Design and Testing of Safer, More Effective Preservatives for Consumer Products

(1)

Citation for this paper:

Buckley, H. L., Hart-Cooper, W. M., Kim, J. H., Faulkner, D. M., Cheng, L. W., Chan, K. L…Mulvihill, M. J. (2017). Design and Testing of Safer, More Effective

Preservatives for Consumer Products, ACS Sustainable Chemistry & Engineering, 5(5), 4320–4331.

https://doi.org/10.1021/acssuschemeng.7b00374

UVicSPACE: Research & Learning Repository

_____________________________________________________________

Faculty of Engineering

Faculty Publications

_____________________________________________________________

This is a post-print version of the following article:

Design and Testing of Safer, More Effective Preservatives for Consumer Products Heather L. Buckley, William M. Hart-Cooper, Jong H. Kim, David M. Faulkner, Luisa W. Cheng, Kathleen L. Chan…Martin J. Mulvihill

2017

The final publication is available at:

(2)

1

Design and Testing of Safer, More E

ﬀective Preservatives for

2

Consumer Products

3

Heather L. Buckley,

*

,¶,†

William M. Hart-Cooper,

‡,†

Jong H. Kim,

§,†

David M. Faulkner,

⊥

4

Luisa W. Cheng,

§

Kathleen L. Chan,

§

Christopher D. Vulpe,

#

William J. Orts,

‡

Susan E. Amrose,

∇ 5

and Martin J. Mulvihill

¶

6¶Berkeley Center for Green Chemistry, College of Chemistry, University of California Berkeley, Berkeley, California 94720, United 7 States

8‡Bioproducts Research Unit, Western Regional Research Center, USDA-ARS, 800 Buchanan St., Albany, California 94710, United 9 States

10§Foodborne Toxin Detection and Prevention Research Unit, Western Regional Research Center, USDA-ARS, 800 Buchanan St., 11 Albany, California 94710, United States

12⊥Molecular Toxicology, Nutritional Sciences and Toxicology, Berkeley Center for Green Chemistry, University of California Berkeley, 13 Berkeley, California 94720, United States

14#Physiological Sciences, Center for Environmental and Human Toxicology, University of Florida, Gainesville, Florida 32611, United 15 States

16∇Civil and Environmental Engineering, University of California Berkeley, Berkeley, California 94720, United States 17

*

S Supporting Information

18 ABSTRACT: Preservatives deter microbial growth, providing 19 crucial functions of safety and durability in composite materials, 20 formulated products, and food packaging. Concern for human 21 health and the environmental impact of some preservatives has led 22 to regulatory restrictions and public pressure to remove individual 23 classes of compounds, such as parabens and chromated copper 24 arsenate, from consumer products. Bans do not address the need for 25 safe, effective alternative preservatives, which are critical for both 26 product performance (including lifespan and therefore life cycle 27 metrics) and consumer safety. In this work, we studied both the 28 safety and efficacy of a series of phenolic preservatives and 29 compared them to common preservatives found in personal care 30 products and building materials. We quantified antimicrobial activity

31 against Aspergillus brasiliensis (mold) and Pseudomonas aeruginosa (Gram negative bacteria), and we conducted a hazard 32 assessment, complemented by computational modeling, to evaluate the human and environmental health impacts of these 33 chemicals. We found that octyl gallate demonstrates better antimicrobial activity and comparable or lower hazards, compared to 34 current-use preservatives. Therefore, octyl gallate may serve as a viable small-molecule preservative, particularly in conjunction 35 with low concentrations of other preservatives that act through complementary mechanisms.

36 KEYWORDS: Preservative, Antimicrobial, Safer alternative, Octyl gallate, Consumer products, Hazard assessment, 37 Computational toxicology

38

■

INTRODUCTION

39Composite materials, formulated products, and prepared foods 40and their packaging all require preservatives to prevent 41microbial degradation. Microbial communities persist in almost 42any environment that oﬀers a carbon source and water. Such 43environments exist nearly everywhere, from bottles of shampoo 44to laminate ﬂooring. Preservatives enhance product value by 45prolonging the shelf life of consumables and decrease life-cycle 46impacts in the built environment by increasing the longevity of 47installed components. As consumer demand increases for bio-48based and naturally derived materials,1−4 technologies that

49

provide safe, eﬀective preservation against microbial attack are

50

essential to avoid compromising shelf life, durability, or

51

performance.5

52

There has been little work on systematically identifying

53

classes of antimicrobial compounds that are both safer than

54

existing options and eﬀective microbiostats or microbiocides.6

55

A patchwork of identiﬁed hazards leading to restrictions or

Received: February 7, 2017

Revised: March 6, 2017

Published: March 15, 2017

Research Article

pubs.acs.org/journal/ascecg

(3)

56marketing/labeling to improve consumer awareness has 57reduced the use of certain hazardous chemicals in prod-58ucts7,8chemicals such as parabens, isothiazolinones, and 59metals such as chromium and arsenic. There is no 60comprehensive approach for proactively identifying and 61introducing safer alternative preservatives using meaningful 62sustainability metrics.9 As health and safety information 63improves, there is increasing consumer and regulatory demand 64for safer alternatives,6 coupled with demand for a compre-65hensive approach to demonstrate that these alternatives are 66both safe and eﬀective.

67 An additional motivation for designing safe and effective 68antimicrobials is the evolution of resistant strains. This issue is 69most widely recognized in the context of concern over 70antibacterial compounds such as triclosan (which led to a 71recent FDA ban of 19 chemicals in topical antiseptics for 72consumers)8and in the emergence of resistant “superbugs” in 73healthcare settings.10−12 Fungal resistance is also a growing 74problem,13−15 and resistant strains of bacterial and fungal 75contaminants pose a challenge to both product formulators and 76manufacturers.16−20Understanding the mechanism of action as 77part of preservative design is one approach to overcoming 78microbial resistance to conventional antimicrobials.21−23 In 79addition, having a broad range of potential preservatives from 80which to choose a synergistic mixture can help product 81formulators avoid inducing antimicrobial resistance.24,25 82 This paper evaluates three classes of phenolic ester/amide 83compounds and, by screening for antimicrobial effectiveness 84and human or environmental hazards, compares them to 85commonly used conventional preservatives. By considering 86chemical safety as a key performance criterion, our approach 87facilitates the direct evaluation of the tradeoffs inherent in 88selecting preservatives.26,27 Furthermore, it provides a model 89for how such a multifaceted screening could be conducted for 90other chemistries used in consumer products, contributing to a 91small but growing body of literature in this area.5,9,28−33 92 There are several mechanisms by which chemical preserva-93tives can act against microbes, which typically consist of Gram-94positive bacteria, Gram-negative bacteria, or fungi (e.g., molds 95and yeasts). These mechanisms include binding to DNA or 96other anionic biomolecules, either covalently (through 97alkylation, e.g. by epoxides or formaldehydes) or noncovalently, 98interfering with transcription or damaging the DNA; protein 99denaturation or coagulation through changes in polarity or 100hydrogen bonding of the local environment (e.g., by alcohols); 101and disruption of redox homeostasis (e.g., by metals such as 102silver, or derivatized phenol compounds).34−36 Disrupting 103redox homeostasis produces many outcomes, including over-104stimulation of oxygen uptake, disrupting ATP synthesis by 105interfering with electron transport chains, and uncoupling 106oxidative phosphorylation or active transport of protons from 107other processes.34,37,38Thefine balance of redox homeostasis in 108cells can be disrupted by introducing or modulating the 109metabolism of free radicals.39For example, phenol compounds 110are potent redox cyclers in cells, which can destabilize cellular 111redox homeostasis and/or antioxidant systems, inhibiting the 112growth of microbial pathogens.35,36,40 In particular, inhibition 113of glutathione reductase or superoxide dismutase enzymes or 114defense pathways such as the mitogen-activated protein kinase 115(MAPK) pathway may effectively prevent fungal growth by 116redox-active compounds.41

117 The inherent challenge of designing or selecting safe and 118eﬀective antimicrobials for consumer products is that, by virtue

119

of their function, antimicrobial compounds or materials must

120

be bioactive and, therefore, often exhibit toxicity to

non-121

microbial organisms. Many common preservatives for food

122

packaging, personal care products, and building materials have

123

known hazards: butylated hydroxyanisole (BHA) is a probable

124

carcinogen;42 parabens are known skin sensitizers43 and have

125

potential endocrine activity; and chromated copper arsenate is

126

highly persistent and has a range of serious human and

127

environmental toxicological eﬀects.44

128

In this paper, we propose a series of potential antimicrobial

129

compounds: phenolic acids, esters, and amides that we

130

postulate should act through the disruption of redox

homeo-131

stasis in microbial metabolism and cell components, including

132

the cell membrane.45,46Our hypothesis is that small structural

133

modifications may have differential impacts on both efficacy

134

and human health hazards. Exploiting these diﬀerences between

135

human and microbial biochemical processes or cell structures

136

could improve antimicrobial potency without adverse human

137

health outcomes. For example, selective toxicity has been

138

achieved in fungal pathogens (ergosterol-based membrane), but

139

not in humans (cholesterol-based membrane) by certain

redox-140

active drugs.47,48 We postulate that the redox activity of

141

phenolic compounds may lead to similar diﬀerentiation of

142

activity/toxicity. We test two classes of phenolic acid

143

derivatives: the esters and amides of salicylic acid

(2-144

hydroxybenzoic acid), and the esters and amides of gallic acid

145

(3,4,5-trihydroxybenzoic acid). [We use the esters and amides

146 f1

of benzoic acid itself as a nonphenolic control (seeFigure 1for

147

chemical structures).] We test a representative range of chain

148

lengths from C₀to C₁₆for these classes of compounds. Gallate

149

esters have been shown to exhibit antimicrobial activity; in

150

several cases, propyl gallate may serve as an eﬀective alternative

151

to salicylhydroxamic acid, which is a likely developmental

152

toxicant.49Salicylhydroxamic acid works by speciﬁcally blocking

153

the activity of alternative oxidase (AOX) in pecan scab50and in

154

ethylene-treated tubers.51

155

We tested all of these compounds for activity against

156

Aspergillus brasiliensis (mold) and Pseudomonas aeruginosa

157

(Gram negative bacteria)representative microorganisms

158

that the health and personal care product industries use for

159

evaluating preservative eﬃcacy52,53over a range of alkyl chain

160

lengths to observe variability in antimicrobial activity, as a

161

function of chain length. Of note, A. brasiliensis has recently

162

been found as a causative agent of keratitis,54 while P.

163

aeruginosa can cause disease in humans,55 thus emphasizing

164

the importance and relevance of testing these microbes. Both of

165

these test organisms, and related species, are known to exhibit

166

exceptional resilience against typical antimicrobial agents.25,56,57

167

Testing a range of ester chain lengths provides useful

168

information about the eﬀectiveness of compounds that, due to

169

diﬀerent physical properties (e.g., water/lipid partitioning), will

170

be compatible with diﬀerent formulations and products.58We Figure 1. General structures of benzoates, salicylates (2-hydrox-ybenzoates), parabens (4-hydrox(2-hydrox-ybenzoates), and gallates (3,4,5-trihydroxybenzoates).

DOI:10.1021/acssuschemeng.7b00374

ACS Sustainable Chem. Eng. XXXX, XXX, XXX−XXX

(4)

171tested the corresponding benzoic acids and alcohols that 172comprise the esters, as well as representative corresponding 173benzamides, to provide insight into whether hydrolyzed 174subcomponents, or diﬀerences in electronegativity (and, 175therefore, radical stabilization) could be responsible for 176antimicrobial activity. Finally, we tested representative preser-177vatives that are currently used in food packaging, personal care 178products, and building materials, and are postulated to operate 179by a similar mechanism (disrupting redox homeostasis in 180microbial cells).38 One of the classes of “control” molecules 181tested is parabens (esters of para-hydroxybenzoic acid); methyl 182and propyl parabens are widely used preservatives with 183structures similar to our proposed alternative compounds, but 184they act as dermal sensitizers.59

185 To complement this evaluation of antimicrobial eﬀectiveness, 186we conducted a hazard assessment of a representative subset of 187our proposed preservatives, and compared them to common 188preservatives, as discussed above. Our evaluation is based on 189the same principles as a GreenScreen assessment,60 drawing 190information from a combination of authoritative lists and 191compiled primary literature on hazard end points. In the 192absence of available data, we used computational tools to make 193structure-related predictions.

194 Overall, this study demonstrates a method for systematically 195comparing both antimicrobial eﬀectiveness and hazards to 196human health and the environment, which are two crucial 197parameters in sustainable material selection. It provides 198information that can help formulators make informed decisions 199about proposed alternative preservatives. It also demonstrates a 200more general strategy for evaluating both the safety and eﬃcacy 201of other ingredients used in materials and consumer products.

202

■

MATERIALS AND METHODS

203Full experimental details for chemical synthesis, microbial assays,

204hazard analysis, and computational toxicology can be found in the

205Supporting Information (SI).

206 Alkyl esters and amides of varying chain lengths of benzoic acid,

207salicylic acid, and gallic acid, along with the acids, alcohols, and amines

208that are their functional constituents, were either synthesized or

209procured from commercial sources. When possible, solventless

210

reactions and puriﬁcations using safer solvents were favored (see the

211

SIfor details).61−63To the extent possible, C0−C6, C8−C10, C12, and

212

C16benzoate esters were prepared or procured, along with C0, C3, and

213

C8amides (those omitted were not readily available and synthetically

214

impractical to prepare, and as such of minimal relevance to possible

215

widespread application). To complement these, three preservatives

216

used in personal care products (phenoxyethanol, methyl paraben, and

217

propyl paraben), octyl paraben, two preservatives used in food and

218

food packaging (sorbic acid and BHA), and two preservatives used in

219

wood and composite material products (chloroxylenol and creosol)

220

were obtained.

221

Esters of salicylic acid and gallic acid were chosen for this study,

222

because they are phenolic compounds, capable of forming a relatively

223

stable phenol radical and, therefore, are capable of acting through

224

disruption of the redox homeostasis.64They are broadly available and

225

therefore realistically applicable to industrial product formulation.

226

Salicylates were speciﬁcally chosen as a complement to widely used

4-227

hydroxybenzoates (parabens) to study whether a structural analogue

228

could be safer and an equally or more eﬀective preservative. Esters of

229

benzoic acid were included as a control, to understand the

230

antimicrobial eﬃcacy of the benzoate group in the absence of a

231

phenol group.

232

Hazard analysis was conducted by systematically reviewing

233

authoritative lists, toxicology literature, and online databases,

234

particularly Pharos (Healthy Building Network)65_{and the Hazardous}

235

Substances Data Bank (HSDB, National Library of Medicine)66,67_for

236

existing information regarding human health and environmental

237

hazard end points for representative compounds. End points were

238

grouped in a similar manner to the end points in GreenScreen,60in

239

keeping with listings by authoritative bodies. In addition, physical

240

properties of note (including log P values) and listings on restricted

241

lists or safer alternative designations are included for reference. In the

242

absence of comprehensive hazard information, computational

243

toxicology methods were used to ﬁll in data gaps with

structure-244

based predictions. Data were collected for a subset of compounds

245

using PBT Proﬁler,68_{the ADMET Predictor,}69_{Lhasa Derek,}70−72_the

246

Endocrine Disruptor Screening Program for the 21st Century,73

247

OECD QSAR Toolbox,74and Toxtree75software suites. See theSIfor

248

additional details.

249

■

RESULTS AND DISCUSSION

250 s1

Antifungal Activity. Scheme 1a displays minimum

251

inhibitory concentrations (MICs) for all of the potential Scheme 1. (a) Antifungal Activity of Compounds Tested inAspergillus brasiliensis ATCC16404;a(b) Antibacterial Activity of Compounds Tested inPseudomonas aeruginosa ATCC9027b,c

a_{MIC (mM) = minimum inhibitory concentration, where no fungal growth was visible in RPMI liquid culture measured up to 6.4 mM, with}

exceptions due to solubility limitations noted inTable S1in the SI.bMIC (mM), where no bacterial growth was detected by Abs600 measurement in Mueller−Hinton liquid culture.c_{In both panels (a0 and (b), phenolic esters and amides, alcohols, and amines are diﬀerentiated according to alkyl}

chain length (C0−C16), while other classes of preservatives are clustered on the right and are labeled.

(5)

252antimicrobial compounds tested against A. brasiliensis 253ATCC16404. Full MIC and minimum fungicidal concen-254trations (MFC) data are shown inTable S1in the SI. 255 At the concentrations tested, none of the esters or amides of 256benzoic acid showed any antifungal activity. This is not 257surprising, because these compounds contain no phenol groups 258to stabilize free radicals that might disrupt redox homeostasis.76 259In addition, none of the esters of salicylic acid exhibited 260antifungal activity; this finding indicates that the presence of 261phenol is not a guarantee of significant disruption of redox 262homeostasis, which is consistent with previous reports.77It is 263possible that this reduced activity can be attributed to steric 264inaccessibility of the hydroxyl group ortho- to the ester, 265reducing the ability to form a phenol radical and influence the 266oxidative stress response. Esters of gallic acid with alkyl chain 267lengths of four or more carbons inhibited growth of A. 268brasiliensis, with the maximum efficacy observed for octyl gallate 269(MIC = 0.1 mM); this MIC was the greatest antifungal efficacy 270observed for any preservative tested in this study, including 271compounds currently in commercial use. The correlation of 272antifungal activity with chain length for short-chain esters is 273consistent with previous studies of 4-hydroxybenzoic acid 274against A. brasiliensis.78 Pentyl gallate and octyl gallate also 275exhibited fungicidal activity (MFC = 6.4 and 0.4 mM, 276respectively; see Figure S1 in the SI for a representative 277display/bioassay, and see theSIfor calculation).

278 While none of the alcohols or benzoic acids tested in this 279study showed fungistatic or fungicidal activity up to the 280concentrations tested, the phenolic alkyl amides (N-propyl 281salicylamide, N-propyl gallamide, and N-octyl gallamide) 282demonstrated inhibition of growth at 1.6, 1.6, and 0.8 mM, 283respectively. None of these compounds exhibited fungicidal 284activity. (See theSIfor calculation.)

285 The better performance of N-propyl amides over the 286corresponding esters can be explained by the following 287arguments:

288 (1) slow rates of amide hydrolysis, which could result in 289 greater bioaccumulation of amides relative to esters (the 290 lack of activity of N-propyl benzamide suggests that the 291 inhibitory activity of these compounds is not simply due 292 to the presence of the amide group),79and

293 (2) enhanced resonance stabilization of a free radical by an 294 amide relative to an ester,58which is a property that is 295 attributable to the inherently greater stability of a 296 nitrogen-based radical over an oxygen-based radical.80 297Similar arguments explain the greater antioxidant capacities of 298amides, compared to their ester-containing analogues.58None 299of these explanations accounts for the greater antifungal activity 300of octyl gallate relative to N-octyl gallamide. However, gallates 301have been shown to inhibit alternative oxidase (AOX) activity 302in the fungal mitochondrial respiratory system, where they 303possess higher binding aﬃnities than the corresponding 304gallamides. This mechanism of action may predominate in 305the interaction of octyl gallates.81−86

306 The antifungal activity of octyl gallate (MIC 0.1 mM, MFC 3070.4 mM) compares favorably to all of the food preservatives 308tested: sorbic acid, gallic acid, and benzoic acid showed no 309activity under the conditions tested, while BHA (butylated 310hydroxyanisole), which is a phenol and therefore presumably 311also acting through the disruption of redox homeostasis,64had 312an MIC value of 0.8 mM (no fungicidal activity by calculation)

313

at the concentrations tested (seeFigure 2andTable S1in the

314

SI).

315

Similarly, octyl gallate and several of the other gallates tested

316

have comparable or better performance than the preservatives

317

conventionally used in personal care products.

2-Phenoxyetha-318

nol shows no activity at the concentrations tested, while methyl

319

and propyl 4-hydroxybenzoate, the two most widely used

320

parabens, have MIC values of 3.2 and 0.8 mM, respectively (no

321

fungicidal activity by calculation) (seeFigure 2andTable S1).

322

Interestingly, octyl paraben shows no fungistatic or fungicidal

323

activity at the concentrations tested; this is in contrast to octyl

324

gallate and other medium-chain gallates, which show signiﬁcant

325

activity. While octyl gallate has been approved for some time as

326

an antioxidant food additive in the United States,87 and the

327

broader antimicrobial activity of this compound has been

328

documented,88−90 the superior antimicrobial potency of

329

gallates, compared to other phenolic esters, is a newﬁnding.

330

Copper arsenate, which is a widely used wood preservative,

331

was not tested in this study, because of the known high acute

332

human health hazard. However, PCMX (4-chloro-3,5-xylenol)

333

and creosol (2-methoxy-4-hydroxybenzoate) were tested and

334

found to have MIC values of 0.4 and 6.4 mM, respectively;

335

PCMX shows a MFC value of 1.6 mM. Both were less eﬀective

336

than octyl gallate in these experiments.

337

Antibacterial Activity. Scheme 1b shows MICs for all of

338

the potential antimicrobial compounds tested against P.

339

aeruginosa. Full MIC data are shown inTable S2in the SI.

340

Using P. aeruginosa as an example of an industrially

341

challenging bacterium,52 we ﬁrst determined diﬀerences in

342

MICs among the three structural classes. Among benzyl, salicyl,

343

and gallyl esters, the gallyl esters exhibited the highest

344

antibacterial potency. This observation parallels the results of

345

A. brasiliensis antifungal assays. These trends are also consistent

346

with previous reports examining the antimicrobial eﬃcacy of

347

substituted benzaldehydes, where increasing−OH substitution

348

generally resulted in greater potency.91

349

We next considered the inﬂuence of R-group chain length on

350

antibacterial eﬃcacy. Within the gallate class, antimicrobial

351

eﬃcacy was high for octyl and nonyl gallate, but low for all

352

other gallates. Although imperfect solubility limited

determi-353

nation of any trends for the lighter gallates (see the SI),

354

structure−activity relationships indicating increased activity at

355

moderate chain lengths have been widely reported for fatty

356

acids and other antimicrobial substances.1,34,92

357

It has previously been reported that carboxylic acids exhibit

358

antimicrobial properties, since bacteria have a tendency to

359

exhibit higher sensitivity to bulk solution acidity than

360

molds.93,94 The three acids exhibited modest activity. This

361

ﬁnding can be attributed to the combined eﬀects of substance

362

action and broth acidiﬁcation, because the addition of acids to

363

broth caused a small but consistent shift in Mueller−Hinton

364

broth acidity, by approximately one pH unit at a concentration

365

of 1 wt %.

366

Parabens generally exhibited greater potency than isomeric

367

salicyl esters (4-hydroxybenzoates vs 2-hydroxybenzoates).

368

This result suggests that either diﬀerences in sterics, or

369

increased potential for hydrogen bonding in a

2-hydroxy-370

substituted benzoate may play a role in antimicrobial activity.

371

Octyl and nonyl gallate performed favorably, in comparison

372

to most antimicrobials used commercially. While the parabens

373

also function as better antimicrobials than other species tested,

374

microbial resistance to parabens has been documented, with

375

active eﬄux of parabens out of the cell as the proposed

(6)

376mechanism of antimicrobial resistance.95The building material 377preservative PCMX (4-chloro-3,5-xylenol) exhibited good 378potency, which was nonetheless exceeded by octyl and nonyl 379gallate. Similarly, preservatives for food packaging and home 380and personal care products such 2-phenoxyethanol, benzoic 381acid, gallic acid, and sorbic acid inhibited the growth of P. 382aeruginosa at comparable concentrations to many gallates with 383the exception of octyl gallate, which was much more potent. 384 Hazard Assessment. Several frameworks exist for compar-385ing chemical hazard to human health and the environment.96 386Our approach to searching for hazard data using authoritative 387lists and toxicology literature closely follows that of Green-388Screen,60 which is a chemical hazard assessment method 389developed by the NGO Clean Production Action. We chose 390GreenScreen as a basis for our hazard analysis because (i) its 391methodology is publically available and (ii) its approach is 392consistent with the European Chemicals Agency (ECHA) 393guidance for alternatives analysis under REACH97and the U.S. 394Environmental Protection Agency’s Design for Environment 395(DfE) chemical assessment framework.98 It uses hazard 396classiﬁcations based on the Globally Harmonized System 397(GHS) of the United Nations.99This hazard-based approach, 398without presuming speciﬁc use cases and therefore limiting

399

exposure estimates, considers a range of ecological and human

400

health hazards.

401

Our hazard assessment compiles available information on

402

human health and environmental hazards, grouped into four

403

categories by type of hazard end point, consistent with

404

groupings established by authoritative bodies.60 Group I end

405

points (carcinogenicity, mutagenicity, reproductive and

devel-406

opmental toxicity, and endocrine toxicity) are those that can

407

have serious chronic eﬀects, some of which may be heritable,

408

Group II (acute) and II* (chronic/sublethal) are hazards that

409

can potentially be moderated through exposure controls or

410

have their impacts be reduced through medical treatment. The

411

environmental fate and toxicity (PBT) category refers to

412

persistence, bioaccumulation, and toxicity in various ecosystem

413

media, with aquatic toxicity being the most commonly

414

highlighted due to the mobility of toxicants in waterways.

415

We focused our hazard assessment on representative

416

compounds from the classes we tested, considering the free

417

acid, propyl, octyl, and dodecyl esters of the three classes of

418

phenolic compounds, as well as all of the commercially used

419 t1

preservatives in the study. Table 1 summarizes the hazard

420

information that we have gathered from the literature and

421

authoritative lists.67,68,100 A full version of our hazard

422

assessment, including information on all end points and Table 1. Summary of Hazard Information for Proposed Alternative Preservatives, Compared to Preservatives Currently Used in Personal Care Products, Building Materials, and Food Packaginga

a_{In this table, data are taken from authoritative lists and literature review. Our full hazard assessment, including information sources, is available in}

the SI (Table S3). The level of hazard in each broad class is determined based on the highest hazard indicated under the subcategories of that class. Level of hazard is denoted by color: (Urgent Concern to Avoid through Low Hazard: Purple, Red, Orange, Yellow, Green); the intensity of the color is a direct indicator of the certainty (the greater the intensity of the color, the greater the certainty of the measurement). Information denoted with a hashtag (#) superscript indicates a hazard designation based primarily on computational toxicology.

(7)

423information sources for all classes of chemicals considered, is 424given in the SI (Table S3).

425 Incomplete hazard information for many of these chemicals 426decreases the certainty of their hazard designations. An absence 427of data should never be taken to imply an absence of hazard. 428However, in situations where information is available, some 429comparisons can be drawn between classes or structural 430features of chemicals, and these comparisons can be used to 431inform decisions about product formulation.

432 In cases where reliable toxicological information is not 433available from the literature or authoritative lists, computational 434toxicology models can supplement existing data with chemical 435hazard predictions.101−104Tables S4 and S5in the SI contain 436complete list of the computational toxicology metrics used and 437their outputs. We include computational results inTable 1only 438in the absence of other data. To draw relevant conclusions 439based on computations, we determined results for the relevant 440end points for several representative gallates and all non-metal-441based commercial preservatives; these are discussed in the 442following section. Chromated copper arsenate was not 443evaluated with computational tools because, in general, tools 444for the evaluation of human health impacts of metals and metal-445containing compounds are still fairly limited, and the majority 446of available tools focus on environmental impacts and the fates 447of metals.105−107 Because toxicological data abound for 448chromated copper arsenate, a full hazard assessment is still 449possible.

450 Computational Toxicology. All of the compounds 451analyzed triggered at least one structural alert or were predicted 452to cause human toxicity in QSAR models during in silico testing. 453This result is not surprising, because these compounds are 454bioactive by design. Trade-off decisions between efficacy and 455hazard reduction are sometimes necessary. When programs 456disagreed with their predictions for toxicity, we prioritized 457more-specific structural alerts (i.e., from Derek). Within those 458results, we gave precedence to the more conservative prediction 459of toxicity.

460 Sorbic acid and phenoxyethanol were the only compounds 461not predicted by any method to have endocrine toxicity, while 462the parabens, chloroxylenol, and 3-tert-butyl-4-methoxyphenol 463were all predicted to have mild to moderate endocrine activity. 464Only the OECD QSAR Toolbox predicted strong estrogen 465receptor binding for the gallates. The Toolbox uses a small set 466of structural criteria (molecular size, number of carbon rings, 467presence of OH or NH₂groups, etc.), which is diﬀerent from 468the more-speciﬁc structural criteria used by Derek to predict 469toxicity, resulting in more conservative predictions than other 470platforms.104The EPA EDSP21 platform also predicted weak 471or very weak estrogen receptor binding for creosol, the gallates, 472the parabens, and chloroxylenol based on QSARs and available 473literature; the platform did not predict estrogen receptor 474binding for phenoxyethanol or sorbic acid, and could not assess 4753-tert-butyl-4-methoxyphenol.

476 Literature evidence on the endocrine activity of gallic acid 477and the corresponding esters varies.73,108−115 Recent reviews 478from the European Food Safety Authority found that, of the 479three gallates we evaluated, only propyl gallate was of potential 480concern as an endocrine disruptor.111,112,116Octyl gallate was 481not found to aﬀect endocrine receptor activity in the 482experimental system used by Amadasi et al., but there is 483some evidence that it can inhibit 5α-reductase and thus 484inﬂuence androgen regulation.115,117 Propyl gallate did not 485inhibit 5α-reductase in this system; lauryl gallate was not

486

studied.115Generally, experimental data agree with the majority

487

of in silico models. Both indicate that, while there is abundant

488

evidence suggesting propyl gallate to be an endocrine disruptor,

489

evidence for octyl gallate is weaker and requires more thorough

490

investigation. There is no evidence that lauryl gallate acts as an

491

endocrine disruptor.110,118−121 However, suﬃcient data gaps

492

exist that no conclusions should be drawn about the relative

493

endocrine disrupting potential of gallates of diﬀerent chain

494

lengths.

495

All compounds tested were predicted to cause sensitization

496

or irritation of dermal or respiratory tissues by at least two of

497

theﬁve in silico platforms used. Phenoxyethanol triggered only

498

the OECD QSAR Toolbox alert (for ethylene glycol ethers)

499

and skin irritation alerts in Toxtree. ADMET predicted that all

500

gallates, sorbic acid, creosol, and chloroxylenol were likely skin

501

irritants and sensitizers; the parabens, phenoxyethanol, and

3-502

tert-butyl-4-methoxyphenol were predicted to be nonsensitizers.

503

Predictions of skin irritation/sensitization by Derek and

504

ADMET were in agreement for creosol and the gallates, except

505

gallic acid, which did not trigger any structural alerts in Derek.

506

Predictions from Derek diﬀered from those of ADMET for

507

sorbic acid, 3-tert-butyl-4-methoxyphenol and the parabens,

508

predicting that sorbic acid and the parabens were less likely to

509

be skin sensitizers than the gallates, while chloroxylenol and

3-510

tert-butyl-4-methoxyphenol were comparable to the gallates.

511

While there is some concern for skin sensitization with the

512

gallates, they are comparable or slightly better than existing

513

antimicrobials for this end point.

514

Overall, computational results for the gallates were

515

comparable to several currently used preservatives that are

516

known to cause skin sensitization, and they are predicted to

517

have less (if any) endocrine activity, compared to the parabens

518

and chloroxylenol.

519

Comparison of Benzoate Esters. In cases where data are

520

available, the structural similarities among various phenolic

521

esters match the hazard properties. The esters of benzoic acid,

522

salicylic acid, and gallic acid, as well as the parabens, diﬀer

523

primarily in the number and position of hydroxyl groups.

524

Where data are available, there is moderate evidence of

525

endocrine disruption (a Group I end point) for many of these

526

compounds.104,110,111,120

527

There is also moderate evidence of skin sensitization, skin

528

irritation, and/or eye irritation for most benzoate esters, based

529

on both authoritative lists and computations. These hazard end

530

points are of particular concern in home and personal care

531

products, but are also of concern for workers who handle

532

building materials or food packaging.112−114,122−124 Our

533

computational modeling supports this evidence; further

534

diﬀerentiation of the degree of skin sensitization expected

535

from a related group of compounds could be obtained with

536

modeling that is speciﬁcally optimized around skin sensitization

537

predictions, such as CADRE-SS developed by Kostal and

538

Voutchkova-Kostal125 or nuclear magnetic resonance (NMR)

539

correlations of skin permeation with spectroscopic

proper-540

ties.126

541

The potential health eﬀects of phenolic esters diﬀer with

542

changing chain length. Bioavailability has a tendency to be

543

lower with increased molecular weight, while increased

544

partitioning to lipids (increased bioconcentration factor) has

545

a tendency to accompany the presence of longer hydrophobic

546

chains. Microbial degradation of related esters can be rapid and

547

oﬀset tendencies to bioaccumulate.127 In addition, molecular

548

weight/chain length are completely independent of some

(8)

549modes of bioactivity; Uramaru et al. have shown that medium-550length parabens demonstrate higher histamine responses, 551compared to short-chain parabens.43Uramaru et al. also find 552significantly lower histamine activity for octyl salicylate and 553octyl-3-hydroxybenzoate, compared to octyl paraben, demon-554strating that substituent position can significantly change 555biological activity in otherwise very similar compounds. These 556types of distinctions underscore the importance of under-557standing the potential health effects of any chemical that may 558be introduced into commerce.

559 Comparison of Octyl Gallate and Current Commercial 560Preservatives. Octyl gallate is highly eﬀective as both an 561antibacterial and antifungal compound, outperforming other 562proposed alternative preservatives, as well as those currently 563widely used in consumer products. As such, the remainder of 564this discussion of hazard will focus on a comparison of octyl 565gallate to the materials currently used in home and personal 566care products (parabens and phenoxyethanol), food packaging 567(sorbic acid, BHA, and benzoic acid), and building materials 568(chromated copper arsenate, PCMX, and creosol) that were f2 569evaluated. Figure 2 shows spider diagrams that visualize the 570relative hazard under each major category of human health/ 571environmental end point, as well as the eﬀectiveness of each 572compound against bacterial and fungal growth as established in 573this study. As outlined above, literature data in this section is 574obtained from sources within the Pharos database65 and the 575Hazardous Substances Database (HSDB).67 Detailed source 576information is found inTable S3in the SI.

577 Home and Personal Care Products. Figure 2a shows 578spider diagram data for home and personal care products. To 579the extent that hazard data are available, the hazard traits of 580gallates mirror the structurally similar parabens. Both show 581some evidence of skin sensitization, as well as skin and eye 582irritation. As discussed above, computational results demon-583strating possible endocrine disruption are mixed for the gallates. 584The balance of evidence suggests that gallates may be less 585hazardous on this end point, compared to parabens.

586 While available data indicate a lack of carcinogenicity, 587mutagenicity, and reproductive/developmental toxicity for 588parabens and for octyl gallate, computational predictions 589suggest potential mammalian carcinogenicity and chromosomal 590damage by gallates. The redox-active structures of these

591

compounds are likely responsible for both this toxicity and

592

their higher antimicrobial activity, postulated to occur through

593

the disruption of redox homeostasis or of redox-sensitive

594

cellular components such as cellular membranes. Previous

595

studies support this hypothesis: antioxidant gene mutants of the

596

yeast S. cerevisae demonstrated high susceptibility to treatment

597

with known disruptors of redox homeostasis,128,129 including

598

octyl gallate.64By taking advantage of biochemical diﬀerences

599

between complex eukaryotes and the microbes responsible for

600

spoilage,130we can adjust molecular properties of redox-active

601

molecules to favor toxicity in simpler organisms but not in

602

humans. This has been achieved with the oxidative antifungal

603

drug amphotericin B (AmB), which binds to fungal but not

604

human cell membranes, allowing for selective toxicity to fungal

605

cells.47,48

606

Characteristics such as potential redox activity of the

607

compounds highlight the challenges and opportunities for

608

making bioactive molecules that are inherently safer for

609

humans.111,113,124 To comprehensively assess all of these

610

compounds for genetic toxicity, OECD guidelines recommend

611

extensive in vitro and in vivo testing, including chromosomal

612

tests and oral dosing in rats;131,132this assessment has not been

613

completed for these compounds.

614

Phenoxyethanol is a widely used “safer” alternative to

615

parabens and other conventional preservatives in home and

616

personal care products; however, it is also a suspected human

617

reproductive and developmental toxicant.133 Unlike octyl

618

gallate, there are no direct indications of mutagenicity, although

619

an Ames test (a predictor of mutagenic activity)134 for

620

analogous butoxyethanol suggests possible mutagenicity.135

621

Unlike the parabens, phenoxyethanol is not ﬂagged on any

622

authoritative list as a skin sensitizer or irritant; computational

623

predictions are consistent with this observation. Skin

624

sensitization/irritation is a major concern to consumers of

625

home and personal care products, but also are a concern for

626

workers in industrial cleaning, as well as food handling and

627

manufacturing. Computational data do not predict that

628

phenoxyethanol would have signiﬁcant estrogenic or

andro-629

genic activity. This information suggests that phenoxyethanol is

630

a safer choice than parabens, in terms of endocrine disruption;

631

it is likely to be comparable to or better than octyl gallate for

632

endocrine disruption as well.

Figure 2.Spider diagrams comparing hazard and eﬀectiveness of octyl gallate to common commercial preservatives in three applications: (a) home and personal care products, (b) food packaging, and (c) building materials. Smaller values (closer to the center of the spiderweb) are indicative of better performance on each of the six metrics, i.e., lower MIC and MFC indicate greater antimicrobial eﬀectiveness, and lower hazard in each broad category.

(9)

633 Food Packaging.Figure 2b shows spider diagram data for 634food packaging. Among the preservatives used in food 635packaging, sorbic acid is the only compound examined that 636has low acute toxicity; octyl gallate is acutely toxic but is also 637approved for use in food (as an antioxidant), suggesting that all 638of these compounds are toxic at concentrations significantly 639above that found in packaging. At least one study indicated that 640potassium sorbate (the potassium salt of sorbic acid) is 641genotoxic to human lymphocytes.136 All of these compounds 642are potential skin sensitizers or irritants. BHA stands out as a 643compound that is a priority for substitution, because of its 644classification as a probable carcinogen by the U.S. National 645Institutes of Health (NIH) and International Agency for 646Research on Cancer (IARC). Computational predictions 647suggest that octyl gallate may plausibly cause chromosome 648damage in mammals.137 These same predictive methods also 649suggest that BHA is a plausible carcinogen; more research is 650needed to understand the relative genotoxic hazards of octyl 651gallate, as opposed to BHA, although other aspects of 652genotoxicity (such as developmental toxicity) are unlikely for 653octyl gallate, based on computational toxicology. BHA and 654benzoic acid both appear on authoritative lists as potential 655endocrine disruptors; computational data support the potential 656for BHA to act as both an estrogen and androgen mimic. 657 Building Materials. Figure 2c shows spider diagram data 658for building materials. If it were used in pure form, creosol 659would potentially be the least harmful of the widely used 660preservatives used in building materials; where information is 661available, it is less hazardous than octyl gallate on nearly all end 662points. However, its typical use is as a component of creosote, 663which is recognized by IARC and the U.S. Environmental 664Protection Agency (EPA) as a probable carcinogen, in addition 665to having acute toxic effects and causing damage to mucous 666membranes.67 Little information is available in the literature 667about the Group I health end points associated with creosol 668itself; computational models indicate that, unlike octyl gallate, it 669does not trigger alerts for potential genotoxicity. Creosol is not 670predicted to be an estrogen or androgen mimic, but it does 671trigger alerts in Derek for possible hepatotoxicity. Like most 672other compounds considered in this study, creosol is a skin and 673eye irritant.

674 From a human and environmental toxicity perspective, 675additional clear-cut cases for the need for safer alternative 676preservatives in building materials are PCMX and chromated 677copper arsenate. Similar to other currently used preservatives, 678PCMX is acutely toxic, a potential endocrine disrupter, and a 679skin sensitizer and irritant. Copper arsenate (used as chromated 680copper arsenate, because chromium improves the binding of 681the copper arsenate to wood and composite materials) is a 682known carcinogen (IARC Group 1), a mutagen, a reproductive 683toxicant, and developmental neurotoxicant, and is acutely toxic. 684In addition to various health concerns, both PCMX and copper 685arsenate are persistent in the environment: PCMX, because of 686the low biodegradability of the organohalogen functionality, 687and copper arsenate, because it is an inorganic compound and, 688therefore, inherently persistent. Replacing each of these 689preservatives with safer alternatives such as octyl gallate or 690phenoxyethanol has the potential for positive human and 691environmental health and safety implications.

692

■

CONCLUSIONS

693As a potential alternative preservative for home and personal 694care products, composite building materials, and food

pack-695

aging, octyl gallate (octyl 3,4,5-trihydroxybenzoate) shows

696

promising antimicrobial activity against representative mold

697

and bacteria, with greater eﬃcacy than common commercial

698

preservatives currently used in these applications. While not as

699

striking, other hydroxyl-substituted benzoic acids also show

700

some antimicrobial activity, particularly against bacteria. Based

701

on these results, we conclude that several design parameters

702

exert a signiﬁcant eﬀect on the antimicrobial potencies of

703

resulting substances:

704

(1) alkyl chain lengththis eﬀect may be due to surfactant

705

action, association with biomolecules, and changes in

706

hydrophilicity or aqueous partitioning;

707

(2) position of phenol substitution (speciﬁcally,

2-hydrox-708

ybenzoates versus 4-hydroxybenzoates); and

709

(3) the number of hydroxybenzoates (singly versus triply

710

hydroxylated benzoates).

711

These results are promising, because, although gallates are

712

not completely free of potential hazards to human health and

713

the environment, a systematic screening of authoritative lists

714

and primary literature, supplemented by computational

715

toxicology, suggests that octyl gallate and its structural

716

analogues have hazard proﬁles that compare favorably to

717

those of many commercial preservatives. Their potent

718

antimicrobial and antifungal properties may make them

719

eﬀective at lower concentrations than current-use preservatives,

720

reducing potential exposure. Taken together, these results

721

indicate that it would be worthwhile to further explore the use

722

of octyl gallate in consumer products, including health and

723

eﬃcacy testing on formulated products as a single ingredient

724

and as part of a mixture of preservatives acting through

725

complementary mechanisms. The propensities of octyl gallate

726

and other compounds for skin sensitization, genotoxicity and

727

endocrine activity through warrant further testing before they

728

can be unconditionally recommended for use in consumer

729

products. However, this evaluation is an important step toward

730

incorporating inherently safer preservatives and other

con-731

stituent chemicals in product design and formulation.

732

■

ASSOCIATED CONTENT

733

*

S Supporting Information

734

The Supporting Information is available free of charge on the

735 ACS Publications website at DOI:

10.1021/acssusche-736 meng.7b00374.

737

Details and results of chemical syntheses, microbial

738

assays, hazard analysis, and computational toxicology

739 (PDF) 740 (XLSX) 741 (XLSX) 742 (XLSX) 743

■

AUTHOR INFORMATION 744 Corresponding Author 745 *E-mail:hbuckley@berkeley.edu. 746 ORCID 747 Heather L. Buckley:0000-0001-7147-0980 748 Author Contributions † 749

These authors contributed equally to this work.

750

Funding

751

This work was supported by the Development Impact Lab

752

(USAID Cooperative Agreement No. AID-OAA-A-13-00002),

753

USAID Higher Education Solutions Network; the U.S.−India

(10)

754Science and Technology Endowment Fund; the United States 755Department of Agriculture (USDA-ARS CRIS Project No. 7562030-42000-039-00); and the National Science Foundation 757(NSF SEES Project No. 1415417; NSF IGERT Systems 758Approach to Green Energy (SAGE) No. 1144885).

759Notes

760The authors declare no competingﬁnancial interest.

761

■

ACKNOWLEDGMENTS

762The authors thank: Dr. John Arnold, Jessica Ziegler, and Dr. 763Ben Kriegel (UC Berkeley) for access to laboratory facilities 764and equipment; Dr. Olivia Lee for helpful conversations on 765chemical synthesis; Dr. Meg Schwarzman (UC Berkeley) for 766helpful review and comments on the manuscript; Emery 767Wilson (UC Berkeley) for chemicals; Raul Leal and the 768Sarpong Group (UC Berkeley) for chemicals and laboratory 769equipment; Dr. Kaj Johnson (Method), Dr. Dominic Wong 770(USDA), and Dr. Larry Weiss for helpful discussions; Ivette 771Quintanilla and Tom McKeag (UC Berkeley) for ongoing 772logistical support; the Greener Solutions Program (UC 773Berkeley); and the Healthy Building Network PHAROS 774Project for database access.

775

■

REFERENCES

(1)

776 Kegel, L. L.; Szabó, L. Z.; Polt, R.; Pemberton, J. E. Alkyl 777Melibioside and Alkyl Cellobioside Surfactants: Effect of Sugar

778Headgroup and Alkyl Chain Length on Performance. Green Chem.

7792016, 18, 4446−4460.

(2)

780 Chen, G.; van Straalen, N. M.; Roelofs, D. The Ecotoxicogenomic

781Assessment of Soil Toxicity Associated with the Production Chain of 7822,5-Furandicarboxylic Acid (FDCA), a Candidate Bio-Based Green

783Chemical Building Block. Green Chem. 2016, 18, 4420−4431. (3)

784 Youssef, A. M.; El-Sayed, S.; El-Sayed, H.; Salama, H.; Dufresne, 785A. Influence of Packaging Material on Shelf Life of Egyptian Soft

786White Cheese. Carbohydr. Polym. 2016, 151, 9−19.

(4)

787 United States Environmental Protection Agency (USEPA). 788Presidential Green Chemistry Challenge: 2014 Designing Greener

789Chemicals Award, The Solberg Company: RE-HEALING Foam 790ConcentratesEﬀective Halogen-Free Fireﬁghting; available via the

791Internet at: https://www.epa.gov/greenchemistry/presidential-green-792chemistry-challenge-2014-designing-greener-chemicals-award (ac-793cessed Nov. 26, 2016).

(5)

794 Anastas, N. D. Connecting Toxicology and Chemistry to Ensure 795Safer Chemical Design. Green Chem. 2016, 18, 4325−4331.

(6)

796 Green Chemistry & Commerce Council (GC3). Need Statement

797& Development Criteria for New Preservatives for Personal Care & 798Household Products; available via the Internet at: http://

799greenchemistryandcommerce.org/projects/preservatives-project, Low-800ell, MA, 2015.

(7)

801 Gramatica, P.; Cassani, S.; Sangion, A. Aquatic Ecotoxicity of

802Personal Care Products: QSAR Models and Ranking for Prioritization 803and Safer Alternatives’ Design. Green Chem. 2016, 18, 4393−4406.

(8)

804 Safety and Eﬀectiveness of Consumer Antiseptics; Topical 805Antimicrobial Drug Products for Over-the-Counter Human Use; U.S. 806Food and Drug Administration (FDA), Department of Health and

807Human Services: Washington, DC, 2016.

(9)

808 Gilbertson, L. M.; Ng, C. A. Evaluating the Use of Alternatives

809Assessment To Compare Bulk Organic Chemical and Nanomaterial 810Alternatives to Brominated Flame Retardants. ACS Sustainable Chem. 811Eng. 2016, 4, 6019−6030.

(10)

812 Gaulke, C. A.; Barton, C. L.; Proffitt, S.; Tanguay, R. L.; 813Sharpton, T. J. Triclosan Exposure Is Associated with Rapid

814Restructuring of the Microbiome in Adult Zebrafish. PLoS One 8152016, 11 (5), e0154632.

(11)

816 McNamara, P. J.; LaPara, T. M.; Novak, P. J. The Impacts of

817Triclosan on Anaerobic Community Structures, Function, and

818

Antimicrobial Resistance. Environ. Sci. Technol. 2014, 48 (13),

819

7393−7400.

(12) Syed, A. K.; Ghosh, S.; Love, N. G.; Boles, B. R. Triclosan 820 821

Promotes Staphylococcus aureus Nasal Colonization. mBio 2014, 5 (2),

822

e01015-13.

(13) Barnes, R. A.; Gow, N. A. R.; Denning, D. W.; May, R. C.; 823 824

Haynes, K. Antifungal Resistance: More Research Needed. Lancet

825

2014, 384 (9952), 1427.

(14) Aigner, M.; Lass-Flörl, C. Treatment of Drug-Resistant826 827

Aspergillus Infection. Expert Opin. Pharmacother. 2015, 16 (15),

828

2267−2270.

(15)Howard, S. J.; Cerar, D.; Anderson, M. J.; Albarrag, A.; Fisher, 829 830

M. C.; Pasqualotto, A. C.; Laverdiere, M.; Arendrup, M. C.; Perlin, D.

831

S.; Denning, D. W. Frequency and Evolution of Azole Resistance in

832

Aspergillus Fumigatus Associated with Treatment Failure 1. Emerging

833

Infect. Dis. 2009, 15 (7), 1068−1076.

(16) Lugauskas, A.; Levinskaite, L.; Peciulyte, D. Micromycetes as 834 835

Deterioration Agents of Polymeric Materials. Int. Biodeterior. Biodegrad.

836

2003, 52, 233−242.

(17) Gu, J.-D. Microbiological Deterioration and Degradation of837 838

Synthetic Polymeric Materials: Recent Research Advances. Int.

839

Biodeterior. Biodegrad. 2003, 52, 69−91.

(18) Shah, A. A.; Hasan, F.; Hameed, A.; Ahmed, S. Biological840 841

Degradation of Plastics: A Comprehensive Review. Biotechnol. Adv.

842

2008, 26, 246−265.

(19)Gu, J. Microbial Colonization of Polymeric Materials for Space 843 844

Applications and Mechanisms of Biodeterioration: A Review. Int.

845

Biodeterior. Biodegrad. 2007, 59, 170−179.

(20)Little, B.; Wagner, P.; Mansfeld, F. Microbiologically Influenced 846 847

Corrosion of Metals and Alloys. Int. Mater. Rev. 1991, 36 (6), 253−

848

272.

(21) Roca, I.; Akova, M.; Baquero, F.; Carlet, J.; Cavaleri, M.; 849 850

Coenen, S.; Cohen, J.; Findlay, D.; Gyssens, I.; Heure, O. E.;

851

Kahlmeter, G.; Kruse, H.; Laxminarayan, R.; Liébana, E.;

López-852

Cerero, L.; MacGowan, A.; Martins, M.; Rodríguez-Baño, J.; Rolain,

J.-853

M.; Segovia, C.; Sigauque, B.; Tacconelli, E.; Wellington, E.; Vila, J.

854

The Global Threat of Antimicrobial Resistance: Science for

855

Intervention. New Microbes New Infect. 2015, 6 (April), 22−29.

(22)Smith, R.; Coast, J. The True Cost of Antimicrobial Resistance. 856 857

BMJ. 2013, 1493, f1493.

(23)Cabello, F. C.; Godfrey, H. P.; Tomova, A.; Ivanova, L.; Dölz, 858 859

H.; Millanao, A.; Buschmann, A. H. Antimicrobial Use in Aquaculture

860

Re-Examined: Its Relevance to Antimicrobial Resistance and to Animal

861

and Human Health. Environ. Microbiol. 2013, 15, 1917−1942.

(24)Campbell, B. C.; Chan, K. L.; Kim, J. H. Chemosensitization as862 863

a Means to Augment Commercial Antifungal Agents. Front. Microbiol.

864

2012, 3, 79.

(25) Livermore, D. M. Multiple Mechanisms of Antimicrobial865 866

Resistance in Pseudomonas Aeruginosa: Our Worst Nightmare? Clin.

867

Infect. Dis. 2002, 34, 634−640.

(26) Anastas, P. T.; Warner, J. C. Green Chemistry: Theory and868 869

Practice; Oxford University Press: New York, 1998.

(27) Anastas, P. T.; Kirchhoff, M. M. Origins, Current Status, and 870 871

Future Challenges of Green Chemistry. Acc. Chem. Res. 2002, 35 (9),

872

686−694.

(28)Melnikov, F.; Kostal, J.; Voutchkova-Kostal, A.; Zimmerman, J. 873 874

B.; Anastas, P. T. Assessment of Predictive Models for Estimating the

875

Acute Aquatic Toxicity of Organic Chemicals. Green Chem. 2016, 18,

876

4432−4445.

(29)Schwarzman, M. R.; Ackerman, J. M.; Dairkee, S. H.; Fenton, S. 877 878

E.; Johnson, D.; Navarro, K. M.; Osborne, G.; Rudel, R. A.; Solomon,

879

G. M.; Zeise, L.; Janssen, S. Screening for Chemical Contributions to

880

Breast Cancer Risk: A Case Study for Chemical Safety Evaluation.

881

Environ. Health Perspect. 2015, 123 (12), 1255−1264.

(30)Kostal, J.; Voutchkova-Kostal, A.; Anastas, P. T.; Zimmerman, J. 882 883

B. Identifying and Designing Chemicals with Minimal Acute Aquatic

884

Toxicity. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, 6289−6294.

(11)

(31)

885 Faludi, J.; Hoang, T.; Gorman, P.; Mulvihill, M. Aiding 886Alternatives Assessment with an Uncertainty-Tolerant Hazard Scoring

887Method. J. Environ. Manage. 2016, 182, 111−125. (32)

888 Api, A. M.; Belsito, D.; Bruze, M.; Cadby, P.; Calow, P.; Dagli, 889M. L.; Dekant, W.; Ellis, G.; Fryer, A. D.; Fukayama, M.; Griem, P.;

890Hickey, C.; Kromidas, L.; Lalko, J. F.; Liebler, D. C.; Miyachi, Y.; 891Politano, V. T.; Renskers, K.; Ritacco, G.; Salvito, D.; Schultz, T. W.;

892Sipes, I. G.; Smith, B.; Vitale, D.; Wilcox, D. K. Criteria for the 893Research Institute for Fragrance Materials, Inc. (RIFM) Safety

894Evaluation Process for Fragrance Ingredients. Food Chem. Toxicol. 8952015, 82, S1−S19.

(33)

896 Coish, P.; Brooks, B. W.; Gallagher, E. P.; Kavanagh, T. J.;

897Voutchkova-Kostal, A.; Zimmerman, J. B.; Anastas, P. T. Current 898Status and Future Challenges in Molecular Design for Reduced

899Hazard. ACS Sustainable Chem. Eng. 2016, 4, 5900−5906. (34)

900 Russell, A. D. Principles of Antimicrobial Activity and 901Resistance. In Disinfection, Sterilization, and Preservation, Fifth Edition;

902Block, S. S., Ed.; Lippincott Williams & Wilkins: Philadelphia, PA, 9032001; pp 31−56.

(35)

904 Guillén, F.; Evans, C. S. Anisaldehyde and Veratraldehyde 905Acting as Redox Cycling Agents for H2O2 Production by Pleurotus

906eryngii. Appl. Environ. Microbiol. 1994, 60, 2811−2817.

(36)

907 Jacob, C. A Scent of Therapy: Pharmacological Implications of

908Natural Products Containing Redox-Active Sulfur Atoms. Nat. Prod. 909Rep. 2006, 23, 851−863.

(37)

910 Chatre, L.; Ricchetti, M. Are Mitochondria the Achilles’ Heel of 911the Kingdom Fungi? Curr. Opin. Microbiol. 2014, 20, 49−54.

(38)

912 Walsh, S. E.; Maillard, J.; Russell, A. D.; Catrenich, C. E.; 913Charbonneau, D. L.; Bartolo, R. G. Activity and Mechanisms of Action

914of Selected Biocidal Agents on Gram-Positive and -Negative Bacteria.

915J. Appl. Microbiol. 2003, 94, 240−247.

(39)

916 Lü, J.-M.; Lin, P. H.; Yao, Q.; Chen, C. Chemical and Molecular 917Mechanisms of Antioxidants: Experimental Approaches and Model 918Systems. J. Cell Mol. Med. 2010, 14 (4), 840−860.

(40)

919 Szentandrássy, N.; Szentesi, P.; Magyar, J.; Nánási, P. P.; 920Csernoch, L. Effect of Thymol on Kinetic Properties of Ca and K

921Currents in Rat Skeletal Muscle. BMC Pharmacol. 2003, 3, 9. (41)

922 Kim, J. H.; Chan, K. L.; Mahoney, N.; Campbell, B. C. 923Antifungal Activity of Redox-Active Benzaldehydes That Target

924Cellular Antioxidation. Ann. Clin. Microbiol. Antimicrob. 2011, 10, 23. (42)

925 Report on Carcinogens, Thirteenth Edition; National Toxicology 926Program (NTP): Research Triangle Park, NC, 2014.

(43)

927 Uramaru, N.; Inoue, T.; Watanabe, Y.; Shigematsu, H.; Ohta, S.;

928Kitamura, S. Structure−Activity Relationship of a Series of 17 Parabens 929and Related Compounds for Histamine Release in Rat Peritoneal Mast

930Cells and Skin Allergic Reaction in Guinea Pigs. J. Toxicol. Sci. 2014, 93139 (1), 83−90.

(44)

932 Ohgami, N.; Yamanoshita, O.; Thang, N. D.; Yajima, I.; Nakano,

933C.; Wenting, W.; Ohnuma, S.; Kato, M. Carcinogenic Risk of 934Chromium, Copper, and Arsenic in CCA-Treated Wood. Environ.

935Pollut. 2015, 206, 456−460. (45)

936 Grady, R. W.; Bienen, E. J.; Clarkson, A. B. J. Esters of 3,4-937Dihydroxybenzoic Acid, Highly Effective Inhibitors of the

Sn-938Glycerol-3-Phosphate Oxidase of Trypanosoma Brucei Brucei. Mol. 939Biochem. Parasitol. 1986, 21, 55−63.

(46)

940 Kubo, I.; Fujita, K.; Nihei, K. Molecular Design of

Multifunc-941tional Antibacterial Agents Against Methicillin Resistant Staphylococcus 942aureus (MRSA). Bioorg. Med. Chem. 2003, 11, 4255−4262.

(47)

943 Huang, W.; Zhang, Z.; Han, X.; Tang, J.; Wang, J.; Dong, S.;

944Wang, E. Ion Channel Behavior of Amphotericin B in Sterol-Free and 945Cholesterol- or Ergosterol-Containing Supported Phosphatidylcholine

946Bilayer Model Membranes Investigated by Electrochemistry and 947Spectroscopy. Biophys. J. 2002, 83, 3245−3255.

(48)

948 González-Párraga, P.; Sánchez-Fresneda, R.; Zaragoza, O.; 949Argüelles, J. C. Amphotericin B Induces Trehalose Synthesis and 950Simultaneously Activates an Antioxidant Enzymatic Response in 951Candida albicans. Biochim. Biophys. Acta, Gen. Subj. 2011, 1810, 777− 952783.

(49) Cao, H.; Jung, M.; Stamatoyannopoulos, G. Hydroxamide953 954

Derivatives of Short-Chain Fatty Acid Have Erythropoietic Activity

955

and Induce G Gene Expression In Vivo. Exp. Hematol. 2005, 33,

956

1443−1449.

(50)Seyran, M.; Brenneman, T. B.; Stevenson, K. L. In Vitro Toxicity 957 958

of Alternative Oxidase Inhibitors Salicylhydroxamic Acid and Propyl

959

Gallate on Fusicladium Effusum. J. Pest. Sci. 2010, 83, 421−427. (51) Janes, H. W.; Wiest, S. C. Inhibition of O2 Consumption960

961

Resistant to Cyanide and Its Development by N-Propyl Gallate and

962

Salicyihydroxamic Acid. Plant Physiol. 1982, 70, 853−857.

(52) Moser, C. L.; Meyer, B. K. Comparison of Compendial 963 964

Antimicrobial Effectiveness Tests: A Review. AAPS PharmSciTech

965

2011, 12, 222−226.

(53)Byrne, A. P.; Hart-Cooper, W. M.; Johnson, K.; Stanker, L. H.;966 967

Wong, D. W. S.; Orts, W. J. Testing for Sustainable Preservatives.

968

HAPPI 2016 (January) (available via the Internet at: http://www.

969

happi.com/issues/2016-01/).

(54) Manikandan, P.; Varga, J.; Kocsubé, S.; Revathi, R.; Anita, R.; 970 971

Dóczi, I.; Németh, T. M.; Narendran, V.; Vágvölgyi, C.; Bhaskar, M.;

972

Manoharan, C.; Samson, R. A.; Kredics, L. Keratitis Caused by the

973

Recently Described New Species Aspergillus Brasiliensis: Two Case

974

Reports. J. Med. Case Rep. 2010, 4, 68.

(55) American Type Culture Collection. ATCC (American Type 975 976

Culture Collection).

(56)Howard, S. J.; Harrison, E.; Bowyer, P.; Varga, J.; Denning, D. 977 978

W. Cryptic Species and Azole Resistance in the Aspergillus Niger

979

Complex. Antimicrob. Agents Chemother. 2011, 55 (10), 4802−4809. (57) Howard, S. J.; Arendrup, M. C. Acquired Antifungal Drug 980

981

Resistance in Aspergillus Fumigatus: Epidemiology and Detection.

982

Med. Mycol. 2011, 49 (Suppl. 1), S90−S95.

(58)Son, S.; Lewis, B. A. Free Radical Scavenging and Antioxidative983 984

Activity of Caffeic Acid Amide and Ester Analogues: Structure−

985

Activity. J. Agric. Food Chem. 2002, 50, 468−472.

(59)Sonnenburg, A.; Schreiner, M.; Stahlmann, R. Assessment of the 986 987

sensitizing potency of preservatives with chance of skin contact by the

988

loose-fit coculture-based sensitization assay (LCSA). Arch. Toxicol.

989

2015, 89, 2339−2344.

(60) Clean Production Action. GreenScreen for Safer Chemicals 990 991

Hazard Assessment Guidance, Version 1.3, March 2016; available via the

992

Internet at: http://www.greenscreenchemicals.org/static/ee_images/

993

uploads/resources/1_GreenScreen_Guidance_v13_2016_3_8.pdf. (61) Prat, D.; Hayler, J.; Wells, A. A Survey of Solvent Selection994

995

Guides. Green Chem. 2014, 16, 4546−4551.

(62) Prat, D.; Wells, A.; Hayler, J.; Sneddon, H.; Mcelroy, C. R.; 996 997

Abou-Shehada, S.; Dunn, P. J. CHEM21 Selection Guide of

Classical-998

and Less Classical-Solvents. Green Chem. 2016, 18, 288−296.

(63)Lipshutz, B. H.; Gallou, F.; Handa, S. Evolution of Solvents in 999 1000

Organic Chemistry. ACS Sustainable Chem. Eng. 2016, 4, 5838−5849. (64)Kim, J. H.; Haff, R. P.; Faria, N. C.; Martins, M.; de, L.; Chan, K. 1001

1002

L.; Campbell, B. C. Targeting the Mitochondrial Respiratory Chain of

1003

Cryptococcus through Antifungal Chemosensitization: A Model for

1004

Control of Non-Fermentative Pathogens. Molecules 2013, 18, 8873−

1005

8894.

(65)The Pharos Project of the Healthy Building Network. Chemical 1006 1007

& Material Library; available via the Internet at:http://pharosproject.

1008

net/material/chemical(accessed Feb. 23, 2016).

(66) Fonger, G. C.; Hakkinen, P.; Jordan, S.; Publicker, S. The 1009 1010

National Library of Medicine’s (NLM) Hazardous Substances Data

1011

Bank (HSDB): Background, Recent Enhancements and Future Plans.

1012

Toxicology 2014, 325, 209−216.

(67) Toxnet Toxicology Data Network. U.S. National Library of 1013 1014

Medicine. Hazardous Substances Data Bank (HSDB); available via the

1015

Internet at:Https://toxnet.nlm.nih.gov/newtoxnet/hsdb.htm; 2016. (68) Environmental Health Analysis Center; Oﬃce of Chemical 1016

1017

Safety and Pollution Prevention; U.S. Environmental Protection

1018

Agency. PBT Proﬁler; available via the Internet at: Http://

1019

pbtproﬁler.net; 2016.