The Journal of Nutrition Vol. 128 No. 10 October 1998, pp. 1737-1744

Confirmation of the Role of Rapidly Fermentable Carbohydrates in the Expression of Swine Dysentery in Pigs after Experimental Infection^1,2,3

John R. Pluske⁴, Zorica Durmic, David W. Pethick, Bruce P. Mullan, and David J. Hampson⁵

Division of Veterinary and Biomedical Sciences, Murdoch University, Murdoch, WA 6150, Australia and Animal Research and Development Services, Agriculture Western Australia, WA 6983, Australia

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Two experiments were conducted to test the hypothesis that soluble non-starch polysaccharides (NSP) and resistant starch (RS) cause swine dysentery (SD) in pigs experimentally infected with the spirochete Serpulina hyodysenteriae. In Experiment 1, a source of soluble NSP (guar gum; GG), insoluble NSP (oat chaff; OC), resistant starch (retrograde cornstarch; RS) or a combination of GG and RS (GG + RS) was added to a diet containing cooked white rice (R), soybean meal (SBM) and animal protein (meat and bone meal, bloodmeal, fishmeal). A diet containing only cooked white rice, SBM and the sources of animal protein (AP) was also fed. In Experiment 2, three rice-based diets containing different levels of RS were fed to pigs. In Experiment 1, the pH of digesta in the cecum, proximal colon and distal colon of pigs fed diets R-GG, R-RS and R-GG + RS was lower (P < 0.001), and volatile fatty acid concentration higher (P < 0.001), than in pigs fed diets R-OC and R-AP. Pigs fed diets with RS and GG + RS had greater (P < 0.05) concentrations of ATP in the large intestine than pigs fed other diets. There were no significant differences in any fermentation indices measured in Experiment 2. In Experiment 1, pigs fed diets R-GG, R-RS and R-GG + RS were colonized with S. hyodysenteriae after experimental infection. However, only pigs consuming diets R-GG (4 of 5) and R-GG + RS (5 of 5) showed clinical signs of SD. Spirochetes were isolated from the feces of all pigs fed diets containing RS in Experiment 2. However, and in contrast to Experiment 1, 80-100% of pigs infected with S. hyodysenteriae displayed clinical signs of SD. These data confirm the role of fermentable carbohydrate in the pathogenesis of SD.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Swine dysentery (SD),⁶ a mucohemorrhagic colitis caused by the anerobic spirochetal bacterium Serpulina hyodysenteriae, is one of the most economically important diseases in pigs throughout the world (Harris and Lysons 1992). The precise pathogenesis of SD is not known, although it is clear that the condition does not always develop clinically in pig herds despite serological evidence that S. hyodysenteriae is present (Hampson et al. 1992, Mhoma et al. 1992). Antimicrobial agents are included in many pig diets to control SD at extra cost to the producer. However, mounting concerns over the emergence of antibiotic-resistant strains of bacteria, the transfer of drug resistance between bacterial strains and the presence of possible drug residues make it clear that alternative approaches to the control of enteric diseases are desirable.

Our previous experiments have established that the clinical expression of SD in pigs after experimental infection with S. hyodysenteriae was reduced by feeding diets low in soluble non-starch polysaccharides (NSP), oligosaccharides and (or) resistant starch (RS) (Pluske et al. 1996, Siba et al. 1996). Although these data provided strong correlative evidence that soluble NSP and RS were implicated in the clinical expression of SD, the precise contribution that these fractions may have played in the pathogenesis of the disease was not tested directly.

In the two experiments described in this paper, sources of soluble NSP, insoluble NSP, RS and a combination of soluble NSP plus RS were added to a diet based on cooked white rice that has been shown previously to prevent the clinical expression of SD in pigs (Pluske et al. 1996, Siba et al. 1996). By replacing a portion of the rice with these carbohydrate sources, we were able to examine directly their role in the pathogenesis of SD. The general hypothesis for the two studies was that diets containing a source of soluble NSP and RS would facilitate the occurrence of SD in experimentally infected pigs.

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

This research was reviewed and approved by the Animal Experimentation and Ethics Committee of Murdoch University in accordance with the NH & MRC/CSIRO/AAC Code of Practice for the Care and Use of Animals for Experimental Purposes.

Experiment 1

Animals and housing.  Fifty pigs (Large White × Landrace) of mixed sex, weaned between 26 and 28 d of age and weighing 8.2 ± 0.24 kg (mean ± SEM), were obtained from a commercial specific-pathogen-free piggery known to be free of SD (Wandalup Farms, Mandurah, WA). Pigs were allocated on the basis of live weight and sex and housed in groups of 10 in pens at the Medina Research Centre. Pigs were kept in a room maintained between 20 and 26°C and had free access to water at all times. Feed was provided at a level of approximately three-time maintenance (3M), according to the equation of Close and Fowler (1985). The average pig weight for each pen was used in the determination of maintenance requirement for digestible energy.

Experimental design and diets.  The experiment was a completely randomized design with five dietary treatments (see below). When pigs reached an average live weight of 34.5 ± 1.79 kg (mean ± SEM), 25 pigs (n = 5/treatment) were randomly selected and killed to examine the effects of the five diets on indices of microbial fermentation in the large intestine. The remaining five pigs in each treatment were then transported to Murdoch University and kept in their same groups. Pigs were again given free access to water and were fed at ~3M. At an average weight of 38.7 ± 0.85 kg (mean ± SEM), pigs were infected with a culture of S. hyodysenteriae (see below). The incidence of SD was noted for the following 28 d.

The five experimental diets are shown in Table 1. Guar gum (GG: NP 3500 Guar Gum, Germantown Australia Company, Botany, NSW, Australia), resistant starch (RS: Novelose, National Starch and Chemical Company, Bridgewater, NJ), a commercially available oat chaff (OC) and a combination of guar gum and resistant starch (GG + RS) were each added to a diet based on cooked white rice (R: Australian Ricegrowers' Cooperative, Leeton, NSW, Australia), soybean meal (SBM), and sources of animal protein (meat and bone meal, bloodmeal, fishmeal). Oat chaff was included as a source of (largely) insoluble NSP. The rice was cooked by adding water at a ratio of 2:1 (v/v) and then placing the mixture in an autoclave for 30 min at 121°C. The final diet used consisted of cooked white rice, SBM and the animal protein (AP) supplement (meat and bone meal, bloodmeal, fishmeal) and has been shown previously to be protective against SD (Pluske et al. 1996, Siba et al. 1996). The gross energy content of the Novelose was 19.5 MJ/kg dry matter. The amount of RS supplied in diets R-RS and R-GG + RS was 8.7 and 4.4 g/100 g, respectively, as measured by the in vitro method of Pluske et al. (1996). Antibiotics were not included in any of the diets.

Experiment 2

Animals and housing.  Thirty pigs (Large White × Landrace) of mixed sex, weaned between 21 and 25 d of age and weighing 7.0 ± 0.16 kg (mean ± SEM), were obtained from a commercial specific-pathogen-free piggery known to be free of SD (Wandalup Farms). Pigs were allocated on the basis of live weight and sex and housed in groups of 10 in pens at the Medina Research Centre. Pigs were kept in a room maintained between 20 and 26°C and had free access to water at all times. Pigs were fed at ~3M (as described previously).

Experimental design and diets.  In Experiment 1, pigs fed diet R + RS showed signs of epithelial colonization by S. hyodysenteriae (positive culture) in the large intestine, but failed to develop clinical signs of SD. Experiment 2 examined further the role of RS in diets on the pathogenesis of SD in pigs by feeding different levels of RS. The experiment was a completely randomized design with three dietary treatments (see below). Increasing concentrations of Novelose (13.2, 20 and 27 g/100 g for diets R-RSa, R-RSb, and R-RSc, respectively) were substituted for cooked white rice and fed to pigs from weaning (Table 2). This supplied 5.1, 8.7 and 10.4 g/100 g RS for diets R-RSa, R-RSb, and R-RSc, respectively (as measured by the in vitro method of Pluske et al. 1996). Antibiotics were not included in any of the diets.

When pigs reached an average live weight of 20.3 ± 0.59 kg (mean ± SEM), half of the pigs from each treatment were randomly selected and killed to examine the effects of these diets on indices of microbial fermentation in the large intestine. The remaining five pigs in each treatment were transported to Murdoch University and kept in their same groups. Pigs were again given free access to water and fed at ~ 3M. At an average weight of 23.2 ± 0.55 kg (mean ± SEM), pigs were infected with a culture of S. hyodysenteriae (see below). The occurrence of SD over the following 28 d was noted.

General

Infection.  Pigs brought to Murdoch University in both experiments were allowed to acclimate in their pens for 7 d. Strain 155/23 (serogroup A) of S. hyodysenteriae, which was isolated from a field case of SD in a Western Australian commercial piggery, was used to infect each pig. To mimic commercial conditions in which pigs are housed in groups and transmission of the disease occurs naturally via the fecal-oral route, pigs were kept together in their respective dietary treatments for the duration of the study.

The bacterium was propagated in a prereduced anerobic medium consisting of Trypticase Soy broth supplemented with 2% fetal bovine serum and 1% ethanolic cholesterol solution (Kunkle et al. 1986), and was incubated at 37°C on a rocking platform until early log-phase growth was achieved. Each pig was challenged orally three times on consecutive days with 100 mL of broth culture containing ~10¹⁰ viable cells. On d 1, feed was withdrawn 18 h before inoculation and then returned 4 h after infection. On the remaining 2 d, feed was withdrawn in the morning 2 h before inoculation and returned within 4 h of oral challenge.

Pigs were checked at least twice daily for clinical signs of SD, i.e., depression, lack of appetite, presence of blood and (or) mucus in feces. Rectal swabs were collected every second day after inoculation and cultured on selective Trypticase Soy agar plates (BBL Microbiology Systems, Cockeysville, MD), according to methods described by Siba et al. (1996).

Post-mortem procedure.  After an overnight fast, all pigs were slaughtered by captive bolt gun stunning followed by exsanguination at a post-mortem facility in the Division of Veterinary and Biomedical Sciences, Murdoch University. The entire gastrointestinal tract was recovered immediately, and the large intestine was identified. The ileocecal junction and rectum were ligated to prevent leakage of digesta, and the cecum was then excised from the colon. The cecum and colon were then weighed with their contents intact. Within 10 min of slaughter, samples of digesta (~5 g) were collected from the cecum, the first loop of the colon (proximal colon) and the apex of the spiral of the colon (distal colon), and placed into scintillation vials for subsequent determination of ATP concentration (Pluske et al. 1996). Another sample of digesta (~10 g) was collected from each of these locations and placed into sterile 100-mL plastic containers. The pH of the digesta was then recorded using a portable pH meter (Orion Research, Boston, MA), and containers were placed immediately on ice for transport to the laboratory where they were frozen at 24°C for subsequent analysis of volatile fatty acids (VFA). In addition, an intact segment ~5 cm in length was excised from the proximal colon for subsequent identification and numeration of the microflora present. The cecum and colon were finally stripped of contents, washed clean with water, blotted dry with soft paper towel and then reweighed.

Similar procedures were used to slaughter pigs infected with S. hyodysenteriae. Weights of the cecum and colon and digesta pH were not measured, and samples of digesta for subsequent ATP and VFA determination were not collected, because the products of colitis (e.g., mucus, blood) caused nonphysiologic values (data not shown). The cecum and colon of all pigs were examined for lesions and the presence of mucus and blood lining the intestinal epithelium. A swab was also taken from each of the cecum, proximal colon and distal colon for confirmation of S. hyodysenteriae by culture.

Sample analysis.  Methodology for the measurement of RS, VFA and digesta ATP is described by Pluske et al. (1996). Non-starch polysaccharides were analyzed according to the technique of Theander and Westerlund (1992).

Statistical analysis.  Data in both experiments were analyzed by one-way ANOVA (Statview v. 4.5, Abacus Concepts, Berkeley, CA) and, where appropriate, treatment means were compared using Fisher's-protected Least Significant Difference procedure (Maindonald 1992). Statistical significance was accepted at P < 0.05.

	RESULTS

Abstract Introduction Methods Results Discussion References

Experiment 1

pH values and large intestinal weights.  Pigs killed before infection and fed diets R-GG, R-RS and R-GG + RS had lower (P < 0.001) pH values in all parts of the large intestine compared with pigs fed diets R-OC and R-AP. Pigs fed diets R-RS and R-GG + RS had higher (P < 0.001) pH values in the proximal and distal colon compared with pigs fed diet R-GG. Differences in pH values were reflected generally in the weights of the cecum, colon and cecum + colon when expressed as a proportion of pig empty-body weight. Pigs fed diet R-AP had the lightest large intestine (1.2%), whereas pigs fed a combination of GG and RS had the heaviest large intestine (2.5%) (Table 3).

VFA and ATP production in the large intestine.  The total VFA concentration was lowest in the cecum of pigs fed diets R-GG, R-OC and R-AP (P < 0.001). In the colon, the concentration of VFA was highest in pigs fed diets R-GG, R-RS and R-GG + RS and lowest in pigs fed diets R-OC and R-AP (P < 0.001). In general, and in the cecum, pigs fed diets R-RS and R-GG + RS produced larger amounts of acetate, propionate and butyrate than pigs fed other diets (Table 4).

The acetate and propionate pools in the colon were similar (P > 0.05) in pigs fed diets R-GG, R-RS and R-GG + RS, and lowest in pigs fed diet R-OC and R-AP. Total pool size of butyrate was highest (P < 0.001) in pigs fed diets R-GG and R-RS and lowest in pigs fed diets R-OC and R-AP (P < 0.001). Pigs fed diet R-GG + RS had a colonic butyrate pool not significantly different (P > 0.05) than that of pigs fed diet R-RS, but lower (P < 0.001) than that of pigs fed diet R-GG (Table 4).

Pigs fed diets R-RS and R-GG + RS had higher (P < 0.05) concentrations of ATP in their ceca than pigs fed all other diets. The concentration of ATP in the proximal and distal colon was highest in pigs fed diet R-GG + RS and remained relatively constant along the entire length of the large intestine. Pigs fed diets R-GG, R-OC and R-AP had, in general, the lowest concentrations of ATP (Fig. 1). The concentration of ATP declined (P = 0.086) from the cecum to the distal colon, but there was no diet × site interaction for ATP levels (P = 0.22).

View larger version (18K): [in this window] [in a new window]   Fig 1. The concentration of ATP in the digesta of the cecum, proximal colon and distal colon in pigs fed rice-based diets containing different sources of carbohydrate or an animal protein supplement (Experiment 1). Values are mean ± SEM, n = 5. Within each region of the large intestine, values without superscripts in common differ, (a, b, c, P < 0.05).

Experiment 2

pH values and large intestinal weights.  There was no significant difference (P > 0.05) in the pH of digesta measured in the cecum and colon of pigs fed diets containing three different levels of RS. The pH of digesta increased as it moved from the cecum to the distal portion of the colon. Similarly, there were no significant differences (P > 0.05) in the weights of the cecum, colon and cecum + colon when expressed as a proportion of pig empty-body weight in pigs fed RS. There was, however, a trend for increased colon and cecum + colon weights with increasing level of RS in the diet (Table 5).

VFA and ATP production in the large intestine.  Total VFA concentration and the individual pool sizes of acetate, propionate and butyrate were similar (P > 0.05) in the cecum and colon of pigs fed three different levels of RS (Table 6).

The concentrations of ATP in the cecum, proximal colon and distal colon were similar in pigs fed all three diets containing RS. There was a trend (P = 0.062) for a decline in the concentration of ATP along the length of the large intestine (Fig. 2).

View larger version (35K): [in this window] [in a new window]   Fig 2. The concentration of ATP in the digesta of the cecum, proximal colon and distal colon in pigs fed rice-based diets containing increasing levels of resistant starch (RS) (Experiment 2). Values are mean ± SEM for 4, 4 and 5 pigs per treatment for diets R-RSa, R-RSb and R-RSc, respectively.

Colonization by spirochaetes and incidence of swine dysentery.  In the 4-wk period after infection in Experiment 1, spirochetes were cultured from rectal swabs taken from pigs eating all diets except R-AP. Only one pig in the group fed diet R-OC was found to be culture-positive for S. hyodysenteriae, and this occurred for 1 d only. Although S. hyodysenteriae colonized the large intestine of pigs fed diet R-RS (as evidenced by positive culture), the clinical expression of SD (confirmed by the presence of lesions in colon and isolation of S. hyodysenteriae upon post-mortem examination) was observed only in pigs fed diets R-GG and R-GG + RS (Table 7).

In Experiment 2, pigs fed all three diets were colonized by S. hyodysenteriae, as evidenced by culture from rectal swabs. In contrast to Experiment 1, pigs fed diets containing resistant starch in this study showed clinical signs of SD (Table 7).

	DISCUSSION

Abstract Introduction Methods Results Discussion References

The data from these two experiments confirm our previous work (Pluske et al. 1996) implicating soluble NSP and RS in the pathogenesis of SD and provide further evidence that the nature of the carbohydrate plays an important role in the clinical expression of the disease. In particular, our data suggest that it is rapidly fermentable carbohydrate entering the large intestine that is linked to the clinical expression of SD. In contrast to pigs fed GG and (or) RS, however, pigs fed oat chaff (diet R-OC, Experiment 1) displayed no clinical signs of SD. Oat chaff was fed as a source of insoluble NSP, containing (on a dry matter basis) 16.1 g/100 g insoluble NSP and 0.52 g/100 g soluble NSP (predominately as arabinose and xylose).

Insoluble NSP such as wheat bran expedites the transit of digesta through the large intestine (Govers et al. 1997, Payler et al. 1975). This phenomenon most likely explains the increased excretion of starch in the feces seen when a source of insoluble NSP is fed (Govers et al. 1997, Key and Mathers 1993, Young et al. 1996). Given this, then it is possible that pigs fed oat chaff may not have succumbed to SD because the digesta moved through at a faster rate, taking any RS contained in the diet with it. Alternatively, and in agreement with our previous work, the low level of soluble NSP may have contributed to the protection afforded against the disease (Pluske et al. 1996). Importantly, this observation supports our previous finding of the lack of a statistical relationship between the level of insoluble NSP in diets and SD (Pluske et al. 1996) and reaffirms the role of soluble NSP and RS in the pathogenesis of SD.

The difference between the incidence of SD in pigs fed RS in the two experiments is difficult to explain, especially because diet R-RS in Experiment 1 and diet R-RSb in Experiment 2 contained both the same source and similar inclusion levels of RS. A possible reason for the difference in the expression of SD may relate to the weight of the pigs at infection and the digestibility of RS in the small intestine. Pigs in Experiment 1 were infected at 38.7 kg, whereas those in Experiment 2 were infected at ~23 kg. Using pigs weighing ~70 kg, Govers et al. (1997) reported that 30-40% of RS (fed as high amylose cornstarch) was recovered in the ileal digesta. In contrast, using pigs weighing ~16 kg, Heijnen and Beynen (1997) reported that between 45 and 70% of RS (fed as either uncooked RS or retrograde RS) was recovered in the ileal digesta, suggesting that digestion of RS in the small intestine increases with age. A reduced quantity of RS entering the large intestine as a consequence of enhanced digestion in the small intestine may explain the apparent anomaly between the two experiments.

It is evident that different sources of "dietary fiber" have different regional effects on fermentation-related indices in the large intestine. This in turn is likely to influence the pathogenesis of SD. Slowly fermentable fiber (e.g., wheat bran) appears to have a greater influence on the distal region of the large bowel than rapidly fermentable fiber, such as oat bran and GG (McIntyre et al. 1991). In these experiments, and based upon ATP concentrations and pH measured along the large intestine, RS was rapidly fermented in the cecum and proximal colon, and decreased distally. These data confirm and extend previous studies in both rats (Lu et al. 1995) and pigs (Govers et al. 1997, Topping et al. 1997). Furthermore, the combination of GG + RS (Experiment 1) appeared to cause a shift in fermentation to more distal parts of the colon, as evidenced by both ATP concentrations and gut weights. However, only pigs fed a combination of GG + RS succumbed to SD (Experiment 1).

In the only other study we are aware of conducted with pigs in which a source of NSP and RS was fed to pigs in combination, Govers et al. (1997) found the highest concentration of butyrate in pigs fed a combination of RS (as high amylose cornstarch, or RS 1) and wheat bran. In this study, feeding GG with a source of RS produced less butyrate compared with feeding GG alone. Evidently, the source of NSP and RS is important in determining the physicochemical behavior of different types of dietary fiber in vivo. This in turn appears to influence the pathogenesis of SD. In addition, McIntyre et al. (1991) showed that in rats, GG was highly fermentable in the cecum and proximal region of the large intestine but that its fermentability reduced distally. This was in contrast to wheat bran, which was fermented more evenly along the large intestine.

Data from the present experiment in pigs fed the diet containing rice plus GG suggest the converse. The pattern of fermentation of GG was constant along the entire length of the large intestine, as evidenced by the production of ATP and the insignificant change in pH (range: 5.8-5.9) from the cecum to the distal colon. A feature in this study was the apparent disparity between fermentation assessed by pH and VFA concentrations, and that based upon ATP concentrations and gut weights. There are several possible reasons for this phenomenon. No one index of fermentation can be considered in isolation; instead, a suite of measurements is more appropriate to allow full interpretation of fermentation status. Using pH of the digesta as an indicator of fermentation can have limitations because the final pH will depend on the pK_a of the VFA accumulating; in addition, any factors that change the buffering capacity of the chyme will alter the final pH achieved at any given VFA concentration. The production of VFA allows the microorganisms in the hind-gut to produce ATP by substrate level phosphorylation for cellular function. The lower numbers of total anerobic bacteria found in pigs fed GG in this study support this notion (Durmic, Z., unpublished results). Given this logic, a high concentration (or total pool size) of VFA in combination with a lower concentration of ATP indicates that VFA is accumulating in the cecum and colon as a result of an imbalance between absorption and production. Guar gum, therefore, may have altered the functionality of the hind-gut by reducing the efficiency of VFA absorption.

Numerous studies (e.g., Govers et al. 1997, Marsono et al. 1993, Noakes et al. 1996, Topping et al. 1993, Weaver et al. 1992) have demonstrated that fermentation of RS, particularly in the cecum and proximal colon, produces more butyrate than does fermentation of NSP. In agreement, we observed a higher concentration of butyrate produced in the cecum of pigs fed the diet containing RS compared with those fed GG, although this was not significant (16 vs. 8 mmol/L, P = 0.130). Similar pool sizes of butyrate were produced in the colon of pigs fed GG and RS (28 vs. 20 mmol/L, P = 0.185). This contrasts with the work of Govers et al. (1997), who reported that the concentration of butyrate in the middle + distal colon of pigs fed a diet based on wheat bran and RS was significantly higher than that in pigs fed diets based on wheat bran and RS alone. Differences between studies are most likely attributable to the type of NSP and RS used. Govers et al. (1997) used wheat bran, a source of insoluble NSP that is known to move fermentation distally, whereas we used GG, a viscous NSP that ferments more evenly along the large intestine.

We have listed previously a number of possible mechanisms to explain the protective effect afforded by some diets against SD (Pluske et al. 1996, Siba et al. 1996). The results of this study confirm that the microbial digestion of fermentable carbohydrates in the large intestine facilitates the expression of SD but provide no insight(s) into the cause of the disease. However, it is possible that the pathogenesis of SD extends beyond simply an effect of fermentation per se. For example, it is recognized that the ability of S. hyodysenteriae to colonize porcine colonocytes containing mucus is an important virulence factor (Milner and Sellwood 1994). Sharma et al. (1995) reported that rats fed a diet containing fermentable fiber showed differences in mucin composition compared with rats fed a purified diet. This, in turn, could make colonocytes more susceptible to colonization by S. hyodysenteriae as a result of changes in chemotaxic-regulated motility (Kennedy et al. 1988).

It is also recognized that several other bacterial species must be present for SD to occur (Whipp et al. 1979). Durmic (unpublished results) observed increased total bacterial counts and stimulation of the growth of gram-negative bacteria when RS was fed to pigs. In contrast, pigs fed a diet containing GG had fewer bacteria (perhaps explaining the reduced ATP concentration observed; Fig. 1), and bacteria known to act synergistically with S. hyodysenteriae to cause SD (Whipp et al. 1979) were detected only in pigs fed this material. These synergistic bacteria were not detected in pigs fed other diets, raising the possibility that RS and soluble NSP have different mechanisms of action in the pathogenesis of SD. Feeding both types of polysaccharide, and the associated changes in the large intestinal microflora, appeared to enhance conditions that favored the growth of S. hyodysenteriae, which in turn assisted in the development of SD.

Given the results of this study and those found previously in our laboratory, we conducted a study aimed at finding an appropriate combination of grain processing and (or) addition of exogenous enzymes to reduce the amount of fermentable substrates entering the large intestine. This in turn should decrease the clinical expression of SD. Durmic et al. (1997) found that the extrusion of wheat to reduce the level of RS in the diet decreased the incidence of SD. However, the addition of an exogenous arabinoxylanase to a wheat-based diet had the opposite effect and increased the occurrence of SD. Studies to find a combination of grain processing and dietary enzymes to prevent SD are continuing.

In conclusion, the results of these two experiments demonstrate that rapidly fermentable carbohydrate, such as GG and a source of retrograde RS, is linked to the clinical expression of SD in experimentally infected pigs. By replacing a portion of a protective diet based on cooked white rice with carbohydrate sources that differed in their fermentability in the large intestine, we were able to examine directly the role of carbohydrates in the pathogenesis of SD. However, it is evident that the mechanism(s) responsible for protection against SD when some diets are fed is (are) presently unknown.

	FOOTNOTES

Manuscript received 14 January 1998. Initial reviews completed 20 February 1998. Revision accepted 27 May 1998.

	ACKNOWLEDGMENTS

Appreciation is extended to staff at Murdoch University and the Medina Research Centre for their assistance in the husbandry and management of experimental animals. The Australian Ricegrowers' Cooperative (Leeton, NSW) is thanked for their generous donation of the rice used in this study.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Close, W. H. & Fowler, V. R. (1985) Energy requirements of pigs. In: Recent Developments in Pig Nutrition (Cole, D.J.A. & Haresign, W., eds.), pp. 1-16. Butterworths, London, UK.
• Durmic, Z., Pethick, D. W., Mullan, B. P., Schulze, H. & Hampson, D. J. (1997) The effects of extrusion and arabinoxylanase in wheat based diets on fermentation in the large intestine and expression of swine dysentery. In: Manipulating Pig Production VI (Cranwell, P. D., ed.), p. 180. Australasian Pig Science Association, Werribee, Victoria, Australia.
• Govers, M.J.A.P., Gannon, N. J., Dunshea, F. R., Fielding, M., Kilias, D., Gibson, P. R. & Muir, J. G. (1997) Altering the site of fermentation in the pig: implications for colon cancer risk in humans. In: Manipulating Pig Production VI (Cranwell, P. D., ed.), p. 181. Australasian Pig Science Association, Werribee, Victoria, Australia.
• Hampson D. J., Cutler R., Lee B. J. Isolation of virulent Serpulina hyodysenteriae from a pig in a herd free from clinical dysentery. Vet. Rec. 1992; 131:318-319[Medline]
• Harris, D. L. & Lysons, R. J. (1992) Swine dysentery. In: Diseases of Swine (Leman, A. D., Straw, B. E., Mengeling, W. L., D'Allaire, S. & Taylor, D. J., eds.), 7th ed., pp. 599-616. Iowa State University Press, Ames, IA.
• Heijnen M.-L.A., Beynen A. C. Consumption of retrograded (RS) but not uncooked (RS) resistant starch shifts nitrogen excretion from urine to feces in cannulated piglets. J. Nutr. 1997; 127:1828-1832[Abstract/Free Full Text]
• Kennedy M. J., Rosnick D. K., Ulrich R. G., Yancey R. J. Jr. Association of Treponema hyodysenteriae with porcine intestinal mucosa. J. Gen. Microbiol. 1988; 134:1565-1576[Abstract/Free Full Text]
• Key F. B., Mathers J. C. Gastrointestinal responses of rats fed on white and wholemeal breads: complex carbohydrate digestibility and the influence on dietary fat content. Br. J. Nutr. 1993; 69:481-495[Medline]
• Kunkle R. A., Harris D. L., Kinyon J. M. Autoclaved liquid medium for propogation of Treponema hyodysenteriae. J. Clin. Microbiol. 1986; 24:669-671[Abstract/Free Full Text]
• Lu Z.-X., Muir J. G., Collier G. R. Comparison of resistant starch with soluble and insoluble non-starch polysaccharides on faecal output and fermentation dependent events in rats. Proc. Nutr. Soc. Aust. 1995; 19:41
• Maindonald J. H. Statistical design, analysis and presentation issues. N. Z. J. Agric. Res. 1992; 35:121-141
• Marsono Y., Illman R. J., Clarke J. M., Trimble R. P., Topping D. L. Plasma lipids and large bowel volatile fatty acids in pigs fed on white rice, brown rice and rice bran. Br. J. Nutr. 1993; 70:503-513[Medline]
• McIntyre A., Young G. P., Taranto T., Gibson P. R., Ward P. B. Different fibers have different regional effects on luminal contents of rat colon. Gastroenterology 1991; 101:1274-1281[Medline]
• Mhoma J.R.L., Hampson D. J., Robertson I. D. A serological survey to determine the prevalence of infection with Treponema hyodysenteriae in Western Australia. Aust. Vet. J. 1992; 69:81-84[Medline]
• Milner J. A., Sellwood R. Chemotactic response to mucin by Serpulina hyodysenteriae and other porcine spirochetes: potential role in intestinal colonization. Infect. Immun. 1994; 62:4095-4099[Abstract/Free Full Text]
• Noakes M., Clifton P. M., Nestel P. J., Le Lieu R., McIntosh G. Effect of high-amylose starch and oat bran on metabolic variables and bowel function in subjects with hypertriglyceridemia. Am. J. Clin. Nutr. 1996; 64:944-951[Abstract/Free Full Text]
• Payler D. K., Pomare E. W., Heaton K. W., Harvey R. F. The effect of wheat bran on intestinal transit. Gut 1975; 16:209-213[Abstract/Free Full Text]
• Pluske J. R., Siba P. M., Pethick D. W., Durmic Z., Mullan B. P., Hampson D. J. The incidence of swine dysentery in pigs can be reduced by feeding diets that limit the amount of fermentable substrate entering the large intestine. J. Nutr. 1996; 126:2920-2933
• Sharma R., Schumacher U., Ronaasen V., Coates M. Rat intestinal mucosal responses to a microbial flora and different diets. Gut 1995; 36:209-214[Abstract/Free Full Text]
• Siba P. M., Pethick D. W., Hampson D. J. Pigs experimentally infected with Serpulina hyodysenteriae can be protected from developing swine dysentery by feeding them a highly digestible diet. Epidemiol. Infect. 1996; 116:207-216[Medline]
• Theander, O. & Westerlund, E. (1992) Determination of individual components of dietary fibre. In: Dietary Fiber in Human Nutrition (Spiller, G. A., ed.), 2nd ed., pp. 77-98. CRC Press, Boca Raton, FL.
• Topping D. L., Gooden J. M., Brown I. L., Biebrick D. A., McGrath L., Trimble R. P., Choct M., Illman R. J. A high amylose (amylomaize) starch raises proximal large bowel starch and increases colon length in pigs. J. Nutr. 1997; 127:615-622[Abstract/Free Full Text]
• Topping D. L., Illman R. J., Clarke J. M., Trimble R. P., Jackson K. A., Marsono Y. Dietary fat and fiber alter large bowel and portal venous volatile fatty acids and plasma cholesterol but not biliary steroids in pigs. J. Nutr. 1993; 123:133-143
• Weaver G. A., Krause J. A., Miller T. L., Wolin M. J. Cornstarch fermentation by the colonic microbial community yields more butyrate than does cabbage fiber fermentation; cornstarch rates correlate negatively with methanogenesis. Am. J. Clin. Nutr. 1992; 55:70-77[Abstract/Free Full Text]
• Whipp S. C., Robinson I. M., Harris D. L., Glock R. D., Matthews P. J., Alexander T.J.L. Pathogenic synergism between Treponema hyodysenteriae and other selected anaerobes in gnotobiotic pigs. Infect. Immun. 1979; 26:1042-1047[Abstract/Free Full Text]
• Young G. P., McIntyre A., Albert V., Folino M., Muir J. G., Gibson P. R. Wheat bran suppresses potato starch-potentiated colorectal tumorigenesis at the aberrant crypt stage in a rat model. Gastroenterology 1996; 110:508-514[Medline]

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences