Although lactase nonpersistence (LNP)⁶ is prevalent worldwide, the symptomatic expression of lactose intolerance is relatively less prevalent. Most LNP people can consume one glass of milk per day asymptomatically (Scrimshaw and Murray 1988). The mechanisms by which a LNP subject can tolerate a certain amount of lactose, which apparently differ from that in lactase-persistent subjects, are not yet completely understood. One mechanism should be attributed to the role of colonic bacteria in carbohydrate malabsorption (Bond et al. 1980). In these LNP individuals, unhydrolyzed lactose passes into the large intestine, where it is fermented by the indigenous microflora into gases (H₂ , CH₄ and CO₂) and easily absorbable metabolites such as short-chain fatty acids (SCFA) (Ruppin et al. 1980). As a result, bacterial fermentation of malabsorbed sugars reduces the osmotic load of the sugar and the amount of the fecal water (Holtug et al. 1992, Saunders and Wiggins 1981), thereby preventing diarrhea. Recent studies have suggested that reduced colonic capacity for fermentation and deceased levels in SCFA are responsible for both carbohydrate-induced and antibiotic-associated diarrhea (Clausen et al. 1991, Holtug et al. 1992). On the other hand, factors that cause an increase in fermentation and SCFA production may contribute to suppression or mitigation of lactose intolerance. It is postulated that lactobacilli supplementation could enhance lactose fermentation and thus improve lactose intolerance.

Improvement of lactose digestion in LNP people by consumption of unfermented dairy products containing lactic acid bacteria has been reported (Jiang et al. 1996, Lin et al. 1991). Lactobacilli have been shown to survive through the gastrointestinal tract and adhere to intestinal tissue and human cells (Goldin et al. 1992, Kleeman and Klaenhammer 1982). Temporary colonization by these bacteria could modulate the metabolic activity of the resident microflora. Studies have shown that the consumption of lactobacilli-containing products reduces the activities of fecal bacterial enzymes, including -glucuronidase, nitroreductase, and azoreductase (Ayebo et al. 1980, Marteau et al. 1990). Further, administration of lactobacilli has resulted in a decrease in the fecal counts of Escherichia coli and other anaerobes (Ayebo et al. 1980, Johansson et al. 1993). Because Lactobacillus acidophilus is indigenous to the large intestine, exhibits high -galactosidase (-gal, EC 3.2.1.23) activity and readily ferments lactose without H₂ production, the survival or colonization of exogenous L. acidophilus in the colon could, theoretically, enhance lactose fermentation by supplementing additional -gal activity or by interacting with the indigenous microflora. However, no information is available regarding the ability of lactic acid bacteria to modify colonic fermentation of lactose by the indigenous microflora.

The large intestine is relatively inaccessible and a difficult area to investigate colonic microbial metabolism in vivo. As an alternative, an in vitro anaerobic continuous culture system has been established with bacterial populations similar to the colonic flora (Edwards et al. 1985, Rumney and Rowland 1992). In the present study, we evaluated the capability of colonic bacteria to ferment lactose and the ability of L. acidophilus to modify lactose fermentation by the colonic microflora using an in vitro continuous culture model. Lactobacillus acidophilus strain LA-1 was selected because it possesses relatively high -gal activity and is capable of improving lactose digestion in vivo as measured breath hydrogen production (Lin et al. 1991).

MATERIALS AND METHODS

Table 1. Composition of medium used for in vitro continuous culture [View Table]

-Galactosidase measurement.

RESULTS

Table 2. -Galactosidase activity and lactobacilli counts in 12-d period continuous cultures of human fecal microflora containing lactose or no lactose1,2 [View Table]

The addition of LA-1 to the continuous culture on d 0 caused a significantly greater decline in lactose concentration on d 1 than the control (40 ± 3.5 vs. 23 ± 4.1 mmol/L, n = 9, n = 13, respectively, P < 0.05, Fig. 2). However, from d 2 through 7, lactose concentrations were not affected by the single addition of LA-1. A similar enhanced decline in lactose concentration on d 1 was observed when LA-1 was introduced daily (d 0 through d 4) to the continuous culture (47 ± 3.8 vs. 27 ± 5.2 mmol/L, n = 5, P < 0.05, Fig. 2). Despite the continued addition of LA-1 through d 4, lactose concentrations from d 2 through d 7 were not different from the control values. Furthermore, in a single continuous culture experiment, the addition of LA-1 only on d 4 did not alter either lactose concentration or -gal activity compared with the control fermentation (data not shown).

An inverse relationship between lactose concentration and -gal activity was observed (Fig. 2 and Fig. 3). -Galactosidase activity in both treatments was significantly increased from d 0 to 1 and increased fivefold at the end (Fig. 3). However, the -gal activity of fecal microflora in continuous culture with LA-1 addition did not differ from that of the control (n = 12 and n = 9, respectively), despite the significant decrease in lactose concentration observed on d 1 (Fig. 3). The increase of -gal activity was similar whether LA-1 was added once or daily. Hence, data were combined from experiments using both single and daily addition of LA-1. Combined results were compared with the control fermentations in which LA-1 was not added (Fig. 3).

The SCFA concentrations in continuous cultures and the effect of daily addition of L. acidophilus strain LA-1 on SCFA production by fecal bacteria are showed in Table 3. Steady production of acetate and propionate was observed after d 1 in the control groups. For the total fermentation period, acetate and propionate concentrations in continuous cultures with added LA-1 were significantly higher than in the control group. This was due to significant differences on d 1. The introduction of LA-1 caused significantly greater productions of acetate and propionate on d 1 compared with the control. However, total butyrate and isobutyrate concentrations in the cultures with added LA-1 did not significantly differ from those in the control culture. Concentrations of butyrate and isobutyrate varied widely (Table 3). Butyrate concentration was diminished gradually, whereas isobutyrate concentration was increased.

Table 3. Short-chain fatty acid (SCFA) concentrations in continuous cultures of human fecal microflora with (+) or without () addition of L. acidophilus strain (LA-1)1,2,3 [View Table]

Bacterial populations seemed to stabilize within the first 2 d of continuous culture. Total lactobacilli, as measured by MRS medium, increased 100-fold from d 0 to 2 (10.55 ± 0.51 and 12.38 ± 0.09 log₁₀ cfu/L, P < 0.01). Gram-negative bacilli showed a similar increase. The addition of LA-1 did not affect the number of lactobacilli or Gram-negative bacilli, as compared with the corresponding control culture. Some reductions in strict anaerobes (as measured by WC medium) and Gram-positive cocci (as measured by CNA medium) were apparent when LA-1 was added to continuous culture, but these values were not significantly different within groups. Viable counts of strict anaerobes and Gram-positive cocci were not measured in continuous cultures without LA-1 addition.

DISCUSSION

The present study further suggests that colonic microflora contribute to the absorption of lactose in maldigesters. Colonic bacteria were able to utilize 15 g of lactose per day and up to 95% of lactose fed to continuous cultures at the dose of 25 g/d by d 7. The capacity of colonic bacteria to ferment lactose might be diminished or saturated when this sugar is ingested at a high dose, such as 50 g per day. Marked reduction of lactose concentrations occurred within 1 d, accompanied by production of SCFA. Production of isobutyrate is due to branched-chain amino acid catabolism. Increased isobutyrate production could result from increased protein degradation by bacteria other than lactobacilli such as bacteroides and bubacterium, paralleled by decreased lactose content in continuous culture.

The potential for lactic acid bacteria to modify the colonic microflora and its metabolism through bacterial enzyme activities is starting to be recognized (Ayebo et al. 1980, Marteau et al. 1990). However, the ability of lactic acid bacteria to modify the colonic bacterial fermentation of lactose and production of SCFA has not been documented. Recently, several studies have been unable to demonstrate the effect of lactic acid bacteria (L. acidophilus, Lactobacillus GG and Bifidobacterium longum) on colonic (fecal) carbohydrate fermentation, using in vitro mixed fecal incubations (Hove et al. 1994) or by directly measuring fecal SCFA concentrations (Bartram et al. 1994, Stansbridge et al. 1993). The models used in these studies have significant limitations. In a simple batch culture, the fermentation conditions (e.g., pH and substrates) are uncontrolled, which may inhibit bacterial metabolism (Rumney and Rowland 1992). In addition, because the vast majority of SCFA produced in the colon are absorbed, measurements of these metabolites in feces do not reflect conditions in the proximal colon. To some extent, these problems can be avoided in continuous cultures. Continuous flow culture methods seem to be excellent tools for in vitro studies of interactions among colonic bacteria (Rumney and Rowland 1992), in which colonic bacteria are maintained in viable and metabolically active steady-state conditions simulating those in the proximal colon (Edwards et al. 1985).

We evaluated the effect of lactobacilli supplementation on lactose fermentation by a mixed microbial community in continuous culture. Our results indicate that the introduction of L. acidophilus into a continuous culture system enhanced lactose utilization and acetate production during the early phase of fermentation, although the corresponding changes in bacterial population and -gal activity by the addition of exogenous lactobacilli were not detected. The latter findings may suggest that some other factors may involve lactose digestion by lactose-fermenting bacteria, including lactose permease, which transports lactose into the cell (Jiang et al. 1996). Moreover, a persistent effect on lactose fermentation was not observed even with daily addition of L. acidophilus. The lack of a long-term effect may be due to the colonic microflora out-competing the added lactobacilli by either rapid adaptation or by clearance mechanisms for exogenous microorganisms in this model.

The study clearly demonstrates that the bacterial capacity for lactose fermentation increases severalfold within 1-2 d with continuous feeding of lactose, whether or not lactobacilli are added. Thus, lactobacilli did not have any detectable effect after adaptation had taken place. Indeed, the lack of improvement in lactose fermentation by exogenous lactobacilli beyond d 1 is probably due to rapid adaptation by the endogenous flora to lactose. The capacity of the colonic flora to metabolize lactose has been considered to be important in moderating intolerance symptoms (Bond et al. 1980). Chronic exposure to lactose (or lactulose) enhances fermentation, resulting in increased SCFA concentrations, a marked fall in pH, and reduced hydrogen production (Florent et al. 1985, Hertzler and Savaiano 1996, Perman et al. 1981). In a recent study, Flourie et al. (1993) confirmed that this adaptation limited the diarrhea induced by a large challenge dose of lactulose. Furthermore, in free-living malabsorbers, chronic ingestion of lactose resulted in an improvement in lactose tolerance (Hertzler and Savaiano 1996). The present study demonstrated that fecal bacteria respond rapidly to the addition of lactose by increasing -gal activity near 20-fold and slightly increasing number of lactobacilli (Table 1). Prolonged feeding of lactose may induce changes in colonic bacterial metabolic pathways and stimulate the growth of all lactose-fermenting organisms at the expense of non-lactose fermenters (Hill 1983). Regardless of the mechanism, adaptation results in a more efficient fermentation of lactose. Thus, an additional effect of lactobacilli supplementation is not expected.

In addition, the gastrointestinal microflora is a complex and stable ecosystem, possessing active clearance mechanisms for exogenous microorganisms (van der Waaij et al. 1971). In our preliminary study, viable counts of L. acidophilus strain LA-1 in pure continuous culture with the same culture variables were 3 × 10¹⁰, 5.5 × 10¹⁰, 1.2 × 10¹⁰ and 2 × 10⁹ cfu/L on d 0, 1, 2, and 3, respectively, suggesting the bacteria grow in the continuous culture system. The ability of the indigenous microflora to prevent the implantation of exogenous bacteria has been well demonstrated (Bouhnik et al. 1992, van der Waaij et al. 1971). After ingestion of a fermented dairy product, only 1.5% of two lactobacilli strains were shown to survive at the terminal ileum (Petterson et al. 1983). Also, Bouhnik et al. (1992) reported that one-third of a dose of ingested bifidobacteria survived and was excreted into the feces. Furthermore, in L. acidophilus feeding trials, both the bacterial populations and enzyme activities returned to presupplementation levels within a few days after treatment was ended (Lidbeck et al. 1987, Marteau et al. 1990). These findings suggest that exogenously administrated lactic acid bacteria do not prosper in the intestinal tract under physiological conditions. Others have assumed that lactobacilli survive in the intestine but do not multiply (Bouhnik et al. 1992, Freter 1992). It has been suggested that lactobacilli should be consumed continuously to influence the intestinal microflora (Johansson et al. 1993, Lidbeck et al. 1987).

The data indicating that the addition of L. acidophilus strain LA-1 improved lactose digestion during the initial phase (i.e., on d 1) may suggest that metabolic alteration by lactobacilli supplementation is most likely to occur when colonization resistance is reduced (i.e., when a stable microflora has not been established). Previous studies (Edwards et al. 1985, Martini et al. 1992) demonstrated that the microbial populations stabilize within 2 d in continuous culture systems. Thus, effective implantation of an exogenous bacterium may require the prior administration of antibiotics to reduce competing organisms. Lactic acid bacteria have been used to treat antibiotic-associated diarrhea with some success (Gotz et al. 1979, Zoppi et al. 1982).

In summary, our study suggests that the colonic flora may adapt quickly to and metabolize most of the lactose malabsorbed in the small intestine. The data may also indicate that supplementation with L. acidophilus results in an improvement of in vitro lactose fermentation when a stable bacterial population and/or an adaptation of colonic flora to lactose load is not established. Accordingly, consumption of lactic acid bacteria may improve colonic lactose fermentation in certain situations in which the endogenous microflora has been disrupted, such as diarrheal disease or antibiotic administration. Further studies are needed to investigate the potential use of lactobacilli supplementation and the adaptive kinetics of the colonic flora in patients with intestinal disturbances.