OUP user menu

Evidence of immunomodulatory effects of a novel probiotic, Bifidobacterium longum bv. infantis CCUG 52486

Jialu You, Parveen Yaqoob
DOI: http://dx.doi.org/10.1111/j.1574-695X.2012.01014.x 353-362 First published online: 1 December 2012

Abstract

Bifidobacterium longum bv. infantis CCUG 52486 was originally isolated from healthy elderly subjects and demonstrated to have particular ecological fitness and anti-pathogenic effects. Bifidobacteria are commonly associated with immuno-modulatory properties, especially in older people, but this strain has not been investigated for effects on immune function. This study aimed to explore the immunomodulatory effects of this novel probiotic, compared with three commercial strains, B. longum SP 07/3, Lactobacillus rhamnosus GG (L.GG) and Lactobacillus casei Shirota (LcS). Human peripheral blood mononuclear cells (PBMCs) were isolated from fasting blood of young or older volunteers and exposed to probiotic strains or Con A. NK activity and activation, and cytokine release was enhanced by all probiotics with strain specificities. The effect of B. infantis on NK activity was influenced by ageing. Except for L.GG, probiotics increased IFN-γ production to a much greater degree in young subjects and increased IL-6 production to a much greater degree in older subjects. Based on IL-10/IL-12 ratios, B. infantis resulted in the most anti-inflammatory profile of all of the probiotics. These results suggest that B. infantis CCUG 52486 has strong immunomodulatory potential compared with well-known commercial strains and that the immune response to probiotics may be influenced by ageing.

Keywords
  • ageing
  • cytokine
  • natural killer cell activity
  • probiotics
  • T lymphocytes

Introduction

Bifidobacterium longum subsp. infantis CCUG 52486 is a little characterized potential probiotic strain, which was originally identified in healthy elderly subjects with an independent life-style, free of chronic disease and aged 90 years or over (Silvi et al., 2003). The strain has shown promising in vitro antimicrobial activity as a growth inhibitor of the pathogen, Clostridium difficile, the main aetiological agent of pseudomembranous colitis and one of the major reasons for antibiotic-associated diarrhoea (Gougoulias et al., 2007). The bacteriocidal effect of B. infantis 52486 is suggested to be due to acidity resulting from secretion of lactic and acetic acids, to antibacterial components of a protein nature and inhibition of C. difficile toxin A release (Gougoulias, 2007). This strain also has high angiotensin-I-converting enzyme inhibitory activity, implicating potential anti-hypertensive activity (Gougoulias et al., 2011). However, there is currently no information on its effects on immune function, which is pertinent, because bifidobacteria and lactobacilli are well known for their immunomodulatory properties (Balejko et al., 2009; Lomax & Calder, 2009; Zhu et al., 2010), and several strains have been commercialized on this basis.

Lactobacillus casei Shirota (LcS) and L. rhamnosus GG (L.GG) are the two most widely tested Lactobacillus strains, and their effects on the immune system have been documented (Nagao et al., 2000; Shida et al., 2006; Balejko et al., 2009; He et al., 2009; Lomax & Calder, 2009; Dong et al., 2010). A number of reports demonstrate upregulation of NK cell activity by LcS in both in vitro models (Shida et al., 2006; Dong et al., 2010) and human studies (Lomax & Calder, 2009), particularly in individuals who have naturally low or declining NK (Nagao et al., 2000). We have recently demonstrated that LcS also selectively promotes activation of T cells and NKT cells in an in vitro model (Dong et al., 2010). There are some data to suggest that bifidobacteria may also enhance NK activity. For example, supplementation with B. lactis HN019 increased NK cell activity in several studies (Arunachalam et al., 2000; Chiang et al., 2000; Gill et al., 2001). However, there is little information on other strains.

There is growing evidence that probiotics modulate immune cell homoeostasis by altering cytokine profiles and that these effects are highly strain dependent (Haller et al., 2000a, b; Lammers et al., 2003; Shida et al., 2006, 2011; Marteau, 2011; Vissers et al., 2011). Furthermore, there appears to be a tendency for some probiotic strains to induce IL-12 production, promoting Th1 cell development, while others tend to induce IL10 production, promoting T regulatory cell development (Shida et al., 2011). As a result, strains have often been compared in terms of their IL-10/IL-12 ratios. LcS and L.GG are reported to be good inducers of pro-inflammatory Th1 cytokines (e.g. IL-1β, IFN-γ, IL-12), both in vitro and in vivo (Helwig et al., 2006; Maragkoudakis et al., 2006; Dong et al., 2010). LcS induces IL-12 more strongly than Lactobacillus plantarum ATCC 14917, Lactobacillus johnsonii JCM 2012 (Shida et al., 2011) and Lactobacillus acidophilus ATCC 4356 (Shida et al., 2006). Bifidobacteria, on the other hand, tend to induce IL-10 production; this has been demonstrated both ex vivo (Lomax & Calder, 2009) and in vitro (Lammers et al., 2003; Helwig et al., 2006; Foligne et al., 2007; Medina et al., 2007; Lopez et al., 2010). However, direct comparisons of the immunoregulatory effects of probiotic strains are relatively limited.

There is particular interest in the positive influence of probiotics in older people, who are subject to alteration in the composition of gut microbial communities [especially numbers of bifidobacteria (Woodmansey, 2007)], as well as immunosenescence (Ponnappan & Ponnappan, 2011). Although it is widely believed that probiotics are of particular benefit to immunocompromised individuals (e.g. older people; Nagao et al., 2000; Gill et al., 2001) and have recently been proposed as prime candidates for ‘anti-immunosenescence’ therapy (Candy et al., 2008), there is little information regarding their influence on those aspects of immunity that are particularly susceptible to immunosenescence, and there is little understanding of the mechanisms underlying these effects.

The aim of the present study was to investigate the immunomodulatory properties of B. infantis 52486 in comparison with three commercial strains (B. longum SP 07/3, L.GG and LcS) in an in vitro model using cells from young and older subjects to characterize strain-specific and age-dependent effects.

Materials and methods

Probiotic bacterial strain preparation

Stock strains (B. infantis 52486, B. longum SP 07/3, L.GG and LcS) were stored frozen at −80 °C in Microbank® mixed vials according to the manufacturer's instructions (ProLab diagnostics). After defrosting, the strains were grown in MRS agar plates (Oxoid Ltd, UK) at 37 °C under anaerobic conditions for 3 days. A single colony from each strain was then transferred to a hungate tube containing 10 mL MRS broth (Oxoid Ltd) with 0.05% l–cysteine hydrochloride (Sigma) and incubated for a further 24 h under the same conditions. Following this, 100 µL of the liquid culture was removed, added to a new MSR broth tube and grown at 37 °C in a shaking incubator (Cooled orbital Incubator, Gallenkamp, Loughborough, UK). Bacteria were harvested in the exponential phase and transferred to centrifuge tubes. After washing twice at 400 g for 10 min, the bacteria were resuspended in 1 mL RPMI 1640 medium (Lonza, UK) and diluted to a concentration of 1 × 107 cfu mL−1.

PBMC preparation

Eight young (23–30 years) and eight older (65–76 years) healthy volunteers were recruited into this in vitro study (gender ratio 1 : 1 in each age group). Exclusion criteria included diabetes requiring medication, asplenia and other acquired or congenital immunodeficiencies, any autoimmune disease, malignancy, cirrhosis, connective tissue diseases, current use of immunomodulating medication (including oral prednisone and inhaled steroids), self-reported symptoms of acute or recent infection (including use of antibiotics within last 3 months), taking lactulose or any other treatment for constipation, alcoholism and drug misuse (University of Reading Ethics Committee project ref 10/05). Fasting blood (approximately 90 mL) was collected from volunteers into sodium heparin-containing Vacutainer tubes (Greiner Bio-One Ltd, UK).

Fresh PBMC were isolated using Lympholyte separation medium (VH BIO Ltd, UK). Briefly, blood was layered over an equal volume of Lympholyte and centrifuged at 930 g for 15 min. PBMC at the interface were transferred to another centrifuge tube and washed once with PBS (Oxoid Ltd, U.K). The cells were resuspended in 5 mL RPMI 1640 medium and then layered over an equal volume of Lympholyte, centrifuged at 930 g and washed once more with PBS. The cells were finally resuspended in 1 mL RPMI medium, counted in a Coulter Z1 Cell Counter (Beckman Coulter Ltd, UK) and adjusted to a concentration of 2 × 106 cells mL−1.

PBMC culture

PBMC (2 × 106 cells mL−1) were incubated with bacteria (1 × 107 cfu mL−1), with addition of 50 µL of autologous plasma (final concentration 2.5%) and RPMI 1640 medium to give a final volume of 2 mL in each well. A negative control containing no bacteria and a positive control containing 5 µg mL−1 Concanavalin A (Con A; Sigma, UK) were also included. Cells were incubated in a 37 °C, 5% CO2 atmosphere for 24 h.

NK cell activity assay

K562 cells (target cells; a leukaemia cell line maintained in RPMI 1640 medium) were counted and assessed for viability using Trypan Blue (Sigma), before transferring 5 × 106 viable cells into 1 mL RPMI complete medium (CM). The K562 cells were labelled with CFDA (1/50 dilution from a stock solution, which was 5 mg of CFDA-SE dissolved in 1 mL of DMSO) and adjusted to a concentration of 5 × 104 cells mL−1 with CM.

PBMC (effector cells), which had been precultured in the presence or absence of probiotics or Con A, were washed and adjusted to 5 × 106 mL−1 with CM and assessed for viability using Trypan Blue. PBMC and K562 cells were then mixed at E/T cell ratios of 100 : 1, 50 : 1, 25 : 1 and 12.5 : 1. Inclusion of a K562 control enabled calculation of the % of dead target cells relative to the total. The cells were incubated for 2 h at 37 °C, 5% CO2. Prior to analysis by flow cytometry, 1 mg mL−1 propidium iodide was added and the samples incubated for a further 5 min at 4 °C. In addition, the activity of NK cells in fresh PBMC was measured (i.e. in the absence of pre-incubation with probiotic).

Samples were analysed by a FACS Canto II flow cytometer (BD Biosciences, Oxford, UK), using flowjo software. Results are expressed as percentage of target cell lysis.

Cytokine analysis

Supernatants were collected from 24-h-incubated PBMC cultures. In a preliminary screening exercise, four representative supernatants were subjected to a human cytokine proteome profile array, which screened for 36 cytokines and chemokines (R&D systems, UK). Cytokines and chemokines produced in significant quantities and demonstrating responsiveness to treatments were then fully analysed using CBA flow cytometric kits (BD Biosciences). Data analysis was carried out using fcap Array software.

PBMC staining and activation

For identification and characterization of NK cells, PBMC were stained with APC-Cy7-labelled anti-CD3, PerCy 5.5–labelled anti-CD8 and PE-Cy7-labelled anti-CD56, in conjunction with the activation markers, CD69 and CD25 (APC labelled). To assess T cell activation, 24-h-cultured PBMC were stained with APC-Cy7-labelled anti-CD3, FITC-labelled anti-CD4, PerCy 5.5–labelled anti-CD8, PE-Cy7-labelled anti-CD45RA, PE-labelled anti-CCR7 and one of the following activation markers (APC labelled): CD25, CD69, CD27, CD28, CD57, CD62L and CD152 (BD Biosciences). Stained cells were incubated at room temperature in the dark for 30–45 min, washed twice, resuspended in 500 µL of Fix solution and kept at 4 °C until analysis by flow cytometry.

Samples were analysed by a FACS Canto II flow cytometer (BD Biosciences), using flowjo 7.6 software. Results are presented as a percentage of cells positive to each surface marker in a specific cell group.

Statistical analysis

Mini Tab 16.0 was used to analyse data. Data were tested for normality, and non-normally distributed data were normalized by the Johnson Transformation. Significant differences were evaluated by two-way anova using the General Linear Model, followed by post hoc t-tests with Bonferroni correction. All data are shown as mean ± SE (standard error). The statistical significance level was set at P < 0.05.

Results

Probiotics stimulate NK activity

There was a significant effect of age (P < 0.05) and treatment (P < 0.001) on NK activity (Fig. 1; data shown for E/T ratio of 100 : 1 only). Ex vivo exposure to B. infantis 52486, B. longum SP 07/3 and LcS for 24 h significantly increased NK activity relative to control in both young and older subjects (Fig. 1). In contrast, L.GG enhanced NK activity only in the older subjects (Fig. 1). For the young subjects, NK activity after exposure to L.GG was significantly lower than that after exposure to the other three strains, but in the older subjects it was only lower than after exposure to B. longum SP 07/3 (Fig. 1). There were trends for higher NK activity in the young compared with the older subjects under all conditions, apart from exposure to L.GG; in the case of B. infantis 52486-treated PBMC, NK activity was significantly greater in the young compared with the older subjects (Fig. 1).

Figure 1

Effects of probiotics on NK activity. C: medium only; S1: B. infantis CCUG 52486; S2: Bifidobacterium longum SP 07/3; S3: L.GG; S4: LcS. Data are mean ± SE for n = 8 samples for each group (100 : 1 ratio only shown) and were normalized by the Johnson Transformation. There was a significant effect of age (P < 0.05) and treatment (P < 0.001) on NK activity (two-way anova). Significant differences are denoted as *P < 0.01 relative to C for young subjects; P < 0.05 at least relative to C for older subjects; ▲, P < 0.05 relative to all the other strains for young subjects; ▵, P < 0.05 relative to C and S2 for older subjects (one-way anova followed by post hoc Bonferroni tests).

Probiotics enhance NK cell activation

NK cell activation was assessed by the expression of the leucocyte activation antigen CD69 and CD25 on CD3CD56+ NK cells and CD8+/CD8 subsets. There was a significant effect of age (P < 0.01) and treatment (P < 0.01) on expression of CD25 on total CD3CD56+ cells and the CD8 subset, and of treatment (P < 0.01) on CD69 and CD25 expression in CD3CD56+ cells and CD8+/CD8 subsets.

All of the probiotics tended to promote CD69 and CD25 expression in both young and older subjects with little strain specificity. In the young subjects, NK cell activation was increased by B. longum SP 07/3 (CD69 and CD25) and LcS (CD25; Fig. 2). In the older subjects, B. longum SP 07/3 upregulated CD69 expression, but not CD25 expression (Fig. 2), while the other three strains increased CD25 expression, but not CD69 expression (Fig. 2).

Figure 2

Effects of probiotics on CD69 and CD25 expressions in CD3CD56+NK cells. C: medium only; S1: B. infantis CCUG 52486; S2: Bifidobacterium longum SP 07/3; S3: L.GG; S4: LcS. Data are mean ± SE for n = 8 samples from each group and were normalized by the Johnson Transformation. There was a significant effect of age (P < 0.01) on CD25 expression (b) and treatment (P < 0.01) on NK cell activation for both CD69 (a) and CD25 (b) (two-way anova). Significant differences are denoted as aP < 0.05, bP < 0.01 relative to the medium control for the same age group (post hoc t-test with Bonferroni correction). Significant age differences are denoted as *P < 0.05 for the same treatment (one-way anova followed by post hoc t-tests with Bonferroni correction).

When CD8+/CD8 subsets were analysed, B. longum SP 07/3 increased CD69 expression in the young subjects in both subsets, but had no effect in the older subjects (Fig. 3a and 3b). CD69 expression was not modified by the other three strains. For CD25, the effects of the probiotic on CD8 subsets were very similar to those on the total CD3CD56+ NK cell population (Fig. 4a and 4b).

Figure 3

Effects of probiotics on CD69 expression of CD8+/CD8 NK cell subsets. C: medium only, S1: B. infantis CCUG 52486; S2: Bifidobacterium longum SP 07/3; S3: L.GG; S4: LcS. Data are mean ± SE for n = 8 samples from each group and were normalized by the Johnson Transformation. There was a significant effect of age (P < 0.01) on expression of CD25 on the CD8 subset (d) and of treatment (P < 0.01) on CD8+ and CD8 NK cell subsets for expression of both CD69 (a, b) and CD25 (c, d) (two-way anova). Significant differences are denoted as aP < 0.05, bP < 0.01 relative to the medium control for the same age group; cP < 0.05 relative to B. longum SP 07/3 for the same age group (post hoc t-tests with Bonferroni correction).

Figure 4

Effects of probiotics on CD25 expression of CD8+/CD8 NK cell subsets. C: medium only, S1: B. infantis CCUG 52486; S2: Bifidobacterium longum SP 07/3; S3: L.GG; S4: LcS. Data are mean ± SE for n = 8 samples from each group and were normalized by the Johnson Transformation. There was a significant effect of age (P < 0.01) on expression of CD25 on the CD8 subset (d) and of treatment (P < 0.01) on CD8+ and CD8 NK cell subsets for expression of both CD69 (a, b) and CD25 (c, d) (two-way anova). Significant differences are denoted as aP < 0.05, bP < 0.01 relative to the medium control for the same age group; cP < 0.05 relative to B. longum SP 07/3 for the same age group (post hoc t-tests with Bonferroni correction). Significant age differences are denoted as *P < 0.05 for the same treatment (one-way anova followed by post hoc t tests with Bonferroni correction).

Particularly notable was the significantly greater activation of CD3CD56+CD8 cells from older subjects by B. infantis 52486 compared with both the control and B. longum SP 07/3 and with the young subjects (Fig. 4b). It is also notable that the probiotics consistently tended to increase CD25 expression in both the CD8+ and the CD8 subsets in young subjects, but only the CD8 subset in the older subjects (Fig. 4).

Probiotics induce cytokine and chemokine production

Following screening of 36 cytokines and chemokines using a Proteome Profile Array (data not shown), the effects of probiotics on selected cytokines and chemokines were further investigated. There was a significant effect of age on the production of IL-6 (P < 0.05), IFN-γ (P < 0.01) and RANTES (P < 0.001), and of probiotics (P < 0.001) on production of IL-1β, IFN-γ, IL-6, IL-10, GM-CSF and RANTES (Figs 5 and 6), but there were no effects of age or probiotics on production of IL-2, IL-4 or IL-5 (data not shown). Probiotics increased production of IL-8, IL-12 and TNF-α in both age groups, with little or no difference between strains and ages (data not shown).

Figure 5

Effects of probiotics on cytokine production. C1: medium only; C2: Con A; S1: B. infantis CCUG 52486; S2: Bifidobacterium longum SP 07/3; S3: L.GG; S4: LcS. Data are mean ± SE for n = 8 samples for each group and were normalized by the Johnson Transformation. There was a significant effect of age on production of IL-6 (P < 0.05), IFN-γ (P < 0.01) and of treatment (P < 0.001) on production of IL-1β, IFN-γ, IL-6, IL-10 (two-way anova). Significant differences are denoted as aP < 0.05 relative to the medium control for the same age group; bP < 0.05 relative to B. infantis 52486 for the same age group; cP < 0.05 relative to B. longum SP 07/3 for the same age group; dP < 0.05 relative to L.GG for the same age group; eP < 0.05 relative to all the other strains for the same age group (post hoc t-tests with Bonferroni correction). Significant age differences are denoted as *P < 0.05, **P < 0.01, ***P < 0.001 for the same treatment (one-way anova followed by post hoc t-tests with Bonferroni correction).

Figure 6

Effect of probiotics on GM-CSF and RANTES production. C1: medium only; C2: Con A; S1: B. infantis CCUG 52486; S2: Bifidobacterium longum SP 07/3; S3: L.GG; S4: LcS. Data are mean ± SE for n = 8 samples for each group and were normalized by the Johnson Transformation. There was a significant effect of age (P < 0.001) on production of RANTES and treatment (P < 0.001) on production of GM-CSF and RANTES (two-way anova). Significant differences are denoted as aP < 0.05, bP < 0.01, cP < 0.001 relative to the medium control for the same age group; dP < 0.01 relative to L.GG for the same age group; eP < 0.01 relative to LcS for the same age group (post hoc t-tests with Bonferroni correction). Significant age differences are denoted as *P < 0.05 for the same treatment (one-way anova followed by post hoc t-tests with Bonferroni correction).

All four probiotics induced the production of IL-1β (Fig. 5a) and GM-CSF (Fig. 6a), and all except L.GG induced the production of IL-10 (Fig. 5d) with no difference between the young and older subjects (Fig. 5a). All four probiotics also induced the production of IFN-γ, but the induction was substantially greater in the young subjects, except for L.GG, which induced a similar low level of IFN-γ production in both young and older subjects (Fig. 5b). The effects of the probiotics on IL-6 production were the reverse of those on IFN-γ production; all probiotics except for L.GG enhanced IL-6 production substantially more in the older subjects than the young subjects (Fig. 5c). RANTES was induced by all probiotics (no differences between strains) in the older subjects, but not the young subjects (Fig. 6b). In general, B. infantis 52486 was the strongest inducer of cytokine production, and L.GG was the weakest (Figs 5a5d and 6a).

Figure 6 shows the IL-10/IL-12 ratio relative to the medium control following exposure to probiotics. There was a significant effect of age (P < 0.001) and treatment (P < 0.001) on the ratio. In the young subjects, all probiotics except for B. infantis 52486 decreased the IL-10/IL-12 ratio (Fig. 7). However, in the older subjects, the Bifidobacterium strains tended to increase the ratio, while the Lactobacillus strains tended to decrease it; these effects were significant in the case of B. infantis 52486 and L.GG (Fig. 7). Overall, all four strains induced a higher IL-10/IL-12 ratio in the older subjects, with B. infantis 52486 having the greatest effect.

Figure 7

Effect of probiotics on ratio of key cytokines. C1: medium only; C2: Con A; S1: B. infantis CCUG 52486; S2: Bifidobacterium longum SP 07/3; S3: L.GG; S4: LcS. Data are mean ± SE for n = 8 samples for each group and are relative to the medium only control. There was a significant effect of age (P < 0.001) and treatment (P < 0.001) on the IL-10/IL-12 ratio (two-way anova). Significant differences are denoted as aP < 0.05, bP < 0.01 relative to the medium control for the same age group; cP < 0.05 relative to L.GG for the same age group; dP < 0.05 relative to LcS for the same age group; eP < 0.05 relative to all the other strains for the same age group (post hoc t-tests with Bonferroni correction). Significant age differences are denoted as *P < 0.05, **P < 0.01, ***P < 0.001 for the same treatment (one-way anova followed by post hoc t-tests with Bonferroni correction).

Effect of probiotics on expression of surface markers on T lymphocytes

The effects of the four probiotic strains on markers of T cell activation (CD69 and CD25) and markers for generation and long-term maintenance of T cell immunity (CD27), transmitting stimulatory (CD28) or inhibitory (CD152) signals for T cells, homing ability (CD62L) and dysfunction of T cells (CD57) were investigated on CD4+, CD8+ and naïve/memory T cell subsets.

There was no effect of any of the probiotics on the expression of CD69, CD27, CD28, CD57 or CD62L (data not shown).

However, compared with the unstimulated PBMCs, the percentage of CD25+CD8+ naïve T cells was increased by B. infantis 52486 in the older subjects (P < 0.05; data not shown). CD152 expression in naïve and memory T cells tended to be increased by all strains in the young subjects; however, this was statistically significant only for TCM cells in the case of B. infantis (data not shown).

Discussion

There is growing evidence that probiotics have immunomodulatory properties and that these properties of probiotics are strain dependent (Haller et al., 2000a, b; Lammers et al., 2003; Shida et al., 2006, 2011; Vissers et al., 2010, 2011; Marteau, 2011). This study explored the immunomodulatory potential of a novel probiotic compared with three commercial strains, B. longum SP 07/3, L. rhamnosus GG (L.GG) and L. casei Shirota (LcS), using an in vitro cell model. There were some marked differences in the responses of ‘young’ and ‘older’ PBMCs to probiotics for some parameters of immune function. The results from the current study may be useful in extrapolating to in vivo studies and for making predictions about the potential clinical uses of probiotics.

Effects of probiotics on NK cell activity and activation

Preservation of NK cell activity is suggested to contribute to healthy ageing (Panda et al., 2009; Le Garff-Tavernier et al., 2010). In the current study, NK activity tended to be lower in the older subjects. All four probiotics enhanced NK activity. However, there was a significant influence of ageing for B. longum subsp. infantis CCUG 52486 only, whereby there was a greater enhancement of NK activity in the young compared with the older subjects. This suggests a greater benefit of this novel probiotic to young subjects. B. longum subsp. infantis CCUG 52486 was originally identified in healthy elderly subjects free of chronic disease on the basis of its promising in vitro inhibition of the growth and toxin release of the pathogen, Clostridium difficile, and its potential for counteracting antibiotic-associated diarrhoea (Gougoulias et al., 2007). It was therefore anticipated that it might have the greatest benefit in older subjects, which, in terms of NK cell activity, is clearly not the case. Interestingly, B. longum subsp. infantis CCUG 52486 increased the expression of CD25 by NK cells significantly more in older subjects compared with young, but it had no effect on expression of CD69 by NK cells. The CD25+ NK cell fraction exhibits a higher proliferation rate than does the CD69+ fraction, so that proliferative potential is indicated by CD25 expression, while CD69 expression is more indicative of cytotoxic potential (Clausen et al., 2003). Thus, the greater enhancement of NK cell activity by B. longum subsp. infantis CCUG 52486 in young subjects cannot be explained by these activation markers, which might have predicted greater benefit in older subjects. For the other three probiotics, as with NK cell activity, there was no difference in activation marker expression by NK cells in young vs. older subjects, suggesting that the age-specific effects on NK cell activity are unique to B. longum subsp. infantis CCUG 52486. The lack of strain-specific effects of the probiotics with respect to NK cell activation is somewhat consistent with previous studies (Haller et al., 2002; Dong et al., 2010; Pérez-Cano et al., 2010).

Within the NK cell population, CD8+ NK cells are more cytotoxic than their CD8 counterparts. Pérez-Cano et al. (2010) reported that the probiotics, L. salivarius and L. fermentum selectively increased activation of CD8+ NK cells relative to CD8 NK cells in vitro (Pérez-Cano et al., 2010). In the current study, all four probiotics tended to increase CD25 expression in both CD8+ and CD8 subsets from young subjects (although this was statistically significant only for B. longum SP 07/3 and LcS), whereas they selectively increased CD25 expression by the less cytotoxic CD8 subset in the older subjects (with the exception of B. longum SP 07/3). Once again, this highlights a stark difference in response between young and old subjects, which could explain inconsistencies in the literature relating to in vitro models of probiotic effects (where subject age is not always defined). The difference in response between young and older subjects is particularly notable for B. longum subsp. infantis CCUG 52486, which increased expression of CD25 in the CD8 subset significantly more in the older subjects than in the young subjects. It is documented that the CD8 NK cell population expands during ageing, whereas the CD8+ subset is relatively unaffected (McNerlan et al., 1998; Borrego et al., 1999). Thus, the greater expression of CD25 in the CD8 NK cell subset in older subjects may simply reflect the expansion of this subset.

Effects of probiotics on cytokine production

Several studies suggest that an in vitro PBMC model is useful as a screening tool to identify characteristic traits of probiotic strains and to select probiotic strains for clinical trials (Helwig et al., 2006; Foligne et al., 2007; Medina et al., 2007; Imaoka et al., 2008; Dong et al., 2010; Pérez-Cano et al., 2010). For example, probiotic strains displaying in vitro potential to induce higher levels of the anti-inflammatory cytokine, IL-10, and lower levels of the proinflammatory cytokine, IL-12, offered the best protection in a murine model of acute TNBS-colitis in vivo (Foligne et al., 2007). The limitation of the model is that it is not fully reflective of the in vivo situation, where intestinal bacteria interact with Peyer's patches, communicate with M cells resident in the epithelium and interact directly with dendritic cells, macrophages and T cells, and there are a number of discrepancies between in vitro and in vivo studies. Nevertheless, there is relatively good agreement regarding the ability of probiotics to enhance both NK cell activity and phagocytosis both in vitro and in vivo (Lomax & Calder, 2009). In the current study, the production of IFN-γ and IL-6 in response to three of the probiotics was dramatically different in the young vs. the older subjects, and the change in IL-10/IL-12 ratio in response to all four probiotics was significantly influenced by age.

For IFN-γ, all four probiotics induced production by PBMCs from young subjects, while induction of this cytokine by PBMCs from older subjects was much lower. L.GG had very little effect on either young or old PBMCs. The effects on the production of IL-6 were almost the reverse of this, with probiotics inducing IL-6 production significantly more in older subjects than in young, with the exception of L.GG, which again had little effect. CD8 engagement in NK cells and NK activity is involved in the production of IFN-γ (Spaggiari et al., 2003), and it is notable that the probiotics increased the activation of the CD8+ NK cell subset preferentially in the young subjects. However, monocytes may also have made a significant contribution to the secretion of IFN-γ in these cultures (Shida et al., 2006). Other in vitro studies have shown, as in the current study, that L.GG is a relatively poor inducer of IL-1β, IL-12, IL-10, IFN-γ and TNF production (Bousvaros et al., 2005; Helwig et al., 2006). Furthermore, L.GG supplementation appears to provide little benefit for inflammatory bowel diseases (Ghadimi et al., 2008). The influence of age on the response to probiotics is intriguing; the fact that induction of IFN-γ by Con A is almost identical in young and older subjects suggests that it is the response to probiotics that is different, not the capacity to produce IFN-γ. For IL-6, however, even Con A stimulates greater production in the older subjects, suggesting that this is an inherent effect of ageing in response to a stimulus (but apparently not L.GG).

Bifidobacteria are well documented to promote the production of IL-10, potentially playing an important role in immune system homoeostasis (Lammers et al., 2003; Helwig et al., 2006; Foligne et al., 2007; Medina et al., 2007; Lopez et al., 2010). In the current study, both bifidobacteria strains increased the IL-10/IL-12 ratio to a greater degree in older subjects than in young subjects. In contrast, the Lactobacillus strains decreased the IL-10/IL-12 ratio, and the decrease was more significant in the young subjects. Overall, this would suggest that probiotics, and bifidobacteria in particular, have greater benefits in terms of cytokine profile for older subjects. It could further be argued that the novel strain, B. longum subsp. infantis CCUG 52486, produced the most favourable cytokine profile, because it was the only strain which increased the IL-10/IL-12 ratio in both young and older subjects. This may have therapeutic implications, for example in inflammatory bowel disease, where the IL-10/IL-12 ratio is abnormally low. Consumption of a similar probiotic, B. infantis 35624, but not L. salivarius UCC4331, has been demonstrated to normalize the IL-10/IL-12 ratio, which was associated with a reduction in symptom scores (O'Mahony et al., 2005). Bifidobacteria have also been reported to attenuate NF-κB activation, the central transcription factor regulating inflammatory responses (Riedel et al., 2006).

The mechanisms by which probiotics alter cytokine profiles, and the influence of ageing on this modulation, remain elusive. Bacterial genomic DNA, cell wall structure and soluble components have all been explored as potential factors. It has been reported that the genomic DNA of L.GG and B. longum SP 07/3 modifies the secretion of IFN-γ and contributes more than 50% of the effects exerted with live strains (Ghadimi et al., 2008). Compared with lactobacilli, the favourable induction of IL-10 by bifidobacteria may be due to the high level of guanine cytosine contained in their DNA (Medina et al., 2007).

Effects of probiotics on T cells surface marker expression

In the current study, B. longum subsp. infantis 52486 increased the activation of naïve CD8+ T cells (CD25 expression) in the older subjects. Previous studies have shown increased CD25 expression in CD4+ T helper cells and CD8+ cytotoxic T cells by several Lactobacillus strains (Castellazzi et al., 2007; Pérez-Cano et al., 2010). Upon activation, naïve CD8+ T cells proliferate rapidly and can expand to an impressive number of effector cells (D'Souza et al., 2008). It is particularly interesting that B. longum subsp. infantis 52486 had this effect in the older, but not in the younger subjects. This increased activation might partly compensate for the decline in the CD8+ naive T cell pool during ageing. Overall, however, none of the probiotics affected the expression of CD25 and CD69 by T cell subsets. Studies with L.GG supernatant, with or without heat-inactivation, have demonstrated inhibition of in vitro CD25 and CD69 expression by T cells and L.GG-degraded casein suppressed T cell activation in healthy donors (Pessi et al., 2001). By contrast, one in vitro study reported increased CD25 and CD69 expression in both CD4+ helper T cells and CD8+ cytotoxic T cells by LcS (Dong et al., 2010). Other studies report increased CD25 and CD69 expression in T cells by other LAB strains or probiotic E. coli (Castellazzi et al., 2007; Pérez-Cano et al., 2010), while one study reported no effect of two Lactobacillus strains (L. johnsonii and L. sakei) on T helper cell activation (CD69/CD25), but slightly enhanced cytotoxic T cell activation (CD25; Haller et al., 2000a, b). Further research is required to clarify these inconsistencies.

In the current study, B. longum subsp. infantis 52486 augmented the expression of CD152 on TCM cells in the young subjects. CD152 counterbalances activatory co-stimulators, such as CD28, and is important for inhibiting T cell activation by either delivery of a negative signal or competitive antagonism of CD28:B7-mediated co-stimulation (Carreno et al., 2000). Effects of probiotics on CD152 expression have been little studied, but in vitro exposure to L. acidophilus has been shown to upregulate the expression of CD152 by CD4+ T cells in allergic patients (Christin, 2009). Allergic diseases are associated with over-activated T cells; consequently, increasing CD152 expression by probiotics is suggestive of a possible anti-allergic effect.

In conclusion, this study presents evidence for immunomodulatory potential of the novel probiotic strain, B. longum subsp. infantis 52486, and demonstrates that the immune response to probiotics is influenced by ageing.

Acknowledgements

This study was funded by the Biotechnology and Biological Sciences Research Council's Diet and Health Research Industry Club (BBSRC-DRINC), UK (grant number BB/H00470X/1).

Footnotes

  • Editor: Willem van Eden

Reference

View Abstract