of Spinach and Lettuce Yukako Komaki, Adam Michael-Anthony Simpson, Jong Kwon Choe, Michael J Plewa, and William A. Mitch Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b03005 • Publication Date (Web): 24 Jul 2018 Downloaded from http://pubs.acs.org on July 26, 2018

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 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 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.

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 31

Environmental Science & Technology

1 2 3 4 5 6 7 8 9 10 11 12

Yukako Komaki1,2,‡, Adam M.-A. Simpson1,‡, Jong Kwon Choe3, Michael J. Plewa4, and

13

William A. Mitch1, *

Chlorotyrosines versus Volatile Byproducts from Chlorine Disinfection During Washing of Spinach and Lettuce

14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

1

Department of Civil and Environmental Engineering, Stanford University, 473 Via Ortega, Stanford, California 94305, United States

2

Present address: Graduate Division of Nutritional and Environmental Sciences, University of Shizuoka, 52-1 Yada, Suruga-ku, Shizuoka, 422-8526, Japan

3

Department of Civil and Environmental Engineering and Institute of Construction and Environmental Engineering, Seoul National University, 1 Gwanak-ro, Gwanak-gu, Seoul 08826, Republic of Korea 4

Department of Crop Sciences and Safe Global Water Institute, University of Illinois at UrbanaChampaign, Urbana, Illinois 61801, United States

‡

These authors contributed equally to this work.

*Contact Information: email: [email protected], Phone: 650-725-9298, Fax: 650-723-7058

1

ACS Paragon Plus Environment


38 39 40 41

Abstract

Following the Food Safety Modernization Act of 2011 in the U.S., guidelines for

42

disinfection washes in food packaging facilities are under consideration to control pathogen

43

risks. However, disinfectant exposures may need optimization because the high concentrations

44

of chlorine disinfectant promote the formation of high levels of disinfection byproducts (DBPs).

45

When chlorine doses up through the 200 mg/L as Cl2 range relevant to current practice were

46

applied to spinach and lettuce, significant DBP formation was observed, even within 5 min at 7

47

°C. Concentrations of volatile chlorinated DBPs in washwater were far higher than typically

48

observed in disinfected drinking waters (e.g., 350 µg/L 1,1-dichloropropanone). However, these

49

DBPs partitioned to the aqueous phase and so represent a greater concern for disposal or reuse of

50

washwater than for consumer exposure via food. The volatile DBPs represent the low-yield,

51

final products of chlorination reactions with multiple biomolecular precursors. The initial, high-

52

yield transformation products of such reactions may represent a greater concern for consumer

53

exposure because they remain bound within the biopolymers in food and would be liberated

54

during digestion. Using protein-bound tyrosine as an example precursor, the concentrations of

55

the initial 3-chlorotyrosine and 3,5-dichlorotyrosine transformation products from this one

56

precursor in the leaf phase were comparable, and in the case of some lettuces exceeded, the

57

aggregate aqueous concentration of volatile DBPs formed from multiple precursors.

58

Chlorotyrosine formation increased when spinach was shredded due to the greater accessibility

59

of chlorine to proteins in the leaf interiors. The cytotoxicity of chlorotyrosines to Chinese

60

hamster ovary cells was higher than any of the trihalomethanes regulated in drinking water.

61 62

2


Page 2 of 31

Page 3 of 31

63 64


Introduction

65

The application of chemical disinfectants to inactivate pathogens in drinking water was one

66

of the most important public health achievements of the past century. As a result, food may now

67

surpass drinking water as a pathogen exposure route in the U.S. An estimated 76 million

68

illnesses are associated with foodborne pathogens per year in the U.S.,1 compared to a

69

conservative estimate of 19.5 million illnesses per year attributable to tap water.2 Leafy

70

vegetables have been associated with the majority of multi-state outbreaks.3 Examples include

71

199 illnesses across 26 states linked to Escherichia coli O157:H7 on spinach in 2006,4 and 19

72

illnesses across 9 states associated with Listeria on lettuce in 2016.5 The resulting interest in food

73

safety culminated in the Food Safety Modernization Act of 2011.6 Washing fresh produce with

74

disinfected water in packing houses represents a critical control measure. Much like drinking

75

water disinfection, guidelines will need to be developed specifying disinfectant exposures, but

76

these guidelines may need to be crop-specific. While not yet codified, produce washing facilities

77

in the U.S. and several other countries (e.g., Spain) frequently employ chlorine at 50-200 mg/L

78

as Cl2 over short timescales (20 mg/L as Cl2 were needed to achieve significant reductions in aerobic plate

80

count bacteria in grated carrots.8

81

Following the discovery in the 1970s that disinfectant reactions with organic precursors in

82

water supplies can produce potentially carcinogenic disinfection byproducts (DBPs),9 drinking

83

water facilities have altered disinfectant strategies to balance the acute risk posed by pathogens

84

and the chronic risk posed by DBPs.10,11 As guidelines are developed for produce disinfection, it

85

will be important to balance these risks. Elevated DBP formation during produce disinfection is

86

expected based upon the high chlorine doses and precursor concentrations (e.g., solid lettuce

3



87

leaves) relative to the lower chlorine doses (~5 mg/L as Cl2) and precursor concentrations (~2

88

mg/L dissolved organic carbon (DOC)) in drinking water. Indeed, certain northern European

89

countries (e.g., Belgium, Germany) have banned chlorinated washwaters for produce

90

decontamination due to these concerns.12,13

91

Research on DBPs related to food has focused on the same low molecular weight, volatile

92

DBPs targeted in drinking water. Shen et al.13 measured up to 858 µg/L chloroform and 1349

93

µg/L bromoacetic acid in a washwater simulated by 13 separate injections of lettuce juice and

94

periodic additions of chlorine to maintain at least a 2 mg/L as Cl2 residual. In a simulated

95

chlorine washwater treatment of cut butterhead lettuce, Van Haute et al.14 measured THMs

96

(mostly chloroform) in the washwater, but not on the lettuce. While volatile DBPs are retained in

97

pressurized drinking water distribution systems, much of those that form from food disinfection

98

may partition to the washwater or disperse via volatilization during transport and storage prior to

99

consumer exposure. Thus, volatile DBPs may represent a greater concern for spent washwater

100

discharge.

101

Moreover, previous research involving chlorination of model biomolecules, the

102

predominant precursors in food, has demonstrated that volatile DBPs generally form at 88% relative to spiking into 60 mL of

194

buffered deionized water in headspace-free glass tubes without filtration for all of the DBPs

195

except chloroform. For chloroform, the recovery was ~70% with the loss mostly attributable to

196

the filtration; the chloroform concentrations reported here were not corrected for this recovery.

197

Filtration was necessary to adequately separate the aqueous phase from shredded lettuce or

198

spinach.

199 200

Analysis of tyrosines: The tyrosine extraction method was adapted from the procedures of

201

Bañuelos et al.25 and Montes-Bayon et al.26 Freeze-dried leaf tissues were homogenized with a

202

ball mill (Retsch MM 400) using 5 mm diameter stainless steel beads, and shaking at 25 Hz for 2

203

min. The homogenized sample was placed in a 11.1-mL glass vial with 6 mL of 4 M

8


Page 8 of 31

Page 9 of 31


204

methanesulfonic acid containing 0.2 wt% tryptamine, and incubated at 110°C for 24 hours in an

205

anaerobic glove box to liberate tyrosines from proteins. After neutralizing the acid-digested

206

sample with 6 mL of 4 N NaOH, 0.5 mL of a 1 mM solution of L-tyrosine-13C9,15N (500 nmol)

207

was spiked into each sample, followed by 17 mL methanol. Each tube was vortexed, allowed to

208

stand for 2 hours, vortexed again, centrifuged at 10,000 rpm for 10 min, and filtered through

209

filter paper (Whatman Qualitative Filter paper, Grade 1 Circles) to remove solids. Then 8.5 mL

210

of chloroform was added and the tubes were capped, shaken vigorously for 2 min, and placed on

211

ice overnight. The upper aqueous/methanol phase was transferred to a round bottom flask and

212

brought nearly to dryness using a rotary evaporator.

213

The sample was reconstituted with 5 mL of 0.1 M pH 5 phosphate buffer. The pH was

214

further titrated to reach pH 5 using NaOH. The sample was then passed over Sep-Pak C18 solid

215

phase extraction (SPE) cartridges (2 g, 55-105 µm) (Waters Corporation, Milford, MA, USA).

216

Each cartridge was pre-treated by flushing with 10 mL of methanol, 10 mL of deionized water

217

and 10 mL of 0.1 M phosphate buffer at pH 5. Then the cartridge was loaded with the sample,

218

rinsed with 3 mL of 0.1 M phosphate buffer at pH 5, and eluted with 10 mL of 40% methanol.

219

For 100% recovery, the stable isotope-labeled tyrosine concentration in this eluent would be 50

220

µM.

221

A 400 µL aliquot of the eluent was spiked with 16.7 µL of 2.5 mM γ-aminobutyric acid

222

(42 nmol) as an internal standard (100 µM final concentration). Then 20 µL of this aliquot was

223

mixed with 60 µL of 0.6 M borate buffer solution (pH 8.8) and 20 µL of 10 mM 6-

224

aminoquinoline-N-hydroxy-succinimidyl ester (AQC) in acetonitrile solution in a

225

microcentrifuge tube, which was then incubated at 55°C for 10 min to derivatize the amino

226

acids.27 Derivatized amino acids were quantified via liquid chromatography mass spectrometry

9



227

(LC-MS; Agilent 1260 HPLC system coupled with a 6460 triple quadrupole mass spectrometer)

228

using electrospray ionization in the negative mode. Text S1 provides additional details regarding

229

the optimization of the SPE protocol and LC-MS parameters. Titrating the pH to 5 prior to SPE

230

extraction was critical for maximizing recovery.

231

To evaluate matrix effects on both derivatization efficiency and LC-MS analysis, a

232

matrix was constituted by processing 1 g of shredded spinach (without exposure to chlorine)

233

according to the protocol described above from freeze-drying of the leaves through the elution

234

from the SPE cartridges. Comparison of signal intensities of analytes (normalized by those of γ-

235

aminobutyric acid as an internal standard) in the matrix to those in deionized water indicated that

236

the matrix suppressed signals by 33% for 3-chlorotyrosine, by 46% for 3,5-dichlorotyrosine and

237

by 53% for mass-labeled tyrosine; mass-labeled tyrosine was used to evaluate matrix effects for

238

tyrosine due to the substantial (~500 µM) concentrations of tyrosine already present in the

239

matrix.

240

To evaluate method detection limits (MDLs) for chlorotyrosines within the post-SPE

241

elution spinach matrix described above, seven different matrix aliquots were spiked with 0.2 µM

242

of 3-chlorotyrosine and 3,5-dichlorotyrosine and analyzed as described above. Based upon the

243

standard deviation of these 7 spiked samples, the MDLs were 0.072 µM for 3-chlorotyrosine and

244

0.085 µM for 3,5-dichlorotyrosine.28 For the 1 g (wet weight) spinach or lettuce samples

245

typically used in our experiments, these concentrations correspond to 0.72 nmol/g for 3-

246

chlorotyrosine and 0.85 nmol/g spinach for 3,5-dichlorotyrosine. The MDL was not determined

247

for tyrosine, because tyrosine was always present in substantial concentrations.

248 249

Method recoveries were evaluated by spiking 20 nmol of 3-chlorotyrosine and 3,5dichlorotyrosine, and 42 nmol of mass-labeled tyrosine into spinach samples directly after acid

10


Page 10 of 31

Page 11 of 31


250

digestion and neutralization. Using standards constituted within the post-SPE elution spinach

251

matrix for comparison, recoveries were 56% (±4%; n = 4) for mass-labeled tyrosine, 48% (±4%;

252

n = 4) for 3-chlorotyrosine and 80% (±4%; n = 4) for 3,5-dichlorotyrosine. Because of the

253

substantial tyrosine concentrations present in the sample matrix, tyrosine was quantified using a

254

standard curve constituted in deionized water using mass-labeled tyrosine for isotope dilution

255

analysis, which accounts for extraction and derivatization efficiencies, and matrix suppression.

256

Chlorotyrosines were quantified using standards constituted in a spinach sample matrix (post-

257

SPE elution) using γ-aminobutyric acid as an internal standard, which accounts for derivatization

258

efficiency and matrix suppression effects on the LC-MS analysis, but not extraction efficiency.

259

These concentrations were then corrected for the recoveries measured above.

260

Chinese Hamster Ovary (CHO) Cell Chronic Cytotoxicity Analyses: For in vitro, chronic

261

quantitative analysis for cytotoxicity, CHO K1 cell line AS52, clone 11-4-8 was used.29 The

262

CHO cells were maintained in Ham’s F12 medium with 5% fetal bovine serum (FBS), 1% L-

263

glutamine, and 1% antibiotics (0.25 µg/mL amphotericin B, 100 µg/mL streptomycin sulfate, and

264

100 units/mL sodium penicillin G in 0.85% saline) at 37°C in a mammalian cell incubator with a

265

humidified atmosphere of 5% CO2. Details of the CHO cell cytotoxicity assay were

266

published.30,31 Each halotyrosine was diluted with F12 plus FBS cell culture medium, and in

267

general, 10 concentrations (with independent replicates) were analyzed within a 96-well

268

microplate. After 72 h, the cell density was measured spectroscopically and expressed as the

269

percentage of the concurrent negative control. These data were used to generate individual

270

cytotoxicity concentration-response curves for 3-chlorotyrosine and 3,5-dichlorotyrosine.

271

Regression analyses of the concentration-response curves were used to generate a LC50 value for

272

each compound. The LC50 value was the concentration of the sample that induced a cell density

11



273

of 50% as compared to the concurrent negative controls. For each halotyrosine, a one-way

274

analysis of variance (ANOVA) statistical test was conducted to determine the lowest molar

275

concentration that induced a statistically significant level of cytotoxicity as compared to their

276

concurrent negative control.32 The power of the test was maintained at >0.8 at α = 0.05.

277 278

Results and Discussion

279

Whole spinach leaf chlorination: Initial experiments involved treatment of whole spinach

280

leaves (~1 g; Sigona’s Market bunch #1) at pH 7 with 0 or 100 mg/L as Cl2 chlorine for 15 min

281

at 21°C. Prior to chlorination, tyrosine was measured at 6.7 µmol/g spinach (Figure S3), along

282

with low levels of 3-chlorotyrosine (2.5 nmol/g spinach) and 3,5-dichlorotyrosine (1.5 nmol/g

283

spinach) (Figure 1). The 3-chlorotyrosine and 3,5-dichlorotyrosine concentrations increased to

284

5.9 nmol/g spinach and 5.5 nmol/g spinach, respectively, after application of 100 mg/L as Cl2

285

free chlorine. While the yields of chlorotyrosines remained low (0.17%) relative to the 6.7

286

µmol/g spinach tyrosine initially measured, all of the chlorotyrosines were retrieved from the

287

spinach rather than the aqueous phase, suggesting their relevance to human exposure.

288

Only the aqueous phase was assayed for volatile DBPs. For a control experiment in

289

which deionized water buffered at pH 7 was spiked with a 10 µg/L mixture of these volatile

290

DBPs and shaken for 15 min in the absence or presence of whole leaf spinach (~1 g), there was

291

no significant difference in the final aqueous volatile DBP concentrations. Concurring with

292

previous results with THMs14, these results demonstrate that the volatile DBPs remain mostly in

293

the aqueous phase, and so are predominantly a concern for washwater disposal. Total volatile

294

DBP concentrations measured in the aqueous phase increased from non-detectable in the absence

295

of free chlorine to 37.5 nmol/g spinach after treatment with 100 mg/L as Cl2 chlorine; although

12


Page 12 of 31

Page 13 of 31


296

measured in the aqueous phase, volatile DBP concentrations are reported as yields relative to

297

spinach wet weight to facilitate comparison of the production of chlorotyrosines and volatile

298

DBPs. The total volatile DBP concentration was 3.3-fold higher than the total chlorotyrosines.

299

While THMs typically predominate among the volatile DBPs in chlorinated drinking waters,33,34

300

1,1-dichloropropanone, was the dominant volatile DBP formed from the chlorinated spinach,

301

followed by dichloroacetonitrile, chloroform and trichloroacetaldehyde (Figure S3).

302 303

Effect of shredding or puréeing spinach: Shredded spinach leaves (from Safeway®) were

304

treated at pH 7 and 21°C for 15 min with 0-200 mg/L as Cl2 chlorine. The tyrosine retrieved

305

from the spinach remained at ~6.6-8.6 µmol/g spinach regardless of free chlorine dose (Figure

306

S4); this tyrosine concentration was similar to the whole leaf experiments (Figure S3), even

307

though the spinach was from a different bunch. Total chlorotyrosine yields retrieved from the

308

leaves increased with chlorine dose from 3.5 nmol/g spinach with no chlorine treatment to 76

309

nmol/g spinach for treatment with 200 mg/L as Cl2 (Figure 2). The 53 nmol/g spinach total

310

chlorotyrosine yield measured for the 100 mg/L as Cl2 chlorine dose represents a 0.79% yield

311

relative to tyrosine, 4.6-fold higher than the 0.17% yield measured for whole spinach leaves. In

312

both cases, the lack of significant transformation of tyrosine upon chlorination (Figures S3 and

313

S4) suggests that the low yields resulted from the inaccessibility of protein-bound tyrosine within

314

leave interiors to chlorine reactions rather than the formation of alternative chlorination products.

315

Among chlorotyrosines, 3-chlorotyrosine dominated at a 20 mg/L as Cl2 chlorine dose, but the

316

concentrations of 3-chlorotyrosine and 3,5-dichlorotyrosine became comparable at the highest

317

chlorine doses.

13



318

Total volatile DBP yields increased with chlorine dose, from non-detectable for no

319

chlorine exposure to 243 nmol/g spinach for 200 mg/L as Cl2 (Figure S4). The concentration of

320

total chlorotyrosines ranged from 21-56% of the total volatile DBP concentration. The 131

321

nmol/g spinach total volatile DBPs measured for the 100 mg/L as Cl2 chlorine dose was

322

significantly higher than the 37.5 nmol/g spinach measured with whole spinach leaves. For

323

chlorine doses above 50 mg/L as Cl2, 1,1-dichloropropanone was the most abundant volatile

324

DBP, followed by dichloroacetonitrile, chloroform, and dichloroacetamide. The aqueous

325

concentrations of some of these DBPs were far higher than typically observed in chlorinated

326

drinking waters33,34; for example, the aqueous concentrations of 1,1-dichloropropanone and

327

dichloroacetonitrile reached 350 µg/L and 190 µg/L, respectively, for 200 mg/L as Cl2 chlorine.

328

The significantly higher chlorotyrosine yields for the shredded relative to whole leaf

329

spinach suggested that shredding the spinach facilitated access of the chlorine to proteins within

330

the interior of the spinach leaves. In whole leaf spinach, proteins within the interior of a leaf

331

may be protected from reactions with chlorine by the waxy cuticles coating the leaf exterior. To

332

test this hypothesis, 200 mg/L as Cl2 chlorine was applied to 20 mg-C/L of puréed and filtered

333

spinach (Safeway®) at pH 7 for 15 min (Figure S5); this DOC level approximates the DOC

334

measured in the shredded spinach or lettuce experiments. The aqueous phase protein remaining

335

after puréeing and filtering was considered to be fully accessible to chlorine reactions. Tyrosines

336

were measured in the aqueous phase after liberation by acid digestion. The 95 µmol/L tyrosine

337

measured in the absence of chlorine was reduced to 8.6 µmol/L by chlorination, yielding 18.4

338

µmol/L 3-chlorotyrosine and 44.6 µmol/L 3,5-dichlorotyrosine, indicating a 67% chlorotyrosine

339

yield. Total volatile DBPs increased from non-detectable to 0.5 µmol/L, far lower than the

14


Page 14 of 31

Page 15 of 31


340

chlorotyrosines. Chloroform was the dominant volatile DBP, followed by dichloroacetonitrile

341

and trichloroacetaldehyde (Figure S5).

342 343

Effect of temperature and contact time: The preceding experiments involved chlorination for

344

15 min at 21 °C. Chlorination at washing facilities may involve shorter chlorine contact times

345

and low temperatures to minimize browning and other aesthetic defects. To characterize the

346

importance of these conditions, DBP formation was measured after treatment of 1 g shredded

347

spinach (Sigona’s Market bunch #2) with 100 mg/L as Cl2 free chlorine for 5 or 15 min at either

348

7 °C or 21°C. The 5.5 µmol/g spinach tyrosine measured after chlorination for 15 min at 21°C in

349

this batch of shredded spinach (Figure S6) was comparable to that measured in shredded leaves

350

from a different batch of spinach after a similar treatment (Figure S4). Total DBP formation

351

increased with both chlorine contact time and temperature (Figure 3), as expected. However,

352

total DBP formation for chlorination for 5 min at 7 °C was still ~50% of that measured after

353

chlorination for 15 min at 21 °C. With respect to the chlorination of shredded spinach for 15 min

354

at 21 °C, the 94 nmol/g spinach yield of total chlorotyrosines (1.7% molar yield relative to

355

tyrosine) was somewhat higher than the 53 nmol/g spinach yield observed with a different batch

356

of spinach for the same treatment conditions (Figure 2). The contributions from 3-chlorotyrosine

357

and 3,5-dichlorotyrosine were approximately equal, and chlorotyrosine formation was more

358

dependent on contact time than temperature. However, the 252 nmol/g spinach yield of total

359

volatile DBPs for chlorination for 15 min at 21 °C (Figure S6) was higher than that observed

360

from a different batch of shredded spinach (Figure S4), and also higher than the total

361

chlorotyrosines. While dichloroacetamide, trichloroacetaldehyde, dichloroacetonitrile and

362

chloroform were the predominant volatile DBPs formed during chlorination of this spinach

15



363

bunch, 1,1-dichloropropanone and dichloroacetonitrile were the predominant volatile DBPs

364

formed with a different shredded spinach bunch (Figure S4); this variability is further evaluated

365

below.

366 367

Shredded lettuce chlorination: Lettuce is more likely to be washed in a shredded state (e.g., for

368

use in restaurants) than spinach. For an initial comparison between spinach and lettuce, 1 g

369

shredded butterhead lettuce was treated at pH 7 and 21 ºC for 15 min with 0-200 mg/L as Cl2

370

chlorine. Approximately 3 µmol/g lettuce of tyrosine was detected in the shredded lettuce before

371

chlorination (Figure S7), two-fold lower than in spinach. Chlorotyrosine formation increased

372

with chlorine dose (Figure 4), reaching 212 nmol/g lettuce after treatment with 200 mg/L as Cl2.

373

The importance of 3,5-dichlorotyrosine increased with chlorine dose, equaling the 3-

374

chlorotyrosine concentration at the highest doses. While the total chlorotyrosine yields per wet

375

weight from lettuce (Figure 4) were two-fold higher than from shredded spinach (Figure 2), the

376

yields for treatment with 200 mg/L as Cl2 chlorine were even higher (6.7% for lettuce vs. 1.2%

377

for spinach) due to the lower tyrosine content in lettuce. The greater molar conversion of

378

tyrosine to chlorotyrosines in lettuce may reflect the higher chlorine-to-tyrosine molar ratio or

379

may suggest greater accessibility of free chlorine to protein in the interiors of shredded lettuce

380

leaves. Alternatively, the high antioxidant content of spinach (carotenes, ascorbic acid, and

381

flavonoids35) could help protect the tyrosines in spinach from reactions with chlorine. However,

382

while raw spinach has roughly twice the total phenolic content (205 mg gallic acid equivalents

383

(GAE)/100 g) compared to butterhead lettuce (100 mg GAE/100 g), their Oxygen Radical

384

Absorbance Capacities are similar (~1500 µmol Trolox equivalents/100 g).36

16


Page 16 of 31

Page 17 of 31


385

Unlike shredded spinach (Figure 2), the yields of volatile DBPs extracted from the

386

aqueous phase from chlorinated shredded lettuce were lower than the yields of chlorinated

387

tyrosines extracted from the shredded lettuce solids (Figure 4). Dichloroacetamide was the most

388

abundant volatile DBP (Figure S7).

389 390

Variability across batches: DBP formation from treatment of shredded leaves from two

391

additional spinach bunches and four additional lettuce varieties (iceberg, romaine, green leaf and

392

red leaf) with 20 mg/L as Cl2 chlorine for 15 min at pH 7 and room temperature was compared to

393

DBP formation measured previously from treatment of shredded spinach (Figures 2 and S4) and

394

butterhead lettuce (Figures 4 and S7) under comparable conditions to characterize variability

395

between batches (Figure 5). The tyrosine content was fairly similar across the three spinach

396

bunches at ~7-9 µmol/g spinach Tyrosine content was lower in the lettuce varieties, ranging from

397

0.7 µmol/g in the green leaf sample to 3 µmol/g in the butterhead lettuce. The total

398

chlorotyrosine content in the shredded spinach samples ranged from 12-18 nmol/g, lower than

399

the 21-56 nmol/g range observed in the shredded lettuce samples. 3,5-Dichlorotyrosine was more

400

significant in lettuce. In addition to higher absolute chlorotyrosine concentrations, shredded

401

lettuce also featured higher chlorotyrosine yields (0.9-5.7%) than shredded spinach (0.14-0.21%)

402

due to the lower tyrosine content in lettuce. Total volatile DBPs ranged from 22-44 nmol/g in

403

shredded spinach samples and 28-81 nmol/g in shredded lettuce samples. The ratio of

404

chlorotyrosines to total volatile DBPs was generally higher in shredded lettuce than shredded

405

spinach, except for red leaf lettuce. The predominant volatile DBP classes also varied widely.

406

Dichloroacetamide and 1,1-dichloropropanone were typically the predominant species, although

407

chloroform was significant for the red leaf, romaine and butterhead lettuces (Figure 5).

17



408 409

Cytotoxicity: The concentration-response curves illustrating CHO cell chronic cytotoxicity for

410

3-chlorotyrosine and 3,5-dichlorotyrosine are presented in Figure S8. Each halotyrosine induced

411

a statistical increase in cytotoxicity as compared to their negative controls (Table S2). Based on

412

the lowest significant cytotoxic concentration, 3,5-dichlorotyrosine was approximately 20× more

413

toxic than 3-chlorotyrosine. Based on the LC50 values, 3-chlorotyrosine (3.17 mM) and 3,5-

414

dichlorotyrosine (0.71 mM) are more cytotoxic (i.e., lower LC50 values) than the four regulated

415

THMs (3.96-11.5 mM).30 These halotyrosines are also comparable in cytotoxicity to chloroacetic

416

acid (0.81 mM), dichloroacetic acid (7.3 mM), trichloroacetic acid (2.4 mM), and dibromoacetic

417

acid (0.59 mM).30,37 Haloaromatic compounds typically are more toxic than haloaliphatic

418

compounds, but toxicity tends to increase with hydrophobicity.38 The comparable toxicity of

419

halotyrosines to haloacetic acids may result from the hydrophilicity of halotyrosines resulting

420

from their charged functional groups.

421 422

Implications: High levels of DBPs can form during food disinfection due to the high chlorine

423

disinfectant doses applied and organic precursor concentrations (e.g., solid leaves). High

424

chlorine doses may be applied (up to 200 mg/L as Cl28,12,13,22,23) even when low chlorine

425

residuals are targeted (~2 mg/L as Cl2) because of the high chlorine demand in these waters, with

426

significant chlorine consumption occurring in

Chlorotyrosines versus Volatile Byproducts from Chlorine Disinfection

Recommend Documents