• Bifidobacterium;
  • inflammatory bowel disease;
  • anti-inflammatory activity;
  • nuclear factor-κ light chain enhancer of activated B cells


  1. Top of page

Certain Bifidobacterium strains have been shown to inhibit inflammatory responses in intestinal epithelial cells. However, the precise mechanisms of these effects, including the chemical nature of the active compounds, remain to be elucidated. Here partial characterization of the anti-inflammatory properties of Bifidobacterium strains isolated from feces of healthy infants is reported. It was found that conditioned media (CM) of all strains studied are capable of attenuating tumor necrosis factor-α (TNF-α) and lipopolysaccharide- (LPS) induced inflammatory responses in the HT-29 cell line. In contrast, neither killed bifidobacterial cells, nor cell-free extracts showed such activities. Further investigations resulted in attribution of this activity to heat-stable, non-lipophilic compound(s) resistant to protease and nuclease treatments and of molecular weight less than 3  kDa. The anti-inflammatory effects were dose- and time-dependent and associated with inhibition of IκB phosphorylation and nuclear factor-κ light chain enhancer of activated B cells (NF-κB)-dependent promoter activation. The combined treatments of cells with CMs and either LPS or TNF-α, but not with CMs alone, resulted in upregulation of transforming growth factor-β1, IκBζ, and p21CIP mRNAs. Our data suggest certain species-specificities of the anti-inflammatory properties of bifidobacteria. This observation should prompt additional validation studies using larger set of strains and employing the tools of comparative genomics.

List of Abbreviations: 

American Type Culture Collection




chemokine (C-C motif) ligand 20


cyclin-dependent kinase inhibitor 1A


cell-free extract

C. glutamicum

Corynebacterium glutamicum


conditioned medium


Dulbecco's modified Eagle medium

E. coli

Escherichia coli


glyceraldehyde 3-phosphate dehydrogenase


gastrointestinal tract








de Man-Rogosa-Sharpe medium


3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide


molecular weight


molecular weight cut-off


Nestlé Culture Collection


nuclear factor-κ light chain enhancer of activated B cells


reinforced clostridial medium


simian virus 40


transforming growth factor-β1


tumor necrosis factor-α

The genus Bifidobacterium is perhaps the most thoroughly studied taxon of the human mutualistic intestinal consortium. The representatives of this genus constitute an obligatory part of distal GIT microbiota in healthy adults and dominate this microbial community during early infancy (1–3). The intestinal bifidobacteria are believed to contribute to maintenance of intestinal homeostasis and host well-being by the digestion and bioconversion of dietary compounds, production of vitamins, competition with pathogens, inhibition of carcinogenesis and modulation of local and systemic immune responses (4–8).

Bifidobacteria-based probiotics and synbiotics have been shown to alleviate gut inflammation and to induce remission in pouchitis and ulcerative colitis patients (9–11). The simplest in vitro model for intestinal inflammation involves treatment of epithelial cell lines of intestinal origin (HT-29, Caco2, etc.) with various pro-inflammatory stimuli, such as cytokines (TNF-α), bacterial cell wall components (LPS, flagellin, muramyl dipeptide) or intact bacterial cells. In particular, probiotic Bifidobacterium strains suppress the stimulus-dependent secretion by epithelial cells of IL-8, an important leukocyte chemokine (12, 13). Such inhibition most likely results from a decrease in NF-κB signal pathway activity, which in turn can be due to down-regulation of IκB phosphorylation and proteosomal degradation, decrease of nuclear translocation of NF-κB, or inhibition of pro-inflammatory ligand binding to its cellular receptor (14–16).

However, the exact molecular pathways and cellular targets associated with the anti-inflammatory activities of bifidobacteria have not yet been identified. In addition, the chemical nature of the active substance(s) and its/their subcellular localization remain to be established. Some data suggest the anti-inflammatory effect is due to secreted factors (17), whereas other results indicate this activity is associated with either intact bacterial cells or bacterial DNA (18, 19).

Furthermore, while many studies conclude that the differences in immunomodulating properties of bifidobacteria are strain-specific (13, 20, 21), other reports suggest the species-specific distribution of certain anti- and pro-inflammatory characteristics within this genus (22–24). Notably, studies on the anti-inflammatory activities of bifidobacteria have often been conducted using well-characterized commercial probiotic or collection strains belonging to a limited number of species characteristic for humans, or industrial strains of non-human origin, such as Bifidobacterium animalis subsp. lactis NCC362 (Bb12) (12, 14, 17, 25).

In this connection, in the present study we aimed to assess the anti-inflammatory properties of several recent, minimally-transferred, human infant fecal isolates belonging to species typically present in healthy infant microbiota. We also attempted to characterize the chemical nature of the active substances and their possible mechanisms of action.


  1. Top of page

Bacterial cultures and preparation of conditioned media

Bifidobacterium strains used in this work (B. bifidum Bif1, Bif2, Bif3, B. longum Lon4, Lon5, Lon6, B. catenulatum Cat7, Cat8, Cat9, B. breve Bre10, Bre11, and B. adolescentis Ado12) were isolated from the feces of nine clinically healthy infants (three boys and six girls) aged 8 to 16 months. Additionally several collection and type strains of bifidobacteria (B. bifidum ATCC 15696, B. bifidum B791, B. longum subsp. infantis ATCC 15697, B. longum NCC 2705, B. longum VMKB44) were included. Non-bifidobacterial strains C. glutamicum ATCC 13032, Lactococcus lactis subsp. cremoris MG 1363, Lactobacillus casei ATCC 334, and E. coli B were used for comparisons. E. coli XL1-Blue was used as a cloning host.

Bifidobacterium strains were grown anaerobically at 37°C using RCM medium (Oxoid, Cambridge, UK). Lactococcus lactis MG1363 and Lactobacillus casei ATCC334 were grown on MRS medium (Himedia, Mumbai, India) aerobically at 30°C, or in a microaerophilic atmosphere at 37°C, respectively. Luria Bertani medium was used to aerobically propagate E. coli strain B and C. glutamicum ATCC13032 at 37°C.

For the preparation of cell suspensions and CFEs to be used for mammalian cell treatments bacteria were grown overnight in broth media. Bacterial cell densities were adjusted to ∼109 cells/mL. The bacterial suspensions were then heated at 55°C for 30  min, centrifuged at 3000  g for 10  min, washed once with DMEM (Biolot, St. Petersburg, Russia) and resuspended in the same medium to the initial volume. Cell-free extracts were prepared from non-heated PBS-washed cells pelleted from 5  mL of broth by vigorous vortexing with 200  mg of acid-washed glass beads (0.1  mm diameter) and 0.25  mL PBS with subsequent centrifugation and collection of supernatant.

For preparation of the CM overnight broth cultures of bacteria were centrifuged at 3000  g for 10  min, washed with DMEM and resuspended in the initial volume of the same medium supplemented with 20  mM of 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid. The cultures were then incubated for 20  hr at 37°C. Cells were removed by centrifugation and the pH of the supernatants adjusted to pH 7.4 followed by 0.22  μm filter sterilization. The size of active CM constituents was characterized by ultrafiltration using a 3  kDa MWCO unit (Millipore, Billerico, MA, USA). The filtrate fractions were collected and used without further treatments, while the retentate fractions (∼10-fold concentrated) were reconstituted to the initial volume with fresh DMEM. Heat treatment of ultrafiltrates was performed at 100°C for 30  min. The solubility of the active substances in organic solvent was determined by extracting the CMs with a chloroform-methanol mixture (CM:chloroform:methanol  = 1:1:0.5) followed by evaporating the organic solvents at 45°C and resuspending the residuals in an initial volume of DMEM. To characterize the sensitivity of the active substance(s) to protease and nuclease, the CMs were incubated for 20  hr at 37°C with either 20  μg/mL of proteinase K (Merck, Darmstadt, Germany) or 10  μg/mL of DNase I/RNase A mixture (9:1, Sigma-Aldrich, St Louis, MO, USA). The enzymes were then removed by passing the CMs through disposable 9  kDa MWCO ultrafiltration units (Pierce, Rockford, IL, USA).

Propagation and treatment of mammalian cell cultures

Mammalian cell culture HT-29 (human colon adenocarcinoma) was propagated on DMEM (Biolot) medium. The medium were supplemented with 10% FCS (HyClone, Logan, UT, USA), penicillin G, 100  U/mL, and streptomycin,100  μg/mL. The cultures were incubated at 37°C in an atmosphere of 5% CO2. The HT-29 cells were seeded into 24-well plates and grown to the desired degree of confluency. Cell viability was assessed using trypan blue staining.

For cytokine induction experiments: HT-29 cells were grown to 100% confluency, washed once with Hank's balanced salt solution and overlaid with a 1:1 mix (unless indicated otherwise) of fresh complete DMEM medium and either intact bacterial CM, treated CM or its fraction, inactivated cell suspension or CFE reconstituted with DMEM to reach the original volume of bacterial culture. After 1  hr pre-incubation, the cells were activated by addition of human recombinant TNF-α (produced by authors, unpublished data), 15  ng/mL final concentration, LPS from E. coli O55:B5 at 10  ng/mL (Sigma-Aldrich, St Louis, MO, USA) or LPS from E. coli 0111:B4, Salmonella typhimurium (Sigma-Aldrich) at the same concentration. The cells were collected for qRT-PCR after 4  hr incubation unless indicated otherwise. The cell supernatants were collected for ELISA cytokine assays after 15  hr incubation.

The cytotoxicity of bifidobacterial CMs was assessed by the MTT test. Briefly, confluent HT-29 cultures in 96-well plates were starved for 24  hr in serum-free medium. After that, the medium was replaced with fresh DMEM/CM (1:1) mix with or without addition of LPS or TNF-α and incubation continued for 24  hr. The cells were stained with 0.5  mg/mL solution of MTT in PBS for 3  hr at 37°C. Finally, the MTT solution was removed and the cells lysed with DMSO. Dye absorbance was read at 595  nm.

Enzyme-linked immunosorbent assay

Concentrations of IL-8 and TNF-α in cell culture supernatants were assayed using human IL-8 and TNF-α ELISA kits (Cytokine, St. Petersburg, Russia) according to the manufacturer's instructions.

Quantitative reverse transcription polymerase chain reaction

Quantitative RT-PCR was used to determine the degree of expression of mRNAs coding for chemokines IL-8 and CCL20, anti-inflammatory cytokine TGF-β1, NF-κB inhibitors IκBα and IκBζ, and cell cycle regulator p21CIP (encoded by CDKN1A gene). Total RNA was isolated from HT-29 cells by using ZR RNA MiniPrep kit (Zymo Research, Irvine, CA, USA). First strand cDNA was synthesized using RevertAid Premium reverse transcriptase (Fermentas, Vilnius, Lithuania) and oligo-dT primer. Real-time PCR was set up in an ANK-32 thermal cycler (Syntol, Moscow, Russia) using a pre-made mastermix containing EvaGreen (Syntol) and appropriate primer pairs at a concentration of 200  nM (Table 1). The reactions were performed using the following program: initial denaturation 95°C 5  min, denaturation 94°C 20  s, annealing 60°C 20  s, extension 72°C 20  s, program duration 40 cycles. The specificity of reaction products was confirmed by melting temperature analysis (from 70°C to 95°C in 0.5°C/15  s increments). Quantification of target transcripts was done by the ΔΔC(t) method using GAPDH as a normalizing house-keeping gene.

Table 1.  Oligonucleotides used in the study
PrimerSequencePCR product size, bp

Western blotting

Cell lysates for Western blotting were prepared from 24-well plates overlaid with 100  μL of lysis buffer (0.15 M NaCl, 5  mM EDTA, 1% SDS, 10  mM Tris-HCl, pH 8.0) mixed 1:1 with gel-loading dye and heated at 100°C for 3  min. The lysates were electrophoresed in 12% SDS-PAGE gel followed by transfer onto Hybond P membrane (GE-Amersham, Piscataway, NJ, USA) using a semi-dry transfer apparatus. The membranes were blocked in a mix of 0.2% Tween-20 and 3% BSA in PBS and stained using goat polyclonal antibodies against GAPDH (dilution 1:1000, catalog number sc-20358, Santa Cruz, Santa Cruz, CA, USA) and phospho-IκB (dilution 1:1000, catalog number sc-7977, SantaCruz), and rabbit anti-goat secondary antibodies (dilution 1:5000, catalog number sc-2768, Santa Cruz). Blots were developed using an ECL kit (GE-Amersham) according to the manufacturer's instructions.

Transient transfection and luciferase assay

Molecular cloning was performed by standard techniques (26). Luciferase reporter plasmid pGL3-promoter-NFκB carrying NF-κB response element was constructed using two complementary oligonucleotides (see Table 1) annealed to give a double-stranded fragment bearing BglII- and KpnI-compatible sticky ends. Vector pGL3-promoter (Promega, Madison, WI, USA) was digested with the corresponding enzymes and ligated with the described insert. The resulting pGL3-promoter-NFκB construct was verified by DNA sequencing. The plasmid was propagated in E. coli XL-1Blue strain and purified using Qiagen Plasmid Midi Kit (Qiagen, Venlo, the Netherlands).

For luciferase reporter experiments, HT-29 cells were grown to 60% confluency and transfected with pGL3-promoter-NFkB plasmid using Metafectene Pro reagent (Biontex, Planegg, Germany). After the cells had reached 100% confluency (approximately 24  hr post-transfection), they were treated with bifidobacterial CM and O55:B5 LPS as described above. At different time points after the application of the stimuli the cells were lysed and assayed for luciferase expression using the Luciferase Assay System (Promega, Madison, WI, USA).

Statistical data analysis

All experiments were performed in three biological replicates and three technical replicates per experiment. The results are expressed as means  ± SD. Values were compared using Student's paired t-test and differences were considered significant where P<  0.05.


  1. Top of page

Conditioned media from infant fecal bifidobacteria inhibit tumor necrosis factor-α and lipopolysaccharide-induced secretion of interleukin-8 by HT-29 cells

The ELISA tests revealed drastically increased amounts of IL-8 upon 16  hr of either TNF-α, or E. coli 055:B5 LPS, or E. coli CM treatment. However, with the exception of Lactobacillus casei, none of the Gram-positive bacteria used in this study induced IL-8 secretion (Fig.  1a).


Figure 1. Effect of bacterial CM on basal and TNF-α- and LPS-stimulated secretion of IL-8 and TNF-α in HT-29 cell line as determined by ELISA of cell culture media.a, effect on basal IL-8 secretion; b, c, effect on IL-8 secretion by TNF-α and LPS-activated cells; d, effect on LPS-driven TNF-α production. Description of bacterial strains used to prepare the CMs is given in the Materials and Methods section. C. glut., C. glutamicum; neg., negative control (non-stimulated cells); pos., positive control (only TNF-α or LPS treatment); **, P  <0.01 versus pos. in panels b, c, d or versus. neg. in panel a; *, P <0,05 versus pos.

Download figure to PowerPoint

Preincubation of the cells with most of the bifidobacterial CMs significantly inhibited both TNF-α-dependent and E. coli 055:B5 LPS-dependent IL-8 production (Fig.  1b, c). Quite expectedly, the extent of these inhibitory effects varied substantially among the strains. For example, different strains inhibited LPS-dependent IL-8 secretion from 9% to 49% as compared to controls. The inhibition was also dependent on the type of pro-inflammatory stimulus used: whereas B. catenulatum strain Cat8 inhibited TNF-α-dependent IL-8 production by 62%, it reduced LPS-dependent production by only 34%. The anti-inflammatory effect was not restricted to bifidobacteria; CM from L. lactis exerted a weak but statistically significant effect on TNF-α-driven IL-8 secretion (Fig.  1c).

Notably, the CMs of all B. bifidum strains used except for Bif2 significantly inhibited LPS-driven IL-8 production. Unlike the other strains tested, the representatives of this species inhibited LPS- and TNF-α-induced responses to virtually the same extent, by an average of 35.2% and 35.6%, respectively, the calculated ratio being 0.97. We found the corresponding ratios to be 0.49, 0.48, 0.47, and 0.47 for the strains of B. longum, B. catenulatum, B. breve, and B. adolescentis (one strain), respectively.

We also noticed that, on average, freshly isolated bifidobacterial strains possessed slightly greater abilities to inhibit both LPS- and TNF-α-induced IL-8 responses than did the strains with a history of in vitro transfers (by 40.9% and 24.6% for the fresh isolates vs. 34.6% and 22.2% for the established strains). This difference was seen more clearly in another experiment, in which we found that CMs from freshly isolated, but not from laboratory or collection Bifidobacterium strains (exception: B. bifidum ATCC15696) or other Gram-positive bacteria, significantly inhibited LPS-driven TNF-α secretion by HT-29 cells (Fig.  1d).

Based on this initial screen, we selected four infant fecal bifidobacteria strains, B. bifidum Bif3, B. longum Lon4, B. breve Bre10, and B.catenulatum Cat7, for further investigation.

Anti-inflammatory activities of infant fecal bifidobacteria depend on low-molecular-weight secreted compound(s)

To characterize the nature of anti-inflammatory activity in bifidobacteria-derived CM we compared the inhibitory actions of the CM, heat-inactivated bifidobacterial cells, and cell-free extracts on IL-8 secretion. In addition, we investigated the properties of active compounds present in CMs such as molecular weight, sensitivity to protease and nuclease, heat stability, and extractability by a non-polar solvent.

We found that, unlike CMs, neither bacterial cells nor CFEs of four bifidobacterial strains belonging to different species (B. bifidum Bif3, B. longum Lon4, B. catenulatum Cat7, and B. breve Bre10) were able to inhibit TNF-α-dependent production of IL-8 (Fig.  2a). Moreover, as judged by ultrafiltration experiments, the anti-inflammatory effect of the CMs are associated with compound(s) of MW less than 3  kDa (Fig.  2b). As shown in Figure  2b, heat-treatment of ultrafiltrated CMs did not inactivate their anti-inflammatory activity. After chloroform/methanol extraction of the CM no activity was found in the organic solvent-soluble fraction for any of the strains (data not shown).


Figure 2. Characterization of chemical properties and sub-cellular localization of B. bifidum Bif3, B. longum Lon4, B. catenulatum Cat7, and B. breve Bre10 anti-inflammatory compound(s).a, effect of inactivated cells, CFE, and CM of bifidobacteria on IL-8 expression in TNF-α-challenged HT-29 cultures (ELISA assay); b, effect of filtrate and retentate fractions of CM, obtained after passing through 3  kDa MWCO membrane, on IL-8 expression in TNF-α-challenged HT-29 cells (ELISA assay); c,d, resistance of anti-inflammatory activity of B. bifidum Bif3, B. longum Lon4, and B. breve Bre10 CMs to proteinase K and Dnase/RNase treatment (IL-8 in HT-29 supernatants was assayed by ELISA in panel c, IL-8 mRNA concentrations were determined by qRT-PCR in panel d). filtrate heat., heated ultrafiltrate; neg, negative control (non-stimulated cells); TNF, positive control (only TNF-α treatment); **, P  <0,01 versus pos.; *, P <0,05 versus TNF.

Download figure to PowerPoint

Treatment of B. bifidum Bif3, B.longum Lon4, and B breve Bre10 -derived CMs with either proteinase K, or a DNase/RNase mix followed by ultrafiltration through a 9  kDa cut-off filter preserved their anti-inflammatory activity, indicating the active compound is neither a protein nor a nucleic acid (Fig.  2c, d). Interestingly, the anti-inflammatory effect of B. bifidum Bif3 CM actually increased after proteinase K treatment.

Cells treated with CM, TNF-α, or LPS alone, or with combinations of CM with either TNF-α or LPS, showed no reduction of mitochondrial enzyme activity as determined by MTT assay (data not shown). In addition, no increase in the percentage of dye accepting cells was detected when cells treated with CM in combination with LPS/TNF-α were subjected to trypan blue staining (data not shown), indicating bifidobacterial CM had no toxicity for the cultured cells used in this study.

Infant fecal Bifidobacterium CMs inhibit nuclear factor-κ light chain enhancer of activated B cells (NF-κB) pathway activation in a dose- and time-dependent manner

It has been reported previously that inhibition of inflammatory response in cell lines of intestinal origin by bifidobacteria is dependent on NF-κB pathway downregulation (27). To confirm these findings we analyzed NF-κB-driven gene expression by using a luciferase reporter assay. A recombinant plasmid was constructed carrying firefly luciferase gene under the control of SV40 basal promoter and an upstream NF-κB response element taken from IL-8 gene promoter. Upon E. coli 055:B5 LPS stimulation, HT-29 cells transfected with this plasmid expressed luciferase, the maximum enzyme activity being seen at 6  hr post stimulation (Fig.  3a). Cells preconditioned with B. breve Bre10 CM showed decreased expression of luciferase 6  hr after LPS treatment. Expression of IL-8 mRNA followed the pattern of luciferase reporter activity (Fig.  3b). However, amounts of IL-8 mRNA peaked at 4  hr after LPS treatment; we also observed the most significant inhibitory effect of B. longum Lon4, B. catenulatum Cat7, and B. breve Bre10-derived CMs at this time point (Fig.  3b). We observed similar inhibitory activities for ultrafiltrate, but not retentate, fractions of CMs (data not shown).


Figure 3. Characterization of time and dose dependence of Bifidobacterium CM anti-inflammatory effect on HT-29 cells.a, analysis of luciferase activity in cells transfected with NF-κB reporter construct after LPS and B. breve Bre10 CM treatment at various time points; b, qRT-PCR assay of IL-8 mRNA dynamics in HT-29 cells treated with TNF-α and B. longum Lon4, B. catenulatum Cat7, and B. breve Bre10 CM; c, dose-dependent inhibition of IL-8 production in TNF-α-treated HT-29 cells determined by serial dilutions of bifidobacterial CM in fresh DMEM (ELISA assay for IL-8); d, dose dependent inhibition of IL-8 mRNA expression in cells treated with LPS and B. catenulatum Cat7 CM (qRT-PCR). neg, negative control (non-stimulated cells); *, P  <0,05 versus LPS in panel a.

Download figure to PowerPoint

Next, we examined the dose-dependence of bifidobacterial CM inhibitory effect on TNF-α-and E. coli 055:B5 LPS-induced IL-8 production at both mRNA and protein levels. We found that B. longum Lon4, B. catenulatum Cat7, and B. breve Bre10-derived CMs are capable of inhibiting IL-8 secretion by HT-29 in a dose-dependent manner (Fig.  3c). Similarly, B. catenulatum Cat7-derived CM dose-dependently inhibited expression of IL-8 at the mRNA level (Fig.  3d). However, a statistically significant decrease in IL-8 production (as determined by ELISA and qRT-PCR) was seen at high CM concentrations only (dilution 1:2 to 1:8 corresponding to ∼5  × 108–1.25  × 108Bifidobacterium cells per well of 24-well plate).

To clarify the molecular mechanism of NF-κB pathway inhibition by bifidobacterial CM we analyzed the rates of IκB phosphorylation in HT-29 cells after LPS and TNF-α stimulation at various time points. Western blotting revealed that, compared to non-treated cells, LPS and TNF-α treated cells contained significantly increased amounts of phospho-IκB shortly after stimulation (Fig.  4a, b). Pre-treatment of HT-29 cell cultures with CMs originating from either B. bifidum Bif3, B. longum Lon4 or B. catenulatum Cat7 strains resulted in decreased IκB phosphorylation at 30  min after LPS stimulation, whereas at 2  hr the difference between CM-pretreated and non-pretreated cells was barely detectable (Fig.  4a). We obtained similar results for TNF-α-stimulated cells pretreated with bifidobacterial CMs with the exception of B. bifidum Bif3 strain, which did not affect IκB phosphorylation at the 30  min time point (Fig.  4b). The latter observation is as expected because B. bifidum Bif3 exerted a more profound inhibitory effect on IL-8 production by LPS-stimulated HT-29 cells than on TNF-α-stimulated cells (Fig.  1b).


Figure 4. Western blotting analysis of IκBα phosphorylation in HT-29 cells treated with Bifidobacterium CM and (a) LPS or (b) TNF-α. p-IκBα, blot developed with anti-phospho-IκBα antibodies; GAPDH, blot developed with anti-GAPDH antibodies; bif, lon, cat, bre, cultures treated with Bif3, Lon4, Cat7, and Bre10 Bifidobacterium strains, respectively. Time of incubation after application of the pro-inflammatory stimuli (LPS or TNF-α) is indicated at the bottom (mins).

Download figure to PowerPoint

Bifidobacterial conditioned medium modulate expression of inflammation- and apoptosis-related genes

In order to establish the effects on gene expression in the HT-29 cell line of bifidobacterial CMs alone or in combination with pro-inflammatory mediators we analyzed mRNA levels of several genes involved in inflammation and apoptosis. While treatment with bifidobacterial CMs alone affected none of the analyzed transcripts, the CMs were capable of modulating cellular transcriptional responses to LPS and TNF-α.

The greatest change we detected in the transcription level after induction with E. coli 055:B5 LPS and human TNF-α was for IL-8 gene. According to the qRT-PCR data all tested bifidobacterial CMs significantly inhibited stimulated IL-8 gene transcription (Fig.  5a). In a similar experiment with B. bifidum Bif3 CM-pretreated cells, substitution of E. coli 055:B5 LPS for either E. coli 0111:B4 or S. typhimurium LPS resulted in a nearly identical extent of inhibition of IL-8 expression (data not shown).


Figure 5. qRT-PCR analysis of gene expression at the mRNA level in HT-29 cells pre-treated with Bifidobacterium CM after LPS or TNF-α stimulation. bif3, lon4, cat7, bre10, cultures pre-treated with Bif3, Lon4, Cat7, and Bre10 Bifidobacterium strains, respectively; neg., negative control (non-stimulated cells); pos., positive control (only TNF-α or LPS treatment); **, P  <0,01 versus pos. in panels a, d, e or versus. neg. in panel f; *, P <0.05 versus pos. in panel a or versus neg. in panel f

Download figure to PowerPoint

Interestingly, another NF-κB-regulated cytokine, CCL20, the expression of which was upregulated at the mRNA level by LPS and TNF-α, did not respond to pretreatment with bifidobacterial CMs in either case (Fig.  5b). Similarly, Bifidobacterium CMs failed to inhibit LPS and TNF-α stimulated expression of the classical NF-κB inhibitor IκBα (Fig.  5c).

Expression of a non-classical NF-κB inhibitor, IκBζ, was induced only by LPS but not by TNF-α (Fig.  5d). Although bifidobacteria did not influence LPS-dependent IκBζ expression significantly, pre-incubation of the cells with B.breve Bre10 and B. bifidum Bif3 CMs followed by TNF-α treatment resulted in a 1.5- to 2-fold upregulation of IκBζ mRNA (Fig.  5d).

Transcription of the anti-inflammatory cytokine TGF-β1 was almost uniformly upregulated by all bifidobacterial CMs when combined with LPS or TNF-α stimulation. However, the CMs alone had no effect on TGF-β1 transcription (Fig.  5e).

A similar synergistic up-regulating effect of the inflammation mediators LPS and TNF-α and bifidobacterial CMs was observed for mRNA encoding for cyclin-dependent kinase inhibitor p21CIP (Fig.  5f). Expression of this transcript was barely affected by TNF-α alone and not affected at all by LPS or any of the four bifidobacterial CMs. However, the combined application of CM and LPS resulted in upregulation of p21CIP when either B. bifidum Bif3-, B. longum Lon4-, or B. catenulatum Cat7- derived CMs were used. Similarly, we observed p21CIP upregulation in cells treated by TNF-α in combination with CMs derived from either B. longum Lon4, or B. catenulatum Cat7 (Fig.  5f).


  1. Top of page

Indigenous microbial populations are involved in complex relationships with the host through interactions with epithelial cells, Paneth cells and gut-associated lymphoid tissue. Microbial symbionts contribute to the maintenance of GIT homeostasis through a variety of mechanisms including stimulation of mucus production in goblet cells (28), up-regulation of defensin synthesis in Paneth cells (29), promotion of plasmocyte differentiation (30), and strengthening the tight junctions between epithelial cells (31). Gnotobiological experiments have established the role of intestinal microbiota in maturation and maintenance of gut-associated lymphoid tissue (32), regulation of T-cell differentiation (33), and induction of several anti-inflammatory proteins, including Toll interacting protein, a TLR pathway inhibitor (34).

Multiple studies have shown that Bifidobacterium strains can inhibit pro-inflammatory signaling pathways in intestinal epithelial cell lines (35–37). In this study, we attempted to characterize the anti-inflammatory properties of 12 freshly isolated infant fecal Bifidobacterium strains belonging to several species.

To estimate the anti-inflammatory activities of bifidobacteria-CMs toward the colonocyte-derived HT-29 cell line we determined the amounts of IL-8 secreted.

We found that, unlike L. casei ATCC 334, C. glutamicum ATCC 13032 and E. coli B, the majority of infant fecal isolates and collection strains of bifidobacteria are able to inhibit TNF-α- and LPS-induced inflammatory responses in HT-29 colon epithelial cells. This activity is restricted to culture supernatants of bifidobacteria; we did not find it in inactivated washed cell suspensions or cell-free extracts. Similarly, previous reports have described unidentified metabolites in Bifidobacterium spent culture broth that were found to stimulate IL-10 production by dendritic cells (38), enhance the function of the intestinal epithelial barrier and reduce systemic inflammatory responses in mice (39). In contrast, other researchers observed anti-inflammatory effects with live and formalin-killed bifidobacterial cells (25, 40, 41).

Although the chemical nature of the anti-inflammatory compound(s) is still has to be determined our results indicate that the principal activity is due to a low molecular weight (less than 3  kDa) substance that is resistant to protease and nuclease treatment. In addition, we found this substance to be thermostable and non-lipophilic. These results are in agreement with previously reported properties of an anti-inflammatory factor in Bifidobacterium CM (17, 42). Ewaschuk et al. proposed that the anti-inflammatory effects of probiotic bacteria observed in tissue culture models might be associated with the ability to produce conjugated linoleic acid isomers (43). However, in our experiments the lipid fraction extracted from bifidobacteral CM had no effect on inflammatory responses in HT-29 cells.

Another important issue to address is the identification of molecular targets for bifidobacterial anti-inflammatory compounds within intestinal epithelial cells. A previous study suggested that signal pathways connected with both NF-κB and AP-1 transcription factors may be affected (14). Specifically, it has been reported that non-pathogenic intestinal Salmonella and Lactobacillus casei can inhibit NF-κB subunit p65 nuclear translocation (44, 45) while Bacteroides thetaiotaomicron stimulates peroxisome proliferator-activated receptorγ-dependent nuclear export of RelA subunit (46). Another group reported that bifidobacteria can inhibit phosphorylation and subsequent degradation of NF-κB inhibitor IκBα along with phosphorylation of the AP-1 pathway component p38-MAPK (14). In contrast, Petrof et al. found that the Bifidobacterium-containing probiotic mixture VSL#3 inhibits proteosomal degradation, but not phosphorylation of IκBα (15). More recently, it was also reported that the same probiotic formula is able to down-regulate the crucial inflammatory bowel disease (IBD)-related cytokine interferon gamma-induced protein-10 (IP-10) at a post-translational level by inhibiting its vesicular trafficking and secretion (47).

In our hands, the CMs produced by several Bifidobacterium strains showed dose- and time-dependent inhibition of LPS and TNF-α induced IL-8 expression at both protein and mRNA levels. These effects were accompanied by a decrease in the NF-κB-dependent expression of a reporter gene, indicating that bifidobacteria exert their anti-inflammatory action at least in part through inhibition of the NF-κB pathway. Indeed, our further experiments revealed that bifidobacterial CMs inhibit phosphorylation of IκBα in a time dependent manner.

Interestingly, three out of five strains of B. bifidum, and none of the representatives of the other Bifidobacterium species used in our study, inhibited LPS-induced IL-8 production more efficiently than that induced by TNF-α. This supports previous observations suggesting that certain immunomodulating properties of bifidobacteria might be species-specific (22, 23).

Analysis of the expression of selected genes in HT-29 cells co-stimulated with bifidobacterial CM and TNF-α or LPS supports the possibility that bifidobacterial CMs affect multiple signaling pathways rather than merely inhibit the activity of a sole transcription factor, NF-κB. If the latter is true then one could expect rather uniform down-regulation of all NF-κB-dependent genes. However, despite inhibiting IL-8 and TNF-α expression, Bifidobacterium CM did not affect the amounts of CCL20 and IκBα mRNA. By contrast, Sibartie and coworkers detected a decrease in the amounts of CCL20 mRNA in cells co-stimulated with B. infantis and various pro-inflammatory stimuli (41).

Further, we found the degree of expression of another NF-κB-dependent gene, a recently characterized NF-κB nuclear inhibitor of IκBζ (48), to be increased by certain bifidobacterial strains (Bif 3 and Bre10 in the presence of TNF-α). Another group previously obtained similar results (17).

Notably, the CMs of all Bifidobacterium strains used in this study significantly increased the expression of TGF-β in TNF-α- or LPS-treated cells. Being a pleiotrophic cytokine, TGF-β is known to participate in the regulation of such processes as apoptosis, angiogenesis, wound healing, and differentiation of Treg lymphocytes (49). Haller et al. proposed a possible role for TGFβ in the anti-inflammatory action of commensal gut bacteria (50).

Unexpectedly, we have detected that certain Bifidobacterium strains, in cooperation with TNF-α and LPS but not alone, were able to induce expression of CDKN1A gene, coding for p21CIP protein, a key inhibitor of cyclin dependent kinases. Overexpression of CDKN1A product may lead to cell cycle arrest and induction of apoptosis. Accordingly, Grimoud et al. reported that live bifidobacteria possess anti-proliferative effects in co-culture with an intestinal epithelial cell line (51).

In summary, our results indicate that so far uncharacterized, low molecular weight substance(s) secreted to the milieu most likely mediate the anti-inflammatory effects of bifidobacteria. These effects involve modulation of several signaling pathways in epithelial cells and are mostly strain-specific, but also exhibit limited species-specificity. However, these observations must be confirmed using more extensive strain collection and following possible changes in bifidobacteria genomes associated with in vitro passaging. Further efforts are also needed to identify the active compounds produced by bifidobacteria, as well as to locate the exact molecular targets in the affected signaling pathways.


  1. Top of page

The authors report that they have no financial relationships or interests to disclose.


  1. Top of page
  • 1
    Magne F., Hachelaf W., Suau A., Boudraa G., Mangin I., Touhami M., Bouziane-Nedjadi K., Pochart P. (2006) A longitudinal study of infant faecal microbiota during weaning. FEMS Microbiol Ecol 58:56371.
  • 2
    Penders J., Thijs C., Vink C., Stelma F.F., Snijders B., Kummeling I., van den Brandt P.A., Stobberingh E.E. (2006) Factors influencing the composition of the intestinal microbiota in early infancy. Pediatrics 118:51121.
  • 3
    Qin J., Li R., Raes J., Arumugam M., Burgdorf K.S., Manichanh C., Nielsen T., Pons N., Levenez F., Yamada T., Mende D.R., Li J., Xu J., Li S., Li D., Cao J., Wang B., Liang H., Zheng H., Xie Y., Tap J., Lepage P., Bertalan M., Batto J.M., Hansen T., Le Paslier D., Linneberg A., Nielsen H.B., Pelletier E., Renault P., Sicheritz-Ponten T., Turner K., Zhu H., Yu C., Li S., Jian M., Zhou Y., Li Y., Zhang X., Li S., Qin N., Yang H., Wang J., Brunak S., Doré J., Guarner F., Kristiansen K., Pedersen O., Parkhill J., Weissenbach J., MetaHIT Consortium, Bork P., Ehrlich S.D., Wang J. (2010) A human gut microbial gene catalogue established by metagenomic sequencing. Nature 464:.5965.
  • 4
    Deguchi Y., Morishita T., Mutai M. (1985) Comparative studies of water soluble vitamins among human species of bifidobacteria. Agric Biol Chem 49:139.
  • 5
    Kim J.E., Kim J.Y., Lee K.W., Lee H.J. (2007) Cancer chemopreventive effects of lactic acid bacteria. J Microbiol Biotechnol 17:122735.
  • 6
    Pompei A., Cordisco L., Amaretti A., Zanoni S., Matteuzzi D., Rossi M. (2007) Folate production by bifidobacteria as a potential probiotic property. Appl Environ Microbiol 73:17985.
  • 7
    de Vrese M., Schrezenmeir J. (2008) Probiotics, prebiotics, and synbiotics. Adv Biochem Eng Biotechnol 111:166.
  • 8
    Fukuda S., Toh H., Hase K., Oshima K., Nakanishi Y., Yoshimura K., Tobe T., Clarke J.M., Topping D.L., Suzuki T., Taylor T.D., Itoh K., Kikuchi J., Morita H., Hattori M., Ohno H. (2011) Bifidobacteria can protect from enteropathogenic infection through production of acetate. Nature 469:5437.
  • 9
    Mimura T., Rizzello F., Helwig U., Poggioli G., Schreiber S., Talbot I.C., Nicholls R.J., Gionchetti P., Campieri M., Kamm M.A. (2004) Once daily high dose probiotic therapy (VSL#3) for maintaining remission in recurrent or refractory pouchitis. Gut 53:10814.
  • 10
    Furrie E., Macfarlane S., Kennedy A., Cummings J.H., Walsh S.V., O’neil D.A., Macfarlane G.T. (2005) Synbiotic therapy (Bifidobacterium longum/Synergy 1) initiates resolution of inflammation in patients with active ulcerative colitis: a randomised controlled pilot trial. Gut 54:2429.
  • 11
    Miele E., Pascarella F., Giannetti E., Quaglietta L., Baldassano R.N., Staiano A. (2009) Effect of a probiotic preparation (VSL#3) on induction and maintenance of remission in children with ulcerative colitis. Am J Gastroenterol 104:43743.
  • 12
    O’Hara A.M., O’Regan P., Fanning A., O’Mahony C., Macsharry J., Lyons A., Bienenstock J., O’Mahony L., Shanahan F. (2006) Functional modulation of human intestinal epithelial cell responses by Bifidobacterium infantis and Lactobacillus salivarius. Immunology 118:20215.
  • 13
    Riedel C.U., Foata F., Philippe D., Adolfsson O., Eikmanns B.J., Blum S. (2006) Anti-inflammatory effects of bifidobacteria by inhibition of LPS-induced NF-kappaB activation. World J Gastroenterol 12:372935.
  • 14
    Heuvelin E., Lebreton C., Grangette C., Pot B., Cerf-Bensussan N., Heyman M. (2009) Mechanisms involved in alleviation of intestinal inflammation by Bifidobacterium breve soluble factors. PLoS One, No. DOI:
  • 15
    Petrof E.O., Kojima K., Ropeleski M.J., Musch M.W., Tao Y., De Simone C., Chang E.B. (2004) Probiotics inhibit nuclear factor-kappaB and induce heat shock proteins in colonic epithelial cells through proteasome inhibition. Gastroenterology 127:147487.
  • 16
    Ménard S., Candalh C., Bambou J.C., Terpend K., Cerf-Bensussan N., Heyman M. (2004) Lactic acid bacteria secrete metabolites retaining anti-inflammatory properties after intestinal transport. Gut 53:8218.
  • 17
    Imaoka A., Shima T., Kato K., Mizuno S., Uehara T., Matsumoto S., Setoyama H., Hara T., Umesaki Y. (2008) Anti-inflammatory activity of probiotic Bifidobacterium: enhancement of IL-10 production in peripheral blood mononuclear cells from ulcerative colitis patients and inhibition of IL-8 secretion in HT-29 cells. World J Gastroenterol 14:25116.
  • 18
    Frick J.S., Fink K., Kahl F., Niemiec M.J., Quitadamo M., Schenk K., Autenrieth I.B. (2007) Identification of commensal bacterial strains that modulate Yersinia enterocolitica and dextran sodium sulfate-induced inflammatory responses: implications for the development of probiotics. Infect Immun 75:34907.
  • 19
    Ghadimi D., Vrese M., Heller K.J., Schrezenmeir J. (2010) Effect of natural commensal-origin DNA on toll-like receptor 9 (TLR9) signaling cascade, chemokine IL-8 expression, and barrier integrity of polarized intestinal epithelial cells. Inflamm Bowel Dis 16:41027.
  • 20
    Medina M., Izquierdo E., Ennahar S., Sanz Y. (2007) Differential immunomodulatory properties of Bifidobacterium longum strains: relevance to probiotic selection and clinical applications. Clin Exp Immunol 150:5318.
  • 21
    Park S.Y., Ji G.E., Ko Y.T., Jung H.K., Ustunol Z., Pestka J.J. (1999) Potentiation of hydrogen peroxide, nitric oxide, and cytokine production in RAW 264.7 macrophage cells exposed to human and commercial isolates of Bifidobacterium. Int J Food Microbiol 46:23141.
  • 22
    He F., Morita H., Ouwehand A.C., Hosoda M., Hiramatsu M., Kurisaki J., Isolauri E., Benno Y., Salminen S. (2002) Stimulation of the secretion of pro-inflammatory cytokines by Bifidobacterium strains. Microbiol Immunol 46:7815.
  • 23
    López P., Gueimonde M., Margolles A., Suárez A. (2010) Distinct Bifidobacterium strains drive different immune responses in vitro. Int J Food Microbiol 138:15765.
  • 24
    Young S.L., Simon M.A., Baird M.A., Tannock G.W., Bibiloni R., Spencely K., Lane J.M., Fitzharris P., Crane J., Town I., Addo-Yobo E., Murray C.S., Woodcock A. (2004) Bifidobacterial species differentially affect expression of cell surface markers and cytokines of dendritic cells harvested from cord blood. Clin Diagn Lab Immunol 11:68690.
  • 25
    Preising J., Philippe D., Gleinser M., Wei H., Blum S., Eikmanns B.J., Niess J.H., Riedel C.U. (2010) Selection of bifidobacteria based on adhesion and anti-inflammatory capacity in vitro for amelioration of murine colitis. Appl Environ Microbiol 76:304851.
  • 26
    Sambrook J., Russell D.W. (2001) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor , NY : Cold Spring Harbor Laboratory Press.
  • 27
    Kim S.W., Kim H.M., Yang K.M., Kim S.A., Kim S.K., An M.J., Park J.J., Lee S.K., Kim T.I., Kim W.H., Cheon J.H. (2010) Bifidobacterium lactis inhibits NF-kappaB in intestinal epithelial cells and prevents acute colitis and colitis-associated colon cancer in mice. Inflamm Bowel Dis 16:151425.
  • 28
    Fukushima K., Sasaki I., Ogawa H., Naito H., Funayama Y., Matsuno S. (1999) Colonization of microflora in mice: mucosal defense against luminal bacteria. J Gastroenterol 34:5460.
  • 29
    Ayabe T., Satchell D.P., Wilson C.L., Parks W.C., Selsted M.E., Ouellette A.J. (2000) Secretion of microbicidal alpha-defensins by intestinal Paneth cells in response to bacteria. Nat Immunol 1:1138.
  • 30
    Macpherson A.J., Gatto D., Sainsbury E., Harriman G.R., Hengartner H., Zinkernagel R.M. (2000) A primitive T cell-independent mechanism of intestinal mucosal IgA responses to commensal bacteria. Science 288:22226.
  • 31
    Hooper L.V., Wong M.H., Thelin A., Hansson L., Falk P.G., Gordon J.I. (2001) Molecular analysis of commensal host-microbial relationships in the intestine. Science 291:8814.
  • 32
    Mueller C., Macpherson A.J. (2006) Layers of mutualism with commensal bacteria protect us from intestinal inflammation. Gut 55:27684.
  • 33
    Di Giacinto C., Marinaro M., Sanchez M., Strober W., Boirivant M. (2005) Probiotics ameliorate recurrent Th1-mediated murine colitis by inducing IL-10 and IL-10-dependent TGF-beta-bearing regulatory cells. J Immunol 174:323746.
  • 34
    Chowdhury S.R., King D.E., Willing B.P., Band M.R., Beever J.E., Lane A.B., Loor J.J., Marini J.C., Rund L.A., Schook L.B., Van Kessel A.G., Gaskins H.R. (2007) Transcriptome profiling of the small intestinal epithelium in germfree versus conventional piglets. BMC Genomics, No. DOI:
  • 35
    Otte J.M., Podolsky D.K. (2004) Functional modulation of enterocytes by Gram-positive and Gram-negative microorganisms. Am J Physiol Gastrointest Liver Physiol 286:61326.
  • 36
    Park M.S., Kim M.J., Ji G.E. (2007) Assessment of lipopolysaccharide-binding activity of Bifidobacterium and its relationship with cell surface hydrophobicity, autoaggregation, and inhibition of interleukin-8 production. J Microbiol Biotechnol 17:11206.
  • 37
    Roselli M., Finamore A., Britti M.S., Mengheri E. (2006) Probiotic bacteria Bifidobacterium animalis MB5 and Lactobacillus rhamnosus GG protect intestinal Caco-2 cells from the inflammation-associated response induced by enterotoxigenic Escherichia coli K88. Br J Nutr 95:117784.
  • 38
    Hoarau C., Lagaraine C., Martin L., Velge-Roussel F., Lebranchu Y. (2006) Supernatant of Bifidobacterium breve induces dendritic cell maturation, activation, and survival through a Toll-like receptor 2 pathway. J Allergy Clin Immunol 117:696702.
  • 39
    Ewaschuk J.B., Diaz H., Meddings L., Diederichs B., Dmytrash A., Backer J., Looijer-van Langen M., Madsen K.L. (2008) Secreted bioactive factors from Bifidobacterium infantis enhance epithelial cell barrier function. Am J Physiol Gastrointest Liver Physiol 295:G102534.
  • 40
    Bai A.P., Ouyang Q., Zhang W., Wang C.H., Li S.F. (2004) Probiotics inhibit TNF-alpha-induced interleukin-8 secretion of HT29 cells. World J Gastroenterol 10:4557.
  • 41
    Sibartie S., O’Hara A.M., Ryan J., Fanning A., O’Mahony J., O’Neill S., Sheil B., O’Mahony L., Shanahan F. (2009) Modulation of pathogen-induced CCL20 secretion from HT-29 human intestinal epithelial cells by commensal bacteria. BMC Immunol, No. DOI:
  • 42
    Broekaert I.J., Nanthakumar N.N., Walker W.A. (2007) Secreted probiotic factors ameliorate enteropathogenic infection in zinc-deficient human Caco-2 and T84 cell lines. Pediatr Res 62:13944.
  • 43
    Ewaschuk J.B., Walker J.W., Diaz H., Madsen K.L. (2006) Bioproduction of conjugated linoleic acid by probiotic bacteria occurs in vitro and in vivo in mice. J Nutr 136:14837.
  • 44
    Neish A.S., Gewirtz A.T., Zeng H., Young A.N., Hobert M.E., Karmali V., Rao A.S., Madara J.L. (2000) Prokaryotic regulation of epithelial responses by inhibition of IkappaB-alpha ubiquitination. Science 289:15603.
  • 45
    Tien M.T., Girardin S.E., Regnault B., Le Bourhis L., Dillies M.A., Coppée J.Y., Bourdet-Sicard R., Sansonetti P.J., Pédron T. (2006) Anti-inflammatory effect of Lactobacillus casei on Shigella-infected human intestinal epithelial cells. J Immunol 176:122837.
  • 46
    Kelly D., Campbell J.I., King T.P., Grant G., Jansson E.A., Coutts A.G., Pettersson S., Conway S. (2004) Commensal anaerobic gut bacteria attenuate inflammation by regulating nuclear-cytoplasmic shuttling of PPAR-gamma and RelA. Nat Immunol 5:10412.
  • 47
    Hoermannsperger G., Clavel T., Hoffmann M., Reiff C., Kelly D., Loh G., Blaut M., Hölzlwimmer G., Laschinger M., Haller D. (2009) Post-translational inhibition of IP-10 secretion in IEC by probiotic bacteria: impact on chronic inflammation. PLoS One, No. DOI:
  • 48
    Totzke G., Essmann F., Pohlmann S., Lindenblatt C., Jänicke R.U., Schulze-Osthoff K. (2006) A novel member of the IkappaB family, human IkappaB-zeta, inhibits transactivation of p65 and its DNA binding. J Biol Chem 281:1264554.
  • 49
    Prud’homme G.J. (2007) Pathobiology of transforming growth factor beta in cancer, fibrosis and immunologic disease, and therapeutic considerations. Lab Invest 87:107791.
  • 50
    Haller D., Bode C., Hammes W.P., Pfeifer A.M., Schiffrin E.J., Blum S. (2000) Non-pathogenic bacteria elicit a differential cytokine response by intestinal epithelial cell/leucocyte co-cultures. Gut 47:7987.
  • 51
    Grimoud J., Durand H., de Souza S., Monsan P., Ouarné F., Theodorou V., Roques C. (2010) In vitro screening of probiotics and synbiotics according to anti-inflammatory and anti-proliferative effects. Int J Food Microbiol 144:4250.