High fat diet impairs the function of glucagon-like peptide-1 producing L-cells

Accepted Manuscript Title: High Fat Diet impairs the function of Glucagon-like Peptide-1 producing L-cells Author: Paul Richards Ramona Pais Abdella M. Habib Cheryl A. Brighton Giles S.H Yeo Frank Reimann Fiona M. Gribble PII: DOI: Reference:

S0196-9781(15)00189-8 http://dx.doi.org/doi:10.1016/j.peptides.2015.06.006 PEP 69496

To appear in:

Peptides

Received date: Revised date: Accepted date:

31-1-2015 19-6-2015 22-6-2015

Please cite this article as: Richards P, Pais R, Habib AM, Brighton CA, Yeo GSH, Reimann F, Gribble FM, High Fat Diet impairs the function of Glucagon-like Peptide-1 producing L-cells, Peptides (2015), http://dx.doi.org/10.1016/j.peptides.2015.06.006 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

*Highlights (for review)

Highlights: Long term dietary changes impair function of the gut endocrine system



High fat diet impairs nutrient-triggered GLP-1 release from murine small intestine



L-cells from HFD-fed mice have reduced expression of many L-cell-specific genes

Ac

ce pt

ed

M

an

us

cr

ip t



Page 1 of 32

*Manuscript Click here to download Manuscript: Version R2 clean.docx

1

High Fat Diet impairs the function of Glucagon-like

2

Peptide-1 producing L-cells

3 a

Paul Richards, aRamona Pais, Abdella M Habib, Cheryl A Brighton, Giles SH Yeo, bFrank Reimann,

5

b

6

Metabolic Research Laboratories and MRC Metabolic Diseases Unit, WT-MRC Institute of Metabolic

7

Science, Addenbrooke's Hospital, Cambridge, CB2 0QQ, UK

8

a

9

b

ip t

4

us

These authors contributed equally to the work

cr

Fiona M Gribble

an

Joint corresponding authors

10 Corresponding authors:

12

Fiona Gribble: [email protected]

13

Frank Reimann: [email protected]

14

Metabolic Research Laboratories and MRC Metabolic Diseases Unit, WT-MRC Institute of Metabolic

15

Science, Addenbrooke's Hospital, Cambridge, CB2 0QQ, UK

16

Tel: 0044 1223 336746, Fax: 0044 1223 330598

ce pt

ed

M

11

17

Ac

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Click here to view linked References

18

Keywords:

19

GLP-1

20

Enteroendocrine

21

L-cell

22

High fat diet

23

1 Page 2 of 32

Abstract

25

Glucagon-like peptide-1 (GLP-1) acts as a satiety signal and enhances insulin release. This study

26

examined how GLP-1 production from intestinal L-cells is modified by dietary changes.

27

Methods: Transgenic mouse models were utilised in which L-cells could be purified by cell specific

28

expression of a yellow fluorescent protein, Venus. Mice were fed on chow or 60% high fat diet (HFD)

29

for 2 or 16 weeks. L-cells were purified by flow cytometry and analysed by microarray and

30

quantitative RT-PCR. Enteroendocrine cell populations were examined by FACS analysis, and GLP-1

31

secretion was assessed in primary intestinal cultures.

32

Results: 2 weeks HFD reduced the numbers of GLP-1 positive cells in the colon, and of GIP positive

33

cells in the small intestine. Purified small intestinal L-cells showed major shifts in their gene

34

expression profiles. In mice on HFD for 16 weeks, significant reductions were observed in the

35

expression of L-cell specific genes, including those encoding gut hormones (Gip, Cck, Sct, Nts),

36

prohormone processing enzymes (Pcsk1, Cpe), granins (Chgb, Scg2), nutrient sensing machinery

37

(Slc5a1, Slc15a1, Abcc8, Gpr120) and enteroendocrine-specific transcription factors (Etv1, Isl1,

38

Mlxipl, Nkx2.2 and Rfx6). A corresponding reduction in the GLP-1 secretory responsiveness to

39

nutrient stimuli was observed in primary small intestinal cultures.

40

Conclusion: Mice fed on HFD exhibited reduced expression in L-cells of many L-cell specific genes,

41

suggesting an impairment of enteroendocrine cell function. Our results suggest that a western style

42

diet may detrimentally affect the secretion of gut hormones and normal post-prandial signalling,

43

which could impact on insulin secretion and satiety.

cr

us

an

M

ed

Highlights:

46



47



48



ce pt

44 45

ip t

24

Ac

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Long term dietary changes impair function of the gut endocrine system High fat diet impairs nutrient-triggered GLP-1 release from murine small intestine L-cells from HFD-fed mice have reduced expression of many L-cell-specific genes

49 50

2 Page 3 of 32

51

1. Introduction Hormones from the gut control food intake and insulin release as well as intestinal motility and

53

secretion [1]. The post-prandial rise in the plasma concentration of GLP-1 signals to the brain that

54

food has arrived in the gastrointestinal tract and to the pancreas that glucose is being absorbed.

55

GLP-1 brings about the sensation of satiety together with the enhancement of insulin release [2]. As

56

dietary habits have changed and people now consume more fat and sugar, this raises concerns

57

about whether dietary intake affects the release of gut hormones, and whether this in turn might

58

contribute to rising levels of obesity and diabetes.

59

GLP-1, like other gut hormones, is released from enteroendocrine cells (EECs) located in the

60

intestinal epithelium. EECs have a life span of only about 5 days, and are constantly replenished from

61

stem cells in the intestinal crypts. Depending on their position along the gastrointestinal tract, EECs

62

exhibit characteristic hormonal signatures. In the duodenum, for example, there is a high number of

63

K-cells producing GIP, whereas in the distal ileum and colon there are more L-cells expressing GLP-1

64

and PYY [3]. A few published studies suggest that the number of EECs can be altered by

65

environmental stimuli. In intestinal organoid cultures in vitro, for example, the number of L-cells is

66

increased by exposure to short chain fatty acids [4], and in vivo, elevated L-cell numbers have been

67

observed in germ-free mice [5], and in rodents on a high-fibre diet [6]. How EECs are affected by

68

dietary exposure is an area of great interest. If it were possible to increase L-cell number, this could

69

lead to increased GLP-1 release and improved glucose tolerance and satiety.

70

Surgical models have been used to investigate how EEC numbers are affected by exposure of

71

different gut segments to luminal nutrients. In rats, gastric bypass surgery has been reported to

72

result in an increase in L-cell numbers in the roux (alimentary) and common limbs [7, 8]. In both

73

cases, it was found that the increase in L-cell numbers was attributable to mucosal hypertrophy, but

74

that there was no change in L-cell density. EECs also developed normally in the biliopancreatic limb.

75

These studies suggest that increased ileal nutrient exposure in the context of a chow diet results in

76

mucosal growth and a corresponding increase in L-cell number, but that the frequency of L-cells is

77

not altered.

78

It is less clear how the intestine responds to diets containing a high fat content. In high-fat diet (HFD)

79

fed rodents, it was reported that there is a reduction in the density of cells staining for chromogranin

80

A and GLP-1, and correspondingly reduced expression of Gcg and Pyy by PCR [9, 10]. In this study we

81

have examined the transcriptomic and secretory properties of L-cells from mice fed on a high fat

82

(60%), high sugar diet.

ce pt

ed

M

an

us

cr

ip t

52

Ac

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

3 Page 4 of 32

83 84

2. Materials and Methods 2.1 Animals:

86

All procedures were approved by the UK Home Office and local ethical committee of the University

87

of Cambridge. Male GLU-Venus mice [11] on a C57Bl6 background were weighed at age 8 weeks and

88

divided into 2 groups (n=4-6 per group), each with a balanced range of body weights. For the 16

89

week study, they were then single housed and placed on either a standard control chow (Special

90

Diets Services, Rat and mouse breeder and grower diet) or high fat diet (Research Diets, #D12331,

91

containing 60% energy from fat). Mice were weighed weekly for 16 weeks. Mice on the 2 week study

92

were treated similarly, but were not single housed. Mice were killed by cervical dislocation 2-4 hours

93

after lights on. Intestinal tissues were harvested and treated as below for flow cytometry (FACS),

94

mRNA extraction or secretion experiments.

95

2.2 Small intestine for FACS sorting

96

For purification of cell populations by FACS, intestinal pieces were stripped of the outer muscle

97

layers. Tissue was chopped into 1–2 mm pieces and digested to single cells with 1 mg/ml collagenase

98

in calcium-free Hanks Buffered Salt Solution (HBSS). Single cell suspensions were separated by flow

99

cytometry using a MoFlo Beckman Coulter Cytomation sorter (FL, USA). Side scatter, forward scatter

100

and pulse width gates were used to exclude debris and aggregates, and venus positive cells collected

101

at ~95% purity, alongside negative (control) cells that comprised a mixed population dominated by

102

other epithelial cell types.

103

2.3 FACS analysis

104

Single cell suspensions were prepared for FACS analysis, as described previously (9). Briefly, cells

105

were fixed with 4% paraformaldehyde (PFA) in phosphate buffered saline (PBS), permeabilised with

106

0.1% v/v Triton X-100, blocked with 10% goat serum and then incubated with primary antibody in

107

PBS-10% goat serum at 4°C overnight for an hour at RT. The primary antibodies used were: anti-GIP

108

(gifted by Prof JJ Holst; 1:1000), anti-CCK (gifted by Prof GJ Dockray; 1:500) and anti-PYY (Progen,

109

London, UK; 16066; 1:100). Cells were rinsed 3 times and then incubated with secondary antibody

110

(Alexa-Fluor 555, 633 or 647 Invitrogen, USA; A-21435, A-21105, A-21428 or A-21070, all at 1:800).

111

Cells were analysed using a LSRCyAn advance digital processing Flow Cytometer (Dako Cytomation,

112

CA, USAFortessa (BD Biosciences)). Events with very low side and forward scatter were excluded as

ce pt

ed

M

an

us

cr

ip t

85

Ac

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

4 Page 5 of 32

these are likely to represent debris, and events with a high pulse width were excluded to eliminate

114

cell aggregates.

115

2.4 RNA extraction and qRT-PCR

116

Tissues used for RNA and protein extraction were washed in PBS and placed in RNAlater (Ambion,

117

Life Technologies, Paisley UK) and frozen until processed. Samples were homogenised in Tri-reagent

118

(Sigma), and then treated as below. Total RNA from FACS-sorted cells prepared from GLU-Venus

119

transgenic mice was isolated using a micro scale RNA isolation kit (Ambion). All samples were

120

reverse transcribed according to standard protocols. Quantitative RT-PCR was performed with 7900

121

HT Fast Real-Time PCR system (Applied Biosystems, Life Technologies), using verified Taqman

122

primer/probe sets supplied by Applied Biosystems. In all cases expression was compared with that of

123

β-actin measured on the same sample in parallel on the same plate, giving a CT difference (ΔCT) for

124

β-actin minus the test gene. Mean, standard error, and statistics were performed on the ΔCT data

125

and only converted to relative expression levels (2ΔCT) for presentation in the figures.

126

2.5 Microarray

127

The quality of RNA samples was determined by Total RNA Pico Chip (Agilent Technologies, Stockport,

128

UK), and only those with RIN values >7.0 selected for microarray analysis. Total RNA from the

129

samples were amplified using the NuGEN Ovation Pico WTA system (NuGEN Technologies, Leek,

130

Netherlands) according to the manufacturer’s protocol, labelled with the NuGEN Exon and Encore™

131

Biotin Modules, and hybridized onto Affymetrix Mouse Gene ST 1.0 arrays. Raw microarray image

132

data was converted to CEL files using Affymetrix GeneChip Operating Software. All the downstream

133

analysis of the microarray data was performed using GeneSpring GX 12.1 (Agilent). The CEL files

134

were used for both the robust multi-array average (RMA) and Probe Logarithmic Intensity Error

135

(PLIER) analyses. After importing the data, each chip was normalized to the 50th centile of the

136

measurements taken from that chip.

137

To assess the purity of L-cell sorts, each probe on the microarray data was assigned a value

138

indicating its relative expression in L-cells vs non L-cells (combining all microarrays from HFD and

139

chow-fed mice), and the data were ordered to indicate the top 50 probes most enriched in the non

140

L-cell population. For each L-cell microarray we then calculated the geometric mean of these 50 non

141

L-cell probes as a measure of contamination by non L-cells. One out of the 4 microarrays from HFD L-

142

cells was excluded from further analysis because its mean expression of non L-cell probes was >2 SD

143

above the mean of all L-cell microarrays. In further analyses, the number of individual microarrays

144

used was: L-cells on chow (3), L-cells on HFD (3), non L-cells on chow (3) and non L-cells on HFD (4).

ce pt

ed

M

an

us

cr

ip t

113

Ac

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

5 Page 6 of 32

2.6 Primary intestinal culture

146

Small intestinal and colonic primary cultures were prepared as previously described [11]. Briefly,

147

mice 3-6 months old were sacrificed by cervical dislocation and the small intestine or colon was

148

excised. Luminal contents were flushed thoroughly with PBS and the outer muscle layer removed.

149

Tissue was minced and digested with Collagenase Type XI and the cell suspension plated onto 24-

150

well plates pre-coated with Matrigel (BD Bioscience, Oxford, UK).

151

2.7 GLP-1 secretion assay

152

18-24 hours after plating, cells were washed and incubated with test agents made up in 0.25 ml

153

saline buffer containing: (in mM: 138 NaCl, 4.5 KCl, 4.2 NaHCO3, 1.2 NaH2PO4, 2.6 CaCl2, 1.2 MgCl2,

154

and 10 HEPES, pH 7.4 with NaOH) supplemented with 0.1% BSA for 2 hours at 37oC. At the end of

155

the 2 hour incubation, supernatants were collected and centrifuged at 2000 rcf for 5 minutes and

156

snap frozen on dry ice. Cells were mechanically disrupted in 0.5 ml lysis buffer containing (mM): 50

157

Tris-HCl, 150 NaCl, 1 % IGEPAL-CA 630, 0.5 % deoxycholic acid (DCA) and complete EDTA-free

158

protease inhibitor cocktail (Roche, Burgess Hill, UK) to extract intracellular peptides, centrifuged at

159

10,000 rcf for 10 minutes and snap frozen. GLP-1 was measured using an electrochemiluminescence

160

total GLP-1 assay (MesoScale Discovery, Gaithersburg, MD, USA) and results expressed as a percent

161

of total (secreted+lysate) GLP-1 and normalised to basal secretion in response to saline measured in

162

parallel on the same day. Chemicals were purchased from Sigma (Poole, UK) unless otherwise

163

indicated.

164

2.8 Data analysis

165

Results are expressed as mean ± SEM. Statistical analysis was performed using GraphPad Prism 5.01

166

(San Diego, CA, USA). For GLP-1 secretion data, one-way ANOVA with post hoc Dunnett’s or

167

Bonferroni’s tests were performed on log-transformed secretion data, as these data were

168

heteroscedastic. Values were regarded as significant when p <0.05.

ce pt

ed

M

an

us

cr

ip t

145

Ac

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

169 170

3. Results

171

3.1 HFD affects the expression of gut hormones mRNAs in tissue homogenates

172

Two weeks on HFD resulted in reduced expression of Gcg and Insl5 in colonic tissue, and a tendency

173

towards reduced expression of Cck and Pyy, suggesting that HFD causes either a reduction in colonic

174

L-cell density or impaired production of hormones by individual enteroendocrine cells (Fig 1). No 6 Page 7 of 32

significant changes were observed in the expression of Gcg, Cck, Gip or Pyy in small intestinal tissue

176

homogenates, although we did detect an increase in somatostatin (Sst) mRNA in the ileum of HFD-

177

fed mice. Hormone expression was unaffected in the rectum.

178

3.2 HFD alters gut peptide production by individual L-cells

179

To evaluate the frequency of L-cells and their individual production of peptide hormones, we

180

performed FACS analysis of cell suspensions from transgenic mice expressing a yellow fluorescent

181

protein Venus driven by the gcg promoter, which were immunostained for different gut hormones.

182

Quantification of Venus-labelled L-cells in different regions of the GI tract revealed that the large

183

intestine (colon+rectum) of HFD-fed mice contained a significantly lower percentage of Venus

184

positive L-cells than their chow-fed littermates (Fig 2A). In large intestinal cell suspensions co-

185

stained with antibodies against CCK and PYY, we observed a particular reduction in the number of L-

186

cells staining strongly for CCK, and a corresponding increase in the intensity of L-cell PYY staining in

187

mice fed on HFD for 2 weeks (Fig 2B-D).

188

In the upper small intestine, the frequency of Venus-labelled L-cells was not markedly affected by

189

the HFD (Fig 2E). Both the frequency of GIP-stained cells (Fig 2E), however, and the proportion of L-

190

cells that were immuno-positive for GIP (data not shown) were reduced in small intestinal cell

191

suspensions from HFD-fed mice. Staining for PYY and CCK were not noticeably different between the

192

chow and HFD groups in the small intestine (Fig 2E and data not shown).

193

3.3 GLP-1 release is altered in intestinal cultures from HFD-fed mice

194

We next evaluated whether the function of L-cells was modified by high fat feeding. GLP-1 secretion

195

was assessed in small intestinal cultures from mice fed for 2 weeks on either HFD or chow (Fig 3).

196

Cultures from chow-fed mice had a basal secretory rate of 2.8 % per 2 hours, which was stimulated

197

3.5-fold by 10 mM glucose, 7.9-fold by 0.5 % peptone, 1.6-fold by 100 M linoleic acid, 20-fold by

198

glucose/fsk/IBMX, 3.2-fold by 20 mM gly-leu, and 5.6-fold by 1 M PMA. Cultures from HFD-fed mice

199

exhibited an increased rate of basal GLP-1 release (5.0 % per 2 hours, cf 2.8 % in chow-fed animals,

200

p<0.05). The amplification of GLP-1 secretion above the basal rate by glucose, peptone,

201

forskolin/IBMX and gly-leu was reduced in the HFD-fed group (Fig 3) although absolute % secretory

202

rates in the presence of different stimuli were similar in chow-fed and HFD-fed mice.

203

Basal and stimulated GLP-1 release from colonic cultures were not altered by placing mice on HFD

204

for 2 weeks (data not shown). In mice fed for 16 weeks on HFD, however, basal GLP-1 secretion from

ce pt

ed

M

an

us

cr

ip t

175

Ac

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

7 Page 8 of 32

colonic cultures was elevated ~2-fold, from 2.4 ± 0.2% per 2 hours in chow fed mice (n=12 wells from

206

n=4 mice) to 5.6 ± 0.9% in HFD-fed mice (n=9 wells from n=3 mice, p=0.0005).

207

3.4 Effect of HFD on L-cell gene expression

208

The effect of HFD on gene expression in L-cells was evaluated using GLU-Venus mice that were fed

209

on a HFD or chow diet for 16 weeks. HFD-feeding in this group resulted in a significant increase in

210

body weight as well as fat-pad weight. Small intestinal L-cells and non-L-cells from 3 chow-fed and 4

211

HFD-fed mice were separated by FACS-sorting and subjected to mRNA microarray analysis. Intensity

212

values calculated by RMA analysis indicate the relative expression of individual genes in L-cells and

213

non L-cells, and were compared in both chow and HFD-fed cohorts (Fig 4).

214

Gcg expression was not different between L-cells from HFD and chow-fed mice. This was expected,

215

as L-cells were collected based on their fluorescence of Venus driven by the gcg promoter.

216

Expression of mRNAs for the gut hormones Gip, Cck, Pyy, secretin (Sct) and neurotensin (Nts) were

217

significantly reduced in small intestinal L-cells from HFD-fed mice (Fig 4A). Corresponding with these

218

reduced mRNA levels for gut hormones, L-cells from HFD mice also exhibited lower expression of the

219

prohormone processing enzymes Pcsk1 (prohormone convertase 1/3, p=0.026) and Cpe

220

(carboxypeptidase E, p=0.034) (Fig 4B), as well as members of the granin group, particularly

221

chromogranin B (Chgb, p=0.009) and secretogranin 2 (Scg2, p=0.019) (Fig 4C).

222

To investigate whether the reduced amplification of GLP-1 secretion by nutrients might correlate

223

with lower expression of nutrient-sensing machinery, we examined expression in L-cells of candidate

224

sensory machinery for glucose (Slc5a1, Kcnj11, Abcc8 and Gck), fatty acids (Gpr120, Ffar1), and

225

oligopeptides (Slc15a1, Casr). Expression of Slc5a1 (sglt1, p=0.017), Abcc8 (sur1, p=0.024), Slc15a1

226

(pept1, p=0.006) and Gpr120 (p=0.017), were significantly lower in L-cells from mice on HFD than

227

chow (Fig 5A). The lower expression of Slc5a1, Slc15a1 and Gpr120 in HFD L-cells was also confirmed

228

by qRT-PCR (Fig 5B).

229

To confirm that the data do not represent a global reduction of gene expression in L-cells from HFD-

230

fed mice, we examined expression of members of the fatty acid binding protein family. Fabp1, 2 and

231

6 were found to be ubiquitously expressed in the different gut cell populations, whereas fabp5

232

expression was highly enriched in L-cells (Fig 4D). Expression of the non L-cell enriched FABPs (1, 2

233

and 6) was high in both L-cells and non-L-cells and was unaffected by HFD. Expression of Fabp5, by

234

contrast, was enriched in L-cells compared with control cells, and was significantly lower in L-cells

235

from mice on HFD (p=0.01). Our data therefore suggest that feeding mice with HFD results in

ce pt

ed

M

an

us

cr

ip t

205

Ac

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

8 Page 9 of 32

reduced expression in L-cells of many genes that determine the characteristics and function of an L-

237

cell.

238

We further examined whether the down-regulation of L-cell genes is accompanied by changes in

239

expression of transcription factors known to be involved in determination of the enteroendocrine-

240

cell lineage. As shown in Fig 5C, L-cells from mice on HFD exhibited a significantly reduced expression

241

of Etv1 (p=0.003), Isl1 (p=0.017), Mlxipl (p=0.009), Nkx2.2 (p=0.045) and Rfx6 (p=0.009).

242

ip t

236

4. Discussion

Our data show that mice fed with a HFD exhibited changes in their gut endocrine system compared

244

with mice on chow diet. There were small changes in enteroendocrine cell number, altered hormone

245

secretion, and reduced expression in L-cells of a number of genes that determine the properties and

246

function of an L-cell.

247

Examination of global gene expression in tissue homogenates from different regions of the guts of

248

mice fed for only 2 weeks on HFD revealed changes in mRNAs for gut hormones, particularly in the

249

colon. The numbers of mice used for these experiments were, however, small and the data do not

250

indicate whether the enteroendocrine cell number has changed, or whether the changes represent

251

alterations in hormonal expression within EECs. FACS analysis with immuno-staining was therefore

252

performed to quantify EECs in different regions of the gut and to confirm the production of gut

253

hormones at the peptide level. This revealed that the number of L-cells was lower in the colons of

254

HFD-fed mice, corresponding with the reductions in gene expression observed in the whole tissue

255

homogenates. By FACS analysis it also appeared that the hormonal signature of colonic L-cells was

256

shifted, with L-cells from the fat-fed mice exhibiting reduced CCK production and a tendency for

257

increased PYY production. In the small intestine, by contrast, L-cell numbers did not appear to be

258

affected by high fat feeding, but we observed a reduced number of GIP-stained cells.

259

As it has been reported that there are sufficient L-cells in the proximal small intestine to account for

260

post-prandial hormone release [12], we examined whether L-cell transcriptomics and hormone

261

release were altered in small intestinal L-cells. A large number of transcriptomic changes were

262

observed in L-cells from mice fed on HFD, with a significant down regulation of expression of mRNAs

263

encoding L-cell specific transcription factors, enteroendocrine hormones, prohormone processing

264

enzymes, granins and nutrient sensing machinery. The data do not appear to represent either a

265

global reduction in gene expression in L-cells, nor contamination of the FACS-purified L-cell pool by

266

non L-cells. We therefore conclude that the L-cells in mice fed HFD had impaired production of many

267

genes required for the normal function of an enteroendocrine cell. This coincided with reduced

ce pt

ed

M

an

us

cr

243

Ac

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

9 Page 10 of 32

expression of a number of EEC transcription factors. Isl1, Nkx2.2 and Rfx6 are transcription factors

269

well known to play a role in the EEC lineage [13-15]. As Rfx6 has been reported previously to

270

promote Gip expression [15], the reduction in Rfx6 might account for our observation of reduced Gip

271

mRNA in L-cells. Etv1 and Mlxipl (encoding CHREBP) were identified in our previous analysis of

272

transcription factors enriched in L-cells [16]. The reduced expression in HFD L-cells of Slc5a1, Abcc8,

273

Slc15a1 and Gpr120 would suggest that L-cells from mice on HFD might exhibit reduced nutrient

274

responsiveness [17-19]. Indeed the changes in gene expression corresponded with an apparent

275

reduced responsiveness of GLP-1 release to L-cell secretagogues in primary small intestinal cultures.

276

Basal GLP-1 release, by contrast was elevated both in small intestinal cultures of mice fed with HFD

277

for 2 weeks, and in colonic cultures of mice fed on HFD for 4 months. It is unclear why the basal rate

278

of GLP-1 release was increased in the HFD-fed models, but we have observed a similar finding in

279

ob/ob mice fed on chow (AMH, FG, FR unpublished observations). Whilst we cannot exclude the

280

possibility that the L-cells from HFD and ob/ob mice are more fragile, resulting in the lysis of a few

281

cells during the incubation period that would mimic a high basal secretory rate, we believe it is more

282

likely that an imbalance of second messenger signalling pathways underlies this increased rate of

283

basal secretion in the absence of added nutritional stimuli.

284

Our data showed that substantial changes in the properties of L-cells arose when mice were fed on a

285

HFD. Compared with chow, the HFD we used contained a higher percentage of fat (60%) and free

286

sugars, and no fibre. We cannot therefore be certain which of these dietary macronutrient changes

287

was responsible for the alterations in L-cell gene expression. Diets rich in fat and sugar and low in

288

fibre are, however, commonly consumed by humans in the western world. Our data would suggest

289

that such diets may have adverse effects on our gut hormones. Whether humans who eat a diet rich

290

in fat and sugar for prolonged periods exhibit increased basal GLP-1 release and lowered

291

responsiveness to food ingestion will be interesting to examine in the future. A reduced post-

292

prandial elevation of GLP-1 would, however, tend to lower the body’s ability to signal satiety, and

293

could result in a vicious cycle of over-eating. Resetting gut hormone responsiveness by a period on a

294

healthy diet might, therefore be a strategy to increase post-prandial satiety and tackle obesity.

ce pt

ed

M

an

us

cr

ip t

268

Ac

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

295 296

5. Acknowledgements

297

This work was funded by grants from the Wellcome Trust (WT088357/Z/09/Z and

298

WT084210/Z/07/Z), the MRC Metabolic Diseases Unit (MRC_MC_UU_12012/3) and Full4Health

10 Page 11 of 32

299

(FP7/2011-2015, grant agreement n° 266408). GLP-1 immuno-assays were performed by Keith

300

Burling at the MRC-MDU.

301 302

6. Figure Legends Figure 1: Effect of 2 weeks HFD on hormone gene expression in the gastrointestinal tract

304

Expression of gut hormone mRNAs was quantified by qRT-PCR in tissue homogenates from the small

305

intestine (SI), colon and rectum of mice fed on chow (white bars) or HFD (black bars) for 2 weeks.

306

Expression of Gcg (A), Gip (B), Cck (C), Pyy (D), Insl5 (E) and Sst (F) is shown relative to that of Actb

307

measured in parallel on the same samples. Columns represent the mean of n=5 (chow) and n=7

308

(HFD) samples, and the error bar represents 1SEM. Statistics were performed on the CT data (actb-

309

gene of interest) and the data transformed by 2^CT for illustration in the figures. * p<0.05, **

310

p<0.01 by Student’s t-test.

311

Figure 2: Analysis of enteroendocrine cell numbers by FACS

312

A. The frequency of L-cells was reduced in the colons of mice fed on HFD for 2 weeks (black bars)

313

compared with chow-fed controls (white bars), as assessed by counting Venus-positive cells by FACS

314

in tissues from GLU-Venus mice. B-D. Colonic cell suspensions, as in A, were fixed and stained for PYY

315

and CCK, using red and far red secondary antibodies. B. The red/far red fluorescence intensity

316

histogram (indicating the intensity of staining for PYY and CCK) of the Venus positive L-cell

317

population is shown for one representative each of a chow-fed (control) and HFD-fed mouse.

318

Increased intensity of PYY staining, and reduced intensity of CCK staining is apparent in the HFD-fed

319

mouse. C. CCK and PYY staining obtained as in B are depicted for a representative chow fed mouse

320

without/with primary antibodies (left and centre, respectively) and a HFD-fed mouse. Each dot

321

represents an individual Venus-positive cell.

322

quadrants 1-3 (Q1, Q2, Q3) for cell suspensions analysed as in C. E. The small intestines of chow-fed

323

and HFD-fed mice were analysed by FACS, as in A,B, and the frequencies of cells positive for Venus,

324

CCK and GIP were quantified as a percentage of the total cell number. Columns represent the mean,

325

and error bars represent 1SEM of n=3 (colon) and n=5 (small intestine). * p<0.05, *** p<0.001, by

326

Student’s t-test.

ce pt

ed

M

an

us

cr

ip t

303

Ac

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

D. Mean percentages of Venus-positive cells in

327 328

Figure 3: Effect of HFD on small intestinal GLP-1 release

11 Page 12 of 32

Small intestinal cultures from mice fed for 2 weeks on chow (open bars) or HFD (black bars) were

330

incubated for 2 hours in control buffer (A), or in buffer containing 10 mM glucose (Glc), 0.5%

331

peptone (Pep), 100 M linoleic acid (LA), 10 mM glucose + 10 M forskolin + 10 M IBMX (F/I/G), 20

332

mM Gly-Leu (G-L) or 1 M phorbol meryistate acetate (PMA) (B). GLP-1 concentrations were

333

measured at the end of the experiment in both the supernatant and cell lysate of each well. In A,

334

basal GLP-1 secretion is represented as the % of the total GLP-1 content released in 2 hours under

335

control conditions (n=9 wells per column, comprising 3 wells each from 3 independent mice). In B,

336

secretion has been normalised to the basal secretory rate measured in control wells from the same

337

culture (n=9, as in A). * p<0.05, *** p<0.001 by 2-way ANOVA with post-hoc Bonferroni test,

338

performed on log transformed data.

cr

ip t

329

us

339 Figure 4: Effect of HFD on L-cell gene expression

341

Small intestinal L-cells and control non-L-cells were purified by FACS from mice fed for 16 weeks on

342

chow or HFD, and analysed using Affymetrix ST1.0 microarrays. Data are represented as the RMA

343

intensity. Non L-cells are represented by striped bars (chow, horizontal, n=3; HFD, diagonal, n=4),

344

and L-cells by filled bars (chow, white, n=3; HFD, black, n=3). Expression of gut hormones (A),

345

prohormone processing enzymes (B), granins (C) and fatty acid binding proteins (D) are illustrated.

346

Data are depicted as the mean and 1SEM. Expression in L-cells from chow and HFD mice was

347

compared by Student’s t-test: * p<0.05, ** p<0.01, *** p<0.001.

ed

M

an

340

348

Figure 5: Effect of HFD on L-cell expression of nutrient sensing genes and transcription factors

350

Small intestinal L-cells and control non-L-cells were purified by FACS from mice fed for 16 weeks on

351

chow or HFD, and analysed using Affymetrix ST1.0 microarrays. Data in A and C are represented as

352

the RMA intensity. Non L-cells are represented by striped bars (chow, horizontal, n=3; HFD, diagonal,

353

n=4), and L-cells by filled bars (chow, white, n=3; HFD, black, n=3). B represents gene expression in L-

354

cells analysed by qRT-PCR, analysed as CT data (actb-gene of interest) and transformed by 2^CT

355

for illustration: chow, white, n=4; HFD, black, n=4. In all panels, expression in L-cells from chow and

356

HFD mice was compared by Student’s t-test: * p<0.05, ** p<0.01, *** p<0.001.

ce pt

349

Ac

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

357 358 359 360 361

7. References [1] Gribble FM. The gut endocrine system as a coordinator of postprandial nutrient homoeostasis. Proc Nutr Soc 2012;71:456-62. [2] Holst JJ. The physiology of glucagon-like peptide 1. Physiological Reviews 2007;87:1409-39.

12 Page 13 of 32

ed

M

an

us

cr

ip t

[3] Sjölund K, Sandén G, Håkanson R, Sundler F. Endocrine cells in human intestine: an immunocytochemical study. Gastroenterology 1983;85:1120-30. [4] Petersen N, Reimann F, Bartfeld S, Farin HF, Ringnalda FC, Vries RG, van den Brink S, Clevers H, Gribble FM, de Koning EJ. Generation of L-cells in mouse and human small intestine organoids. Diabetes 2013. [5] Wichmann A, Allahyar A, Greiner TU, Plovier H, Lundén G, Larsson T, Drucker DJ, Delzenne NM, Cani PD, Bäckhed F. Microbial modulation of energy availability in the colon regulates intestinal transit. Cell Host Microbe 2013;14:582-90. [6] Everard A, Lazarevic V, Derrien M, Girard M, Muccioli GM, Neyrinck AM, Possemiers S, Van Holle A, François P, de Vos WM, Delzenne NM, Schrenzel J, Cani PD. Responses of gut microbiota and glucose and lipid metabolism to prebiotics in genetic obese and diet-induced leptin-resistant mice. Diabetes 2011;60:2775-86. [7] Mumphrey MB, Patterson LM, Zheng H, Berthoud HR. Roux-en-Y gastric bypass surgery increases number but not density of CCK-, GLP-1-, 5-HT-, and neurotensin-expressing enteroendocrine cells in rats. Neurogastroenterol Motil 2013;25:e70-9. [8] Hansen CF, Bueter M, Theis N, Lutz T, Paulsen S, Dalbøge LS, Vrang N, Jelsing J. Hypertrophy dependent doubling of L-cells in Roux-en-Y gastric bypass operated rats. PLoS One 2013;8:e65696. [9] Sakar Y, Duca FA, Langelier B, Devime F, Blottiere H, Delorme C, Renault P, Covasa M. Impact of high-fat feeding on basic helix-loop-helix transcription factors controlling enteroendocrine cell differentiation. Int J Obes (Lond) 2014;38:1440-8. [10] Kappe C, Zhang Q, Nyström T, Sjöholm A. Effects of high-fat diet and the anti-diabetic drug metformin on circulating GLP-1 and the relative number of intestinal L-cells. Diabetol Metab Syndr 2014;6:70. [11] Reimann F, Habib AM, Tolhurst G, Parker HE, Rogers GJ, Gribble FM. Glucose Sensing in L Cells: A Primary Cell Study. Cell Metabolism 2008;8:532-9. [12] Theodorakis MJ, Carlson O, Michopoulos S, Doyle ME, Juhaszova M, Petraki K, Egan JM. Human duodenal enteroendocrine cells: source of both incretin peptides, GLP-1 and GIP. Am J Physiol Endocrinol Metab 2006;290:E550-9. [13] Li HJ, Ray SK, Singh NK, Johnston B, Leiter AB. Basic helix-loop-helix transcription factors and enteroendocrine cell differentiation. Diabetes Obes Metab 2011;13 Suppl 1:5-12. [14] May CL, Kaestner KH. Gut endocrine cell development. Mol Cell Endocrinol 2010;323:70-5. [15] Suzuki K, Harada N, Yamane S, Nakamura Y, Sasaki K, Nasteska D, Joo E, Shibue K, Harada T, Hamasaki A, Toyoda K, Nagashima K, Inagaki N. Transcriptional regulatory factor X6 (Rfx6) increases gastric inhibitory polypeptide (GIP) expression in enteroendocrine K-cells and is involved in GIP hypersecretion in high fat diet-induced obesity. J Biol Chem 2013;288:1929-38. [16] Habib AM, Richards P, Cairns LS, Rogers GJ, Bannon CA, Parker HE, Morley TC, Yeo GS, Reimann F, Gribble FM. Overlap of endocrine hormone expression in the mouse intestine revealed by transcriptional profiling and flow cytometry. Endocrinology 2012;153:3054-65. [17] Gorboulev V, Schürmann A, Vallon V, Kipp H, Jaschke A, Klessen D, Friedrich A, Scherneck S, Rieg T, Cunard R, Veyhl-Wichmann M, Srinivasan A, Balen D, Breljak D, Rexhepaj R, Parker HE, Gribble FM, Reimann F, Lang F, Wiese S, Sabolic I, Sendtner M, Koepsell H. Na(+)-D-glucose cotransporter SGLT1 is pivotal for intestinal glucose absorption and glucose-dependent incretin secretion. Diabetes 2012;61:187-96. [18] Diakogiannaki E, Pais R, Tolhurst G, Parker HE, Horscroft J, Rauscher B, Zietek T, Daniel H, Gribble FM, Reimann F. Oligopeptides stimulate glucagon-like peptide-1 secretion in mice through proton-coupled uptake and the calcium-sensing receptor. Diabetologia 2013;56:2688-96. [19] Parker HE, Adriaenssens A, Rogers G, Richards P, Koepsell H, Reimann F, Gribble FM. Predominant role of active versus facilitative glucose transport for glucagon-like peptide-1 secretion. Diabetologia 2012;55:2445-55.

ce pt

362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410

Ac

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

411

13 Page 14 of 32

*Marked Manuscript Click here to view linked References

1

High Fat Diet impairs the function of Glucagon-like

2

Peptide-1 producing L-cells

3 a

Paul Richards, aRamona Pais, Abdella M Habib, Cheryl A Brighton, Giles SH Yeo, bFrank Reimann,

5

b

6

Metabolic Research Laboratories and MRC Metabolic Diseases Unit, WT-MRC Institute of Metabolic

7

Science, Addenbrooke's Hospital, Cambridge, CB2 0QQ, UK

8

a

9

b

ip t

4

us

These authors contributed equally to the work

cr

Fiona M Gribble

an

Joint corresponding authors

10 Corresponding authors:

12

Fiona Gribble: [email protected]

13

Frank Reimann: [email protected]

14

Metabolic Research Laboratories and MRC Metabolic Diseases Unit, WT-MRC Institute of Metabolic

15

Science, Addenbrooke's Hospital, Cambridge, CB2 0QQ, UK

16

Tel: 0044 1223 336746, Fax: 0044 1223 330598

ce pt

ed

M

11

Ac

17 18

Keywords:

19

GLP-1

20

Enteroendocrine

21

L-cell

22

High fat diet

23

1 Page 15 of 32

Abstract

25

Glucagon-like peptide-1 (GLP-1) acts as a satiety signal and enhances insulin release. This study

26

examined how GLP-1 production from intestinal L-cells is modified by dietary changes.

27

Methods: Transgenic mouse models were utilised in which L-cells could be purified by cell specific

28

expression of a yellow fluorescent protein, Venus. Mice were fed on chow or 60% high fat diet (HFD)

29

for 2 or 16 weeks. L-cells were purified by flow cytometry and analysed by microarray and

30

quantitative RT-PCR. Enteroendocrine cell populations were examined by FACS analysis, and GLP-1

31

secretion was assessed in primary intestinal cultures.

32

Results: 2 weeks HFD reduced the numbers of GLP-1 positive cells in the colon, and of GIP positive

33

cells in the small intestine. Purified small intestinal L-cells showed major shifts in their gene

34

expression profiles. In mice on HFD for 16 weeks, significant reductions were observed in the

35

expression of L-cell specific genes, including those encoding gut hormones (Gip, Cck, Sct, Nts),

36

prohormone processing enzymes (Pcsk1, Cpe), granins (Chgb, Scg2), nutrient sensing machinery

37

(Slc5a1, Slc15a1, Abcc8, Gpr120) and enteroendocrine-specific transcription factors (Etv1, Isl1,

38

Mlxipl, Nkx2.2 and Rfx6). A corresponding reduction in the GLP-1 secretory responsiveness to

39

nutrient stimuli was observed in primary small intestinal cultures.

40

Conclusion: Mice fed on HFD exhibited reduced expression in L-cells of many L-cell specific genes,

41

suggesting an impairment of enteroendocrine cell function. Our results suggest that a western style

42

diet may detrimentally affect the secretion of gut hormones and normal post-prandial signalling,

43

which could impact on insulin secretion and satiety.

cr

us

an

M

ed

46



47



48



Ac

Highlights:

ce pt

44 45

ip t

24

Long term dietary changes impair function of the gut endocrine system High fat diet impairs nutrient-triggered GLP-1 release from murine small intestine L-cells from HFD-fed mice have reduced expression of many L-cell-specific genes

49 50

2 Page 16 of 32

51

1. Introduction Hormones from the gut control food intake and insulin release as well as intestinal motility and

53

secretion [1]. The post-prandial rise in the plasma concentration of GLP-1 signals to the brain that

54

food has arrived in the gastrointestinal tract and to the pancreas that glucose is being absorbed.

55

GLP-1 brings about the sensation of satiety together with the enhancement of insulin release [2]. As

56

dietary habits have changed and people now consume more fat and sugar, this raises concerns

57

about whether dietary intake affects the release of gut hormones, and whether this in turn might

58

contribute to rising levels of obesity and diabetes.

59

GLP-1, like other gut hormones, is released from enteroendocrine cells (EECs) located in the

60

intestinal epithelium. EECs have a life span of only about 5 days, and are constantly replenished from

61

stem cells in the intestinal crypts. Depending on their position along the gastrointestinal tract, EECs

62

exhibit characteristic hormonal signatures. In the duodenum, for example, there is a high number of

63

K-cells producing GIP, whereas in the distal ileum and colon there are more L-cells expressing GLP-1

64

and PYY [3]. A few published studies suggest that the number of EECs can be altered by

65

environmental stimuli. In intestinal organoid cultures in vitro, for example, the number of L-cells is

66

increased by exposure to short chain fatty acids [4], and in vivo, elevated L-cell numbers have been

67

observed in germ-free mice [5], and in rodents on a high-fibre diet [6]. How EECs are affected by

68

dietary exposure is an area of great interest. If it were possible to increase L-cell number, this could

69

lead to increased GLP-1 release and improved glucose tolerance and satiety.

70

Surgical models have been used to investigate how EEC numbers are affected by exposure of

71

different gut segments to luminal nutrients. In rats, gastric bypass surgery has been reported to

72

result in an increase in L-cell numbers in the roux (alimentary) and common limbs [7, 8]. In both

73

cases, it was found that the increase in L-cell numbers was attributable to mucosal hypertrophy, but

74

that there was no change in L-cell density. EECs also developed normally in the biliopancreatic limb.

75

These studies suggest that increased ileal nutrient exposure in the context of a chow diet results in

76

mucosal growth and a corresponding increase in L-cell number, but that the frequency of L-cells is

77

not altered.

78

It is less clear how the intestine responds to diets containing a high fat content. In high-fat diet (HFD)

79

fed rodents, it was reported that there is a reduction in the density of cells staining for chromogranin

80

A and GLP-1, and correspondingly reduced expression of Gcg and Pyy by PCR [9, 10]. In this study we

81

have examined the transcriptomic and secretory properties of L-cells from mice fed on a high fat

82

(60%), high sugar diet.

Ac

ce pt

ed

M

an

us

cr

ip t

52

3 Page 17 of 32

83 84

2. Materials and Methods 2.1 Animals:

86

All procedures were approved by the UK Home Office and local ethical committee of the University

87

of Cambridge. Male GLU-Venus mice [11] on a C57Bl6 background were weighed at age 8 weeks and

88

divided into 2 groups (n=4-6 per group), each with a balanced range of body weights. For the 16

89

week study, they were then single housed and placed on either a standard control chow (Special

90

Diets Services, Rat and mouse breeder and grower diet) or high fat diet (Research Diets, #D12331,

91

containing 60% energy from fat). Mice were weighed weekly for 16 weeks. Mice on the 2 week study

92

were treated similarly, but were not single housed. Mice were killed by cervical dislocation 2-4 hours

93

after lights on. Intestinal tissues were harvested and treated as below for flow cytometry (FACS),

94

mRNA extraction or secretion experiments.

95

2.2 Small intestine for FACS sorting

96

For purification of cell populations by FACS, intestinal pieces were stripped of the outer muscle

97

layers. Tissue was chopped into 1–2 mm pieces and digested to single cells with 1 mg/ml collagenase

98

in calcium-free Hanks Buffered Salt Solution (HBSS). Single cell suspensions were separated by flow

99

cytometry using a MoFlo Beckman Coulter Cytomation sorter (FL, USA). Side scatter, forward scatter

100

and pulse width gates were used to exclude debris and aggregates, and venus positive cells collected

101

at ~95% purity, alongside negative (control) cells that comprised a mixed population dominated by

102

other epithelial cell types.

103

2.3 FACS analysis

104

Single cell suspensions were prepared for FACS analysis, as described previously (9). Briefly, cells

105

were fixed with 4% paraformaldehyde (PFA) in phosphate buffered saline (PBS), permeabilised with

106

0.1% v/v Triton X-100, blocked with 10% goat serum and then incubated with primary antibody in

107

PBS-10% goat serum at 4°C overnight for an hour at RT. The primary antibodies used were: anti-GIP

108

(gifted by Prof JJ Holst; 1:1000), anti-CCK (gifted by Prof GJ Dockray; 1:500) and anti-PYY (Progen,

109

London, UK; 16066; 1:100). Cells were rinsed 3 times and then incubated with secondary antibody

110

(Alexa-Fluor 555, 633 or 647 Invitrogen, USA; A-21435, A-21105, A-21428 or A-21070, all at 1:800).

111

Cells were analysed using a LSRCyAn advance digital processing Flow Cytometer (Dako Cytomation,

112

CA, USAFortessa (BD Biosciences)). Events with very low side and forward scatter were excluded as

Ac

ce pt

ed

M

an

us

cr

ip t

85

4 Page 18 of 32

these are likely to represent debris, and events with a high pulse width were excluded to eliminate

114

cell aggregates.

115

2.4 RNA extraction and qRT-PCR

116

Tissues used for RNA and protein extraction were washed in PBS and placed in RNAlater (Ambion,

117

Life Technologies, Paisley UK) and frozen until processed. Samples were homogenised in Tri-reagent

118

(Sigma), and then treated as below. Total RNA from FACS-sorted cells prepared from GLU-Venus

119

transgenic mice was isolated using a micro scale RNA isolation kit (Ambion). All samples were

120

reverse transcribed according to standard protocols. Quantitative RT-PCR was performed with 7900

121

HT Fast Real-Time PCR system (Applied Biosystems, Life Technologies), using verified Taqman

122

primer/probe sets supplied by Applied Biosystems. In all cases expression was compared with that of

123

β-actin measured on the same sample in parallel on the same plate, giving a CT difference (ΔCT) for

124

β-actin minus the test gene. Mean, standard error, and statistics were performed on the ΔCT data

125

and only converted to relative expression levels (2ΔCT) for presentation in the figures.

126

2.5 Microarray

127

The quality of RNA samples was determined by Total RNA Pico Chip (Agilent Technologies, Stockport,

128

UK), and only those with RIN values >7.0 selected for microarray analysis. Total RNA from the

129

samples were amplified using the NuGEN Ovation Pico WTA system (NuGEN Technologies, Leek,

130

Netherlands) according to the manufacturer’s protocol, labelled with the NuGEN Exon and Encore™

131

Biotin Modules, and hybridized onto Affymetrix Mouse Gene ST 1.0 arrays. Raw microarray image

132

data was converted to CEL files using Affymetrix GeneChip Operating Software. All the downstream

133

analysis of the microarray data was performed using GeneSpring GX 12.1 (Agilent). The CEL files

134

were used for both the robust multi-array average (RMA) and Probe Logarithmic Intensity Error

135

(PLIER) analyses. After importing the data, each chip was normalized to the 50th centile of the

136

measurements taken from that chip.

137

To assess the purity of L-cell sorts, each probe on the microarray data was assigned a value

138

indicating its relative expression in L-cells vs non L-cells (combining all microarrays from HFD and

139

chow-fed mice), and the data were ordered to indicate the top 50 probes most enriched in the non

140

L-cell population. For each L-cell microarray we then calculated the geometric mean of these 50 non

141

L-cell probes as a measure of contamination by non L-cells. One out of the 4 microarrays from HFD L-

142

cells was excluded from further analysis because its mean expression of non L-cell probes was >2 SD

143

above the mean of all L-cell microarrays. In further analyses, the number of individual microarrays

144

used was: L-cells on chow (3), L-cells on HFD (3), non L-cells on chow (3) and non L-cells on HFD (4).

Ac

ce pt

ed

M

an

us

cr

ip t

113

5 Page 19 of 32

2.6 Primary intestinal culture

146

Small intestinal and colonic primary cultures were prepared as previously described [11]. Briefly,

147

mice 3-6 months old were sacrificed by cervical dislocation and the small intestine or colon was

148

excised. Luminal contents were flushed thoroughly with PBS and the outer muscle layer removed.

149

Tissue was minced and digested with Collagenase Type XI and the cell suspension plated onto 24-

150

well plates pre-coated with Matrigel (BD Bioscience, Oxford, UK).

151

2.7 GLP-1 secretion assay

152

18-24 hours after plating, cells were washed and incubated with test agents made up in 0.25 ml

153

saline buffer containing: (in mM: 138 NaCl, 4.5 KCl, 4.2 NaHCO3, 1.2 NaH2PO4, 2.6 CaCl2, 1.2 MgCl2,

154

and 10 HEPES, pH 7.4 with NaOH) supplemented with 0.1% BSA for 2 hours at 37 oC. At the end of

155

the 2 hour incubation, supernatants were collected and centrifuged at 2000 rcf for 5 minutes and

156

snap frozen on dry ice. Cells were mechanically disrupted in 0.5 ml lysis buffer containing (mM): 50

157

Tris-HCl, 150 NaCl, 1 % IGEPAL-CA 630, 0.5 % deoxycholic acid (DCA) and complete EDTA-free

158

protease inhibitor cocktail (Roche, Burgess Hill, UK) to extract intracellular peptides, centrifuged at

159

10,000 rcf for 10 minutes and snap frozen. GLP-1 was measured using an electrochemiluminescence

160

total GLP-1 assay (MesoScale Discovery, Gaithersburg, MD, USA) and results expressed as a percent

161

of total (secreted+lysate) GLP-1 and normalised to basal secretion in response to saline measured in

162

parallel on the same day. Chemicals were purchased from Sigma (Poole, UK) unless otherwise

163

indicated.

164

2.8 Data analysis

165

Results are expressed as mean ± SEM. Statistical analysis was performed using GraphPad Prism 5.01

166

(San Diego, CA, USA). For GLP-1 secretion data, one-way ANOVA with post hoc Dunnett’s or

167

Bonferroni’s tests were performed on log-transformed secretion data, as these data were

168

heteroscedastic. Values were regarded as significant when p <0.05.

Ac

ce pt

ed

M

an

us

cr

ip t

145

169 170

3. Results

171

3.1 HFD affects the expression of gut hormones mRNAs in tissue homogenates

172

Two weeks on HFD resulted in reduced expression of Gcg and Insl5 in colonic tissue, and a tendency

173

towards reduced expression of Cck and Pyy, suggesting that HFD causes either a reduction in colonic

174

L-cell density or impaired production of hormones by individual enteroendocrine cells (Fig 1). No 6 Page 20 of 32

significant changes were observed in the expression of Gcg, Cck, Gip or Pyy in small intestinal tissue

176

homogenates, although we did detect an increase in somatostatin (Sst) mRNA in the ileum of HFD-

177

fed mice. Hormone expression was unaffected in the rectum.

178

3.2 HFD alters gut peptide production by individual L-cells

179

To evaluate the frequency of L-cells and their individual production of peptide hormones, we

180

performed FACS analysis of cell suspensions from transgenic mice expressing a yellow fluorescent

181

protein Venus driven by the gcg promoter, which were immunostained for different gut hormones.

182

Quantification of Venus-labelled L-cells in different regions of the GI tract revealed that the large

183

intestine (colon+rectum) of HFD-fed mice contained a significantly lower percentage of Venus

184

positive L-cells than their chow-fed littermates (Fig 2A). In large intestinal cell suspensions co-

185

stained with antibodies against CCK and PYY, we observed a particular reduction in the number of L-

186

cells staining strongly for CCK, and a corresponding increase in the intensity of L-cell PYY staining in

187

mice fed on HFD for 2 weeks (Fig 2B-D).

188

In the upper small intestine, the frequency of Venus-labelled L-cells was not markedly affected by

189

the HFD (Fig 2E). Both the frequency of GIP-stained cells (Fig 2E), however, and the proportion of L-

190

cells that were immuno-positive for GIP (data not shown) were reduced in small intestinal cell

191

suspensions from HFD-fed mice. Staining for PYY and CCK were not noticeably different between the

192

chow and HFD groups in the small intestine (Fig 2E and data not shown).

193

3.3 GLP-1 release is altered in intestinal cultures from HFD-fed mice

194

We next evaluated whether the function of L-cells was modified by high fat feeding. GLP-1 secretion

195

was assessed in small intestinal cultures from mice fed for 2 weeks on either HFD or chow (Fig 3).

196

Cultures from chow-fed mice had a basal secretory rate of 2.8 % per 2 hours, which was stimulated

197

3.5-fold by 10 mM glucose, 7.9-fold by 0.5 % peptone, 1.6-fold by 100 M linoleic acid, 20-fold by

198

glucose/fsk/IBMX, 3.2-fold by 20 mM gly-leu, and 5.6-fold by 1 M PMA. Cultures from HFD-fed mice

199

exhibited an increased rate of basal GLP-1 release (5.0 % per 2 hours, cf 2.8 % in chow-fed animals,

200

p<0.05). The amplification of GLP-1 secretion above the basal rate by glucose, peptone,

201

forskolin/IBMX and gly-leu was reduced in the HFD-fed group (Fig 3) although absolute % secretory

202

rates in the presence of different stimuli were similar in chow-fed and HFD-fed mice.

203

Basal and stimulated GLP-1 release from colonic cultures were not altered by placing mice on HFD

204

for 2 weeks (data not shown). In mice fed for 16 weeks on HFD, however, basal GLP-1 secretion from

Ac

ce pt

ed

M

an

us

cr

ip t

175

7 Page 21 of 32

colonic cultures was elevated ~2-fold, from 2.4 ± 0.2% per 2 hours in chow fed mice (n=12 wells from

206

n=4 mice) to 5.6 ± 0.9% in HFD-fed mice (n=9 wells from n=3 mice, p=0.0005).

207

3.4 Effect of HFD on L-cell gene expression

208

The effect of HFD on gene expression in L-cells was evaluated using GLU-Venus mice that were fed

209

on a HFD or chow diet for 16 weeks. HFD-feeding in this group resulted in a significant increase in

210

body weight as well as fat-pad weight. Small intestinal L-cells and non-L-cells from 3 chow-fed and 4

211

HFD-fed mice were separated by FACS-sorting and subjected to mRNA microarray analysis. Intensity

212

values calculated by RMA analysis indicate the relative expression of individual genes in L-cells and

213

non L-cells, and were compared in both chow and HFD-fed cohorts (Fig 4).

214

Gcg expression was not different between L-cells from HFD and chow-fed mice. This was expected,

215

as L-cells were collected based on their fluorescence of Venus driven by the gcg promoter.

216

Expression of mRNAs for the gut hormones Gip, Cck, Pyy, secretin (Sct) and neurotensin (Nts) were

217

significantly reduced in small intestinal L-cells from HFD-fed mice (Fig 4A). Corresponding with these

218

reduced mRNA levels for gut hormones, L-cells from HFD mice also exhibited lower expression of the

219

prohormone processing enzymes Pcsk1 (prohormone convertase 1/3, p=0.026) and Cpe

220

(carboxypeptidase E, p=0.034) (Fig 4B), as well as members of the granin group, particularly

221

chromogranin B (Chgb, p=0.009) and secretogranin 2 (Scg2, p=0.019) (Fig 4C).

222

To investigate whether the reduced amplification of GLP-1 secretion by nutrients might correlate

223

with lower expression of nutrient-sensing machinery, we examined expression in L-cells of candidate

224

sensory machinery for glucose (Slc5a1, Kcnj11, Abcc8 and Gck), fatty acids (Gpr120, Ffar1), and

225

oligopeptides (Slc15a1, Casr). Expression of Slc5a1 (sglt1, p=0.017), Abcc8 (sur1, p=0.024), Slc15a1

226

(pept1, p=0.006) and Gpr120 (p=0.017), were significantly lower in L-cells from mice on HFD than

227

chow (Fig 5A). The lower expression of Slc5a1, Slc15a1 and Gpr120 in HFD L-cells was also confirmed

228

by qRT-PCR (Fig 5B).

229

To confirm that the data do not represent a global reduction of gene expression in L-cells from HFD-

230

fed mice, we examined expression of members of the fatty acid binding protein family. Fabp1, 2 and

231

6 were found to be ubiquitously expressed in the different gut cell populations, whereas fabp5

232

expression was highly enriched in L-cells (Fig 4D). Expression of the non L-cell enriched FABPs (1, 2

233

and 6) was high in both L-cells and non-L-cells and was unaffected by HFD. Expression of Fabp5, by

234

contrast, was enriched in L-cells compared with control cells, and was significantly lower in L-cells

235

from mice on HFD (p=0.01). Our data therefore suggest that feeding mice with HFD results in

Ac

ce pt

ed

M

an

us

cr

ip t

205

8 Page 22 of 32

reduced expression in L-cells of many genes that determine the characteristics and function of an L-

237

cell.

238

We further examined whether the down-regulation of L-cell genes is accompanied by changes in

239

expression of transcription factors known to be involved in determination of the enteroendocrine-

240

cell lineage. As shown in Fig 5C, L-cells from mice on HFD exhibited a significantly reduced expression

241

of Etv1 (p=0.003), Isl1 (p=0.017), Mlxipl (p=0.009), Nkx2.2 (p=0.045) and Rfx6 (p=0.009).

242

ip t

236

4. Discussion

Our data show that mice fed with a HFD exhibited changes in their gut endocrine system compared

244

with mice on chow diet. There were small changes in enteroendocrine cell number, altered hormone

245

secretion, and reduced expression in L-cells of a number of genes that determine the properties and

246

function of an L-cell.

247

Examination of global gene expression in tissue homogenates from different regions of the guts of

248

mice fed for only 2 weeks on HFD revealed changes in mRNAs for gut hormones, particularly in the

249

colon. The numbers of mice used for these experiments were, however, small and the data do not

250

indicate whether the enteroendocrine cell number has changed, or whether the changes represent

251

alterations in hormonal expression within EECs. FACS analysis with immuno-staining was therefore

252

performed to quantify EECs in different regions of the gut and to confirm the production of gut

253

hormones at the peptide level. This revealed that the number of L-cells was lower in the colons of

254

HFD-fed mice, corresponding with the reductions in gene expression observed in the whole tissue

255

homogenates. By FACS analysis it also appeared that the hormonal signature of colonic L-cells was

256

shifted, with L-cells from the fat-fed mice exhibiting reduced CCK production and a tendency for

257

increased PYY production. In the small intestine, by contrast, L-cell numbers did not appear to be

258

affected by high fat feeding, but we observed a reduced number of GIP-stained cells.

259

As it has been reported that there are sufficient L-cells in the proximal small intestine to account for

260

post-prandial hormone release [12], we examined whether L-cell transcriptomics and hormone

261

release were altered in small intestinal L-cells. A large number of transcriptomic changes were

262

observed in L-cells from mice fed on HFD, with a significant down regulation of expression of mRNAs

263

encoding L-cell specific transcription factors, enteroendocrine hormones, prohormone processing

264

enzymes, granins and nutrient sensing machinery. The data do not appear to represent either a

265

global reduction in gene expression in L-cells, nor contamination of the FACS-purified L-cell pool by

266

non L-cells. We therefore conclude that the L-cells in mice fed HFD had impaired production of many

267

genes required for the normal function of an enteroendocrine cell. This coincided with reduced

Ac

ce pt

ed

M

an

us

cr

243

9 Page 23 of 32

expression of a number of EEC transcription factors. Isl1, Nkx2.2 and Rfx6 are transcription factors

269

well known to play a role in the EEC lineage [13-15]. As Rfx6 has been reported previously to

270

promote Gip expression [15], the reduction in Rfx6 might account for our observation of reduced Gip

271

mRNA in L-cells. Etv1 and Mlxipl (encoding CHREBP) were identified in our previous analysis of

272

transcription factors enriched in L-cells [16]. The reduced expression in HFD L-cells of Slc5a1, Abcc8,

273

Slc15a1 and Gpr120 would suggest that L-cells from mice on HFD might exhibit reduced nutrient

274

responsiveness [17-19]. Indeed the changes in gene expression corresponded with an apparent

275

reduced responsiveness of GLP-1 release to L-cell secretagogues in primary small intestinal cultures.

276

Basal GLP-1 release, by contrast was elevated both in small intestinal cultures of mice fed with HFD

277

for 2 weeks, and in colonic cultures of mice fed on HFD for 4 months. It is unclear why the basal rate

278

of GLP-1 release was increased in the HFD-fed models, but we have observed a similar finding in

279

ob/ob mice fed on chow (AMH, FG, FR unpublished observations). Whilst we cannot exclude the

280

possibility that the L-cells from HFD and ob/ob mice are more fragile, resulting in the lysis of a few

281

cells during the incubation period that would mimic a high basal secretory rate, we believe it is more

282

likely that an imbalance of second messenger signalling pathways underlies this increased rate of

283

basal secretion in the absence of added nutritional stimuli.

284

Our data showed that substantial changes in the properties of L-cells arose when mice were fed on a

285

HFD. Compared with chow, the HFD we used contained a higher percentage of fat (60%) and free

286

sugars, and no fibre. We cannot therefore be certain which of these dietary macronutrient changes

287

was responsible for the alterations in L-cell gene expression. Diets rich in fat and sugar and low in

288

fibre are, however, commonly consumed by humans in the western world. Our data would suggest

289

that such diets may have adverse effects on our gut hormones. Whether humans who eat a diet rich

290

in fat and sugar for prolonged periods exhibit increased basal GLP-1 release and lowered

291

responsiveness to food ingestion will be interesting to examine in the future. A reduced post-

292

prandial elevation of GLP-1 would, however, tend to lower the body’s ability to signal satiety, and

293

could result in a vicious cycle of over-eating. Resetting gut hormone responsiveness by a period on a

294

healthy diet might, therefore be a strategy to increase post-prandial satiety and tackle obesity.

Ac

ce pt

ed

M

an

us

cr

ip t

268

295 296

5. Acknowledgements

297

This work was funded by grants from the Wellcome Trust (WT088357/Z/09/Z and

298

WT084210/Z/07/Z), the MRC Metabolic Diseases Unit (MRC_MC_UU_12012/3) and Full4Health

10 Page 24 of 32

299

(FP7/2011-2015, grant agreement n° 266408). GLP-1 immuno-assays were performed by Keith

300

Burling at the MRC-MDU.

301 302

6. Figure Legends Figure 1: Effect of 2 weeks HFD on hormone gene expression in the gastrointestinal tract

304

Expression of gut hormone mRNAs was quantified by qRT-PCR in tissue homogenates from the small

305

intestine (SI), colon and rectum of mice fed on chow (white bars) or HFD (black bars) for 2 weeks.

306

Expression of Gcg (A), Gip (B), Cck (C), Pyy (D), Insl5 (E) and Sst (F) is shown relative to that of Actb

307

measured in parallel on the same samples. Columns represent the mean of n=5 (chow) and n=7

308

(HFD) samples, and the error bar represents 1SEM. Statistics were performed on the CT data (actb-

309

gene of interest) and the data transformed by 2^CT for illustration in the figures. * p<0.05, **

310

p<0.01 by Student’s t-test.

311

Figure 2: Analysis of enteroendocrine cell numbers by FACS

312

A. The frequency of L-cells was reduced in the colons of mice fed on HFD for 2 weeks (black bars)

313

compared with chow-fed controls (white bars), as assessed by counting Venus-positive cells by FACS

314

in tissues from GLU-Venus mice. B-D. Colonic cell suspensions, as in A, were fixed and stained for PYY

315

and CCK, using red and far red secondary antibodies. B. The red/far red fluorescence intensity

316

histogram (indicating the intensity of staining for PYY and CCK) of the Venus positive L-cell

317

population is shown for one representative each of a chow-fed (control) and HFD-fed mouse.

318

Increased intensity of PYY staining, and reduced intensity of CCK staining is apparent in the HFD-fed

319

mouse. C. CCK and PYY staining obtained as in B are depicted for a representative chow fed mouse

320

without/with primary antibodies (left and centre, respectively) and a HFD-fed mouse. Each dot

321

represents an individual Venus-positive cell.

322

quadrants 1-3 (Q1, Q2, Q3) for cell suspensions analysed as in C. E. The small intestines of chow-fed

323

and HFD-fed mice were analysed by FACS, as in A,B, and the frequencies of cells positive for Venus,

324

CCK and GIP were quantified as a percentage of the total cell number. Columns represent the mean,

325

and error bars represent 1SEM of n=3 (colon) and n=5 (small intestine). * p<0.05, *** p<0.001, by

326

Student’s t-test.

Ac

ce pt

ed

M

an

us

cr

ip t

303

D. Mean percentages of Venus-positive cells in

327 328

Figure 3: Effect of HFD on small intestinal GLP-1 release

11 Page 25 of 32

Small intestinal cultures from mice fed for 2 weeks on chow (open bars) or HFD (black bars) were

330

incubated for 2 hours in control buffer (A), or in buffer containing 10 mM glucose (Glc), 0.5%

331

peptone (Pep), 100 M linoleic acid (LA), 10 mM glucose + 10 M forskolin + 10 M IBMX (F/I/G), 20

332

mM Gly-Leu (G-L) or 1 M phorbol meryistate acetate (PMA) (B). GLP-1 concentrations were

333

measured at the end of the experiment in both the supernatant and cell lysate of each well. In A,

334

basal GLP-1 secretion is represented as the % of the total GLP-1 content released in 2 hours under

335

control conditions (n=9 wells per column, comprising 3 wells each from 3 independent mice). In B,

336

secretion has been normalised to the basal secretory rate measured in control wells from the same

337

culture (n=9, as in A). * p<0.05, *** p<0.001 by 2-way ANOVA with post-hoc Bonferroni test,

338

performed on log transformed data.

cr

ip t

329

us

339 Figure 4: Effect of HFD on L-cell gene expression

341

Small intestinal L-cells and control non-L-cells were purified by FACS from mice fed for 16 weeks on

342

chow or HFD, and analysed using Affymetrix ST1.0 microarrays. Data are represented as the RMA

343

intensity. Non L-cells are represented by striped bars (chow, horizontal, n=3; HFD, diagonal, n=4),

344

and L-cells by filled bars (chow, white, n=3; HFD, black, n=3). Expression of gut hormones (A),

345

prohormone processing enzymes (B), granins (C) and fatty acid binding proteins (D) are illustrated.

346

Data are depicted as the mean and 1SEM. Expression in L-cells from chow and HFD mice was

347

compared by Student’s t-test: * p<0.05, ** p<0.01, *** p<0.001.

ed

M

an

340

348

Figure 5: Effect of HFD on L-cell expression of nutrient sensing genes and transcription factors

350

Small intestinal L-cells and control non-L-cells were purified by FACS from mice fed for 16 weeks on

351

chow or HFD, and analysed using Affymetrix ST1.0 microarrays. Data in A and C are represented as

352

the RMA intensity. Non L-cells are represented by striped bars (chow, horizontal, n=3; HFD, diagonal,

353

n=4), and L-cells by filled bars (chow, white, n=3; HFD, black, n=3). B represents gene expression in L-

354

cells analysed by qRT-PCR, analysed as CT data (actb-gene of interest) and transformed by 2^CT

355

for illustration: chow, white, n=4; HFD, black, n=4. In all panels, expression in L-cells from chow and

356

HFD mice was compared by Student’s t-test: * p<0.05, ** p<0.01, *** p<0.001.

Ac

ce pt

349

357 358 359 360 361

7. References [1] Gribble FM. The gut endocrine system as a coordinator of postprandial nutrient homoeostasis. Proc Nutr Soc 2012;71:456-62. [2] Holst JJ. The physiology of glucagon-like peptide 1. Physiological Reviews 2007;87:1409-39.

12 Page 26 of 32

ce pt

ed

M

an

us

cr

ip t

[3] Sjölund K, Sandén G, Håkanson R, Sundler F. Endocrine cells in human intestine: an immunocytochemical study. Gastroenterology 1983;85:1120-30. [4] Petersen N, Reimann F, Bartfeld S, Farin HF, Ringnalda FC, Vries RG, van den Brink S, Clevers H, Gribble FM, de Koning EJ. Generation of L-cells in mouse and human small intestine organoids. Diabetes 2013. [5] Wichmann A, Allahyar A, Greiner TU, Plovier H, Lundén G, Larsson T, Drucker DJ, Delzenne NM, Cani PD, Bäckhed F. Microbial modulation of energy availability in the colon regulates intestinal transit. Cell Host Microbe 2013;14:582-90. [6] Everard A, Lazarevic V, Derrien M, Girard M, Muccioli GM, Neyrinck AM, Possemiers S, Van Holle A, François P, de Vos WM, Delzenne NM, Schrenzel J, Cani PD. Responses of gut microbiota and glucose and lipid metabolism to prebiotics in genetic obese and diet-induced leptin-resistant mice. Diabetes 2011;60:2775-86. [7] Mumphrey MB, Patterson LM, Zheng H, Berthoud HR. Roux-en-Y gastric bypass surgery increases number but not density of CCK-, GLP-1-, 5-HT-, and neurotensin-expressing enteroendocrine cells in rats. Neurogastroenterol Motil 2013;25:e70-9. [8] Hansen CF, Bueter M, Theis N, Lutz T, Paulsen S, Dalbøge LS, Vrang N, Jelsing J. Hypertrophy dependent doubling of L-cells in Roux-en-Y gastric bypass operated rats. PLoS One 2013;8:e65696. [9] Sakar Y, Duca FA, Langelier B, Devime F, Blottiere H, Delorme C, Renault P, Covasa M. Impact of high-fat feeding on basic helix-loop-helix transcription factors controlling enteroendocrine cell differentiation. Int J Obes (Lond) 2014;38:1440-8. [10] Kappe C, Zhang Q, Nyström T, Sjöholm A. Effects of high-fat diet and the anti-diabetic drug metformin on circulating GLP-1 and the relative number of intestinal L-cells. Diabetol Metab Syndr 2014;6:70. [11] Reimann F, Habib AM, Tolhurst G, Parker HE, Rogers GJ, Gribble FM. Glucose Sensing in L Cells: A Primary Cell Study. Cell Metabolism 2008;8:532-9. [12] Theodorakis MJ, Carlson O, Michopoulos S, Doyle ME, Juhaszova M, Petraki K, Egan JM. Human duodenal enteroendocrine cells: source of both incretin peptides, GLP-1 and GIP. Am J Physiol Endocrinol Metab 2006;290:E550-9. [13] Li HJ, Ray SK, Singh NK, Johnston B, Leiter AB. Basic helix-loop-helix transcription factors and enteroendocrine cell differentiation. Diabetes Obes Metab 2011;13 Suppl 1:5-12. [14] May CL, Kaestner KH. Gut endocrine cell development. Mol Cell Endocrinol 2010;323:70-5. [15] Suzuki K, Harada N, Yamane S, Nakamura Y, Sasaki K, Nasteska D, Joo E, Shibue K, Harada T, Hamasaki A, Toyoda K, Nagashima K, Inagaki N. Transcriptional regulatory factor X6 (Rfx6) increases gastric inhibitory polypeptide (GIP) expression in enteroendocrine K-cells and is involved in GIP hypersecretion in high fat diet-induced obesity. J Biol Chem 2013;288:1929-38. [16] Habib AM, Richards P, Cairns LS, Rogers GJ, Bannon CA, Parker HE, Morley TC, Yeo GS, Reimann F, Gribble FM. Overlap of endocrine hormone expression in the mouse intestine revealed by transcriptional profiling and flow cytometry. Endocrinology 2012;153:3054-65. [17] Gorboulev V, Schürmann A, Vallon V, Kipp H, Jaschke A, Klessen D, Friedrich A, Scherneck S, Rieg T, Cunard R, Veyhl-Wichmann M, Srinivasan A, Balen D, Breljak D, Rexhepaj R, Parker HE, Gribble FM, Reimann F, Lang F, Wiese S, Sabolic I, Sendtner M, Koepsell H. Na(+)-D-glucose cotransporter SGLT1 is pivotal for intestinal glucose absorption and glucose-dependent incretin secretion. Diabetes 2012;61:187-96. [18] Diakogiannaki E, Pais R, Tolhurst G, Parker HE, Horscroft J, Rauscher B, Zietek T, Daniel H, Gribble FM, Reimann F. Oligopeptides stimulate glucagon-like peptide-1 secretion in mice through proton-coupled uptake and the calcium-sensing receptor. Diabetologia 2013;56:2688-96. [19] Parker HE, Adriaenssens A, Rogers G, Richards P, Koepsell H, Reimann F, Gribble FM. Predominant role of active versus facilitative glucose transport for glucagon-like peptide-1 secretion. Diabetologia 2012;55:2445-55.

Ac

362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411

13 Page 27 of 32

Figure

B

0.06

Cck/actb

Gip/actb

0.08

0.006 0.005 0.004 0.002

0

0 Colon

Rectum

Upper SI

E

0.045 0.04

Insl5/actb

0.035 0.03 0.025 0.02

Lower SI Colon

Rectum

Upper SI

F

0.06

0.014

0.05

0.012

0.04

0.01

Sst/actb

Lower SI

0.03

Lower SI

Colon

Rectum

Colon

Rectum

*

0.008

us

Upper SI

0.006

0.002

0.001

0

0.008

0.004

0.003

0.02

0.006

0.015

0.02

0.01

*

0.01

0

0.002

0 Colon

Rectum

Upper SI

Lower SI

Colon

Rectum

0 Upper SI

Lower SI

ed

M

Lower SI

pt

Upper SI

an

0.005

0.004

Ac ce

Pyy/actb

0.01

0.007

0.1

0.04

D

0.012

0.008

0.12 Gcg/actb

C 0.009

ip t

**

0.14

cr

Figure 1 A

Page 28 of 32

Figure 2 A 1.0

B

*

% Venus cells

0.8 Control

0.6

HFD HFD

0.4

Control

0.2 0.0

Secondary only Q2

4

HFD Q2

Q2

cr

3

Q3

D 80

2

3

***

4

Q1

Q3

1 2 3 4 Log PYY fluorescence

***

60

E

Q1

Q3 1

20

3

0.2 *

0.1

0.05

0

0

CCK

GIP

Venus (gcg)

ed

Q3

pt

Q2

Ac ce

Q1

4

0.25

0.15

40

2

an

Q1

M

1

us

2

1

% Venus Cells

1 2 3 4 Log CCK fluorescence

Chow diet

Cell frequency (%)

Log CCK fluorescence

C

1 2 3 4 Log PYY fluorescence

ip t

Control HFD

Page 29 of 32

Figure 3 B

25

***

***

*

0 Glc

Pep

LA

F/I/G

G-L

PMA

us

cr

control

an

chow HFD

*** 5

M

0

*** 10

ed

2

15

pt

4

20

ip t

Relative GLP-1 release

6

Ac ce

Basal GLP-1 secretion (% of content)

A

Page 30 of 32

Figure 4 A

B

***

6000

*

***

5000 4000 3000 2000 1000 0 Gcg

Gip

Cck

Pyy

Sct

*

4000 3000

*

2000 1000 0 Pcsk1

Nts

Pcsk2

Cpe

D

3000

2000

1000

0 Scg2

Scg3

5000 4000 3000

*

2000 1000 0

Scg5

Fabp1

Fabp2

Fabp5

Fabp6

pt

ed

M

Chgb

6000

Ac ce

Chga

7000

cr

*

us

**

an

4000

Relative expression by microarray

C Relative expression by microarray

Relative expression by microarray

Relative expression by microarray

7000

5000

**

8000

ip t

*

9000

Page 31 of 32

Etv1 Hopx

**

Insm1 Isl1 Mlxipl

500

Neurod1 Ngn3

1500

1000 **

Nkx2.2

Pax4

2000

cr

*

Pax6

*

1000

800

*

400

*

0 mRNA expression by qPCR vs actb

1200

Prox1

ip t

Gpbar1

Gpr119

200

Gpr120

Ffar3

Ffar2

Ffar1

Casr

Slc6a19

Slc15a1

Abcc8

600

us

C Kcnj11

1600

*

an

M

Arx

ed

0

pt

2500 Gck

Slc5a1

Relative expression by microarray 1400

Ac ce

Relative expression by microarray

Figure 5 A **

B 0.3 *

0.2 ***

0.1 *

0.0

sglt1

pept1 gpr120

**

Rfx6

Page 32 of 32