Many countries are faced with the challenge of maintaining good health and quality of life in an ageing population living longer than ever before. Physiological phenomena associated with ageing include impaired memory acquisition, decreased immune responses, increased peroxidation in vivo and the loss of bone density. Intestinal microbiota also displays changes with increasing age, for example, the presence of Bifidobacterium sp. in the elderly may decrease(Reference Mitsuoka and Hayakawa1). In addition, ageing compromises the intestinal mucosal immune responses in animals and human subjects(Reference Schmucker and Owen2). Consequences of this immune dysfunction include increases in the incidences of infectious diseases and in the associated rates of morbidity and mortality in the elderly(Reference Schmucker and Owen2).
The probiotic properties of lactic acid bacteria have garnered particular interest. Probiotics have been defined as ‘live micro-organisms which when administered in adequate amounts confer a health benefit to the host’ by the FAO/WHO. For example, some lactic acid bacteria and bifidobacteria are immunostimulatory in elderly humans and animals. It has been reported that the Lactobacillus casei strain Shirota activates immune responses in aged mice and ameliorates influenza viral infections(Reference Hori, Kiyoshima and Shida3). The term ‘immunobiotics’ has been proposed to identify bacteria that promote health by driving mucosal immune mechanisms(Reference Chancy4). Many studies of the immunomodulatory activity of lactic acid bacteria include both viable and non-viable micro-organisms(Reference Salminen, Ouwehand and Benno5). These and other studies have shown that the anti-inflammatory effects of probiotics are mediated by the bacterial DNA and that live micro-organisms are not required to attenuate murine experimental colitis(Reference Rachmilewitz, Katakura and Kaemeli6). Currently, little is known about the immunomodulatory activity of viable and non-viable lactic acid bacteria and their effect on the physiological changes associated with increasing age.
The senescence-accelerated mouse is a good model for studying the physiological phenomena associated with increasing age(Reference Takeda, Hosokawa and Takeshita7). Senescence-accelerated mice develop normally, and then start senescence at approximately 6 months of age and irreversible advancement of senescence, manifested by signs such as loss of hair and increased lordokyphosis. Eight senescence-accelerated mouse strains are senescence prone, each of which develops specific age-related conditions. The senescence-accelerated-prone mouse (SAMP)6 strain provides a model of senile osteoporosis because after 4–5 months of age, these mice exhibit decreasing bone mass and reduced bone remodelling(Reference Takeda8). Few reports address the physiological effects of administering lactic acid bacteria to senescence-accelerated mouse strains(Reference Matsumoto, Watanabe and Imaoka9). Previously, we reported that oral administration of Lactococcus lactis H61 to aged SAMP6 mice was associated with reduced bone-density loss, improved gross lesion and enhanced immune responses, compared with those of controls (without receiving strain H61)(Reference Kimoto-Nira, Suzuki and Kobayashi10).
L. lactis G50 modulates immune responses in young Balb/c mice(Reference Kimoto, Mizumachi and Okamoto11). We surmised that strain G50 could modulate immune responses in aged mice also, and lead to beneficial effects on the ageing process. The present study investigated the effects of orally administered strain G50 on phenotypic changes such as the immune responses and bone density of SAMP6 mice. Here, we focused on the aspects of intestinal immunity, such as IgA antibody production and numbers and types of intestinal bacteria, because oral agents directly affect intestinal function. For example, some lactic acid bacteria enhance faecal IgA levels(Reference Perdigon, Alvarez and de Macias12). We used heat-killed cells of strain G50 for oral administration to facilitate comparing the results with those of heat-killed cells of strain H61 in our previous study and because non-viable micro-organisms are thought to be as immunomodulatory as live organisms, as mentioned previously. Furthermore, we prolonged the G50 administration period in the present study, compared with that in our previous study using strain H61 in the hope of increasing any beneficial effects.
Materials and methods
Mice
Male, 1-month-old, SAMP6 mice (eighteen per experimental group) were purchased from SLC Japan (Shizuoka, Japan). Two to four mice were housed in each cage (27 × 15 × 10 mm) and maintained at 22 ± 2°C with a 24 h light/dark cycle. Mice were fed MM-3 diet (Funabashi Farm, Chiba, Japan), the composition of which has been described previously(Reference Kimoto-Nira, Suzuki and Kobayashi10). Food and water were provided ad libitum. Animal experiments followed the animal experimentation guidelines of our institute.
Preparation of bacterial cells
L. lactis ssp. lactis G50 was maintained by the International Patent Organism Depositary (FERM P-18 415, Tsukuba, Japan). The strain was cultured in M17 broth (Becton Dickinson, Sparks, MD, USA) supplemented with 0·5 % glucose by subculturing 1 % inocula for 18 h at 30°C. Heat-killed G50 cells were prepared as described previously(Reference Kimoto-Nira, Suzuki and Kobayashi10) and were added to MM-3 to a final concentration of 0·5 g/kg (w/w)(Reference Kimoto-Nira, Suzuki and Kobayashi10). The control mice received MM-3 without G50.
Quantification of IgA in faeces
Faeces were taken from 2, 3, 4, 7, 9 and 11-month-old mice. Faecal samples from two SAMP6 mice were pooled to obtain sufficient material for the analysis of the IgA levels. Pooled faeces were suspended in PBS (1·9 mm-NaH2PO4, 8·1 mm-Na2HPO4, 154 mm-NaCl (pH 7·2)) supplemented with 1 % fetal calf serum (Sigma Chemical Co., St Louis, MO, USA) to a final concentration of 0·1 g/ml (w/v) and vortexed until suspended. Supernatants were obtained by centrifugation at 10 000 g for 10 min. Total IgA levels in the supernatants were determined by ELISA using biotinylated rat anti-mouse IgA antibody (PharMingen, San Diego, CA, USA) and a streptavidin–horseradish peroxidase conjugate (PharMingen).
Detection of enteric bacteria in faeces
Faeces were obtained from 2-, 7- and 10-month-old control and G50-fed mice. Faeces from three or four mice of the same mice were pooled, mixed with sterile 0·85 % NaCl solution to a concentration of 0·1 g/ml (w/v) and vortexed until suspended. Serial dilutions of stool samples were streaked on to selective agar for Lactobacillus sp., Bacteroides sp., Bifidobacterium sp., Staphylococcus sp., Enterobacteriaceae sp., Enterococcus/Streptococcus sp. and H2S-producing bacteria(Reference Kimoto, Ohmomo and Nomura13). Plates for Bacteroides sp., Bifidobacterium sp. and Lactobacillus sp. were incubated at 37°C anaerobically; those for Staphylococcus sp., Enterococcus/Streptococcus sp., Enterobacteriaceae sp. and H2S-producing bacteria were incubated at 37°C aerobically. Colonies were counted after 24–48 h incubation.
Bone-density analysis
The right femurs of the 12-month-old G50-fed and control mice (eight per group) were removed and immersed in 70 % ethanol for 1 week. Bone density was determined by single-energy X-ray absorption (DCS-600R; Aloka, Tokyo, Japan), which measures bone mineral and bone area.
Statistical analysis
Statistical analyses were done using Statistical Analysis System software version 9.1 (SAS Institute, Cary, NC, USA). Bone mineral densities were analysed according to the general linear model procedure with one-way allocation. Changes in the body weight, faecal IgA level and viable count of intestinal bacteria with increasing age were analysed by the general linear model procedure, in which the diet was designated as the main plot, and age and diet × age interaction as the sub-plot; the error term for the main plot was animal or cage within the diet, whereas the error term for the sub-plot was the residual error. Comparisons of the least-squares means were analysed using the PDIFF option of the general linear model procedure. Data are expressed as means with their standard errors. P < 0·05 was considered statistically significant.
Results and discussion
Body weight
Before the start of the experiment, body weight did not differ between groups (the mean body weight of the mice ranged from 20·2 to 21·4 g). However, thereafter body weight (mean (se)) began to differ between the G50 and control animals. At 2 months of age, the mice fed G50 weighed 29·7 (se 0·59) g (n 18) and the control mice weighed 27·5 (se 0·59) g (n 18; P = 0·0087). Body weight at 7 months was 39·4 (se 0·67) g for the G50 group (n 15) and 34·6 (se 0·65) g for the control group (n 15; P < 0·0001). Among the 10-month-old mice, the body weight of the G50 group was 39·6 (se 0·80) g (n 10) and 35·7 (se 0·67) g for the control group (n 14; P = 0·0002). Body weight was 39·1 (se 1·72) g for the 12-month-old mice fed G50 (n 8) and 34·7 (se 1·35) g for their age-matched controls (n 13; P = 0·0634). The number of mice was changed during the experiment because three mice from each group were killed at 6 months of age for bone-density analysis. In addition, several mice died due to increased age.
Immune response
Faecal IgA levels (Fig. 1) were significantly (P = 0·0252) higher in the mice fed G50 than in the controls at 3 months of age (i.e. after the G50 animals had been eating the supplemented diet for 2 months), but thereafter no significant difference between the G50 and control mice was found. The mucosal immune responses to antigenic challenges reportedly decreases with increasing age(Reference Paganelli, Scala and Quinti14), as did IgA production in the present study. Therefore, the intestinal immunity of aged mice likely would be poorly susceptible to stimulation by oral administration of strain G50. A similar result occurred in the rats fed kefir milk fermented by using several bacteria and yeast. In that study, enhanced intestinal immunity occurred in young adult, but not old, rats(Reference Thoreux and Schmucker15).
Changes in total faecal IgA level with increasing age (control, ○; G50, ●). Values are the means of 5–9 independent trials, each with two mice, with their standard errors depicted by vertical bars. * Mean values were significantly different between the two groups on the corresponding age (P = 0·0252).
The immunomodulatory factors of lactic acid bacteria include the slime products produced by L. lactis ssp. cremoris, which are B-cell mitogens(Reference Kitazawa, Yamaguchi and Itoh16). Although strain G50 does not produce slime products, cell wall fractions prepared from the strain in our preliminary experiments stimulated cytokine production (such as IL-6) from a macrophage cell line (data not shown). IL-6 can enhance intestinal immunity by stimulating IgA antibody production(Reference Muraguchi, Hirano and Tang17). Various cell wall components, present in heat-killed cells, may be immunomodulatory factors of strain G50.
Effect on growth of enteric bacteria
In humans, the viable counts of Clostridium sp. Enterococcus sp., Lactobacillus sp. and Enterobacteriaceae sp. increase with age, and those of Bifidobacterium sp. decrease(Reference Mitsuoka and Hayakawa1). Table 1 shows that the viable counts of Lactobacillus sp., Bacteroides sp., Enterococcus/Streptococcus sp. and Enterobacteriaceae sp. did not differ with increasing age in either control or G50-fed mice; Bifidobacterium sp. was not ever detected in either group. In terms of the effect of diet, there was no difference in the viable counts of those bacteria between the control and G50-fed mice at any age. This result might reflect the use of heat-killed cells that are unable to multiply and cannot produce antimicrobial substances such as lactic acid or bacteriocin. However, viable H2S-producing bacteria were significantly (P = 0·0289) fewer in the G50 than control mice. The H2S-producing intestinal bacteria include Salmonella sp., Shigella sp. and Citrobacter sp., all of which are associated with various chronic diseases including inflammatory bowel disease and colorectal cancer(Reference Huycke and Gaskins18). Furthermore, heat-killed Lactobacillus acidophilus strain LB inhibits the adhesion of enteroinvasive pathogens to human intestinal Caco-2 cells(Reference Coconnier, Bernet and Kerneis19). Adhesion ability and hydrophobicity are correlated in some lactobacilli(Reference Del Re, Sgorbati and Miglioli20), and hydrophobicity increased when the bacterial cells were heat inactivated(Reference Collado, Surono and Meriluodo21). In addition, some lactic acid bacteria co-aggregate with pathogen(Reference Collado, Surono and Meriluodo21). Elucidating whether strain G50 adheres to the intestinal cells or whether it co-aggregates with pathogen in vitro will help clarify the mechanism through which the heat-killed cells of strain G50 inhibit the H2S-producing intestinal bacteria in vivo.
(Means with their standard errors for two independent trials, each with three or four mice)
Difference between values in the same diet or age column is NS.
* P = 0·0289, significance of effect.
Bone-density analysis
The heat-killed G50 cells were administered to 1-month-old SAMP6 mice for 11 months, after which their bone density was measured. Body density (mean (se)) was 39·7 (se 0·88) mg/cm2 for the 12-month-old mice fed G50, and 38·7 (se 0·88) mg/cm2 for their controls; these values do not differ significantly. Similarly, in our preliminary experiment, bone density did not differ between the 6-month-old control mice (n 3, 40·6 (se 0·91) mg/cm2) and their age-matched G50-fed cohort (n 3, 42·8 (se 0·91) mg/cm2), which had received the supplemented diet for 5 months.
In our previous study, oral administration of L. lactis ssp. cremoris H61 suppressed bone-density loss with increasing age(Reference Kimoto-Nira, Suzuki and Kobayashi10). However, in that study, compared with that of the controls, the greater mean body weight of the H61 group might account for their higher bone density, because bone density usually increases with increasing body weight(Reference Wildner, Peters and Raghuvanshi22). In the present study, to exclude a perceived effect of strain G50 on bone density because of the organism's effect on weight gain, we only compared mice whose body weights were similar to the mean value for their group. Alternatively, the differing effects of strains G50 and H61 on bone density may reflect species (ssp. lactis or ssp. cremoris) or strain specificity. Furthermore, because osteoporosis is associated with immune dysfunction(Reference Blair, Robinson and Zaidi23), the effects of strains G50 and H61 on the immune system should be studied.
In our previous study, oral administration of the heat-killed strain H61 reduced grading scores following senescence-associated changes such as hair loss, skin ulceration and altered immune responses such as enhancement of Th1-type immune response(Reference Kimoto-Nira, Suzuki and Kobayashi10). In the present study, the 12-month-old G50-fed mice showed some of the same manifestations of increasing age as did the control mice. We hypothesised that prolonged supplementation with strain G50 would have a stronger anti-ageing effect than that of strain H61, because strain G50 had higher immunomodulatory activity in vitro than did strain H61(Reference Kimoto, Mizumachi and Okamoto11). However, although administered to young (1 month old) through to aged (12 months old) mice in the present study, strain G50 lacked the anti-ageing effect of strain H61, which was administered to 5-month-old to aged (9 months old) mice. During the analysis of the mechanism of the anti-ageing effect of strain H61, G50 may be an effective negative control for that process.
Strain G50 enhanced the intestinal immunity of young mice and suppressed the growth of enteric H2S-producing bacteria. Strain G50 may be useful in ameliorating the disease caused by the intestinal bacterial pathogens. In addition, supplementation with strain G50 increased the body weight of SAMP6 mice. Similarly, oral administration of probiotic bacteria including lactic acid bacteria has a positive but poorly understood effect on body weight in the animal industry context(Reference Pollmann, Danielson and Wren24). Perhaps, strain G50 can help maintain the health in or suppress some diseases of young livestock, which often have decreased resistance to pathogens. The present study used the heat-killed cells of strain G50; as such, these organisms do not meet the accepted definition of probiotics. However, non-viable micro-organisms with beneficial effects on animal health, like strain G50, might also be useful as food supplements for humans or livestock.
Acknowledgements
The present study was supported in part by a grant for the project ‘Development of evaluation and management methods for supply of safe, reliable and functional food and farm procedure’ from the Japan Ministry of Agriculture, Forestry, and Fisheries. H. K.-N. designed the study and did the data analysis. K. M., T. O. and J. K. helped to write the manuscript. K. S. contributed to data analysis. The authors have no conflicts of interest to report.
