Specific Microbiota Dynamically Regulate the Bidirectional GutâBrain

Subscriber access provided by University of South Dakota

Omics Technologies Applied to Agriculture and Food

Specific microbiota dynamically regulate the bidirectional gut– brain axis communications in mice fed meat protein diets Yunting Xie, Guanghong Zhou, Chao Wang, Xinglian Xu, and Chunbao Li J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.8b05654 • Publication Date (Web): 12 Dec 2018 Downloaded from http://pubs.acs.org on December 17, 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 46

Journal of Agricultural and Food Chemistry

Specific microbiota dynamically regulate the bidirectional gut-brain axis communications in mice fed meat protein diets Yunting Xie#, Guanghong Zhou#,+, Chao Wang#, Xinglian Xu#,+, Chunbao Li#, +, §*

#Key

Laboratory of Meat Processing and Quality Control, MOE; Jiangsu Collaborative

Innovation Center of Meat Production and Processing, Quality and Safety Control; Key Laboratory of Meat Products Processing, MOA; Nanjing Agricultural University; Nanjing 210095, P.R. China +Joint

International Research Laboratory of Animal Health and Food Safety, MOE, Nanjing

Agricultural University; Nanjing 210095, P.R. China §National

Center for International Research on Animal Gut Nutrition, Nanjing Agricultural

University; Nanjing 210095, P.R. China

*Corresponding author: Dr. Chunbao Li E-mail: [email protected]; Tel/fax: 86 25 84395679

1

ACS Paragon Plus Environment


1

Abstract

2

The purpose of this study was to characterize the dynamic changes of different protein diets to

3

gut microbiota and explore the influence on communications between the gut and the brain.

4

C57BL/6J mice were fed casein, soy protein and four kinds of processed meat proteins at a

5

normal dose of 20% for 8 months. Bacteroidales S24-7 abundance increased from 4 to 8 months,

6

whereas the abundances of six genera including Akkermansia decreased remarkably.

7

Lachnospiraceae Unclassified abundance in the emulsion-type sausage protein and stewed pork

8

protein groups showed an opposite change from 4 to 8 months. Twenty-eight and 48 specific

9

operational taxonomy units in cecum and colon respectively involved in regulating serotonin,

10

peptide YY, leptin and insulin levels. Specific microbiota was involved, directly or indirectly

11

through signaling molecules, in the regulation of body metabolism, which may affect the

12

communications between the gut and brain and cause different growth performances.

13 14

Keywords: meat proteins, 16S rRNA sequencing, fecal microbiota, gut-brain axis, signaling

15

molecules

16

2


Page 2 of 46

Page 3 of 46


17

Introduction

18

In recent years, the gut-brain axis has attracted great interest, and previous studies have shown

19

that the gut microbiota plays an important role in the bidirectional communications between the

20

gut and the brain1. The brain ensures proper maintenance and coordination of gastrointestinal

21

functions. In turn, the gut microbiota has a great influence on central nervous system activities

22

and host behavior, with chemical signaling of the gut-brain axis being involved. The trillions of

23

microbes in the gastrointestinal tract are considered a complex and dynamic ecosystem that has

24

coevolved with the host2. Many factors have a certain impact on gut microbiota, e.g., genetics,

25

geographic origin, age, medication and diet3, 4, among which diet is the dominant modulator of

26

the composition and function of gut microbiota5. The majority of dietary proteins are digested

27

into peptides and free amino acids in the small intestine, but some proteins cannot be digested

28

and absorbed, and thus they enter the large intestine for microbial fermentation6, 7. High-protein

29

diets have been shown to alter the gut microbial composition of mice or rats8, 9. The temporal

30

microbial changes were also observed in the feces of rats after 6 weeks of a high-protein diet

31

intake10. Moreover, some studies indicated that dietary protein sources affect the gut microbial

32

composition11, 12.

33

Meat is known to be an important source of high-quality protein that contains all essential amino

34

acids. In processed meat, the processing methods may lead to different degrees of protein

35

oxidation and denaturation, which cause protein aggregation and changes in secondary

36

structures13, 14. Moderate denaturation will increase the degradation of meat proteins, but various

37

amino acid modifications might lead to the formation of “limit peptides”, which are not further 3



38

broken down and thus result in a reduction of protein bioavailability15, 16. Our in vitro studies

39

showed that protein digestibility and digested products differed among cooked pork,

40

emulsion-type sausage, dry-cured pork and stewed pork17. Most studies have focused on the

41

short-term effects of dietary proteins, and few data are available on the temporal variations in gut

42

microbial composition. This study aimed to investigate whether a relatively long-term intake of

43

proteins from processed meat affects the gut microbiota and the bidirectional communications

44

between the gut and the brain.

45 46

Materials and Methods

47

Animals and diets

48

All experiments were carried out in compliance with the relevant guidelines and regulations of

49

the Ethical Committee of Experimental Animal Center of Nanjing Agricultural University. A

50

total of 60 4-week-old male C57BL/6J mice were housed in a specific pathogen-free animal

51

center (SYXK 2011-0037, Nanjing, Jiangsu, China). The temperature (20.0 ± 0.5°C)

52

and relative humidity (60 ± 10%) were kept constant during the experiment, with a 12-h light

53

cycle. Mice were fed a standard chow diet during a 2-week acclimation period. Then, animals

54

were assigned to one of six diet groups (ten mice in each group), that is, casein (C), emulsion

55

sausage protein (ESP), dry-cured pork protein (DPP), stewed pork protein (SPP), cooked pork

56

protein (CPP) and soy protein (SP) groups. Diet preparations were shown in supplementary file.

57

Isolated soy protein and casein were commercially obtained. Isoflavones were removed from soy

58

protein in 80% methanol. Meat proteins were isolated from processed meat products by 4


Page 4 of 46

Page 5 of 46


59

removing fat and water in methylene chloride/methanol solvent (2:1, v:v). The amino acid

60

profile in protein powders were shown in tables S1. The diets were prepared according to the

61

AIN-93G diet formulation18 and the diet formulation was shown in tables S2. Mice were allowed

62

to access water and diets ad libitum for 8 months. Body weight and feed intake of mice were

63

routinely recorded for calculating the average daily gain (ADG) and average daily feed intake

64

(ADFI).

65

Sample collection

66

At 4 and 8 months of feeding, the feces and blood of mice were collected without fasting. The

67

fecal samples were stored at −80°C for the microbial composition analysis. Blood samples were

68

collected in Eppendorf tubes. The tubes stood at room temperature for 45 min and then were

69

centrifuged at 3,000 g for 30 min to pellet the blood cells. Serum samples were collected and

70

stored at −80°C. At 8 months, all the mice were euthanized by cervical dislocation, and the

71

epididymal adipose and liver tissues were taken and weighed. Relative weights of epididymal

72

adipose and liver tissues were calculated according to body weight.

73

Serum biochemical indicators, signaling molecules of the gut-brain axis and short chain fatty

74

acids

75

Serum biochemical indicators containing triglycerides (TG), total cholesterol (T-Cho), total

76

protein (T-Pro), and urea nitrogen (UN) were detected by automated chemical analyzer

77

(Spotchem EZ, Nakagyo-ku, Kyoto, Japan) with kits (Arkray, Nakagyo-ku, Kyoto, Japan)

78

according to the manufacturer's instructions. The signaling molecules of the gut-brain axis,

79

including peptide YY (PYY), leptin and insulin, were measured using the Milliplex magnetic 5



80

bead mouse metabolic hormone multiplex panel (Mmhmag-44K; EMD-Millipore, Billerica MA),

81

and serotonin (5-hydroxytryptamine, 5-HT) was quantified using a serotonin ELISA kit

82

(KA2518, Abnova, Taipei, Taiwan, China) according to the manufacturer’s protocols. The short

83

chain fatty acids (SCFAs) were characterized by gas chromatograph (GC-2010 Plus, Shimadzu,

84

Japan) as previously described19. Briefly, fecal samples (50mg) were suspended in 250 μl ddH2O

85

and were centrifuged at 12,000 g for 5 min in a micro-centrifuge (Microfuge 22R, Beckman

86

coulter, California). The SCFAs analysis was carried out on the supernatants (200 μl), with

87

crotonic acid as an internal standard. A flame ionization detector and a capillary column

88

( Agilent Technologies, HP-INNOWax, 30 m×0.25 mm×0.25 μm, California ) were used, with

89

an injector/detector temperature of 180°C/180°C, a column temperature of 130°C and a gas flow

90

rate of 30 ml/min.

91

16S rRNA gene sequencing

92

Total genomic DNA in fecal samples was extracted using the QIAamp DNA Stool Mini Kit

93

(Qiagen 51504, Dusseldorf, Nordrhein-westfalen, Germany) according to the manufacturer’s

94

instructions. The DNA was quantified by a Nanodrop® spectrophotometer (Nanodrop2000,

95

ThermoFisher Scientific, Waltham, MA). Purified DNA was used to amplify the V4 region of

96

16S rRNA, which is associated with the lowest taxonomic assignment error rate20. Polymerase

97

chain reaction (PCR) was performed in triplicate. Amplicons were extracted from 2% agarose

98

gels and purified using the AxyPrep DNA Gel extraction kit (Axygen Biosciences, Union City,

99

CA) according to the manufacturer’s instructions and quantified using QuantiFluor™ -ST

100

(Promega, Madison, WI). The pooled DNA product was used to construct Illumina Pair-End 6


Page 6 of 46

Page 7 of 46


101

library following Illumina’s genomic DNA library preparation procedure. Then the amplicon

102

library was paired-end sequenced (2 × 250) on an Illumina MiSeq platform (San Diego, CA)

103

according to the standard protocols.

104

Bioinformatics analysis

105

Raw fastq files were trimmed and chimeric sequences were identified and removed from all

106

samples to reduce noise, and operational taxonomic units (OTUs) were clustered with ≥97%

107

similarity cutoff using UPARSE (version 7.1, http://drive5.com/uparse/). Then community

108

richness estimator (Chao and ACE), diversity indices (Shannon and Simpson), and Good’s

109

coverage were calculated21. Principal coordinate analysis (PCoA) and clustering analysis

110

(http://sekhon.berkeley.edu/stats/html/hclust.html) were applied on the basis of the OTUs to

111

offer an overview of the fecal microbial composition22. Multivariate analysis of variance

112

(MANOVA) was conducted to further confirm the observed differences. Linear discriminant

113

analysis effect size (LEfSe) analysis (http://huttenhower.sph.harvard.edu/galaxy/) was carried

114

out to discover biomarkers for fecal bacteria and to distinguish between biological conditions

115

among different groups23. Besides, the Spearman’s correlation coefficients were assessed to

116

determine the relationships between microbiota and signaling molecules of the gut-brain axis.

117

Correlation was considered significant when the absolute value of Spearman’s rank correlation

118

coefficient (Spearman’s r) was >0.6 and statistically significant (P < 0.05). The R (pheatmap

119

package) and Cytoscape (http://www.cytoscape.org/）were applied to visualize the relationships

120

through correlation heatmap and network diagrams respectively.

121

Functional prediction of the microbial genes 7



122

Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt)

123

(http://picrust.github.io/picrust/tutorials/genome_prediction.html#genome-prediction-tutorial)

124

program based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database was used to

125

predict the functional alteration of fecal microbiota in different samples24. The OTU data

126

obtained were used to generate BIOM files formatted as input for PICRUSt v1.1.09 with the

127

make.biom script usable in the Mothur. OTU abundances were mapped to Greengenes OTU IDs

128

as input to speculate about the functional alteration of microbiota.

129

Statistical analysis

130

The diet effect was evaluated by one-way ANOVA with SAS software (SAS Institute Inc., Cary,

131

NC). Means were compared and the significance threshold was set at 0.05 for statistical analyses.

132

Figures were constructed using the GraphPad Prism (version 5.0.3, San Diego, CA).

133

More details about materials and methods can be found in the supplementary file.

134 135

Results

136

Composition and functions of gut microbiota

137

Richness and diversity. Before diet change, fecal samples were selected for sequencing

138

randomly from 10 mice to check the variations in gut microbiota composition among individuals.

139

The PCoA showed a relatively small variation in fecal microbiota among individuals (Fig. S2).

140

However, the gut microbiota in the C, SP and ESP groups was distinct from other meat protein

141

diet groups at 4 months. The gut microbiota of diet groups were also well separated at 8 months,

142

except the CPP and SPP groups. In addition, the gut microbiota showed a time-dependent change 8


Page 8 of 46

Page 9 of 46


143

(Fig. 1A to C).

144

The community richness estimators (Chao and ACE), and diversity indices (Shannon and

145

Simpson) were calculated in order to evaluate the alpha diversity (Table 1). The protein diets

146

significantly affected ACE and Chao values at both time points. The ACE and Chao values of

147

the C and SPP groups were significantly lower than those of other groups at 4 months, while the

148

values were the highest for the SP group at 8 months. The Shannon and Simpson values were not

149

affected by diets at 4 months, but the Shannon value of the SP group was higher than those of

150

other groups at 8 months. In addition, the Shannon values decreased with feeding time,

151

indicating that the microbial diversity may be reduced during a long term feeding of the same

152

diet.

153

Composition of gut microbiota. At the phylum level, Bacteroidetes and Firmicutes were the

154

predominant phyla (Fig. 2A and B). Hierarchical clustering analysis indicated that the microbial

155

composition of the SP group was different from other groups at 4 months, but the DPP and CPP

156

groups showed a significant difference from other groups at 8 months. Furthermore,

157

Bacteroidetes abundance increased but Verrucomicrobia abundance remarkably decreased during

158

feeding (Fig. 2C).

159

At the genus level, Bacteroidales S24-7 was the most abundant genus at 4 months, accounting

160

for 31.43% of the fecal microbial population, and followed by Rikenellaceae RC9 gut (9.75%).

161

At 8 months, Bacteroidales S24-7 and Faecalibaculum were the most prevalent genera,

162

accounting for 44.88% and 9.81% of the total microbial population in all diet groups,

163

respectively (Fig. 2D and E). Moreover, seven species showed time-dependent changes. The 9



164

abundance of Bacteroidales S24-7 increased from 4 to 8 months, whereas those of Rikenellaceae

165

RC9 gut, Akkermansia, Alistipes, Clostridiales vadinBB60, Clostridium sensustricto 1 and

166

Anaerotruncus were dramatically reduced (Fig. 2F). These data provide a general overview of

167

gut microbiota composition at two time points.

168

Linear discriminant analysis of fecal microbiota. LEfSe analysis revealed significant

169

differences in 35 and 47 OTUs among six groups at 4 and 8 months, respectively. As shown in

170

the biological cladograms, taxonomic distributions confirmed specific microbial taxa from

171

phylum to genus associated with protein diets (Fig. 3A and B). At 4 months, Coriobacteriaceae,

172

Erysipelotrichaceae, Comamonadaceae and Oxalobacteraceae were more dominant in the ESP

173

group. Clostridiales vadinBB60 and Anaeroplasmataceae were more abundant in the DPP group,

174

with Defluviitaleaceae in the SP and Hyphomicrobiaceae in the CPP group. However, at 8

175

months, Rikenellaceae, Aerococcaceae, Peptostreptococcaceae, Desulfovibrionaceae and

176

Moraxellaceae were more enriched in the SP group, with Clostridiales vadinBB60 and

177

Deferribacteraceae in the ESP group and also Erysipelotrichaceae and Anaeroplasmataceae in

178

the SPP group. These results strongly suggest time-specific alterations of gut microbiota that is

179

mediated by protein diets.

180

Few data are available on the differences of gut microbiota between the diets based on processed

181

meat. Thus the top 20 genera in the four meat protein groups were further analyzed. The

182

abundances of 3 and 6 dominant genera were observed to differ significantly with diets at 4 and

183

8 months, respectively (Figs. 4A and B). At 4 months, Anaerotruncus, Lachnospiraceae

184

NK4A136 group and Lachnospiraceae Unclassified were less abundant in the ESP and CPP 10


Page 10 of 46

Page 11 of 46


185

groups than in the SPP group. However, at 8 months, Faecalibaculum in the SPP group, and

186

Alistipes and Clostridiales vadinBB60 group_norank were more abundant in the DPP group than

187

those in the CPP group. Rikenellaceae RC9 gut group, Lachnospiraceae Unclassified and

188

Coriobacteriaceae UCG-002 were the most abundant in the DPP, ESP and CPP groups,

189

respectively.

190

Functional prediction of microbial genes. The diet has been known to affect the composition

191

and structure of microbiota, and thus, the PICRUSt package was applied to predict microbial

192

gene functions. Protein diets significantly changed the microbial functions related to metabolism

193

(Fig. 5A and B). At 4 months, the ESP diet upregulated microbial carbohydrate metabolism

194

compared with the C, SP and DPP groups. At 8 months, the SPP, CPP and DPP diets

195

downregulated the microbial xenobiotics biodegradation and metabolism, while they upregulated

196

the microbial carbohydrate metabolism, enzyme families, and nucleotide metabolism compared

197

with the C group. Different microbes encode different enzymes. These results suggested that

198

dietary proteins may mediate microbial metabolism by altering the microbial composition.

199

SCFAs. SCFAs are important metabolites of the gut microbiota. Dietary proteins significantly

200

affected the levels of SCFAs at both time points (Table 2). At 4 months, the levels of acetate,

201

butyrate and total SCFAs in the DPP and ESP groups and the levels of isobutyrate and branch

202

chain fatty acids (BCFAs) in the SPP and C groups were significantly lower than these in the

203

CPP group. Propionate and isovalerate levels were the highest in the C and CPP groups,

204

respectively. At 8 months, the levels of acetate, isobutyrate and BCFAs were dramatically higher

205

in the CPP group than in the other groups except the SPP group. The butyrate and valerate levels 11



206

were significantly higher in the CPP group than in the SPP and DPP groups. The level of total

207

SCFAs was dramatically lower in the C group than in the CPP group, however, isovalerate levels

208

showed the opposite pattern in the two groups. The propionate levels were significantly lower in

209

the processed meat protein groups compared with the SP group. These results showed that

210

dietary proteins significantly affected the metabolism of gut microbiota.

211

Variations in signaling molecules of the gut-brain axis

212

To further explore the effect of protein diets on bidirectional communications between the gut

213

and the brain via the peripheral circulatory system, several signaling molecules, that is, leptin,

214

insulin, PYY, and serotonin in serum were quantified. The protein diets significantly affected the

215

concentrations of these signaling molecules. The leptin level did not differ with diets before

216

changing diets but increased greatly afterwards (Fig. 6). At 4 months, the leptin level was lower

217

in the SPP group than in the DPP and SP groups (P

Specific Microbiota Dynamically Regulate the Bidirectional GutâBrain

Recommend Documents

Specific Microbiota Dynamically Regulate the Bidirectional GutâBrain