PeskyPeater
Member
The Journal of Nutrition, Volume 130, Issue 3, March 2000, Pages 585–593, https://doi.org/10.1093/jn/130.3.585
Published:
01 March 2000
Increased amounts of short-chain fatty acids (SCFA)5 and unfermented material in the colonic lumen are two processes by which dietary fiber is suggested to function in human health and disease prevention. Butyrate, one of the SCFA generated during fermentation of dietary fiber, has been shown to have trophic properties (Sakata 1987). In addition, it has been shown to be an antitumor agent using several cultured cell lines (Bugaut and Bentejac 1993) and a rat model (McIntyre et al. 1993). Molecular bases for this antitumor property are being defined (Smith et al. 1998, Velazquez et al. 1997). Wheat bran (WB) has been proposed as protective in colon cancer because a higher proportion of butyrate is generated by its fermentation, compared to the fermentation of other fibers (Bugaut and Bentejac 1993, Klurfeld 1997). Elevated concentrations of n-butyrate have been reported during in vitro fermentation of WB by some investigators (McBurney and Thompson 1990, McIntyre et al. 1993, Salvador et al. 1993), but not by others (Bourquin et al. 1992).
A second possible protective mechanism for WB is that if it were incompletely fermented, the unfermented WB-derived residue would increase colonic lumenal contents and increase rate of transit of material through the colon, thereby reducing the exposure of the mucosa to carcinogens (Bugaut and Bentejac 1993, Klurfeld 1997). The slow and incomplete fermentation of WB, but not of oat bran (OB) (Chen et al. 1998, Nyman et al. 1986, Stephen and Cummings 1980), supports the role for WB of diluting carcinogenic agents in the intestinal lumen. Lupton (1995) proposed after a careful comparison of results from in vitro and in vivo studies that WB is protective against colon cancer because it is incompletely fermented and not because fermentation of WB yields a higher proportion of butyrate.
Propionate, another SCFA generated during fermentation of dietary fiber, has been proposed as an inhibitor of hepatic synthesis of cholesterol (Anderson et al. 1990). Various clinical and experimental studies have demonstrated that OB, but not WB, reduces serum cholesterol levels (Anderson et al. 1990, Shinnick and Marlett 1993). However, the many experiments, using a variety of animal and in vitro models and protocols, that have been conducted to evaluate the possible hypocholesterolemic action of propionate have yielded conflicting and inconsistent results (Bugaut and Bentejac 1993).
One objective of this research was to compare and contrast the rate and extent of fermentations of OB and WB and the accompanying SCFA production to distinguish their hypothesized antineoplastic and hypocholesterolemic functions. Fermentation in the large intestine of monogastric species is a dynamic process that involves over 400 species of microbes (Savage 1983) and material much more complex than the test fibers usually employed in in vitro fermentation studies (Cummings 1981). Further, a single measurement of fermentation, typical of most in vivo studies, might not detect elevations in specific SCFA in an ongoing process. Therefore, we also used this experiment to evaluate the ability of an in vitro system using physiological substrate and microflora to predict what is known about fiber digestibility in humans. The physiological system used ileal digesta as the substrate, and cecal microflora previously exposed to the substrate as the source of inoculum, because prior exposure of the inoculum to the substrate to be fermented, as well as the source of inoculum, has been shown to influence fermentation (Monsma and Marlett 1995, 1996).
DISCUSSION
This experiment was designed with the intent of using animals to model fermentation in the human large intestine. Ileal digesta was used because it is the substrate for microflora in the large intestine; protein, fat and other carbohydrates, besides the dietary fiber, in ileal digesta are rarely considered in in vitro fermentation studies, even though they are considered important sources of fermentable material (Cummings 1981). For example, we found that ileal digesta from both rats (Monsma et al. 1992) and swine (Monsma and Marlett, unpublished data) fed purified fiber-free diets contained ∼500 μmol carbohydrate/g of dry weight ileal digesta. Further, as much as 20% of the total SCFA generated during microbial fermentation of the glycoprotein mucin, produced endogenously by the gastrointestinal tract, were propionate and n-butyrate (Monsma and Marlett, unpublished data). Ileal digesta was collected from pigs because the digested remnants of defined meals could be collected in sufficient quantity to analyze and ferment from this animal model.
The similarity among the proximate compositions of the swine ileal digesta used in this study, other swine ileal digesta we have analyzed and ileal effluent from humans supports our decision to use swine ileal digesta as the substrate to model human ileal digesta. McBurney et al. (1988) fed a human ileostomate a basal diet containing 13 g/d of fiber and the same diet supplemented with white bread, OB, kidney beans or red lentils. In these studies, the mean (±SEM) content of the ileal effluent was (g/kg dry digesta) 217 ± 18 protein, 27 ± 6 fat, 143 ± 17 ash and 611 ± 37 carbohydrate by difference. Lia et al. (1996) fed nine ileostomy subjects a basal diet supplemented with bread containing only white flour or supplemented with OB, OB and β-glucanase or a barley fraction. In these studies, the mean content of the ileal effluent was (g/kg dry digesta) 236 ± 9 protein, 39 ± 12 fat and 725 ± 18 carbohydrate by difference and ash combined. The mean content of ileal digesta collected from pigs fed a fiber-free test meal or one containing 5% dietary fiber from canned peas was (g/kg dry digesta) 215 protein, 45 fat, 264 ash and 476 carbohydrate by difference (Marlett and Longacre, unpublished), similar to the composition of the swine ileal digesta containing WB or OB used in this study.
The rat was the inoculum source as it was more cost- effective and justifiable than using swine for this purpose. The cecum was the inoculum source because that is the microbial population that is first exposed to ileal residue in vivo. Our experience indicates that rat fecal inoculum does not ferment to the same extent as cecal inocula from the same animal prefed the test material to be fermented (Monsma and Marlett 1995, and 1996). Savage (1983) views feces as a waste product and proposed that determining the effect of diet on biochemical activities in the proximal colon lumen from biochemical activities in feces may be misleading. Actual measurements made by MacFarlane et al. (1992) using human colonic contents support Savage's contention. MacFarlane et al. (1992) reported that bacteria from the ascending colon of two sudden death victims generated five to eight times more SCFA than did bacteria from the sigmoid-rectum region of the same subjects. However, rat cecal inocula production of SCFA from pectin and purified soybean fiber was not different than what was produced when human fecal inocula was used to ferment the same substrates (Barry et al., 1995). In vitro fermentation using human fecal inocula of 11 of 12 dietary fiber concentrates or fiber extracted from mixed diets were generally similar to the net digestibilities of the sugars from the same fibers in the same humans from which the fecal inocula were collected (Daniel et al. 1997, Wisker et al. 1998). The one fiber source that was fermented much more extensively in vitro vs. in vivo was a barley concentrate that contained a high proportion of total fiber as cellulose.
Our in vitro fermentation results are similar to net fiber digestibility in vivo in humans and rats. The apparent digestibilities of WB fiber in rats (Hansen et al. 1992, Nyman et al. 1986), of 41 and 49%, are similar to digestibilities reported in humans, of 34% (Nyman et al. 1986) and 56% (Chen et al. 1998), and to the disappearance of WB-derived sugars, of 47%, in our in vitro fermentation system. Likewise, the disappearance of OB-derived sugars in our in vitro system, of 84%, is similar to apparent digestibility of OB in the rat (Hansen et al. 1992), of 93% and humans (Chen et al. 1998), of 96%. These in vitro and in vivo results of WB and OB fibers across species are remarkably similar, in light of the facts that they were conducted by different laboratories using different substrates. In humans, OB dietary fiber increases fecal bacterial mass (Chen et al. 1998). Changes in the muramic acid content of the in vitro system we used also are consistent with these in vivo observations.
The initial increase in propionate in our study corresponded with the rapid disappearance of β-glucan in the OB digesta fermentation, suggesting that fermentation of β-glucan is at least one source of the increase in propionate. This observation supports those by Bach Knudsen et al. (1993) who reported in vivo increases in propionate in the ceca of swine fed diets containing either OB or β-glucan-enriched fractions, compared to diets containing the insoluble residue of OB. The larger molar proportion of propionate we measured during in vitro OB fermentation is consistent with the data of Jackson and Topping (1993). They observed a larger proportion of propionate was generated in the ceca of rats consuming OB, compared to cecal SCFA composition of the group fed WB. Thus, some in vitro and in vivo data are consistent with the proposal that a component of the hypocholesterolemic effect of some dietary fibers could be caused by propionate inhibition of hepatic cholesterol synthesis (Anderson et al. 1990). However, other studies (McBurney and Thompson 1990, McIntyre et al. 1993) did not observe elevated molar proportions of propionate when OB was fermented, relative to when WB was fermented in vitro. As comprehensively reviewed by Bugaut and Bentejac (1993), data to support this hypothesis are not compelling. Rather, the hypocholesterolemic action of some fibers appears to be related to their effects on sterol balance (Marlett 1997). Viscous soluble fibers decrease bile acid absorption in the terminal small bowel (Marlett et al. 1994) which stimulates hepatic bile acid synthesis that uses LDL-cholesterol as its primary substrate (Schwartz et al. 1982).
Our observation that fermentation of WB ileal digesta did not produce a greater proportion of SCFA as n-butyrate than OB digesta, but rather a smaller absolute amount of n-butyrate, agrees with those of Bourquin et al. (1992), who fermented dietary fiber either isolated from WB or OB, or the fiber-derived polysaccharides extracted and subsequently recombined. Neither our findings or those of Bourquin et al. (1992) support the hypothesis of McIntyre et al. (1993) that n-butyrate from WB fermentation is a significant protective mechanism against tumor formation. Others (McBurney and Thompson 1990, Salvador et al. 1993) who observed increased n-butyrate production during in vitro fermentation of WB used WB that contained residual starch. In vitro fermentation of starch (Englyst and Macfarlane 1986) produces a significant proportion of SCFA as n-butyrate, and it is possible that fermentation of the starch contaminating the WB, and not the WB fiber, was responsible for the production of more butyrate in these studies.
The greater microbial efficiency of carbohydrate utilization we observed during fermentation of WB ileal digesta, compared to OB ileal digesta, in conjunction with the less complete fermentation of WB carbohydrate, suggests that the microflora used additional sources of carbon for growth. Protein fermentation has been estimated to account for 17% of the SCFA in the cecum to 38% in the sigmoid/rectum of humans (Macfarlane et al. 1986), and it likely occurred during our studies. At every time point in our study at least 1.5 to 2 times more SCFA were being produced than what was predicted by the stoichiometric equation for carbohydrate fermentation developed by Miller and Wolin (1979).
... In summary, although we observed a larger proportion of propionate produced during OB fermentation, the majority of the evidence suggests that the hypocholesterolemic mechanism for viscous, soluble dietary fibers is not propionate inhibition of cholesterol synthesis (Bugaut and Bentejac 1993, Marlett 1997). The lack of increase in butyrate production during WB fermentation in our studies supports the contention that poorly-fermented WB is protective against colon cancer because it dilutes lumenal contents, not because it provides butyrate for the colonic mucosa (Bugaut and Bentejac 1993, Klurfeld 1997, Lupton 1995).
