The Journal of Nutrition Vol. 128 No. 12 December 1998, pp. 2313-2318

Probiotics Reinforce Mucosal Degradation of Antigens in Rats: Implications for Therapeutic Use of Probiotics^1,2

Tanja Pessi, Yelda Sütas, Aulis Marttinen^*, and Erika Isolauri³

Department of Pediatrics, University of Turku, Finland and ^* Medical School, University of Tampere, Finland

	ABSTRACT

Abstract Introduction Methods Results Discussion References

The effects of probiotics, administered with different diets, i.e., unhydrolyzed or hydrolyzed dietary antigens, on macromolecular degradation in the gut mucosa were studied. Rat pups were divided into five feeding groups at the age of 14 d. In addition to maternal milk, the milk group was gavaged daily with cows' milk and the hydrolysate group with extensively hydrolyzed whey formula, while controls received sterile saline. In addition to these diets, the milk-GG group and the hydrolysate-GG group were given probiotic bacteria, Lactobacillus GG ATCC 53103 (10¹⁰ colony-forming units per day). At 21 d, the absorption of macromolecules, horseradish peroxidase and -lactoglobulin across patch-free jejunal segments was studied in Ussing chambers. The degree of macromolecular degradation was studied by means of HPLC gel filtration. The absorption rate of intact horseradish peroxidase differed among the feeding groups (P = 0.038). This was due to the high median (interquartile range) absorption of intact horseradish peroxidase (ng × h¹ × cm²) in the milk group [255 (14-1332)] and supplementation with L. GG in the milk-GG group [35 (8-233)] restoring the status to the control level [22 (0-116)]. A parallel effect was seen in the hydrolysate group [100 (9-236)] vs. the hydrolysate-GG group [1 (0- 13)]. A gel filtration study confirmed that larger molecules were absorbed across the mucosa in the milk group compared to the other groups. The absorption of degraded horseradish peroxidase differed between the feeding groups (P = 0.005). L. GG had a distinct effect when administered with unhydrolyzed, native protein vs. hydrolyzed protein: it increased absorption of degraded horseradish peroxidase in the milk-GG group [7310 (4763-8228)] vs. the milk group [3726 (2423-5915)], while reducing it in the hydrolysate-GG group [2051 (1463-2815)] vs. the hydrolysate group [4573 (3759-9620)]. Our results showed that probiotics not only restore aberrant macromolecular transport, but they also have a specific effect on mucosal degradation depending on dietary antigen: adjuvant-like properties (unhydrolyzed antigen) and immunosuppressive-like properties (hydrolyzed antigen). The antigenicity of the diet therefore should be taken into consideration, when introducing novel probiotic functional foods for the management of gastrointestinal disorders.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

The gut mucosa plays a central role in the exclusion and elimination of potentially harmful dietary antigens and microorganisms, while providing selective absorption of nutrients (Brandzaeg 1995). Antigen exclusion has been associated with factors such as the capacity of the gut mucosa to produce secretory IgA and mucus (Slomiany et al. 1987, Stokes et al. 1975). Secretory IgA prevents the adherence of enteral antigens to the mucosal surface, and mucus prevents microbial infestation. The intestinal microflora has been shown to contribute to antigenic exclusion (Bengtmark and Jeppsson 1995). The resident microflora prevents adherence of antigens by competing for nutrients and adhesion sites, producing antimicrobial agents, and by increasing production of specific antibody secreting cells and mucus.

Early exposure of the gut mucosa to live microorganisms and bacterial colonization, together with the introduction of dietary antigens, play an important role in the development of gut barrier functions and unresponsiveness to ingested antigens at neonatal stage (Helgeland et al. 1996, Sudo et al. 1997). The microflora enhances the development of the barrier by increasing the population of duodenal IgA-plasmocytes (Moreau et al. 1987) and the number of enteroendocrine cells in the jejunal and colonic epithelium (Sharma and Schumacher 1995), indicating enhancement of the production of secretory IgA and mucus. The microflora have also been shown to stimulate proliferation of epithelial cells and increase the total intestinal surface (Heyman et al. 1986). Demonstration that the intestinal microflora constitutes an important part of the exclusive component of the mucosal barrier has led to the introduction of novel modes of therapeutic intervention, using specific strains of microflora as probiotics. Probiotics are "a live microbial food ingredient that is beneficial to health" (Salminen et al. 1998). Beneficial effects have been reported in clinical and in experimental studies (Isolauri et al. 1993, Kaila et al. 1992, Malin et al. 1996a, Perdigón et al. 1988).

Whether probiotics contribute to the capability of mucosa to degrade antigens, i.e., to antigen elimination, has not been addressed. Antigen elimination involves the ability of the mucosa to process antigens and decrease the antigen load to the immune system. Processed antigens are presented to the underlying lymphoid tissue so modified that inflammatory responses and ensuing barrier disruption are prevented (Weiner et al. 1994). An answer to the question was sought here by introducing probiotics to suckling rat pups before gut closure, i.e., at a stage when the mechanisms of antigen exclusion and elimination are not fully developed (Brandzaeg 1995, Teichberg et al. 1990). A probiotic strain, L. GG (ATCC 53103), was introduced to rat pups concomitantly with unhydrolyzed or hydrolyzed dietary antigens, mimicking different feeding regimens. The degree of macromolecular degradation across the gut mucosa was studied using the Ussing chamber method and HPLC gel filtration.

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Animals and study design.  Wistar rat pups (University of Tampere, Tampere, Finland) of either sex were reared at a density of 8-10 pups per dam in an experimental animal laboratory with barriers. At the age of 14 d the pups were randomly assigned to five feeding groups until 21 d of age. The animals were weighed daily. All pups remained with the dam and received maternal milk. The dams received standard rodent diet (Altromin 1310, CRH Petersen A/S, Ringsted, Denmark; crude proteins 22.5%) free of cows' milk proteins to eliminate the effects of maternal antibodies as antigen transport in pups.

All pups were gavaged daily with bicarbonate buffer (0.9 mol/L NaHCO₃, pH 7.9) at 0.01 mL/1 g body weight immediately before feedings to reduce the acidity of stomach contents. In addition, the milk group (n = 10) was gavaged with cows' milk (osmolality, 602 mOsm/kg of H₂O; Valio, Helsinki, Finland) at 1 mg milk protein/1 g body wt in 0.01 mL sterile water; the hydrolysate group (n = 10) was gavaged with extensively hydrolyzed whey formula, Peptidi-Tutteli (osmolality, 580 mOsm/kg of H₂O; Valio, Helsinki, Finland), at 1 mg protein hydrolysate/1 g body wt in 0.01 mL sterile water. The milk-GG group (n = 10) and the hydrolysate-GG group (n = 10) were gavaged with 0.4 mg of L. GG plus 1 mg milk protein/extensively hydrolyzed whey formula/1 g body wt in 0.01 mL sterile water. Daily dose of L. GG, freeze-dried powder, contained 10¹⁰ colony-forming units. Control rats (n = 10) were gavaged with bicarbonate buffer only.

The study was approved by the Institute of Animal Care and Use at the University of Tampere.

Measurement of intestinal function.  The Ussing chamber method was used to measure the quantity of transmural protein transfer in its intact and degraded forms. Antigens are transported across the intestinal epithelium by endocytosis along two pathways: a minor route allowing transfer of the intact protein and a major route involving lysosomal degradation during the transport. Both pathways are transcellular. Protein transfer was measured described elsewhere (Heyman et al. 1982). Horseradish peroxidase (HRP)4 was chosen to represent the properties of dietary antigens by means of its molecular weight as well as by means of its intestinal transport (Isolauri et al. 1990, Marcon-Genty et al. 1991).

The reagents used in the chamber studies were HRP (Sigma type VI; Sigma Chemical, St. Louis, MO, MW 40,000) and -lactoglobulin (-LG) (Sigma; MW 36,000). HRP and -LG were iodinated with 18,500 kBq of ¹²⁵I/1.0 mg protein according to the method of Markwell (1982). Labeled HRP and -LG were separated from free iodine using gel filtration. All reagents used in HPLC and Ussing chamber studies were diluted in Ringer's solution containing (in mmol/L) 140 Na⁺, 5.2 K⁺, 1.2 Ca²⁺, 1.2 Mg²⁺, 120 Cl, 25 HCO₃, 2.4 HPO₄² and 0.4 H₂PO₄.

At 21 d of age, the pups were anaesthetized with 30-40 mg/kg pentobarbital intraperitoneally, and laparotomy was immediately performed. After removal of the intestine, the animals were decapitated by cervical dislocation. For each group, eight to 10 jejunal segments without Peyer's patches were excised and mounted flat on filters in an Ussing chamber (exposed area 0.15 cm²). HRP was added to the mucosal compartment at a final concentration of 0.4 g/L. Samples of 800 µL from the serosal compartment were collected at 10-20 min intervals. The studies were performed as previously described (Isolauri et al. 1993). Separate Ussing chamber studies were made with -LG using the same protocol as in the HRP experiment. The final concentration of -LG in the mucosal compartment was 1.0 g/L. Double-antibody enzyme-linked immunosorbent assay was used to measure -LG concentrations as described previously (Mäkinen-Kiljunen and Palosuo 1992). The results from different analyses were equalized employing the comparison of standard curves and were expressed as µg/L. The sensitivity of the assay was 0.04 µg/L.

A voltage clamp was used to control the transmembrane potential difference (PD) in the chambers. Tissues were short-circuited using an automatic voltage clamp (World Precision Instruments, Aston, UK) after appropriate correction for fluid and system resistance. The tissue was pulsed at 1.0 mV every 50 s. The short circuit current (I_sc) was continuously recorded by a plotter, and the electrical conductance (G) of the tissue was calculated according to Ohm's law.

Measurement of intestinal degradation with HPLC.  TSK SW3000 and SW4000 (Beckman, Beckman Instruments, San Remon, Canada) columns connected in series were used to fractionate macromolecules of serosal samples from the Ussing chamber study. Potassium phosphate buffer (60 mmol K₂HPO₄-KH₂PO₄, pH 6.9) containing 0.01% 3-[3-(cholamidopropyl) dimethyl-ammonio]-1-propanesulfonate (CHAPS) and 1% 2-propanol was used as elution buffer. To prevent aggregation of molecules in samples, they were treated with 0.03% n-octyl--D-glycopyranoside, 0.06 CHAPS, 10% formic acid and 2 mmol DL-dithiothreitol. After 20 min incubation at room temperature, the samples were injected into the column. The absorbance was monitored at a wavelength of 280 nm. First, the elution volumes of HRP and -LG were estimated. Second, to assess the amount of macromolecules released from the jejunal segment itself, separate experiments were performed without adding HRP or -LG to the mucosal compartment. Third, the amount of ¹²⁵I-HRP or ¹²⁵I--LG in serosal samples was estimated according to the radioactivity of their fractions. Fractions (0.9 mL) were collected after reaching 36 mL elution volume until an elution volume of 58.5 mL. The amount of ¹²⁵I-HRP or ¹²⁵I--LG in fractions was expressed in counts per minute (cpm) in elution volume profiles.

Statistical analysis.  Electrical parameters and the absorption rate of HRP in its intact and degraded forms are presented as medians with interquartile ranges during the steady-state period, i.e., between 50 and 70 min. The proportion of degradation is expressed as means 95% CI during the steady-state period. The quantities of ¹²⁵I-HRP or ¹²⁵I--LG in fractions expressed in cpm are given as geometric means of individual experiments. Statistically significant differences between the groups were designated when P < 0.05 by Kruskal-Wallis test and between two groups by ANOVA with post hoc comparisons using Fisher's PLSD.

	RESULTS

Abstract Introduction Methods Results Discussion References

Effect of Lactobacillus GG supplementation on epithelial functions in Ussing chamber.  The viability of jejunal tissue was observed by potential difference, an index of paracellular integrity by conductance and active electrolyte transport capacities by short circuit current (Armstrong 1987). During the Ussing chamber study none of the electrical parameters PD, G or I_sc was modified by supplementation with L. GG or by dietary antigens. They varied within physiological limits in all feeding groups.

The absorption rate of intact HRP across the jejunum was significantly different in the feeding groups (P = 0.04, Kruskal-Wallis) (Fig. 1a). This could be attributed to the high absorption rate of intact HRP in the milk group, while supplementation with L. GG in the milk-GG group restored the rate to control level (P = 0.03, ANOVA with Fisher's PLSD). The same trend was observed with the hydrolysate group and the hydrolysate-GG group (Fig. 1a). Parallel results were obtained with -LG: there was a greater absorption rate of intact -LG in the milk group compared to controls, while the other groups, milk-GG, hydrolysate and hydrolysate-GG, did not differ from controls.

View larger version (23K): [in this window] [in a new window]   Fig 1. Effect of L. GG supplementation on absorption rate of antigens in Ussing chamber. Absorption of intact (A) and degraded (B) horseradish peroxidase (HRP) across patch-free jejunal segments from rat pups at the age of 21 d. Data are the medians and interquartile ranges of measurements in control rats (n = 10), rats fed with cows' milk (milk, n = 10), rats fed with cows' milk and L. GG (milk-GG, n = 10), rats fed with extensively hydrolyzed whey formula (hydrolysate, n = 10) and rats fed with extensively hydrolyzed whey formula and L. GG (hydrolysate-GG, n = 10). Significant differences between two groups were designated when P < 0.05 by ANOVA with post hoc comparisons using Fisher's PLSD test.

The absorption rate of HRP in its degraded form differed in the feeding groups (P = 0.005, Kruskal-Wallis) (Fig. 1b). The absorption rate of degraded HRP was not different from controls in the milk group and the hydrolysate group. The absorption rate of HRP in its degraded form was different between the hydrolysate and hydrolysate-GG group (P = 0.005, ANOVA with Fisher's PLSD).

Supplementation with L. GG had a distinct effect on macromolecular transport depending on the type of dietary antigen: unhydrolyzed in the milk-GG group and hydrolyzed in the hydrolysate-GG group. In the milk-GG group, degraded HRP absorption was much greater, whereas it was lower in the hydrolysate-GG group than in their unsupplemented counterparts. Similar results were obtained with -LG used as tracer (data not shown).

The proportion of HRP transported in degraded form differed among the groups (P = 0.04, Kruskal-Wallis). In experiments with control rats it was 97.6% (94.6-100) of total HRP absorption: mean (95% CI). The proportion of degraded HRP absorption was lower in the milk group [83.1% (69.7-96.5); P = 0.01, ANOVA with Fisher's PLSD], while in the milk-GG group it was 94.8% (85.1-100). In the hydrolysate and hydrolysate-GG groups the proportions of degraded HRP of the total absorption were 97.8% (95.9-99.7) and 99.1% (97.1-100), respectively. The trial was repeated with -LG with similar results (data not shown).

Degree of degradation by jejunal segments.  According to HPLC gel filtration chromatograms (Fig. 2), the elution volume of native HRP was less than in experiments for serosal HRP, i.e., serosal HRP is more degraded and its molecular weight is smaller than that of native HRP. Experiments with control rats without addition of HRP showed that there was a release of molecules from the jejunum itself. In experiments using HRP, the elution volume of serosal molecules was greater in controls than that in the milk group. Larger molecules and a greater amount of molecules were transported across the gut mucosa in the milk group than in controls. Elution volumes and an amount of molecules in serosal samples in the milk-GG, the hydrolysate and hydrolysate-GG groups were comparable to controls (data not shown).

View larger version (20K): [in this window] [in a new window]   Fig 2. Evaluation of degradation of antigens by jejunal segments. HPLC elution volume profile of molecules in the serosal compartment and of native horseradish peroxidase (HRP) after Ussing chamber study. Arrows denote the elution volumes in three different experiments: 1) with jejunal segments from control rats without addition of HRP [without HRP (controls)], 2) with jejunal segments from control rats with addition of HRP [with HRP (controls)], 3) with jejunal segment from the milk group with addition of HRP [with HRP (milk)].

The elution volume of native ¹²⁵I-HRP was less than that of serosal transported ¹²⁵I-HRP (Fig. 3). In experiments with control rats, the elution volume of ¹²⁵I-HRP in serosal samples was greater than that in the milk group. This means that HRP was degraded more by gut mucosa in controls than in the milk group. The amount of ¹²⁵I-HRP in serosal samples was also greater in the milk group compared to controls. In the milk-GG group the elution volume of ¹²⁵I-HRP was comparable to that in controls (Fig. 3a). In the hydrolysate group and in the hydrolysate-GG group, the elution volumes of ¹²⁵I-HRP and the amounts of ¹²⁵I-HRP were comparable to control levels (Fig. 3b). Similar results were obtained with -LG (Fig. 4a, b).

View larger version (18K): [in this window] [in a new window]   Fig 3. Degree of degradation of antigens by jejunal segments. HPLC elution volume profile of 125I-horseradish peroxidase (HRP) degradation products by the gut mucosa in serosal compartment and of native 125I-HRP fractions after Ussing chamber study. The counts per minute (cpm) of elution volume fractions were expressed as the geometric mean of eight individual experiments in four different groups: pups fed with cows' milk (milk, A), pups fed with extensively hydrolyzed whey formula (hydrolysate, B), pups fed with cows' milk + L. GG (milk-GG, A) or pups fed with extensively hydrolyzed whey formula + L. GG (hydrolysate-GG, B).

View larger version (19K): [in this window] [in a new window]   Fig 4. Degree of degradation of antigens by jejunal segments. HPLC elution volume profile of 125I--lactoglobulin (-LG) degradation products by the gut mucosa in serosal compartment and of native 125I--LG fractions after Ussing chamber study. The counts per minute (cpm) of elution volume fractions were expressed as the geometric mean of six individual experiments in four different groups: pups fed with cows' milk (milk, A), pups fed with extensively hydrolyzed whey formula (hydrolysate, B), pups fed with cows' milk + L. GG (milk-GG, A) or pups fed with extensively hydrolyzed whey formula + L. GG (hydrolysate-GG, B).

	DISCUSSION

Abstract Introduction Methods Results Discussion References

Antigen exclusion and elimination and immune regulation mechanisms are incompletely developed at early stages of postnatal life (Brandzaeg 1995). Immaturity of the gut barrier predisposes to aberrant antigen uptake (Teichberg et al. 1990). Before gut closure, antigen exposure itself may enhance transport of intact macromolecules across the gut mucosa (Teichberg et al. 1990). The antigenicity of macromolecules is characterized by foreignness to the host, by their molecular complexity, solubility and stability, and by their concentration (Thompson and Stewart 1993). Our results show that in early life, encountered cows' milk protein may enhance transport of bystander macromolecules in their intact forms, whereas exposure to cows' milk protein enzymatically hydrolyzed to molecular weight less than 2500 does not do so.

In health, mucosal processing of enteral antigens modifies most of their antigenic properties, thus rendering them partially inert to the systemic immunity (Peng et al. 1990). Environmental factors, particularly those associated with mucosal inflammation, may interfere with this particular phenomenon of systemic hyporesponsiveness, the result being induction of a systemic immune response to the same antigens (Macpherson et al. 1996). This is hypothesized to perpetuate the damage to the mucosa and impairment of mucosal barrier function. Disruption of the mucosal barrier and the consequent increase in permeability to bystander antigens in the gut implies a positive feedback on mucosal inflammation (Wyatt et al. 1993). Also an increase in the uptake of luminal antigens stimulates the immune compartment of the gut mucosal barrier and induces the local secretion of proinflammatory cytokines such as interferon- and tumor necrosis factor- (Strobel and Ehrhardt 1993, Rugtveit et al. 1997). These proinflammatory cytokines can directly disrupt the tight junctional integrity and thus have the potential to cause a leakage of macromolecules in their intact forms (Marano et al. 1993). In this manner patients with intestinal inflammation tend to manifest increased mucosal immune response against their nonpathogenic commensal intestinal bacteria (Macpherson et al. 1996) and a proportional change in the composition of gut microflora (Drasar and Shiner 1969, Malin et al. 1996b). These earlier observations sum up a trilogy in gut inflammation which involves changes in mucosal permeability and immune response to luminal antigens as well as changes in the composition of the intestinal microflora.

The resident intestinal microflora reinforces the barrier function concomitantly with counteracting inflammation-induced dysfunctions. Experimental studies with rotavirus-infected mice have shown that in the absence of intestinal microflora, disturbance in intestinal uptake of macromolecules is more severe than in its presence (Heyman et al. 1987). Also an exposure to intestinal microflora at the neonatal stage may regulate the induction of unresponsiveness by affecting the maturation process in the gut (Sudo et al. 1997). Early dietary manipulation modifies the maturation of gut-associated lymphoid tissue as well as the content of microflora (Guihot et al. 1997). These recent data further underline the crucial role of the diet and microflora in the gut maturation and counteracting intestinal inflammation. It is possible to modulate the dietary antigen structure and to implement specific strains of intestinal microorganisms as probiotics. There is hitherto little information as to how such therapeutic means affect the mucosal response to inflammation. Administration of probiotics is known to modify the intestinal composition of microflora by increasing the numbers of other lactobacilli and anaerobes (Salminen et al. 1998). Dietary implementation of probiotic bacteria has been shown to promote IgA immune response (Kaila et al. 1992, Malin 1996a) and target antigen transport across the Peyer's patches in the gut (Isolauri et al. 1993). Peyer's patches are one of the primary areas in the gut mucosa where specific immune responses are generated (Weiner et al. 1994). Luminal antigens are mainly transported into Peyer's patches and presented to major histocompatibility complex (MHC) class II bearing antigen presenting cells (Weiner et al. 1994). Therefore targeting the transfer of antigens across Peyer's patches may be important in generating local secretory immune responses. Probiotics have been shown to stimulate the production of interferon- (DeSimone et al. 1986), which also contributes to antigen presentation via MHC II expression and confers IgA responses. Unresponsiveness to ingested antigen is probably a combined effect of immune exclusion performed by such antibodies and suppression of systemic antibody and T cell-mediated responses (Husby et al. 1994). The present results suggest that probiotics effect mucosal permeability by restoring aberrant macromolecular transport and may thus be beneficial in reversing disturbances induced by mucosal inflammation: increased permeability, immune response to antigen and composition of microflora.

The present study contributes new data on macromolecular degradation by the gut mucosa, based on results of Ussing chamber studies and HPLC studies. The Ussing chamber method allowed quantitative measurement of macromolecular transport across the gut mucosa, and the HPLC study allowed comparison between sizes of molecules absorbed across the gut in different feeding groups. To evaluate the degree of degradation, elution volumes of tracer molecules were run through columns separately. There was a slight difference between elution volumes of native HRP according to HPLC gel filtration chromatograms (48.6 mL) and radioactivity of ¹²⁵I-HRP fractions (46.8 mL). This may be due to the fact that labeling (¹²⁵I) increased the molecular size of HRP. Probiotic bacteria were shown to enhance the macromolecular degradation by the gut mucosa, thereby reducing the antigen load. This is reflected in the result indicating that both the proportion of degraded macromolecules of the total protein transport and the degree of degradation of macromolecules were both enhanced.

Another novel finding in the present study was that probiotic bacteria generated their effect on mucosal antigen handling in association with the type of dietary content and thereby the antigenic milieu in the gut lumen. L. GG administered together with unhydrolyzed antigen increased the transport of degraded macromolecules, but when administered with hydrolyzed antigen, it reduced the transport. Increased degraded macromolecular absorption, after administered L. GG with unhydrolyzed antigen, was not associated with dysfunction of the gut mucosa or tissue damage, since absorption of intact protein was not increased and electrical parameters remain unchanged. Increased absorption may have pronounced the immunostimulatory effect of probiotics as an adjuvant: adherence of adjuvant to the epithelium enhances antigen absorption and antigen-specific immune responses to antigens encountered (Lycke and Holmgren 1986, Perdigón et al. 1991). Diminished degraded macromolecular absorption is seen in nude mice, where the limiting step is not intracellular degradation but rather occurs at the luminal membrane (Heyman et al. 1986). The contrary effect of L. GG on absorption with hydrolyzed antigen may be due to reduced antigenicity. Lack of antigen stimulus, the complex of hydrolyzed antigen and L. GG may have similar immunomodulatory activity as earlier shown in dietary peptides generated using L. GG, i.e., the downregulation of proliferative (Sütas et al. 1996) and interleukin-4-producing (Sütas et al. 1997) immune responsiveness. This kind of suppressive effect of L. GG may be a useful tool for alleviating hyperresponsiveness in the gut.

Our results provide possibilities for exploiting natural defense mechanisms in the prevention and the treatment of gut barrier dysfunction by probiotic therapy. Our primary result indicates that probiotics restore aberrant macromolecular transport, suggesting their potential use in reversing increased mucosal permeability in gut inflammation. In addition, probiotic bacteria were here identified with a contributory role in mucosal degradation of antigens when the antigen content of the diet is enriched. It is intriguing to find that the more reduced the antigen content in the diet, the less pronounced is the effect of probiotic bacteria on mucosal degradation. These opposite effects depending on the dietary antigen content should be taken into consideration in the generation of the clinical applications of novel functional foods including probiotics.

	FOOTNOTES

Manuscript received 31 March 1998. Initial reviews completed 5 June 1998. Revision accepted 11 August 1998.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Armstrong, W. M. (1987) Cellular mechanisms of ion transport in the small intestine. In: Physiology of the gastrointestinal tract (Johnson, L. R., ed.), pp. 1251-1265. Raven Press, New York.
• Bengtmark S., Jeppsson B. Gastrointestinal surface protection and mucosa reconditioning. J. Parent. Enteral. Nutr. 1995; 19:410-415[Abstract/Free Full Text]
• Brandzaeg P. Molecular and cellular aspects of the secretory immunoglobulin system. APMIS 1995; 103:1-19[Medline]
• DeSimone C., Bianchi-Salvadori B., Negri R., Ferrazi M., Baldinelli L., Vesely R. The adjuvant effect of yogurt on production of gamma-interferon by ConA-stimulated human peripheral blood lymphocytes. Nutr. Rep. Int. 1986; 33:419-433
• Drasar B. S., Shiner M. Studies on the intestinal flora. II. Bacterial flora of the small intestine in patients with gastrointestinal disorders. Gut 1969; 10:812-819[Abstract/Free Full Text]
• Guihot G., Merle V., Lebotgne M., Pivert G., Corriol O., Brousse N., Ricour C., Colomb V. Enteral nutrition modifies gut-associated lymphoid tissue in rat regardless of the molecular form of nitrogen supply. J. Pediatr. Gastroenterol. Nutr. 1997; 24:153-161[Medline]
• Helgeland L., Vaage J. T., Rolstad B., Midvedt T., Brandtzaeg P. Microbial colonization influences composition and T-cell receptor V beta repertoire of intraepithelial lymphocytes in rat intestine. Immunology 1996; 89:494-501[Medline]
• Heyman M., Ducroc R., Desjeux J. F., Morgat J. L. Horseradish peroxidase transport across adult rabbit jejunum in vitro. Am. J. Physiol. 1982; 242:G558-G564[Abstract/Free Full Text]
• Heyman M., Crain-Denoyelle M., Corthier G., Desjeux J. F. Postnatal development of protein absorption in conventional and germ-free mice. Am. J. Physiol. 1986; 251:G326-G331
• Heyman M., Gorthier G., Petit A., Meslin J. C., Moreau C., Desjeux J. F. Intestinal absorption of macromolecules during viral enteritis: an experimental study on rotavirus-infected conventional and germ-free mice. Pediatr. Res. 1987; 22:72-78[Medline]
• Husby S., Metecky J., Moldoveanu Z., Holland S., Elson C. O. Oral tolerance in humans. T cell but not B cell tolerance after antigen feeding. J. Immunol. 1994; 152:4663-4670[Abstract]
• Isolauri E., Gotteland G., Heyman M., Porchart P., Desjeux J. F. Antigen absorption in rabbit bacterial diarrhea (RDEC-1). In vitro modifications ileum and Peyer's patches. Dig. Dis. Sci. 1990; 35:360-366
• Isolauri E., Majamaa H., Arvola T., Rantala I., Arvilommi H. Lactobacillus strain GG reverses increased intestinal permeability induced by cow milk in suckling rats. Gastroenterology 1993; 105:1643-1650[Medline]
• Kaila M., Isolauri E., Soppi E., Virtanen E., Laine S., Arvilommi H. Enhancement of the circulating antibody secreting cell response in human diarrhoea by a human Lactobacillus strain. Pediatr. Res. 1992; 32:141-144[Medline]
• Lycke N., Holmgren J. Strong adjuvant properties of cholera toxin on gut mucosal immune responses to orally presented antigens. Immunology 1986; 59:301-308[Medline]
• Macpherson A., Khoo U. Y., Forgacs I., Philpott-Howard J., Bjarnason I. Mucosal antibodies in inflammatory bowel disease are directed against intestinal bacteria. Gut 1996; 38:365-375[Abstract/Free Full Text]
• Mäkinen-Kiljunen S., Palosuo T. A sensitivity enzyme-linked immunosorbent assay for determination of bovine -lactoglobulin in infant feeding formulas and in human milk. Allergy 1992; 47:347-352[Medline]
• Malin M., Suomalainen H., Saxelin M., Isolauri E. Promotion of IgA immune responses in patients with Crohn's disease by oral bacteriotherapy with Lactobacillus GG. Ann. Nutr. Metab. 1996a; 40:137-145[Medline]
• Malin M., Verronen P., Mykkänen H., Salminen S., Isolauri E. Increased bacterial urease activity in faeces in juvenile chronic arthritis: evidence of altered intestinal microflora? Br. J. Rheumat. 1996b; 35:689-694[Abstract/Free Full Text]
• Marano C. W., Laughlin K. V., Russo L. M., Peralta-Soler A., Mullin J. M. Long term effects of tumor necrosis factor on LCC-PK1 transepithelial resistance. J. Cell Physiol. 1993; 157:519-527[Medline]
• Marcon-Genty D., Krempf M., Tome D. Transmucosal transport of phytohemagglutinin, concanavalin A and alpha-lactalbumin across rabbit ileum in vitro. Gastroenterol. Clin. Biol. 1991; 15:10-15[Medline]
• Markwell M. A. A new solid state reagent to iodinate proteins. I. Conditions for efficient labelling of antiserum. Anal. Biochem. 1982; 125:427-432[Medline]
• Moreau M. C., Ducluzeau R., Guy-Grand D., Muller M. C. Increase in the population of duodenal immunoglobulin A plasmocytes in axenic mice associated with different living or dead bacterial strains of intestinal origin. Infect. Immun. 1987; 21:532-539
• Peng H. J., Turner M. W., Strobel S. The generation of a tolerogen after ingestion of ovalbumin is time dependent and unrelated to serum levels of immunoreactive antigen. Clin. Exp. Immunol. 1990; 81:520-525
• Perdigón G., de Macías M. E. N., Alvarez S., Oliver G., de Ruiz Holgadol, A. P. Systemic augmentation of immune response in mice by feeding fermented milks with Lactobacillus casei and Lactobacillus acidophilus. Immunology 1988; 63:17-23[Medline]
• Perdigon G., Alvarez S., de Ruiz Holgado, A. P. Immunoadjuvant activity of oral Lactobacillus casei: influence of dose on the secretory immune response and protective capacity in intestinal infections. J. Dairy Res. 1991; 58:485-496[Medline]
• Rugtveit J., Nislen E. M., Bakka A., Carlsen H., Brantzaeg P., Scott H. Cytokine profiles differ in newly recruited and resident subset of mucosal macrophages from inflammatory bowel disease. Gastroenterology 1997; 112:1493-1505[Medline]
• Salminen S., Bouley C., Boutron-Ruault M. C., Cummings J., Franck A., Gibson G., Isolauri E., Moreau M. C., Roberfroid M., Rowland I. R. Functional food science and gastrointestinal physiology and function. Br. J. Nutr. 1998; 80:5147-5171
• Sharma R., Schumacher U. Morphometric analysis of intestinal mucins under different dietary conditions and gut flora in rats. Dig. Dis. Sci. 1995; 40:2532-2539[Medline]
• Slomiany B. L., Sarosiek Y., Slomiany A. Gastric mucus and the mucosal barrier. Dig. Dis. Sci. 1987; 5:125-145
• Stokes C. R., Soothill J. F., Turner M. W. Immune exclusion is a function of IgA. Nature 1975; 255:745-746[Medline]
• Strobel W., Ehrhardt R. O. Chronic intestinal inflammation: an unexpected outcome in cytokine or T cell receptor mutant mice. Cell 1993; 75:203-205[Medline]
• Sudo N., Sawamura S. A., Tanaka K., Aiba Y., Kubo C., Koga Y. The requirement of intestinal bacterial flora for the development of an IgE production system fully susceptible to oral tolerance induction. J. Immunol. 1997; 157:1739-1745
• Sütas Y., Soppi E., Korhonen H., Syväoja E. L., Saxelin M., Rokka T., Isolauri E. Suppression of lymphocyte proliferation in vitro by bovine caseins hydrolysed with Lactobacillus GG-derived enzymes. J. Allergy Clin. Immunol. 1996; 98:216-224[Medline]
• Sütas Y., Hurme M., Isolauri E. Down-regulation of antiCD3 antibody-induced IL-4 production by bovine caseins hydrolysed with Lactobacillus GG-derived enzymes. Scand. J. Immunol. 1997; 43:687-689
• Teichberg S., Isolauri E., Wapnir R. A., Roberts B., Lifshitz F. Development of neonatal rat small intestinal barrier to non-specific macromolecular absorption: effect of early weaning to artificial diets. Pediatr. Res. 1990; 28:31-37[Medline]
• Thompson, P. J. & Stewart, G. A. (1993) Allergens. In: Allergy (Holgate, S. T. & Church, M. K., eds.), pp. 1.1-1.14. Gower Medical Publishing, London, UK.
• Weiner H. L, Friedman A., Miller A., Khoury S. J., Al-Sabbagh A., Santos L., Sayegh M., Nussenblatt R. B., Trentham D. E., Hafler D. A. Oral tolerance: immunologic mechanisms and treatment of animal and human organ-specific autoimmune diseases by oral adminstration of autoantigens. Annu. Rev. Immunol. 1994; 12:809-837[Medline]
• Wyatt J., Vogelsang H., Hübl W., Waldhöer T., Lochs H. Intestinal permeability and the prediction of relapse in Crohn's disease. Lancet 1993; 341:1437-1439[Medline]

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