Zn Supplement-Antagonized Cadmium-Induced Cytotoxicity in

Apr 3, 2019 - By continuing to use the site, you are accepting our use of cookies. Read the ACS privacy policy. CONTINUE. pubs logo. 1155 Sixteenth St...
0 downloads 0 Views 1MB Size
Subscriber access provided by UNIV OF LOUISIANA

Food Safety and Toxicology

Zn supplement antagonizes cadmium induced cytotoxicity in macrophages in vitro: involvement of cadmium bioaccumulation and metallothioneins regulation Ding Zhang, Ting Zhang, Jingying Liu, Jianshan Chen, Ying Li, Guanbao Ning, Nairui Huo, Wenxia Tian, and Haili Ma J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.9b00232 • Publication Date (Web): 03 Apr 2019 Downloaded from http://pubs.acs.org on April 4, 2019

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 53

Journal of Agricultural and Food Chemistry

Zn supplement antagonizes cadmium induced cytotoxicity in macrophages in vitro: involvement of cadmium bioaccumulation and metallothioneins regulation Ding Zhanga; Ting Zhanga; Jingying Liub; Jianshan Chena; Ying Lia; Guanbao Ninga; Nairui Huoa; Wenxia Tiana; Haili Ma a,* a

College of Animal Science and Veterinary Medicine, Shanxi Agricultural

University, Taigu, PR China;

b

Function Laboratory, Shanxi Medical

University, Taiyuan, PR China.

* Corresponding author: Name: Haili Ma Email: [email protected] Telephone: +86 150 0344 5899

1

ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry

1

Abstract: Cadmium (Cd) is a toxic metal leading to multiple organ

2

damages. Zinc (Zn) was reported as a potential antagonist against Cd

3

toxicity. The present study investigates the antagonistic effect of Zn (20

4

μM) on Cd (20 or 50 μM) cytotoxicity in macrophages in vitro. Results

5

shows that Cd exposure caused dose-dependent morphologic and

6

ultrastructural alterations in RAW 264.7 macrophages. Zn supplement

7

significantly inhibited Cd cytotoxicity in RAW 264.7 or HD-11

8

macrophages by mitigating cell apoptosis, excessive ROS output, and

9

mitochondrial membrane depolarization. Notably, Zn supplement for 12 h

10

remarkably prevented intracellular Cd2+ accumulation in 20 μM (95.99±

11

9.93 vs 29.64±5.08 ng/106 cells; P=0.0008) or 50 μM Cd (179.78±28.66

12

vs 141.62 ± 22.15 ng/106 cells; P=0.003) exposed RAW 264.7 cells.

13

Further investigation found that Cd promoted metallothioneins (MTs) and

14

metal regulatory transcription factor 1 (MTF-1) expression in RAW 264.7

15

macrophages. 20 μM Zn supplement dramatically enhanced MTs and

16

MTF-1 levels in Cd exposed RAW 264.7 macrophages. Intracellular Zn2+

17

chelation or MTF-1 gene silencing inhibited MTs synthesis in Cd exposed

18

RAW 264.7 macrophages, which was accompanied by the declined

19

expression of MTF-1, indicating that regulation of Zn on MTs was partially

20

achieved by MTF-1 mobilization. In conclusion, this study demonstrates

21

the antagonism of Zn against Cd cytotoxicity in macrophages and reveals

22

its antagonistic mechanism by preventing Cd2+ bioaccumulation and 2

ACS Paragon Plus Environment

Page 2 of 53

Page 3 of 53

Journal of Agricultural and Food Chemistry

23

promoting MTs expression.

24

Key words: Cadmium; Zinc; Macrophage; Cytotoxicity; Metallothioneins

25

Introduction

26

Cadmium (Cd) is widely used in the industrial manufacturing because

27

of its good malleability, ductility and resistance to corrosion 1. However,

28

as a toxic metal, extensive Cd utilization leads to serious Cd

29

ecological environment pollution which poses a threat to public health.

30

Food chain is the main source of Cd for humans and animals because

31

cadmium ions (Cd2+) is easily absorbed by plants and cereals 2. Apart from

32

food source, occupational Cd exposure, cosmetic products and tobacco

33

smoking significantly add to the body burden of Cd. One cigarette was

34

detected to contain 1-2 μg Cd and approximately 50% of the inhaled Cd

35

can be absorbed in the body 3. Cd exposure gives rise to a wide range of

36

toxic effects to the body such as hepatotoxicity, nephrotoxicity,

37

neurotoxicity, carcinogenicity, immunotoxicity and metabolic disorders 4.

38

Cd exerts biological toxicity mainly in the form of Cd2+. Cd2+ interferes

39

with the homeostasis of key intracellular molecules that elicit cellular

40

signaling cascades dictating cell function and fate. For example, high

41

magnitude or persistence of Cd stress induces a significant increase of

42

cytosolic calcium ion (Ca2+) and reactive oxygen species (ROS), which are

43

two key second messengers 5. Ca2+ and Cd2+ have quite similar ionic radius.

44

Increased intracellular Cd2+ could displace Ca2+ from its physiological 3

ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry

45

binding sites and results in cytosolic Ca2+ accumulation 6. Cd2+ is not a

46

Fenton metal, while can induce ROS production through mediating

47

endogenous iron-dependent hydroxyl radical generation 7. Moreover, Cd2+

48

shows high affinity to thiols and can alter the cellular redox status by

49

reacting with exogenous and endogenous antioxidants 8. Severe ROS/Ca2+

50

signals activate cell death effectors (e.g., caspases, Jun N-terminal kinase

51

(JNK), p38 MAP-kinase (p38) and ceramides), causing irreversible

52

damages of organelles like mitochondria and endoplasmic reticulum (ER)

53

5, 9.

54

mitochondrial ROS formation which contribute to amplify the pathological

55

signaling pathways 10. All these processes disrupt cellular function, gene

56

transcription, transport and metabolism, causing manifestation as oxidative

57

stress, apoptosis or necrosis in Cd treated cells 9.

ER stress and mitochondria dysfunction facilitate ER Ca2+ release and

58

Metallothioneins (MTs) are a family of low molecular weight proteins

59

found in various mammalian and non-mammalian tissues. MTs are

60

characterized as metal-binding protein due to their numerous cysteine

61

residues. In mammals, MTs predominantly bind to zinc ion (Zn2+) in

62

physiological condition and serve as Zn2+ reservoirs for apoprotein that act

63

in cellular signaling and transcriptional regulation 11. As a result, MTs may

64

affect numerous cellular processes including gene transcription, apoptosis,

65

proliferation and differentiation by maintaining intracellular Zn2+

66

homeostasis. Under oxidative stress conditions, MTs provide cells with a 4

ACS Paragon Plus Environment

Page 4 of 53

Page 5 of 53

Journal of Agricultural and Food Chemistry

67

primitive antioxidant defense. Sulfhydryl clusters of MTs are suggested to

68

be primary targets for the reaction of oxidants and electrophiles

69

addition, mobilization of Zn2+ from MTs by oxidative stress was also

70

considered as the nonnegligible factor of protection

71

major role in the detoxification of Cd poisoning. Environmentally exposure

72

to Cd induced significant up-regulation of MTs gene expression in people

73

peripheral lymphocytes 14. Transgenic mice overexpressing MTs presented

74

more resistance to Cd lethality and hepatotoxicity, whereas MT-knockout

75

mice showed higher susceptibility to Cd induced liver injury compared to

76

wild-type mice 15. Research in fish demonstrated that Cd exposure induced

77

profound MTs protein synthesis in the liver, kidney, gill and brain of O.

78

mossambicus (Tilapia) 16. MTs induction by Cd were also fully observed

79

in other organs or cells on different experimental models 17-18. MTs show

80

high binding affinity for Cd2+ and could convert toxic Cd2+ into nontoxic

81

MT-Cd 19. In addition to metal sequestration, MTs may inhibit the toxic

82

effects of Cd by decreasing the uptake of Cd2+ into cells 20. In mammals,

83

four isoforms of MTs designated MT-1 to MT-4 have been identified.

84

Among these isoforms, MT-1 and MT-2 represent the most prevalent

85

thionein isoforms expressed in mammalian cells and been most extensively

86

studied, particularly in relation to their roles in essential metals

87

homeostasis and response to toxic metal stimuli 21.

5

ACS Paragon Plus Environment

13.

12.

In

MTs also play a

Journal of Agricultural and Food Chemistry

Page 6 of 53

88

Zn is an essential trace element ubiquitously distrubuted in all living

89

cells. Zn exerts rich physiological functions due to its structural constituent

90

and functional activation of Zn-dependent proteins 22. Zn deficiency causes

91

a series of pathologic consequences, whereas proper Zn supplement is

92

suggested to enhance the resistance to Cd toxicity. Dietary intake of Zn

93

was reported to decrease oxidative damages caused by Cd 23. Zn reduces

94

oxidative stress associated with ROS through the restoration of key

95

antioxidants (e.g., superoxide dismutase, glutathione) and inhibition of

96

nicotinamide adenine dinucleotide phosphate (NADPH) oxidase

97

Intracellular Zn availability is linked directly to MTs synthesis

98

expression is regulated by the metal regulatory transcription factor 1

99

(MTF-1) protein which contains six zinc fingers of the Cys-2-His type.

100

Zn2+ is an indispensable cofactor for MTF-1 protein activity. After binding

101

to Zn2+, MTF-1 protein is activated and mediates DNA-binding to metal

102

response element (MRE) which assists the initiation of MTs mRNA

103

transcription 26. Intracellular Zn2+ metabolism is mainly regulated by two

104

important metal ion transporter families: SLC39A (ZIP) and SLC30A

105

(ZnT). The ZIP family have 14 members (ZIP1~Zip14), being responsible

106

for transporting Zn2+ into the cytoplasm from the extracellular environment

107

or intracellular organelles. In contrast, the ZnT family, which consists of

108

10 members (ZnT1~ZnT10), reduce the cytoplasmic Zn2+ accumulation by

109

shipping cytoplasmic Zn2+ out of cells or into intracellular organelles 27. 6

ACS Paragon Plus Environment

25.

24.

MTs

Page 7 of 53

Journal of Agricultural and Food Chemistry

110

Some members of ZIP and ZnT were reported to mediate the Cd

111

metabolism in cells. For example, Zip8 strikingly possesses high affinity

112

for Cd2+ and facilitates Cd2+ uptake into cells, subsequently contributing to

113

cell death, cancer and other diseases 28. Transgenic mice carrying 2.5-fold

114

more Zip8 demonstrated enhanced Cd accumulation in liver and kidney 29.

115

ZIP8 and ZIP14 display very similar cation substrate specificities.

116

Similarly, ZIP14-mediated Cd uptake is proportional to cell toxicity 30. It

117

was also demonstrated that Cd stress decreased the gene expression of

118

ZnT1 and caused up-regulation of ZIP10, changes of which were

119

accompanied by increased Cd accumulation 17. These indicated that levels

120

of Zn2+ and Cd2+ in the extracellular environment have profound influence

121

on cell fate by competitive binding with metal ion transporters 31.

122

Macrophages are innate immune cells with well-established roles in

123

the immune system. Macrophages directly fight against invading

124

microorganisms as sentinels of innate immunity. Meanwhile, as antigen-

125

presenting cells, macrophages participate in the initiation of inflammation

126

and coordination of the adaptive immunity by activating lymphocytes 32.

127

Cd damage was reported to decrease phagocytosis, antigen presenting as

128

well as cytokine secretion abilities of macrophages, which eventually

129

suppress the adaptive immune response of body against pathogens

130

34.Thus,

131

marcophages and the effect of Zn supplement on Cd exposed macrophages

33-

this study was carried out to assess the Cd cytotoxicity on

7

ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry

132

as well as to explore the mechanisms underlying these effects.

133

Materials and Methods

134

Chemicals

135

Cadmium chloride (CdCl2) (Fuchen, Tianjin, China) and zinc

136

gluconate (C12H22O14Zn) (Aladdin, Shanhai, China) were separately

137

dissolved in distilled water at 50 mM as the stock solution and diluted to

138

obtain working solution with desired final concentrations. Rigorous pre-

139

experiments were based on bibliographic reference to select the optimal

140

doses of Cd2+ (20 or 50 μM) and Zn2+ (20 μM) which were used in the

141

present study 35.

142

Cell culture and treatment

143

The murine RAW 264.7 macrophage cell line, obtained from Zhong

144

Qiao Xin Zhou Biotechnology, Shanhai, China, and chicken HD-11

145

macrophage cell line (gift from biological products laboratory of Shanxi

146

Agricultural University, China) were maintained in dulbecco's modified

147

eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS)

148

(Gibco, South America) and 100 U/mL penicillin–streptomycin at 37°C in

149

a humidified incubator with 5% CO2. Treated macrophages with Cd2+

150

and/or Zn2+ were collected at different intervals (0~24 h).

151

Scanning electron microscopy (SEM) detection

152

RAW 264.7 treated cells were fixed for 4 h at 4°C in a solution

153

containing 2.5% glutaraldehyde and 0.1 M sodium cacodylate. Samples 8

ACS Paragon Plus Environment

Page 8 of 53

Page 9 of 53

Journal of Agricultural and Food Chemistry

154

were then fixed in 1% OsO4 in 0.2 M sodium cacodylate solution for 1 h

155

and dehydrated in an ascending series of ethanol. All samples were dried

156

using a critical point drier and sputtered with gold-palladium. Pictures were

157

captured using a scanning electron microscope (JSM 6510, JEOL, Japan)

158

at an electron accelerating voltage of 20 kV.

159

Transmission electron microscopy (TEM) detection

160

TEM was performed to reveal ultrastructural details of the

161

macrophages. RAW 264.7 treated cells were suspended by a cell scraper

162

and fixed for 2 h at 4°C in a solution containing 2% glutaraldehyde and

163

100 mM cacodylate. Samples were dehydrated through an ascending

164

concentration of ethanol and embedded in araldite. Ultrathin sections were

165

then stained with 2% methanolic uranyl acetate and examined under a

166

transmission electron microscope (JEM-1010, JEOL, Japan).

167

Cytoskeleton staining

168

HD-11 treated cells were rinsed by phosphate-buffered saline (PBS)

169

and fixed with 4% neutral formaldehyde for 20 min. After intensive

170

washing, membrane permeabilization with 0.1% Triton X-100 containing

171

3% bovine serum albumin (BSA) for 20 min was followed. Cells were then

172

incubated with fluorescein isothiocyanate (FITC)-labeled phalloidin

173

(Invitrogen, USA) for 30 min in dark. After gentle rinsing with PBS, cells

174

were stained with 100 ng/mL DAPI (4',6-diamidino-2-phenylindole)

175

solution for the visualization of nuclei. Images were captured using a 9

ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry

Page 10 of 53

176

confocal fluorescence microscopy (NOL-LSM710, Carl Zeiss Jena,

177

Germany) and analyzed by ZEN 2009 light edition software.

178

Apoptosis detection

179

Apoptotic macrophages in each group were examined with an

180

Annexin V-EGFP/PI Apoptosis Detection Kit (KGA101, Nanjing

181

Jiancheng Bioengineering Institute, Nanjing, China). Briefly, RAW 264.7

182

and HD-11 cells treated with Cd2+ and/or Zn2+ were suspended in 100 μL

183

binding buffer added with 5 μL Annexin V-EGFP and 5 μL propidium

184

iodide (PI). After incubation at room temperature for 20 min in dark, the

185

labelled macrophages were detected via a FACSCalibur flow cytometer

186

(BD Biosciences, CA, USA). Results were given according to the

187

fluorescence intensity of probes in macrophages, with 20,000 cells

188

analyzed for each sample.

189

Intracellular ROS detection

190

Intracellular ROS levels were detected with a ROS Assay Kit (S0033,

191

Beyotime Biotechnology, Shanghai, China). Briefly, RAW 264.7 and HD-

192

11

193

dihydrodichlorofluorescein diacetate) in serum-free medium in 37 °C for

194

30 min. After washing cells with serum-free medium for three times, ROS

195

were monitored by detecting the fluorescent signal of DCF at the excitation

196

wavelength of 488 nm using a flow cytometry (BD Biosciences, CA, USA).

treated

cells

were

incubated

with

10

ACS Paragon Plus Environment

DCFH-DA

(2',7'-

Page 11 of 53

Journal of Agricultural and Food Chemistry

197

Results were given according to the mean fluorescence intensity of probes

198

in macrophages, with 20,000 cells being analyzed for each sample.

199 200

Mitochondrial membrane potential (MMP) measurement

201

Changes of MMP in RAW 264.7 and HD-11 cells were detected with

202

a MMP Assay Kit (C2006, Beyotime Biotechnology, Shanghai, China).

203

Treated macrophages of both lines were incubated with 5 mg/mL of JC-1,

204

a cationic lipophilic fluorescent probe monomer which can selectively

205

enter mitochondria matrix, for 20 min at 37°C in dark. After intensive

206

washing, samples were detected by either FACSCalibur flow cytometer

207

(BD Biosciences, USA) or fluorescence microscope (IX-70, Olympus,

208

Tokyo, Japan).

209

Real-time PCR examination

210

RNA of treated RAW 264.7 macrophages was extracted using the TRI

211

Reagent (Invitrogen, USA) and reversely transcribed using a cDNA

212

Synthesis Kit (TaKaRa, Dalian, China). Primers of target genes including

213

MTF-1, MT-1 and MT-2 were given in table S1. Each assay was carried out

214

in a 20 μL reaction mixture containing 50 ng of cDNA, 0.2 μM of each

215

primer and 10 μL of reaction mix adding with SYBR-Green dye (TaKaRa,

216

Dalian, China). Quantitative PCR was performed in a stepone plus real-

217

time PCR system (Applied Biosystem, USA). PCR program initiated by a

218

5-min denaturation step at 94°C, followed by 40 amplification cycles of 11

ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry

219

92°C for 1 min, 58°C for 1 min and 72°C for 1 min. GAPDH

220

(Glyceraldehyde 3-phosphate dehydrogenase) was chosen as the reference

221

gene to adjust the threshold cycle (Ct) values of the target genes. Fold

222

change for target gene compared to the control was analyzed by the 2-ΔΔCt

223

method 36.

224

Immunofluorescence staining

225

Treated RAW 264.7 cells were fixed with 4% neutral formaldehyde

226

for 30 min and then permeabilized with 0.1% Triton X-100 for 20 min.

227

After intensive washing, cells were treated with 10% BSA for 30 min to

228

block unspecific binding sites. Incubating cells with mouse anti-

229

metallothioneins antibody (1:100, ab12228, Abcam, England) or rabbit

230

anti-MTF-1 antibody (1:200, NBP1-86380, Novus, USA) overnight at 4ºC,

231

which was followed by incubation with corresponding FITC-conjugated

232

secondary antibody (Proteintech, Wuhan, China ) for 30 min in dark.

233

Images of cells were captured using a confocal fluorescence microscopy

234

(NOL-LSM710, Carl Zeiss Jena, Germany) and analyzed with ZEN 2009

235

light edition software.

236

Western blot analysis

237

Treated RAW 264.7 cells from each group were harvested and lysed in

238

50 μL of lysis buffer (25 mM Tris, 150 mM NaCl, 1% NP-40, and 0.1 mM

239

SDS). Cell lysates were centrifuged at 15, 000 rpm for 10 min at 4°C.

240

Protein content in cell supernatant was determined using a BCA kit 12

ACS Paragon Plus Environment

Page 12 of 53

Page 13 of 53

Journal of Agricultural and Food Chemistry

241

(P0010S, Beyotime, China). Protein was denatured in boiling water for 5

242

min and then subjected to polyacrylamide gel electrophoresis (7.5%

243

stacking gel and 15% separation gel). Protein was moved onto

244

polyvinylidene fluoride (PVDF) membrane by transfer electrophoresis,

245

followed by incubation with anti-metallothioneins antibody (1:400,

246

ab12228, Abcam, England) or anti-MTF-1 antibody (1:500, NBP1-86380,

247

Novus, USA) overnight at 4ºC. Peroxidase conjugated goat anti-mouse or

248

goat anti-rabbit IgG (Proteintech, Wuhan, China) were used as the

249

secondary antibody. Signals were detected using an Enhanced

250

Chemiluminescence (ECL) Kit (P0018, Beyotime, Shanghai, China). β-

251

actin was used as the reference protein (60008-1-lg, Wuhan, China).

252

Intracellular Zn2+ chelation

253

N,N,N',N'-tetrakis (2-pyridylmethyl) ethylenediamine (TPEN) was

254

purchased from Sigma, USA. RAW 264.7 cells were incubated with 2.5

255

μM/L TPEN for 1 h in serum-free culture medium. Then, cells were

256

cultured with new medium containing 10% FBS and treated with Cd2+

257

and/or Zn2+ for specified time.

258

MTF-1 siRNA interference

259

Mouse MTF-1 (NM_008636.4) specific siRNA was designed by

260

GenePharma (Shanghai, China) and the target sequence was as follow:

261

1242CCAGCAAUCAUCUUUGAAUtt1260. Control groups were transfected

262

with a negative control siRNA (UUCUUCGAACGUGUCACGUTT) 13

ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry

263

obtained from GenePharma, Shanghai, China. RAW 264.7 cells were

264

seeded into 6-well plates at a density of 5×105 per well before transfecting

265

with MTF-1 siRNA. MTF-1 siRNA transfection was achieved using

266

Lipofectamine 2000 according to manufacturer’s instruction at a final

267

concentration of 100 nM. Interference efficiency of MTF-1 siRNA was

268

fully evaluated at the gene and protein level. After transfection with MTF-1

269

or negative control siRNA for 8 h, RAW 264.7 cells were cultured with

270

fresh medium and subjected to Cd2+ and/or Zn2+ for specified time.

271

Quantification of intracellular Cd2+ and Zn2+

272

Contents of intracellular Cd2+ and Zn2+ in RAW 264.7 cells were

273

detected by inductively coupled plasma mass spectrometry (ICP-MS). Cd2+

274

and/or Zn2+ treated RAW 264.7 cells were intensively washed with PBS to

275

rinse out the residual metal ion in culture medium. Cells were quantified

276

and digested with HNO3/HCl (1:3) in acid-washed high-density

277

polyethylene (HDPE) bottles. Heating up HDPE bottles slowly until the

278

liquid was completely evaporated. Sediments in the bottle were re-

279

dissolved with 0.1% HNO3. Metal ions were measured using the following

280

operating conditions: gas box nebulizer flow (0.8 L/min); gas box auxiliary

281

flow (0.6 L/min); acquisition time (10 s); replicates times (3); and radio

282

frequency power (2400 V).

283

Statistical analysis

284

Each experimental project in the present study was performed with 14

ACS Paragon Plus Environment

Page 14 of 53

Page 15 of 53

Journal of Agricultural and Food Chemistry

285

three replications and each replication contains four samples. Statistical

286

analysis of the data was performed using the Statistical Package for the

287

Social Sciences (SPSS; version 20.0 software, SPSS Inc., IL, USA).

288

Comparisons between two groups were performed using one-way ANOVA

289

followed by Duncan’s test. Date were presented as the mean±SD and

290

differences were considered statistically significant at P