Comparison of mucosal adhesion and species identification of bifidobacteria isolated from healthy and allergic infants


*Corresponding author. Tel.: +358 (2) 3336894; Fax: +358 (2) 3336860, E-mail address:


Fifty bifidobacteria strains were isolated from fecal samples of allergic and age matched healthy infants. Allergic infants were found to have an adult type Bifidobacterium flora with high levels of Bifidobacterium adolescentis. Healthy infants had a typical infant Bifidobacterium flora with high levels of Bifidobacterium bifidum. These isolates were tested for their adhesive properties to human intestinal mucus. The adhesion of the fecal bifidobacteria from healthy infants was significantly higher (P<0.0001) than for allergic infants. This suggests a correlation between allergic disease and the composition of the intestinal bifidobacteria flora which has reduced adhesive abilities to the intestinal mucus. Therefore, dietary supplementation of bifidobacteria typical for healthy infants, may be beneficial in the treatment of allergic disorders.


Bifidobacteria are among the predominant bacteria in the human intestinal microflora, they colonize the neonatal intestine from the first week after birth and inhabit the gastrointestinal tract throughout life [1,2]. However, it has been shown that the mode of delivery has a significant influence on the rate of colonization and the composition of the microflora colonizing the gastrointestinal tract of an infant [3]. The same researchers have shown that microbial colonization also influences the immune development of the infant. The presence of bifidobacteria in the human intestine is believed to contribute to human health and well being, and many health benefits of these microorganisms have been documented [4–8]. However, the available information on the adhesion properties of bifidobacteria is still limited.

The presence of allergic diseases is characterized by enhanced immunoglobulin (Ig) E responses to common environmental antigens, is increasing world wide, especially in Western industrialized countries [9]. Atopic dermatitis and food allergy represent the earliest manifestations of allergic disease, occurring in infancy and early childhood [10]. Although the etiology of such allergic disease is still not completely understood, a disturbance of the intestinal microflora, especially of endogenous lactic acid bacteria, has been reported as one of the important events related to allergic diseases [11–13]. The immune development may be related to the development and composition of the intestinal microflora. Since bifidobacteria are a major member of the intestinal microflora, it is important to understand their role in allergic and healthy infants.

Adhesion to intestinal mucus is regarded as a prerequisite for colonization by microorganisms. Therefore, mucosal adhesion has been proposed as one of the main selection criteria for probiotic strains [7,14]. Adhesion of probiotics to the intestinal mucosa is also considered important for modulation of the immune system [15]. The different capacities to adhere to human mucus have been reported for some probiotic, dairy and clinical lactic bacteria strains [16–19]. However, no studies have been performed to compare the adhesive capacity of human endogenous bifidobacteria from healthy and allergic infants to human intestinal mucus.

The present study was conducted to investigate the differences between Bifidobacterium strains in the feces of allergic and healthy, age matched, infants. The main goal was to establish if differences exist in the composition and the adhesion properties of bifidobacterial strains isolated from healthy and allergic infants. In the present study, 50 strains of bifidobacteria were isolated from the feces of healthy and allergic infants. The strains were identified to species level and the adhesion to human intestinal mucus was examined. During the adhesion studies, the probiotic strains Lactobacillus rhamnosus GG (ATCC 53103) and Bifidobacterium lactis Bb12 were included for comparison.

2Material and methods


Fecal samples were collected from four infants diagnosed with food allergy and from six age matched healthy infants (2–7 months of age), all infants were breast fed. All allergic infants manifested atopic eczema according to the Hanifin criteria [20]. The collected fecal samples were kept at 4°C and analyzed within 6 h. Fecal bifidobacteria were isolated and identified using the methods reported by Mitsuoka [5] and Hosoda and co-workers [21]. In brief, the fecal samples were homogenized, and a 10-fold serial dilution was prepared with an anaerobic buffer solution. Fifty microliters of appropriate dilutions were plated on the Blood Liver agar (BL agar; Nissui Seiyaku Co., Tokyo, Japan). The inoculum was incubated under anaerobic conditions at 37°C for 72 h. The characteristic brown colonies formed on BL agar were randomly picked as potential bifidobacteria and further confirmed by Gram stain, cell morphology, absence of growth under aerobic conditions and spore formation. The strains were identified to species level based on their sugar fermentation pattern, as described by Mitsuoka [5]. Isolated bifidobacteria were stored at −75°C in 50% glycerol until use.

B. lactis Bb12 and L. rhamnosus GG (ATCC 53103) were obtained from Chr. Hansen Ltd. (Hørsholm, Denmark) and Valio Ltd. (Helsinki, Finland) respectively.

2.2Mucus preparation

Mucus was isolated from the feces by extraction and dual ethanol precipitation [16,22]. In short, feces from healthy adult volunteers were suspended in ice cold PBS containing protease inhibitors and sodium azide. Fecal extracts were prepared by centrifuging the suspensions at 15 000×g at 4°C. Mucus was isolated from the fecal extract by dual ethanol precipitation. The pooled mucus was dissolved in HEPES (N-(2-hydroxyethyl)piperazine N-2(ethane sulfonic acid)) buffered Hanks’ balanced salt solution (HH; 10 mM HEPES; pH 7.4) at a concentration of 1 mg ml−1. Any particulate material was removed from the suspension by centrifugation (2000×g for 10 min).

2.3In vitro adhesion assay

The tested bacteria were grown in 1 ml GAM broth (Nissui Seiyaku Co., Tokyo, Japan) from frozen stocks. To metabolically radiolabel the bacteria, 20 μl (methyl-1,2-3H)thymidine (117 Ci mmol−1) was added to the medium. After approximately 36 h of anaerobic growth at 37°C, the bacteria were harvested by centrifugation (2000×g, 7 min), washed twice and resuspended in HH buffer. The absorbance at 600 nm was adjusted to 0.25±0.01 in order to standardize the number of bacteria (approximately 107 CFU ml−1).

The adhesion assay was performed essentially as reported earlier [16,18]. In short, the intestinal mucus was passively immobilized on polystyrene microtiter plate wells by overnight incubation at 4°C. The wells were washed with HH buffer and the radioactively labelled bacteria were added. After 1.5 h incubation at 37°C, the wells were washed and the adhered bacteria were lysed with 1% SDS in 0.1 M NaOH at 60°C for 1 h. The radioactivity of the lysed bacteria was assessed by liquid scintillation. Adhesion is expressed as the percentage of radioactivity recovered from the wells compared to radioactivity added to the wells.

2.4Statistical analysis

The results from the adhesion experiments are expressed as the average of three or four independent experiments. Each experiment was performed with four parallels to correct for intra-assay variation. The Mann–Whitney U-test was used to evaluate the statistical significance (P<0.05) of the differences in the ability to adhere to intestinal mucus of different Bifidobacterium species and between bifidobacteria from the two different infant groups.


A total of 50 strains of the predominant bifidobacteria in the fecal samples from healthy and allergic infants were isolated. The strains were identified to species level (Tables 1 and 2) and a distinct profile of the bifidobacterial flora was detected. Among the isolates from the allergic infants, Bifidobacterium adolescentis was the most predominant species (20 out of 24 isolates), while in the healthy infants Bifidobacterium bifidum was found to be the most common species (15 out of 26 isolates).

Table 1. Bifidobacterium strains isolated from healthy infants and the adhesion of these strains to human intestinal mucus
Species identificationAdhesion (%) (mean±S.D.)
B. infantis H-1(1)5.3±5.1
B. infantis H-1(2)10.4±6.1
B. breve H-1(3)12.8±1.4
B. breve H-1(4)8.7±5.6
B. breve H-1(5)11.3±4.5
B. breve H-1(6)5.6±7.5
B. infantis H-1(7)8.0±4.1
B. breve H-1(8)14.3±6.0
B. infantis H-1(9)7.0±8.4
B. breve H-1(10)13.6±3.9
B. bifidum H-2-19.2±4.9
B. bifidum H-2-213.3±8.4
B. bifidum H-2-38.1±5.1
B. bifidum H-2-414.6±4.8
B. bifidum H-2-59.4±5.2
B. bifidum H-2-611.0±3.0
B. bifidum H-3-19.6±6.3
B. bifidum H-3-411.7±14.1
B. bifidum H-3-55.6±4.6
B. bifidum H-3-62.6±2.2
B. bifidum H-3-75.6±2.6
B. bifidum H-3-89.1±3.0
B. bifidum H-3-98.1±5.8
B. bifidum H-3-1010.7±9.9
B. bifidum H-3-119.7±3.3
B. breve H-3-125.6±4.8
Table 2. Bifidobacterium strains isolated from allergic infants and the adhesion of these strains to human intestinal mucus
Species identificationAdhesion (%) (mean±S.D.)
B. infantis 12(1)9.2±1.3
B. infantis 12(2)3.6±4.1
B. infantis 12(3)9.3±6.3
B. infantis 12(4)2.9±1.3
B. adolescentis b 12(5)6.6±2.2
B. adolescentis b 28(1)2.6±1.9
B. adolescentis b 28(2)2.9±3.7
B. adolescentis a IS-8(1)6.1±1.4
B. adolescentis a IS-8(2)1.9±1.4
B. adolescentis a IS-8(4)4.7±2.5
B. adolescentis a IS-8(5)0.9±0.2
B. adolescentis a IS-8(6)2.8±1.9
B. adolescentis a IS-8(7)6.6±4.9
B. adolescentis a IS-8(8)5.4±3.8
B. adolescentis c 66(1)3.7±3.4
B. adolescentis c 66(5)4.5±3.9
B. adolescentis c 66(6)4.2±2.3
B. adolescentis c 66(7)2.3±0.6
B. adolescentis c 66(8)2.1±1.2
B. adolescentis c 66(9)2.0±0.8
B. adolescentis c 66(10)0.3±0.2
B. adolescentis c 66(12)5.5±8.7
B. adolescentis c 66(13)3.5±2.4
B. adolescentis c 66(17)1.3±0.7

The isolated bifidobacteria were tested for their ability to adhere to human fecal mucus. Each of the tested fecal bacteria exhibited their characteristic adhesion to fecal mucus ranging from 0.9 to 14.6% (Tables 1 and 2). The adhesion of the fecal bifidobacteria from healthy infants (9.3%; S.D. 3.1%) was significantly higher (P<0.0001) than the adhesion of the isolates from allergic infants (3.9%; S.D. 2.3%). Bifidobacterium infantis was the only Bifidobacterium species allergic and healthy infants had in common. No significant difference in adhesion was observed for this species between the two groups. B. adolescentis was found to be the least adhesive species, 3.5% (S.D. 1.9%); P=0.0044. None of the other isolated species adhered significantly different from each other. The probiotic strains, Lactobacillus GG and B. lactis Bb12 which were included for comparison, adhered 38.5% (S.D. 7.8%) and 19.0% (S.D. 7.7%) respectively.


Several studies have indicated that physiological disorders and infectious diseases are associated with a disturbance of human intestinal microflora characterized by low numbers of bifidobacteria and lactobacilli. For example, in infective and antibiotic associated diarrhea, constipation, Crohn's disease, gastritis and in infants with biliary atresia, the intestinal bifidobacteria were found to be significantly decreased [5,6,23,24]. A similar tendency has also been observed in some cases of allergic disease. The microflora of children suffering from food allergy and atopic dermatitis were described by low numbers of bifidobacteria and lactobacilli [11,12]. However, no direct evidence on the role of intestinal microbes in the mechanism of food allergy exists [25]. In the present study, the composition of the fecal bifidobacteria flora from healthy and allergic infants was determined. Also, the adhesion of these bifidobacteria and two probiotic bacteria, Lactobacillus GG and B. lactis Bb12, to human intestinal mucus, was examined.

Bifidobacteria have been reported to first appear in the neonatal intestine between days 2 and 5 after birth, and reaching a level of up to 99% of the fecal flora within one week [1,2]. The implantation and development of bifidobacteria in the neonate are greatly influenced by prematurity, method of birth, early feeding regime, endogenous substrate availability from breast milk or formula and the environment [3,26,27]. Upon weaning, the numbers of bifidobacteria decrease and stay at a more constant level in the adult, then decreasing further among the elderly [2,28,29].

In the current study, the composition of the fecal bifidobacteria flora was found to be very different between the healthy and allergic infants. Healthy infants had a typical infant bifidobacteria flora with B. bifidum, B. breve and B. infantis. Allergic infants, in contrast, were mainly colonized with B. adolescentis and some B. infantis, resembling a more adult-like Bifidobacterium flora [1,30].

The probiotic strains, Lactobacillus GG and B. lactis Bb12, exhibited the highest adhesion to human mucus. The adhesion of these probiotic strains is in agreement with previous reports, in which Lactobacillus GG and B. lactis Bb12 showed high adhesive capacities to human intestinal mucus and Caco-2 tissue culture cell line [16–18]. Conversely, most of the fecal bifidobacteria exhibited a relatively low affinity for human intestinal mucus. These results suggest that for persistence of the endogenous bifidobacteria, factors other than adhesion are likely to be involved. The colon is the main habitat of intestinal bifidobacteria, the relatively low flow in this part of the gastrointestinal tract may explain the persistence of low adhering strains.

The present work is the first to determine the adhesive abilities of endogenous bifidobacteria from allergic and healthy infants. In the current study, the bifidobacteria isolated from allergic infants were found to adhere significantly less to the human mucus compared to the strains from healthy infants (P<0.0001). Also the composition of the Bifidobacterium flora was different for the two study groups. These results indicate a correlation between allergic diseases and the human intestinal Bifidobacterium flora. Based on these results, it may be suggested that the decrease of the intestinal bifidobacteria in the allergic disorders previously reported could be explained by the loss of the mucus adhesive endogenous bifidobacteria, due to the presence of low adhesive B. adolescentis strains. Adhesive strains have also been observed to be more likely to modulate the immune system [31]. The low adhesion of the predominant bifidobacteria in the feces of allergic infants may thereby be related to the aberrant immune response in allergic infants. However, more detailed investigations should be performed to characterize the mucosal adhesion of endogenous bifidobacteria and their role in the etiology of allergic disorders.

The results from the current study suggest that specific species of more adhesive bifidobacteria, typically for healthy infants, may need to be administered to allergic infants in order to ‘normalize’ the intestinal Bifidobacterium flora. Merely increasing the level of colonization by bifidobacteria may not be enough, the levels of specific Bifidobacterium species may need to be altered.


Ms. Karita Peltonen, M.Sc., Research Scientist at the Department of Biochemistry and Food Chemistry, University of Turku, is thanked for skilful technical assistance in the identification of the bifidobacteria. Financial support was obtained from the Academy of Finland.