| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ANIMAL NUTRITION |
Department of Animal Health, Welfare and Nutrition, Danish Institute of Agricultural Sciences, Research Centre Foulum, DK-8830 Tjele, Denmark
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
Key Words: Morphology • Mucin • Particle Size • Pig • Salmonella
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
). Furthermore, feeding pellets increased the risk of Salmonella infections (Jørgensen et al., 1999). Recent investigations have shown that in pigs fed a coarsely ground meal feed, the stomach acts as a barrier that decreases the occurrence of pathogenic bacteria (Mikkelsen et al., 2004). It is, however, unknown whether feeding coarsely ground feeds changes the gut epithelium in a way that decreases the binding ability of Salmonella.
Salmonella bind to ileal tissue (Ewen et al., 1997; Naughton et al., 2001). There is considerable evidence that bacterial adherence is mediated by glycoconjugates, lectin-like substances, on the surface of bacteria. The binding characteristics of Salmonella differ among different Salmonella serotypes. Mannose and galactosyl residues were suggested to be receptors for Salmonella choleraesuis and Salmonella typhimurium (Giannasca et al., 1996; Meng et al., 1998).
Feeding a coarsely ground diet affects the mucosal architecture, epithelial cell proliferation, production and composition of the mucins, and the lectin binding pattern in the large intestine of pigs (Brunsgaard, 1998), but the effect of particle size on the small intestinal morphology remains to be elucidated.
The purpose of the present study was to investigate whether feeding diets either finely or coarsely ground with or without pelleting affects the mucosal architecture and mucin characteristics in the pig intestine and the adhesion of Salmonella in the ileum in vitro.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
Animals and Feed
Ninety-six crossbreed Danish Landrace x Yorkshire pigs (average weight 33 ± 7 kg) were obtained from the herd at the Research Centre Foulum, Denmark, and assigned to one of four treatment groups in a 2 x 2 factorial arrangement of treatments in a randomized complete block design. The pigs were allocated to the treatment groups with an equal distribution for litter and sex. The experiment was carried out in six replicates of four pigs per group housed together in a pen with no physical or visual contact between animals in different pens. The pigs were allowed ad libitum access to feed and water. The composition of the experimental diets is shown in Table 1. The feed was ground at two levels to obtain a fine or coarse feed. Grinding was accomplished using a hammer mill with a 2-mm screen for the fine diets, whereas a 5-mm screen was used for the coarse diets. Half of each batch of feed was pelleted. In the pellet press, the holes of the matrix had a diameter of 3 mm. The feed was steam-conditioned and reached a temperature of approximately 83°C during pelleting.
|
) using a wet-sieve method. A volume of 400 mL of deionized water, 20°C, was added to 100 g of feed on an as-fed basis. The solution was kept for 1 h at room temperature without stirring. The suspension was then sieved for 2 min, at a 3-mm amplitude, using a Retsch AS200 control "g" sieve shaker (Retsch GmbH and Co., Haan, Germany). The sieves (200 i.d. x 50 mm DIN ISO 3310/1) were made of stainless steel, and the following mesh sizes (mm) were used: 3,150; 2,000; 1,000; 500; and 250. The different particle fractions were then dried to constant weight at 100°C and weighed.
|
Collection of Samples
After 28 d on the experimental diets, one pig per replicate pen was killed (n = 6) with a lethal injection of pentobarbital sodium (40 mg/kg of BW). The pigs selected were littermates; hence, the effect of litter and replicate was the same. The gastrointestinal tract was immediately removed. The length of the small intestine and the colon was measured and tissue samples for morphological measurements were taken at 50 and 90% of the length of the small intestine (SI50 and SI90), in the cecum and at 75% of the length of the colon. The samples were immediately transferred to 10% (vol/vol) neutral buffered formaldehyde. A piece of the ileum (11 cm) was taken 30 cm from the ileocecal valve for the pig intestine organ culture model. The sample was immersed in Dulbecco’s modified Eagle’s medium (DMEM; Invitrogen, Carlsbad, CA), and kept on ice. After the tissue samples were taken, the gastrointestinal tract was emptied of luminal contents and the small intestine, cecum, and colon were weighed.
Samples for Microscopy
After 24 h in the 10% neutral buffered formaldehyde, the tissue samples were carefully cleaned of remaining digesta with deionized water and then transferred to a fresh solution of 10% neutral buffered formaldehyde. Subsequently, the samples were dehydrated and infiltrated with paraffin wax. Three slides were prepared from each sample, and each slide contained a minimum of four sections cut at 4 µm, at least 50 µm apart. The slides were processed for carbohydrate histochemistry using either the periodic acid-Schiff (PAS) reaction or the Alcian blue reaction at either pH 2.5 or pH 1.0 (Kiernan, 1990). The PAS reaction stains for neutral mucins, the Alcian blue pH 2.5 stains for carboxylated or sulfated types of acidic mucins, and the Alcian blue pH 1.0 stains for sulfomucins (Kiernan, 1990).
Carbohydrate histochemistry on the PAS- and Alcian blue-stained samples was evaluated as described previously (Brunsgaard, 1997). Briefly, 15 well-oriented villi and crypts were selected on each slide, and for each villi and crypt, the area of mucin granules with a clear positive reaction for either neutral mucins, acidic mucins, and sulfomucins was determined with a computer-integrated microscope and an image analysis system (Quantimet 500MC, Leica, Cambridge, U.K.) with a monitor. This area included the mucus material present in the crypt lumen. As the histochemical procedure used in our study stains the granules of all mucous cells (goblet cells and crypt secretory cells), as well as the apical secretion of these cells, these are all included in the measures.
The slides processed for neutral mucins were further used to determine the area and the height of the villi and the crypts, and the thickness of the muscularis externa using the image analyses system. The villi area was determined on 15 well-oriented villi. The height was determined on the same villi as the distance from the villi tip to the bottom of the villi. The crypt area was determined on 15 well-oriented crypts as the area encircled by the basement membrane and the crypt mouth including the crypt lumen. The height was determined on the same crypts as the distance from the crypt base at the basement membrane to the crypt mouth. All measures were made with a light microscope at 10x magnification.
Lectin Histochemistry
The samples from the distal small intestine and the cecum were processed for lectin histochemistry. Sections were deparaffinized in xylene and then hydrated through a series of alcohols to straight distilled water. Endogenous peroxidase activity was blocked by incubation in 0.3% (vol/vol) hydrogen peroxide in methanol. Trypsinization was carried out in 0.1% (wt/vol) trypsin at 37°C for 30 min. The sections were then incubated with the specific biotinylated lectin (Table 3) for 60 min at room temperature. Lectin binding was detected by use of the Vectastain ABC kit (Vector Laboratories, Burlingame, CA) using diaminobenzidine as the peroxidase substrate. Slides were counterstained with Mayer’s hematoxylin, dehydrated, and covered with a cover slip.
|
Pig Intestine Organ Culture Model
The pig intestine organ culture model study was performed as described in detail by Naughton et al. (2001). Briefly, the tissue was washed with phosphate-buffered saline (PBS; pH 7.2) to remove the luminal contents. Then, 10 mL of DMEM containing S. enterica serovar Typhimurium DT12 (5 x 108 cfu/mL) was added to the segment. The segment was sealed and the organ culture was immersed in DMEM in a 300-mL infusion bottle in a shaking water bath at 37°C in a 10% CO2 atmosphere. After 60 min, the tissue was removed from the infusion bottle and the tissue was washed with PBS. The tissue was homogenized using a Janke-Kunkel (Staufen, Germany) Ultra Turrax T25 homogenizer in PBS plus Triton X-100 (1% vol/vol). A 10-fold dilution series was prepared from the homogenate to a final dilution of 10–6. The enumeration of Salmonellae was performed on Brilliant phenol lysine sucrose agar (Merck 1.10747, Damstadt, Germany) after aerobic incubation at 38°C for 16 h.
Statistical Analyses
The statistical analyses were performed using the Mixed procedure of SAS (SAS Inst., Inc., Cary, NC). All statistics on epithelial morphology were done using the sample means generated from the 15 individual measurements. Data were analyzed using the following model:
 |
where Yfgh is the dependent variable, µ is the overall mean, 
f is the effect of form (nonpelleted or pelleted), ßg is the effect of grinding (fine or coarse), ßfg is the interaction between form and grinding, Uh is the random effect of litter, and 
fgh is the error term. Data on epithelial morphology were analyzed by segment.
Results are expressed as least squares means and SEM. Treatment differences were determined with orthogonal contrasts for a 2 x 2 factorial arrangement of treatments. Treatment differences were considered significant at = 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
The relative stomach weight was higher in pigs fed the coarse diets, whereas the relative weight of the small intestine was lower in these pigs (Table 4). The effect of grinding and form of the feed (nonpelleted vs. pelleted) on the relative weight of the cecum was interactive. The cecum of pigs fed the fine nonpelleted diet weighed 1.78 g/kg of BW, which was less than the cecum of the pigs fed the other diets (2.19 to 2.43 g/kg of BW). A significant interaction between form and grinding also was observed for colon weight. Pigs fed C-NP and F-P had the highest relative colon weight (13.3 and 13.0 g/kg of BW, respectively) compared with pigs fed F-NP and C-P (12.1 and 12.0 g/kg of BW, respectively). Pigs fed the fine diets had longer small intestines relative to the BW than pigs fed the coarse diets. The form of the diet affected relative colon length; pigs fed nonpelleted diets had longer colons than pigs fed the pelleted diets.
|
). Pigs fed the coarse nonpelleted diet had longer villi (527 µm) than pigs fed the coarse pelleted diet (442 µm), whereas the villus height of the pigs fed the fine diets did not differ from the others. The villus height remained unaffected by diet at SI90. The villus area at SI50 tended (P = 0.07) to be greater in pigs fed the nonpelleted diet, which can be related to the increased villus height in these pigs. No difference in villus area was observed at SI90. Feeding pigs a coarse diet tended (P = 0.10) to increase crypt depth at SI50, and feeding the coarse diet increased the crypt depth in the colon. In the distal small intestine and the cecum, no effect of the experimental diets was observed on the crypt depth. The differences in crypt depths did not translate into differences in the crypt areas, as they remained unaffected by the diets. The particle size and feed processing did not induce any changes in the thickness of muscularis externa at any of the positions examined.
|
); the staining area was larger for the C-P diet than for the C-NP and F-P diets, whereas the F-NP diet did not differ from the others. At SI90, the area of neutral mucins on the villi was larger in pigs fed pelleted diets. Grinding affected the area of neutral mucins in the crypts at SI50. Pigs fed coarse diets had a greater area than pigs fed fine diets. In the distal small intestine, cecum, and colon, no effects of the experimental diets on the area of neutral mucins in the crypts were observed. The effect of form and grinding was interactive with respect to the area of the acidic mucins on the villi at SI50. The area was greatest in pigs fed C-P and least in pigs fed C-NP or F-P. In the distal small intestine, the area of acidic mucins on the villi was greatest in pigs fed the pelleted diets. In the crypts the area of the acidic mucins was only affected by the experimental diets at SI50, where pigs fed coarse diets had the greatest area. At the other positions investigated, the area of acidic mucins in the crypts remained unaffected by the experimental diets. For the area of sulfomucins on the villi at SI50, the interaction between form and grinding approached significance (P = 0.06), and the greatest area was observed in pigs fed C-P. At SI90, the area of sulfomucins on the villi was affected both by form and grinding; fine diets resulted in a greater area than the coarse diets, and pigs fed pelleted diets had a greater area than pigs fed meal. The area of sulfomucins in the crypts was only affected by experimental diets in the cecum, where the effect of form and grinding was interactive. Pigs fed C-NP had the largest area (7.1 x 103 µm2), and pigs fed C-P had the smallest (5.8 x 103 µm2).
|
). Furthermore, a total lack of response (lectin score = 1) was observed in many of the observed structures for the lectins Canavalia ensiformis (Con A), Galanthus nivalis (GNA), and Maackia amurensis (MAA).
|
Reactivity with the lectin GNA was only observed in the cytoplasm. In the distal small intestine, a high score was observed on the villi, but no staining was observed in the crypts. In the cecum, a difference between the experimental diets was observed. In the group fed the nonpelleted coarse diet, five of six pigs showed GNA reactivity compared with the other groups, where only one or two pigs showed GNA reactivity.
The mucous cells in the distal small intestine showed MAA reactivity, but no reactivity was observed in the cecum. The apical surface of the crypts in the cecum had moderate reactivity toward MAA, whereas the crypts in the distal small intestine showed no MAA reactivity; however, the apical membrane of the villi had a low MAA reactivity. The reactivity toward Ulex europaeus (UEA-I) was similar in the distal small intestine and the cecum.
Association of Salmonella in the Ileum In Vitro
In ileal tissues challenged with Salmonella 7.66 ± 0.21 and 7.73 ± 0.21 cfu/g of wet tissue was recovered from pigs fed F-NP and C-NP, respectively, whereas 8.05 ± 0.21 and 8.12 ± 0.21 cfu/g of intestinal tissue was recovered from pigs fed F-P and C-P, respectively. The overall effect of form (nonpelleted vs. pelleted) was significant (P < 0.05). Feeding nonpelleted diets resulted in a 60% decrease in the adherence of Salmonella to the ileal tissue. No effect of grinding of the feed was observed.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
; Eisemann and Argenzio, 1999; Jørgensen et al., 1999). Particle size and pelleting can affect the gastric health of slaughter pigs (Eisemann and Argenzio, 1999; Nielsen and Ingvartsen, 2000). These studies have shown, in agreement with the present investigation, that pigs fed a coarse diet have heavier stomachs than pigs fed a fine diet, which probably reflects that coarse diets require more muscular action for processing by the stomach than fine diets.
The content of dietary fiber did not differ between the diets in the present study; however, in the coarse diets, more starch probably was rendered inaccessible to enzymatic digestion than in the fine diets. More accessible starch may explain why pigs fed the finely ground diets had heavier and longer small intestines than pigs fed the coarsely ground diets. Colon weight did not differ between the experimental groups, which agrees with the findings of Brunsgaard (1998), who reported no effect of feeding wheat or barley either finely or coarsely ground on the weight of hindgut tissue.
An increased crypt depth in the colon was observed in pigs fed coarse diets, which agrees with previous studies (Brunsgaard, 1998). Increased crypt depth indicates that the proliferative activity in the crypt is augmented. Growth of epithelial cells is promoted by short-chain fatty acids that are the products of bacterial degradation of unabsorbed starch and fiber (Scheppach, 1994). Butyrate has been shown to be the preferred substrate for the colonocyte, and higher concentrations of butyric acid were observed in the cecum and the colon of pigs fed the coarsely ground feed compared with pigs fed the finely ground feed (Mikkelsen et al., 2004).
A continuous mucus gel that varies in thickness covers the epithelial lining from the stomach to the large intestine. The mucus gel is composed predominantly of mucin glycoproteins secreted by goblet cells. Goblet cells differentiate and mature as they migrate up the villus. The process is affected by a number of factors, such as the age of the animal (Bruininx et al., 2002), diet (More et al., 1987), and composition of the microflora (Sharma and Schumacher, 2000). The greater mucin-staining area on the villi in the distal small intestine in pigs fed the pelleted diets indicates a greater production and secretion of mucus in this region in these pigs compared with pigs fed nonpelleted diets. In chickens, a decrease in digesta viscosity has been shown to increase the amount of mucins in the jejunum (Sharma et al., 1997). In addition, recent studies have shown that piglets fed diets containing pectin had a smaller mucin-staining area than piglets fed barley hulls (Hedemann, unpublished). Fiber supplementation has been shown to increase mucin secretion in hamsters (Lundin et al., 1993), and the same observation has been made in pigs (Lien et al., 2001). The diets used in the present study create digesta with differing physicochemical properties, especially in the stomach (Mikkelsen et al., 2004), but how these differences extend to the distal small intestine is unknown. In rats, the villus region accounts for 55% of the total ileal mucin store (Phillips, 1992). But in the present study, the mucin store of the villi accounted for only approximately 16% of the total mucin store in the mid and distal small intestine (results not shown), and the relative contribution of the goblet cells on the villi to the mucus layer is unknown.
The mucin-staining area in the cecum and the colon was not affected by the experimental diets in the present study. Brunsgaard (1998) observed that feeding coarse wheat or barley diets induced a larger mucin-staining area. The difference in particle size distribution in the present study may have been too small to elicit changes in mucin production and secretion.
The lectin staining characteristics of the goblet cells on the villi did not indicate any differences between the experimental groups. The lectins used in the present study were chosen according to their relevance to binding of Salmonella and other bacteria to the epithelium. Mannose has been identified as one of the carbohydrates recognized by the Type 1 fimbriae on Salmonella (Baba et al., 1993). Terminal mannose was detected with Con A and GNA in the present investigation. The results showed that although both lectins detect mannose, they have different staining characteristics. A high lectin score was observed for Con A on the apical membrane of the villi, whereas no reactivity was observed with GNA. This finding illustrates the high specificity of GNA in contrast to Con A, which recognizes mannose, glucose, and N-acetyl-glucosamine, whereas GNA only binds to -1-3 D-mannose. In the crypts in the cecum, a higher proportion of the pigs fed the coarse nonpelleted diet showed GNA reactivity in the cytoplasm compared with the other groups. This result may indicate augmented cellular turnover in this group, which increases the proportion of less differentiated, immature epithelial cells with the consequent increase in the concentration of terminally mannosylated cytoplasmic glycoconjugates (Pusztai et al., 1995). This group had a numerically larger crypt depth (Table 5
) and was influenced by a higher concentration of butyrate in the cecum and the colon (Mikkelsen et al., 2004), which may indicate that the cellular turnover was indeed increased.
The sialic acid, -2,3 neuraminic acid, which has been shown to inhibit bacteria from binding to epithelial cells, is recognized by MAA (Simon et al., 1997). Increased MAA reactivity has been observed in the colon in pigs fed a coarse barley diet (Brunsgaard, 1998), whereas the reactivity did not differ in the cecum, which agrees with the findings of the present study. In guinea pigs, aggregation of Salmonella typhimurium was abolished by fucose and the lectin UEA-I, specific for fucose, inhibited bacterial binding (Ensgraber et al., 1992). In contrast, Salmonella pullorum exhibited no binding to UEA, and no lectin staining of the ileal epithelium of chicks was observed with UEA (Zhou et al., 1995). Pigs exhibit UEA-I reactivity of goblet cells, apical membrane, and cytoplasm, but no effect of the experimental diets was observed in the current investigation. Thus, the role of fucose as inhibitor of binding of Salmonella to the epithelium seems to depend on both animal species and the strain of Salmonella.
Regional differences are observed in lectin binding reactivity (Brunsgaard, 1998). These differences could contribute to differences in the location of infections along the intestinal tract of pigs (Siba et al., 1996). In the present study, differences between the distal small intestine and the cecum were observed, as were differences between the lectin reactivity of the villi and the crypts in the small intestine. Several factors influence the glycosylation pattern in the intestine (e.g., site in the intestine, position along the crypt-villus axis, state of differentiation and maturation, diet and bacterial status; Pusztai and Bardocz, 1996).
Salmonella binds predominantly to the distal small intestine of pigs (Naughton et al., 2001). In the present study, pelleting of the diets more than doubled the association of Salmonella in the culture model. One of the earliest events in Salmonella typhimurium pathogenesis seems to be the interaction of the bacterium with the mucus of the gut (Ensgraber et al., 1992). Pigs fed pelleted diets had a larger mucin-staining area on the villi in the distal small intestine. The combination of these results suggests that pigs fed pelleted diets secrete mucins that favor the binding of Salmonella. Thus, pigs fed a nonpelleted diet may be better protected against Salmonella infections than pigs fed a pelleted diet.
|
Implications |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
|
Footnotes |
|---|
3 Current address: School of Rural Sci. and Agric., Univ. of New England, Armidale, Australia.
4 Current address: School of Biomed. Sci., Univ. of Ulster, Cromore Rd, Coleraine, U.K.
2 Correspondence; P.O. Box 50 (phone: +45 89 99 11 18; fax: +45 89 99 13 78; e-mail:mette.hedemann{at}agrsci.dk).
Received for publication September 6, 2004. Accepted for publication April 5, 2005.
|
Literature Cited |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. S. Hedemann, B. B. Jensen, and H. D. Poulsen Influence of dietary zinc and copper on digestive enzyme activity and intestinal morphology in weaned pigs J Anim Sci, December 1, 2006; 84(12): 3310 - 3320. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Lesniewska, I. Rowland, P. D. Cani, A. M. Neyrinck, N. M. Delzenne, and P. J. Naughton Effect on Components of the Intestinal Microflora and Plasma Neuropeptide Levels of Feeding Lactobacillus delbrueckii, Bifidobacterium lactis, and Inulin to Adult and Elderly Rats. Appl. Envir. Microbiol., October 1, 2006; 72(10): 6533 - 6538. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. S. Hedemann, M. Eskildsen, H. N. Laerke, C. Pedersen, J. E. Lindberg, P. Laurinen, and K. E. B. Knudsen Intestinal morphology and enzymatic activity in newly weaned pigs fed contrasting fiber concentrations and fiber properties J Anim Sci, June 1, 2006; 84(6): 1375 - 1386. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V Lesniewska, I Rowland, H. N Laerke, G Grant, and P. J Naughton Relationship between dietary-induced changes in intestinal commensal microflora and duodenojejunal myoelectric activity monitored by radiotelemetry in the rat in vivo Exp Physiol, January 1, 2006; 91(1): 229 - 237. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
