soluble fiber viscosity affects both ... -...

The Journal of Nutrition

Nutrient Physiology, Metabolism, and Nutrient-Nutrient Interactions

Soluble Fiber Viscosity Affects Both GobletCell Number and Small Intestine MucinSecretion in Rats1–3

Hiroyuki Ito,4 Mitsuru Satsukawa,5 Eiko Arai,6 Kimio Sugiyama,5 Kei Sonoyama,7 Shuhachi Kiriyama,8

and Tatsuya Morita5*

4Graduate School of Science and Technology and 5Department of Applied Biological Chemistry, Faculty of Agriculture, and 6Department

of Home Economics, Faculty of Education, Shizuoka University, 836 Ohya, Shizuoka 422-8529, Japan; 7Division of Applied

Bioscience Graduate School of Agriculture, Hokkaido University, Sapporo 060-8589 Japan; and 8Luminacoids Laboratory,

Mure 4-1-4, Mitaka, Tokyo 181-0002, Japan

Abstract

We examined the role of soluble fiber viscosity in small intestinal mucin secretion. Viscosities were defined as the area

under the viscosity curve (VAUC). Rats were fed a control diet or diets containing konjac mannan (KM) [low, medium, or

high molecular weight (LKM, MKM, HKM), respectively] at 50 g/kg diet for 10 d. Luminal mucin content and goblet cell

number increased in proportion to the molecular weight of KM. Such effects with the HKM diet were nullified by the

concurrent ingestion of 2 g cellulase/kg diet. Diet containing LKM, MKM, HKM, guar gums (high or low molecular weight;

HGG, LGG), psyllium (PS), or pectin (PC) at 50 g/kg was fed to rats. Fibers with higher VAUC (MKM, HKM, HGG, and PS)

increased goblet cell numbers, but not those with lower VAUC (LKM, LGG, and PC). Luminal mucins were greater in rats

fed HKM, PC, and PS diets. Goblet cell numbers and VAUC were correlated (r = 0.98; P, 0.01). In rats fed the HKM diet,

ilealMuc2 gene expression was not affected, but that ofMuc3 was lower than in those fed the control diet, indicating that

the increase in luminal mucins after ingestion of HKM diet occurred independently of enhancedMuc gene expression. An

incorporation study of 59-bromo-deoxyuridine (BrdU) showed the position of the uppermost-BrdU labeled cell along the villi

was higher in rats fed the HKM diet than in those fed the control diet. The results suggest that soluble fibers, except PC,

upregulate baseline secretion of luminal mucins by increasing goblet cell numbers in proportion to fiber VAUC. J. Nutr.

139: 1640–1647, 2009.

Introduction

The absorptive surface of the small intestine is covered by a layerof mucus composed predominantly of mucin glycoproteins thatare synthesized and secreted by goblet cells. This layer can serveas a barrier between the luminal contents and the absorptivesystem of the intestine and protects the epithelial surface frompotential luminal insults (1). Changes in the properties of thisbarrier may affect the absorption of both dietary nutrients andendogenous macromolecules and ions in the small intestine (2,3).

Dietary features can positively influence characteristics of theintestinal mucus in vivo. Certain milk peptides were shownrecently to stimulate mucin secretion in ex vivo rat intestinalsections (4). Additionally, consumption of dietary fiber appearsto enhance the total capacity for mucin secretion in the small

intestinal lumen, although the stimulatory effect on mucinsecretion was dependent on the quantity, as well as the quality,of dietary fiber ingested (5,6). Our previous studies showed thatsmall intestinal mucins were secreted in proportion to thesettling volume in water (a numerical representation for bulk-forming properties) of dietary indigestible components in rats inso far as water-insoluble dietary fibers (IDF)9 are concerned (6).The mucin secretory effect of IDF was linked to epithelial cellturnover and the subsequent increase in goblet cell numbers (7).Meanwhile, the relationship between small intestinal mucinsecretion and water-soluble dietary fiber (SDF) has not been fullyelucidated.

Previous studies showed that supplementation with 5%citrus fiber in a purified diet produced a significant increase insmall intestinal mucin secretion (5), but supplementation with

1 Supported in part by a Grant-in-Aid for Scientific Research from the Ministry of

Education, Science, Sports and Culture of Japan.2 Author disclosures: T. Morita, M. Satsukawa, H. Ito, E. Arai, K. Sugiyama,

K. Sonoyama, and S. Kiriyama, no conflicts of interest.3 Supplemental Figure 1 is available with the online posting of this paper at jn.

nutrition.org.

* To whom correspondence should be addressed. E-mail: [email protected].

ac.jp.

9 Abbreviations used: BrdU, 59-bromo-deoxyuridine; CL, cellulase; GAPDH,

glyceraldehyde 3-phosphate dehydrogenase; HGG, guar gum of high molecular

weight; HKM, konjac mannan of high molecular weight; IDF, water-insoluble

dietary fiber; KM, konjac mannan; LGG, guar gum of lowmolecular weight; LKM,

konjac mannan of low molecular weight; MKM, konjac mannan of medium

molecular weight; PC, pectin; PS, psyllium; PSF, polystyrene foam; SDF,

water-soluble dietary fiber; VAUC, area under the viscosity curve.

0022-3166/08 $8.00 ã 2009 American Society for Nutrition.

1640 Manuscript received May 15, 2009. Initial review completed June 4, 2009. Revision accepted June 22, 2009.

First published online July 15, 2009; doi:10.3945/jn.109.110171.Downloaded from https://academic.oup.com/jn/article-abstract/139/9/1640/4670525by gueston 18 May 2018

carrageenan, guar gum (8), or psyllium (PS) had no effect (9).Recent studies by Piel et al. (10) showed that ingestion of highlyviscous carboxymethylcellulose increased the number of ilealgoblet cells and luminal crude mucin (determined as a 60%ethanol precipitate) in pigs. Larsen et al. (11) indicated thatincreases in ileal effluent sialic acid concentrations in rats weredependent on the viscosities of the ingested carboxymethylcel-lulose. These observations may suggest that SDF viscosity mightbe a contributing factor in small intestinal mucin secretion. How-ever, a generalized concept regarding the relationship betweenthe viscosity of naturally occurring SDF and small intestinalmucin secretion has not yet been established.

Our primary objective in this study was to precisely examinewhether it is possible to uniformly explain the role of SDF viscosityin small intestinal mucin secretion in relation to epithelial cellturnover, focusing particularly on the change in the number ofmucin-producing goblet cells. For this purpose, we selected severalSDF with different viscosities in solution, and measurementsof reliablemucinmarkers (O-linkedoligosaccharidechains,mucin-protein by ELISA, and sialic acid) were employed using a crudemucin fraction prepared from intestinal contents.

Mucin production is affected by both the increase in numbersof terminally differentiated goblet cells and the increase in Mucgene expression within these cells (12). A previous study showedthat 10% dietary supplementation with wheat bran, as IDF, didnot affect the number of goblet cells per villus, but causedsignificant increases in [3H]-glucose and [35S]-sulfate incorpo-ration into intestinal glycoproteins (13), suggesting that mucinsecretory effects by the ingestion of wheat bran might be partlyexplained by the stimulation of mucin synthesis. Also, lessdramatic results were obtained in rats fed diets containingcellulose (13). To better understand the mechanism by whichviscous fibers increase mucin secretion, we further examinedwhether konjac mannan (KM) and powdered polystyrene foam(PSF), as IDF, share a common effect on epithelial cell migrationand Muc gene expression, regardless of their different physico-chemical properties, such as bulk (settling volume) and viscosity.

Methods

Materials. KM [a copolymer of glucose and mannose (1:1.6) joined

through b-1,4-glucosidic linkages] of various molecular weights was

provided by Shimizu Chemical, i.e. PROPOL A (1000–2000 kDa),

RHEOLEX LM (,1000 kDa), and RHEOLEX RS (1.5 kDa). Thesewere referred to as low (LKM), medium (MKM), and high molecular

weight (HKM) KM. Guar gum of high molecular weight (HGG;

Sunfiber) and its partially hydrolyzed product (LGG; Sunfiber R) wereprovided by Taiyo Kagaku and low methoxy-pectin (PC) and PS were

also fromDainippon Sumitomo Pharm and Bizen Chemical, respectively.

Dietary fiber contents, as determined by the Prosky method (14), were:

LKM (97%), MKM (93%), HKM (95%), LGG (86.6%), HGG(81.5%), PC (81.0%), and PS (90%). The viscosity of a 1.0% solution

of each dietary fiber, preincubated for 6 h at 378C, was measured with a

rotational viscometer (PK100, Haake) at 37.06 0.28C by measuring the

sheer stress developed at sheer rates ranging from 50 to 500 (1/s). A cone-plate sensor of type PK5/18 (diameter, 49.991 mm; cone angle, 1.0048)was used. The data were analyzed by software of Haake RheoWin

(Haake; version, 3.0). Thus, obtained viscosities were expressed as thearea under the viscosity curve (VAUC) described by Dikeman et al. (15)

and were 34.0 Pa (LKM), 421.1 Pa (MKM), 599.3 Pa (HKM), 1.1 Pa

(LGG), 165.8 Pa (HGG), 2.6 Pa (PC), and 67.6 Pa (PS), respectively

(Supplemental Fig. 1). PSF, with an experimentally determined expan-sion ratio (6) of 54.9, was provided by JSP. PSF was powdered using a

Wiley mill with an adjusted mesh size of 30–100. The settling volume (as

a determinant of bulk) of PSF in water was determined to be 11.0 mL/g

by the method described previously (6). Crude cellulase (CL) powder

(CELLULASE NAGASE, prepared from Aspergillus niger) was provided

by Nagase ChemteX. This preparation showed a cellulase activity of 600

units/g. The unit was defined as the activity that liberates 1 mol ofglucose from a 0.625% sodium carboxymethylcellulose solution (pH

4.5) per minute at 408C.

Care of animals. Male Wistar rats (purchased from Shizuoka Labora-

tory Animal Center, Hamamatsu, Japan) were housed in individual wirescreen-bottomed, stainless steel cages in a temperature (23 6 28C) andlight (lights on from 0800–2000) controlled room. For adaptation, rats

were fed a control diet for at least 5 d. This diet (6) was formulated from

250 g/kg of casein, 652.25 g/kg of cornstarch, and 50 g/kg of corn oil.The remainder of the diet consisted of vitamins and minerals. Sub-

sequently, rats were allocated to groups on the basis of body weight and

allowed free access to experimental diets and water. The addition of eachdietary fiber or CL was performed at the expense of an equal amount of

cornstarch. Body weight and food intake were recorded every morning

before replenishing the diet. The study was approved by the Animal Use

Committee of Shizuoka University and animals were maintained inaccordance with the guidelines of Shizuoka University for the care and

use of laboratory animals.

Expt. 1. Rats (130–157 g) were allocated to 5 groups (8 rats each) and

were allowed free access to the control diet or diet containing either 50 gLKM, MKM, or HKM/kg or 80 g PSF/kg for 10 d. We used PSF as a

positive reference showing increasing effects on intestinal goblet cell

numbers and luminal mucin content. From our previous findings (6),

small intestinal mucins are secreted in proportion to the settling volumeof IDF and dietary level of PSF at 8% is enough to increase luminal

mucins when we use the present PSF preparation (SV value = 11.0 mL/g).

Finally, diets were withdrawn overnight, and rats were killed bydecapitation and the small intestine was excised. Luminal contents

were collected by flushing with 15 mL of ice-cold PBS (pH 7.4)

containing 0.02 mol sodium azide/L and the same volume of air. The

contents were freeze-dried and stored for luminal mucin analysis. Forhistologic evaluation, the upper half of the small intestine was defined as

the jejunum and the lower half was defined as the ileum. The mid-

portions of jejunum and ileum segments (~5 cm) were removed and

placed in 10% buffered formalin.


were allowed free access to the control diet or diet containing 50 g of

HKM/kg for 10 d in the presence or absence of 2 g of CL/kg. There were

thus 4 groups: control, control + CL, HKM, and HKM + CL. Thecollection and preparation of luminal contents and intestinal segments

were conducted in the same manner as for Expt. 1.


were allowed free access to the control diet or diet containing either 50 gof LKM, MKM, HKM, LGG, HGG, PC, or PS/kg for 10 d. The

collection and preparation of luminal contents and intestinal segments

were conducted in the same manner as for Expt. 1.

Expt. 4. Rats (129–154 g) were allocated to 3 groups (12 rats each) andwere allowed free access to the control diet or diet containing either 50 g

of HKM/kg or 80 g of PSF/kg for 10 d. The collection and preparation of

luminal contents and ileal segments were conducted in the same manneras for Expt. 1, without overnight food deprivation. A part of the ileal

segment (~10 cm) was opened longitudinally and the mucosa was

scrapped with a glass slide and used for total RNA isolation.


the same dietary treatments as for Expt. 4 were conducted. For the

examination of epithelial cell migration, 59-bromo-deoxyuridine (BrdU)

(50 mg/kg body weight) was intraperitoneally injected into rats at d 9(1000–1100 h). At 24 h after administration, without food deprivation,

rats were killed by decapitation and the jejunum and ileum segments

were removed and placed in 10% buffered formalin.

Mucin analysis. The mucin fraction was isolated by the method of Lien

et al. (16), with some modification, as described previously (6). Finally,

Viscous fiber and mucin secretion in rats 1641

Downloaded from https://academic.oup.com/jn/article-abstract/139/9/1640/4670525by gueston 18 May 2018

the mucin fraction was dissolved in 5.0 mL of distilled water foranalyses. After an appropriate dilution of the mucin fraction, O-linked

oligosaccharide chains were measured using a fluorimetric assay (17)

that discriminated O-linked glycoproteins (mucin) from N-linked

glycoproteins as described by Bovee-Oudenhoven et al. (18). Standardsolutions of N-acetylgalactosamine (Sigma) were used to calculate the

amount of oligosaccharide chains liberated from mucins during the

procedure. Mucin-ELISA was performed as previously described (19).Sialic acid was determined by the method described previously (6). N-

acetylneuramic acid was used as a standard.

Goblet cell staining. Six 5-mm-thick cross-sections were prepared perrat from paraffin-embedded samples and stained with periodic acid

Schiff, counter-stained with hematoxylin. Five complete villi (entire

crypt/villus axis) per section were selected and villus length and the

number of goblet cells per villus (left side) were determined. Twoobservers unaware of the treatments independently analyzed each

section at the light microscopic level (Olympus BH2) with a micrometer

eyepiece.

RNA isolation and quantitative real-time PCR. Total RNA was

isolated using Takara RNAiso reagent (Takara Bio) according to the

manufacturer’s instructions. The concentration and purity of total RNAwas estimated by measuring its OD at 260 and 280 nm. One microgram

of total RNAwas reverse transcribed by Takara Prime Script RT reagent

(Takara Bio) and RT-PCR was performed at 378C for 15 min. The

synthesized cDNA was amplified by PCR using a LightCycler System(Roche Applied Science). The primer pairs for the amplification ofMuc2,Muc3 (20), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH)

(21), which was used as an endogenous reference gene, were used. For

LightCycler reaction, a mastermix of the following reaction componentswas prepared to the indicated end-concentration: 6.4 mL of water, 0.8 mL

of forward primer (0.4 mmol/L), 0.8 mL of reverse primer (0.4 mmol/L),

and 10 mL of SYBR Premix Extaq II (Takara Bio). LightCyclermastermix (18 mL) was filled in the LightCycler glass capillaries and

2 mL cDNAwas added as PCR template. The following PCR conditions

were employed: initial denaturation at 958C for 1 min and 40 cycles with

denaturation (958C, 5 s), annealing [628C: Muc2 (10 s), Muc3 (10 s);578C: GAPDH (10 s)], and elongation [728C: Muc2 (10 s), Muc3 (20 s),

GAPDH (10 s)]. To confirm amplification specificity, the PCR products

from each primer pair were subjected to a melting curve analysis and

subsequent agarose gel electrophoresis. Gene expression was quantifiedusing the comparative threshold of cycles method (22) and the data were

expressed as the relative value to the control group. In the present study,

threshold of cycles values of GAPDH gene among the dietary treatmentswere 18.66 0.12 (control), 18.66 0.06 (PSF), and 18.56 0.12 (HKM)

and did not differ.

BrdU staining. Per animal, 6 5-mm-thick cross-sections were collectedon aminopropyltriethoxysilane-coated slides. After deparaffinization

and rehydration, the sections were immersed in preheated 10 mmol/L

citrate buffer (pH 6.0) and heated at 1008C for 20 min for antigen

retrieval. Then, the sections were treated with 3% H2O2/PBS to blockendogenous peroxidase activity for 10 min. After blocking with 2.5%

horse serum (Vectastain Elite ABC kit, Vector Laboratories) for 30 min,

the sections were incubated with anti-BrdU (1:50, clone: ZBU30, EMDChemicals) for 60 min. The sections were incubated with biotinylated

horse anti-mouse/rabbit IgG (Vectastain Elite ABC kit, Vector Labora-

tories) for 30 min and subsequently incubated with avidin-biotin

peroxidase complex reagent (Vectastain Elite ABC kit, Vector Labora-tories) for 30 min. The section was visualized with 3, 3-diaminobenzidine

tetrahydrochloride substrate followed by counterstaining with hema-

toxylin. BrdU-positive cells were subsequently counted in the same

manner for goblet cell staining.

Statistical analyses.Data were analyzed by 1-way ANOVA except the

results in Expt. 2 (2-way ANOVA). In Expt. 2, interaction of dietary

treatments (HKM 3 CL) was analyzed. Significant differences amongmeans were identified by Tukey-Kramer. Results were expressed as

mean6 SEM and a 5% level of probability was considered a significant

difference in all statements. When variances were not homogenous bythe Bartlett test, data were logarithmically transformed. When vari-

ances were not homogenous even after logarithmic transformation, the

data were presented as medians with range and then analyzed by

Kruskal-Wallis ANOVA followed by Kolmogorov-Smirnov 2-sampletests. The statistical calculations were carried out using Stat View 5.0

computer software (SAS Institute). Regression analyses were per-

formed using the Stat Cel 2 program (Tokyo Shoseki). When correla-

tion coefficients were calculated, nwas at the level of the diet, not at thelevel of the individual rat.

Results

Stimulatory effects of KM on small intestinal mucinsecretion (Expt. 1). Food intake and body weight gain inHKM and MKM groups were significantly lower than in thecontrol and PSF groups, and those in the LKM group wereintermediate (Table 1). Both the amounts of ELISA-detectedmucin and O-linked oligosaccharide chains in the small intes-tinal contents were significantly higher in the HKM and PSFgroups than in the control and LKM groups, whereas those intheMKM group were intermediate among the groups. The sametendency was observed in the amount of sialic acid (P , 0.05).

TABLE 1 Food intake, body weight gain, amount of mucin in the small intestinal contents, andhistologic variables in the small intestinal tissues in rats fed the control diet or a diet eithercontaining 80 g PSF/kg or 50 g LKM, MKM, or HKM/kg for 10 d (Expt. 1)1

Groups Control 8% PSF 5% LKM 5% MKM 5% HKM

Food intake, g/10 d 163 6 7cd 164 6 5d 143 6 5bc 135 6 4ab 123 6 2a

Body weight gain, g/10 d 47 6 2b 47 6 3b 43 6 2ab 36 6 2a 35 6 3a

Small intestinal mucin

O-linked oligosaccharide chains, mmol 1.4 6 0.2a 2.6 6 0.2b 1.4 6 0.2a 2.0 6 0.2ab 2.4 6 0.3b

ELISA-detected mucin,2 mg 1.0 (0.7–1.7) 2.0 (1.6–2.6)* 1.1 (0.03–1.5) 1.4 (1.0–1.8) 1.8 (1.2–6.2)*

Sialic acid, mmol 0.82 6 0.12a 1.57 6 0.17b 0.71 6 0.14a 1.10 6 0.11ab 1.33 6 0.13b

Villus height, mm

Jejunum 528 6 16 532 6 24 574 6 11 550 6 22 560 6 18

Ileum 394 6 7 398 6 8 394 6 13 427 6 18 432 6 19

Goblet cells (left side), n/villus

Jejunum 11.5 6 0.4a 17.5 6 0.8c 13.4 6 0.6ab 14.3 6 0.5b 14.5 6 1.0b

Ileum 11.1 6 0.5a 14.9 6 0.4b 11.0 6 0.4a 14.5 6 0.5b 14.4 6 0.6b

1 Data are means 6 SE, n = 8 unless otherwise noted. Means in a row with superscripts without a common letter differ, P , 0.05.2 Data are medians (range), n = 8. *Different from control, P , 0.05.

1642 Ito et al.


Villus height in the jejunum and ileum were comparable amongthe groups, but the number of goblet cells in the jejunum andileum differed among the groups (Fig. 1A,B). In the jejunum, thenumber of goblet cells was highest in the PSF, intermediate in theHKM and MKM, and lowest in the control and LKM groups,whereas in the ileum, the number of goblet cells was equallyhigher in the PSF, HKM, andMKM groups than in the LKM andcontrol groups. The small intestinal mucins (ELISA-detectedmucin or O-linked oligosaccharide chains) and goblet cellnumbers in the ileum (ELISA, r = 0.892, P = 0.042;O-linked, r =

0.920, P = 0.027) were correlated. These variables also tended tobe correlated in the jejunum (ELISA, r = 0.900, P = 0.037;O-linked, r = 0.862, P = 0.060).

Effects of concurrent ingestion of CL on mucin secretoryeffect of HKM (Expt. 2). Food intake (control, 144; HKM, 115;HKM + CL, 132; control + CL, 144; pooled SE, 5 g) and bodyweight gain (control, 52; HKM, 40; HKM + CL, 47; control +CL, 49 g; pooled SE, 3 g) were significantly lower in the HKMgroup than in the control group, but the other groups did not

FIGURE 1 Light micrographs of jejunum

(A) and ileum (B) from rats fed the control diet

or a diet either containing 80 g PSF/kg or 50 g

LKM, MKM, or HKM/kg for 10 d (Expt. 1).

Micrographs are stained with PAS reagent

and hematoxylin. Magnification = 2003.

FIGURE 2 ELISA-detected mucin (A) and O-linked oligosaccharide chains (B) in the small intestinal contents and number of ileal goblet cells (C)

in rats fed the control diet or a diet containing 50 g HKM/kg in the presence or absence of 2 g CL/kg for 10 d (Expt. 2). Values are means6 SE, n =

6. Means without a common letter differ, P , 0.05.



differ. Both the amounts of ELISA-detected mucin and O-linkedoligosaccharide chains in the small intestinal contents weregreater in the HKM group than in the control group, but theseeffects were nullified by the concurrent addition of CL (Fig. 2A,B). The number of ileal goblet cells tended to have the sameeffects (P , 0.05; Fig. 2C).

Various SDF, with different VAUC, affect small intestinalmucin secretion (Expt. 3). Food intake (control, 146; LKM,138; MKM, 124; HKM, 121; LGG, 144; HGG, 111; PC, 147;PS, 130 g; pooled SE, 2) and body weight gain (control, 49;LKM, 47;MKM, 35; HKM, 38; LGG, 47; HGG, 30; PC, 47; PS,44 g; pooled SE, 2) were significantly different among thegroups. In general, SDF with higher viscosities, such as HKM,MKM, and HGG, significantly reduced food intake and bodyweight gain compared with those in the control. The amount ofO-linked oligosaccharide chains in the small intestinal contentswas significantly greater in the HKM, PC, and PS groups than inthe control group. When a direct comparison was made byStudent’s t test, differences between the MKM and HGG groupscompared the control were also significant (Fig. 3A). Thenumber of goblet cells in the ileum was equally higher in theMKM, HKM, HGG, and PS groups than the others (Fig. 3B).This was also the case for the jejunum goblet cell numbers (datanot shown). The logarithm of VAUC and the number of gobletcells in the ileum were also significantly correlated. The numberof ileal goblet cells was also correlated with the amount ofO-linked oligosaccharide chains in the small intestinal contents.The correlation between the logarithm of VAUC and the amountof O-linked oligosaccharide chains in the small intestinalcontents was not significant (Fig. 4A–C).

Effects of HKM and PSF on Muc 2 and Muc 3 geneexpression (Expt. 4). Food intake (control, 155; HKM, 122;

PSF, 167 g; pooled SE, 5 g) and body weight gain (control, 57;HKM, 35; PSF, 48 g; pooled SE, 4) were significantly lower in theHKM group than in the control and PSF groups. The amount ofO-linked oligosaccharide chains in the small intestinal contents,as well as the number of goblet cells in the ileum, weresubstantial in the PSF and HKM groups. However, whereasMuc2 gene expression was comparable among the groups, Muc 3gene expression was significantly lower in the PSF and HKMgroups compared with the control (Fig. 5).

Effects of HKM and PSF on epithelial cell migration in thesmall intestine (Expt. 5). Lower food intake and body weightgain were repeatedly observed in the HKM group comparedwith the control and PSF groups (data not shown). Both in thejejunum and ileum, villus length and total epithelial cells/villuswere comparable among the groups, but the position of theuppermost BrdU-labeled cell from the bottom of the villus wassignificantly different among the groups (Fig. 6; Table 2). In thejejunum, these indices were higher in the HKM group than inthe others, whereas in the ileum, these were significantly higherin both the HKM and PSF groups than in the control groupand the difference between the HKM and PSF groups was alsosignificant.

Discussion

In accordance with our previous studies (6,7,19,23,24), anincrease in small intestinal mucin contents was repeatedlyobserved after PSF ingestion. In the same experimental condi-tion, when KM (LKM, MKM, or HKM) of different molecularweights was added to a fiber-free purified diet (control diet), allof the mucin markers (ELISA-detected mucin, O-linked oligo-saccharide chains, and sialic acid) in the small intestinal contentstended to increase in proportion to KM molecular weight.

FIGURE 3 O-linked oligosaccharide chains (A) in

the small intestinal contents and number of ileal

goblet cells (B) in rats fed the control diet or a diet

containing 50 g LKM, MKM, HKM, LGG, HGG, PS,

or PC/kg for 10 d (Expt. 3). Values are means 6 SE,

n = 12. Means without a common letter differ, P ,0.05.

FIGURE 4 Correlations among vis-

cosities, ileal goblet cells, and O-linked

oligosaccharide chains (Expt. 3).

1644 Ito et al.


Apparent increases in luminal mucin contents in rats fed MKM,HKM, and PSF diets were highly correlated with the concom-itant augmentation in goblet cell numbers in both jejunum andileum tissues. Further, both the mucin secretory effect and theincrease in goblet cells after HKM ingestion were completelynullified by the concurrent addition of 0.2% CL, which has ahigh specificity for splitting b-1, 4-glucosidic linkages of KM(25), thereby yielding lower molecular products. These findingsclearly show that KM must possess a higher molecule weight toincrease mucin secretory activity. Therefore, higher viscosityappears to be required for the increased secretory activity,because the viscosity of solutions or suspensions of certain typesof polysaccharides depends on mean molecular weight (26). Thepresent results regarding KM coincide with those by Larsen et al.

(11), indicating that the ileal flow of sialic acid increased withincreasing viscosity of the ingested carboxymethylcellulose.

To extend the findings obtained with KM, we preparedseveral SDF in which viscosities were defined as VAUCaccording to the method of Dikeman et al. (15). Calculation ofVAUC allowed not only for the interpretation of entire flowcurve characteristics but also for simplified data presentationsfor statistical comparisons (15,27). As expected, SDF withhigher VAUC (MKM, HKM, HGG, and PS) resulted in anincrease in the number of goblet cells, but not SDF with lowerVAUC (LKM, LGG, and PC), and there was a highly significantcorrelation between the VAUC and the number of goblet cells.The correlation between goblet cell number and luminal mucincontent was also significant, but the correlation between VAUCand luminal mucin content was not (P = 0.13). This is probablybecause PC ingestion caused an exceptionally significant in-crease in luminal mucin contents, irrespective of the number ofileal goblet cells. At present, the reason for the mucin secretoryeffect by PC ingestion remains unclear. However, this may bepartly explained by the enhanced mucin secretion in thestomach, because little mucin digestion occurs prior to the largeintestine (28) and an estimate of the mucins in the smallintestinal contents is necessary to account for the sum of bothgastric and small intestinal mucins. Another possibility issuggested by the study involving isolated ferret trachea, whichshowed sodium alginate secreted by Pseudomanas aeruginosawas a potent mucin secretagogue (29), implicating a possiblecontribution of uronic acid-polymers. Further, PC forms gelsrather than viscous solutions and this might be in part associatedwith the increased mucin secretion. In the present study, highlyviscous fibers such as HKM, MKM, and HGG significantlyreduced food intake and body weight gain. Nunez et al. (30)showed that dietary restriction caused a significant decrease inthe number of goblet cells in the small intestine of nursingpiglets. Therefore, it is not likely that food restriction per se maycause an increase in the number of goblet cells even in rats.However, to clarify this point, a pair feeding study is required forfurther experiment.

The increase in luminal mucin secretion can be partiallyexplained by the increase in terminally differentiated gobletcells, but studies in cell culture models also suggest thatdifferentiation factors can directly affect mucin gene expression(12). In the small intestine, the mainMuc genes areMuc2 and, toa lesser extent, Muc3. Muc2 codes for the main secreted mucin

FIGURE 5 O-linked oligosaccharide chains (A) in the small intestinal

contents, number of ileal goblet cells (B), and ileal mRNA expression

of Muc2 (C) and Muc3 (D) in rats fed the control diet or a diet either

containing 80 g PSF/kg or 50 g HKM/kg for 10 d (Expt. 4). Values are

means6 SE, n = 12. Means without a common letter differ, P, 0.05.

FIGURE 6 Light micrographs indicating incorpo-

ration of BrdU into small intestine epithelial cells in

rats fed the control diet (A) or a diet either containing

80 g PSF/kg (B) or 50 g HKM/kg (C) for 10 d (Expt.

5). Arrows indicate the uppermost migrated BrdU-

positive cells on the villi. Magnification = 2003.



and is confined to goblet cells, whereas Muc3 mainly codes formembrane-located mucins and is expressed by absorptive as wellas goblet cells (31). In the present study, we assessed theexpression of theseMuc genes in the ileal tissue of rats fed HKMand PSF diets compared with that of rats fed the control diet.The expression of Muc2 did not differ among the groups,whereas the expression of Muc3 was significantly decreased inthe HKM and PSF diet groups. The reason for the significantdecrease in Muc3 expression remains unclear. Chang et al. (32)showed that as the proliferating cells migrate in the crypt andalong the villi and differentiate, Muc3 expression is turned off,indicating a maturational gradient for Muc3 expression. Thepresent results in the BrdU incorporation study showed thatposition of the uppermost-BrdU labeled cell along the villi washigher in rats fed the HKMdiet than in those fed the control diet.This accelerated epithelial cell migration might be in partassociated with decreased Muc3 expression. Nevertheless, it isclear that the increase in luminal mucin contents by PSF andHKM ingestion occurred independently of enhanced Muc geneexpression. Taken together, it is reasonable to contend thatviscous fibers, except PC, upregulate baseline secretion of smallintestinal mucins by increasing the number of goblet cells inproportion to fiber viscosity. However, we should keep in mindthat although the correlations provide evidence that viscosity isrelated to mucin secretion and goblet cell numbers, a correlationis not necessarily a cause-and-effect relationship.

The underlying mechanism(s) responsible for the increase ingoblet cells after ingestion of viscous SDF and high bulk-formingIDF is unclear. However, any proposed mechanisms mustoriginate in some modification of the luminal environmentthrough the physico-chemical properties of the dietary fiberingested, because dietary fibers are considered to be non-digestible and nonabsorbable. Apparent intestinal viscositylargely reflects the viscosity of SDF determined in vitro (33).Increased intestinal viscosity is likely to assist the more bulkyand viscous digesta along the small intestinal tract and affect theresistance of digesta to peristaltic contractions (34), naturallyleading to an increase in “intra-luminal pressure.” One mightexpect that such an increase in the intraluminal pressure, or“differential stretching force” as Piel et al. (10) suggested, has aninfluence on stem cells or on immature crypt cells committedto be goblet cells through a Notch signal (35). Finally, we

hypothesize that viscous fibers enhance goblet cell differentia-tion in the small intestine, through increased intraluminalpressure. This might be also the case for high-bulk forming IDF.

In conclusion, the ingestion of viscous fibers, except PC,increased the number of small intestine goblet cells in proportionto their own viscosities and enhanced luminal mucin secretion.The increase in luminal mucin secretion after ingestion ofviscous SDF occurred independently of enhanced Muc geneexpression. Accordingly, it is reasonable to consider that viscousfibers upregulate baseline secretion of small intestinal mucins byincreasing the number of goblet cells.

Literature Cited

1. Forstner JF, Forstner GG. Gastrointestinal mucus. In: Johnson LR,editor. Physiology of the gastrointestinal tract. New York: Raven Press;1994. p. 1255–83.

2. Dryden PA, Jones GP, Burcher E, Read RSD. Effect of chronic ingestionof dietary fiber on the rate of glucose absorption in rats. Nutr Rep Int.1985;31:609–14.

3. Quarterman J. Metal absorption and intestinal mucus layer. Digestion.1987;37:1–9.

4. Trompette A, Claustre J, Caillon F, Jourdan G, Chayvialle JA, PlaisancieP. Milk bioactive peptides and b-casomorphins induce mucus release inrat jejunum. J Nutr. 2003;133:3499–503.

5. Satchithanandam S, Vargofcak-Apker M, Calvert RJ, Leeds AR,Cassidy MM. Alteration of gastrointestinal mucin by fiber feeding inrats. J Nutr. 1990;120:1179–84.

6. TanabeH, SugiyamaK,MatsudaT,KiriyamaS,MoritaT. Small intestinalmucins are secreted in proportion to the settling volume in water ofdietary indigestible components in rats. J Nutr. 2005;135:2431–7.

7. Morita T, Tanabe H, Ito H, Yuto S, Matsubara T, Matsuda T, SugiyamaK, Kiriyama S. Increased luminal mucin does not disturb glucose orovalbumin absorption in rats fed insoluble dietary fiber. J Nutr.2006;136:2486–91.

8. Satchithanandam S, Jahangeer S, Cassidy MM, Floor MK, Carvert RJ,Leeds AR, Alabaster O. Quantitative effects of wheat bran feeding onrat intestinal mucin. FASEB J. 1989;3:A1066.

9. Satchithanandam S, Klurfeld DM, Calvert RJ, Cassidy MM. Effects ofdietary fibers on gastrointestinal mucin in rats. Nutr Res. 1996;16:1163–77.

10. Piel C, Montagne L, Seve B, Lalles J-P. Increasing digesta viscosity usingcarboxymethylcellulose in weaned piglets stimulates goblet cell numbersand maturation. J Nutr. 2005;135:86–91.

11. Larsen FM, Moughan PJ, Wilson MN. Dietary fiber viscosity andendogenous protein excretion at the terminal ileum of growing rats.J Nutr. 1993;123:1898–904.

12. Theodoropoulos G, Carraway KL. Molecular signaling in the regulationof mucins. J Cell Biochem. 2007;102:1103–16.

13. Vahouny GV, Le T, Ifrim I, Satchithanandam S, Cassidy MM.Stimulation of intestinal cytokinetics and mucin turnover in rats fedwheat bran or cellulose. Am J Clin Nutr. 1985;41:895–900.

14. Prosky L, Asp NG, Schweizer TF, Devries JW, Furda I. Determination ofinsoluble, soluble, and total dietary fiber in food and food products:interlaboratory study. J Assoc Off Anal Chem. 1988;71:1017–23.

15. Dikeman CL, Murphy MR, Fahey GC Jr. Dietary fibers affect viscosityof solutions and simulated human gastric and small intestinal digesta.J Nutr. 2006;136:913–9.

16. Lien KA, McBurney MI, Beyde BI, Thomson ABR, Sauer WC. Ilealrecovery of nutrients and mucin in humans fed total enteral formulassupplemented with soy fiber. Am J Clin Nutr. 1996;63:584–95.

17. CrowtherRS,WetmoreRF. Fluorometric assay ofO-linked glycoproteinsby reaction with 2-cyanoacetamide. Anal Biochem. 1987;163:170–4.

18. Bovee-Oudenhoven IMJ, Termont DSML, Heidt PJ, Van der Meer R.Increasing the intestinal resistance of rats to the invasive pathogenSalmonella enteritidis: additive effects of dietary lactulose and calcium.Gut. 1997;40:497–504.

19. Tanabe H, Ito H, Sugiyama K, Kiriyama S, Morita T. Estimation ofluminal mucin content in rats by measurement of O-linked oligosac-charide chains and direct ELISA. Biosci Biotechnol Biochem. 2007;71:575–8.

TABLE 2 Food intake, body weight gain, and incorporation ofBrdU into epithelial cells of the small intestine inrats fed the control diet or a diet either containing80 g PSF/kg or 50 g HKM/kg for 10 d (Expt. 5)1

Groups Control 8% PSF 5% LKM

Food intake, g/10 d 155 6 4b 168 6 5b 122 6 4a

Body weight gain, g/10 d 57 6 2 c 48 6 1b 35 6 2a

Jejunum

Villus length, mm 408 6 19 424 6 14 410 6 22

Total epithelial cells (left side), n/villus 92.8 6 3.1 108.4 6 10.5 98.8 6 3.2

Position of uppermost-BrdU labeled cell2 32.0 6 0.9a 37.5 6 1.3a 46.7 6 2.4b

Ileum

Villus length, mm 314 6 16 311 6 16 307 6 15

Total epithelial cells (left side), n/villus 72.8 6 2.2 68.0 6 3.2 78.3 6 3.5

Position of uppermost-BrdU labeled cell 19.9 6 1.3a 26.9 6 1.7b 35.5 6 2.3c

1 Data are means 6 SE, n = 8. Means in a row with superscripts without a common

letter differ, P , 0.05.2 Values indicate the highest position of BrdU-labeled cell from the bottom of the villus

at 24h after BrdU injection.

1646 Ito et al.


20. Tsuboi Y, Kim Y, Paparella MM, Chen N, Schachern PA, Lin J. Patternchanges of mucin gene expression with pneumococcal otitis media. Int JPediatr Otorhinolaryngol. 2001;61:23–30.

21. Kesavalu L, Bakthavatchalu V, Rahman MM, Su J, Raghu B, DawsonD, Fernandes G, Ebersole JL. Omega-3 fatty acid regulates inflamma-tory cytokine/mediator messenger RNA expression in Porphyromonasgingivalis-induced experimental periodontal disease. Oral MicrobiolImmunol. 2007;22:232–9.

22. Heid CA, Stevens J, Livak KJ, Williams PM. Real time quantitativePCR. Genome Res. 1996;6:986–94.

23. Tanabe H, Ito H, Sugiyama K, Kiriyama S, Morita T. Dietaryindigestible components exert different regional effects on luminalmucin secretion through their bulk-forming property and fermentabil-ity. Biosci Biotechnol Biochem. 2006;70:1188–94.

24. Morita T, Tanabe H, Ito H, Sugiyama K, Kiriyama S. Long-termingestion of insoluble dietary fiber increases luminal mucin content, buthas no effect on nutrient absorption in rats. Biosci Biotechnol Biochem.2008;72:767–72.

25. Kato K, Watanabe T, Matsuda K. Studies on the chemical structure ofkonjac mannan. Part 2. Isolation and characterization of oligosaccha-rides from the enzymatic hydrolysate of the mannan. Agr Biol Chem.1970;34:532–9.

26. Morris ER. Assembly and rheology of non-starch polysaccharides. In:McClearly BV, Prosky L, editors. Advanced dietary fibre technology.London: Blackwell Science; 2001. p. 30–41.

27. Dikeman CL, Fahey GC Jr. Viscosity as related to dietary fiber: a review.Crit Rev Food Sci Nutr. 2006;46:649–63.

28. Hoskins LC. Mucin degradation by enteric bacteria: ecological aspectsand implications for bacterial attachment to the gut mucosa. In:Boedeker EC, editor. Attachment of organisms to the gut mucosa. 2ndvol. Florida: CRC Press; 1984. p. 51–67.

29. Kishioka C, Okamoto K, Hassett DJ, de Mello D, Rubin BK.Pseudomonas aeruginosa alginate is a potent secretagogue in theisolated ferret trachea. Pediatr Pulmonol. 1999;27:174–9.

30. Nunez MC, Bueno JD, Ayudarte MV, Almendros A, Rios A, Suarte MD,Gil A.Dietary restriction induces biochemical andmorphometric changesin the small intestine of nursing piglets. J Nutr. 1996;126:933–44.

31. Van Klinken BJ, Dekker J, Buller HA, Einerhand AWC. Mucin genestructure and expression: protection vs. adhesion. Am J Physiol.1995;269:G613–27.

32. Chang SK, Dohrman AF, Basbaum CB, Ho SB, Tsuda T, Toribara NW,Gum JR, Kim YS. Localization of mucin (MUC2 and MUC3) messengerRNA and peptide expression in human normal intestine and coloncancer. Gastroenterology. 1994;107:28–36.

33. McDonald DE, Pethick DW, Mullan BP, Hampson DJ. Increasingviscosity of the intestinal contents alters small intestinal structure andintestinal growth, and stimulates proliferation of enterotoxigenicEscherichia coli in newly-weaned pigs. Br J Nutr. 2001;86:487–98.

34. Cherbut C, Albina E, Champ M, Coublier JL, Lecannu G. Action ofguar gums on the viscosity of digestive contents and on the gastroin-testinal motor function in pigs. Digestion. 1990;46:205–13.

35. Schonhoff SE, Giel-Moloney M, Leiter AB. Minireview: developmentand differentiation of gut endocrine cells. Endocrinology. 2004;145:2639–44.



soluble fiber viscosity affects both ... -...

Documents