International Journal of Food Microbiology 138 (2010) 157–165
`
`Contents lists available at ScienceDirect
`
`International Journal of Food Microbiology
`
`j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / i j f o o d m i c r o
`
`Distinct Bifidobacterium strains drive different immune responses in vitro
`Patricia López a,b, Miguel Gueimonde b, Abelardo Margolles b,⁎, Ana Suárez a
`a Department of Functional Biology, Immunology Area, University of Oviedo, C/ Julián Clavería s/n, 33006 Oviedo, Asturias, Spain
`b Department of Microbiology and Biochemistry of Dairy Products, Instituto de Productos Lácteos de Asturias (CSIC), Ctra. Infiesto s/n, 33300 Villaviciosa, Asturias, Spain
`
`a r t i c l e
`
`i n f o
`
`a b s t r a c t
`
`Article history:
`Received 3 October 2009
`Received in revised form 18 November 2009
`Accepted 19 December 2009
`
`Keywords:
`Bifidobacterium
`Probiotics
`Dendritic cells
`Cytokines
`T helper lymphocytes
`
`In this work we evaluated the specific immune activation properties of different Bifidobacterium strains,
`some of the most relevant intestinal microorganisms. To this end, we examined the in vitro effect of 12
`Bifidobacterium strains belonging to 4 different species, Bifidobacterium longum, Bifidobacterium breve, Bifi-
`dobacterium bifidum and Bifidobacterium animalis subsp.
`lactis, on the maturation pattern of human
`monocyte-derived dendritic cells (DCs), as well as in their ability to induce cytokine secretion. In addition,
`we determined peripheral blood mononuclear cell (PBMC) proliferation and cytokine expression after
`exposure to bacterial strains.
`All bifidobacteria tested were able to induce full DC maturation but showed differences in the levels of
`cytokine production, especially IL-12, IL-10, TNFα and IL-1β, suggesting that specific cytokine ratios could be
`used to predict the type of Th response that they may promote. In fact, analysis of cytokine production by
`PBMC showed that most of the tested B. animalis and B. longum strains induced the secretion of large
`amounts of IFNγ and TNFα, in agreement with the Th1 profile suggested by DC cytokine production.
`Remarkably, three of four B. bifidum strains induced poor secretion of these cytokines and significant
`amounts of IL-17, the main product of Th17 cells, in accordance with the high IL-1β/IL-12 ratio observed after
`DC stimulation.
`In conclusion, this work shows species and strain-specific immune effects of bifidobacteria and describes a
`valuable method for screening possible probiotic strains with different immunomodulatory properties.
`Notably, some B. bifidum strains seem to promote Th17 polarization, which could be useful for future
`probiotic applications.
`
`© 2009 Elsevier B.V. All rights reserved.
`
`1. Introduction
`
`Bifidobacterium sp. are one of the most relevant probiotic micro-
`organisms since they colonize the intestinal tract soon after birth and are
`present at high population levels in both infants and adults (Guarner
`and Malagelada, 2003). These bacteria have a long history of being safely
`consumed (Salminen et al., 1998). Although there is a lot of information
`about the healthy properties of probiotics (Kopp-Hoolihan, 2001;
`Marteau et al., 2001), it is not always based on proven evidence and
`little is known about the precise mechanisms of action by which such
`bacteria may exert their beneficial effects. The events underlying these
`healthy effects are now beginning to be understood mainly from in vitro
`studies of host intestinal epithelial cell or immune cell responses to
`probiotic strains. Thus, probiotics are known to beneficially modulate
`several host cell functions, the most prevalent of which are immune
`responses and intestinal barrier integrity.
`Host defense against foreign challenge is elicited by the innate and
`the acquired immune systems, that induce both the systemic and the
`
`⁎ Corresponding author. Tel.: +34 985892131; fax: +34 985892233.
`E-mail address: amargolles@ipla.csic.es (A. Margolles).
`
`0168-1605/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
`doi:10.1016/j.ijfoodmicro.2009.12.023
`
`mucosal immune responses. At the gut mucosal level, the innate
`immune response provides the first line of defense against pathogenic
`microorganisms but also provides the signals that instruct the adaptive
`immnune response (Macpherson et al., 2005). In this respect, dendritic
`cells (DCs) are considered as the link between innate and adaptive
`immunity. They are the most powerful antigen-presenting cells,
`promoting specific adaptive immune responses, but they also play a
`central role in the innate defense (Steinman and Hemmi, 2006). It has
`been shown that these antigen-presenting cells open the tight junctions
`between intestinal epithelial cells, send dendrites outside the epithe-
`lium and directly sample bacteria from the intestinal lumen (Rescigno
`et al., 2001). Contact with antigens or inflammatory stimuli can induce
`the maturation of DCs, accompanied by functional and phenotypic
`changes like the upregulation of costimulatory molecules and cytokine
`production (Joffre et al., 2009; Reis E Sousa, 2006), thus acquiring the
`ability to induce naive T cell proliferation and polarization towards Th1,
`Th2 or Th17 effector cells or, alternatively, to regulatory T cells (Zhu and
`Paul, 2008).
`A number of works indicated that commensal intestinal bacteria
`administered orally, such as probiotics, have the potential to modulate
`and regulate the immune response, at least in part, through their effects
`on intestinal mucosa DCs. In this respect, it has been described that
`
`Genome Ex. 1038
`Page 1 of 9
`
`

`

`158
`
`P. López et al. / International Journal of Food Microbiology 138 (2010) 157–165
`
`different lactic acid bacteria and bifidobacteria induced very different
`maturation and cytokine production patterns on DCs, even generating
`opposite T cell responses. Certain probiotics were demonstrated to
`induce IL-12 production by macrophages and DCs, thus promoting IFNγ
`secretion and inflammatory Th1 responses (Fujiwara et al., 2004;
`Mohamadzadeh et al., 2005; Pochard et al., 2005), whereas other reports
`have shown that probiotic-induced IL-10 levels play a main role limiting
`Th1-mediated proinflammatory responses (Hart et al., 2004; Zeuthen
`et al., 2006). Furthermore, recent studies showed that Th17 cells are
`essential for the first-line defense against intestinal infection, although
`little is known about the effects of probiotic bacteria on human Th17
`polarization, a not well-known phenomenon in which IL-1β and IL-23
`seem to play an important role (Annunziato et al., 2008; Boniface et al.,
`2008; Mills, 2008). Finally, DCs exposed to probiotic bacteria may
`acquire tolerogenic properties and then T cell activation could be biased
`towards the generation of regulatory T cells (Smits et al., 2005).
`Like other probiotic bacteria, it has been described that Bifidobac-
`terium sp. may present distinct immunomodulatory effects both in vitro
`(Baba et al., 2008; Hart et al., 2004; Latvala et al., 2008; Medina et al.,
`2007; Niers et al., 2007; Young et al., 2004; Zeuthen et al., 2006) and in
`vivo (Fujiwara et al., 2004; Isolauri et al., 2000; Marteau et al., 2001;
`Myllyluoma et al., 2005; Schiffrin et al., 1995). However, comparative
`studies on the immunological traits of different strains of bifidobacteria
`that could support a rationale selection of probiotic strains for specific
`immunomodulatory benefits are very limited. For this reason, the aim of
`this study was to determine and compare the specific immune
`activation properties of different Bifidobacterium strains in order to
`establish useful criteria for their evaluation and selection which could be
`applied for further possible biotechnological or clinical applications. To
`this end, we determined maturation and cytokine production of human
`monocyte-derived DCs after exposure to different bifidobacterial
`strains. In addition, since most Bifidobacterium species are commensal
`microorganisms usually present in the gut of adult individuals, and thus
`interacting with immune cells, we analyzed the pattern of cytokine
`production by PBMCs after bifidobacterial stimulation to estimate the
`type of Th profile that could be induced in healthy individuals by each
`bacterial strain.
`
`2. Materials and methods
`
`2.1. Bacterial strains and culture conditions
`
`0.25% L-cysteine (Sigma Chemical Co, St. Louis, MO) (MRSc) and
`incubated at 37 °C in anaerobic conditions (10% H2, 10% CO2 and 80% N2)
`in a Mac 500 chamber (Don Whitley Scientific, West Yorkshire, UK).
`Escherichia coli was grown in LB medium at 37 °C under aerobic
`conditions. For Lactococcus lactis MRSc was used and incubations were
`carried out at 30 °C under aerobic conditions. Planococcus antarcticus
`DSM 14505 was grown in medium 92 (trypticase soy broth 30.0 g, yeast
`extract 3.0 g, agar 15.0 g, distilled water 1,000 ml) at 4 °C.
`Overnight bacterial cultures were used to inoculate (1%) 50 ml of the
`corresponding fresh media. After incubation (24 h for E. coli and L. lactis,
`24–48 h for bifidobacteria and 10–15 days for P. antarcticus) cultures
`were harvested by centrifugation. Supernatants (25 ml for each strain)
`were dialysed using 2 kDa cut-off dialysis pipes (Sigma) against 46
`volumes of PBS (Oxoid LTD, Basingstoke, Hampshire, England).
`Subsequently, the dialysed material was concentrated using 5 kDa
`Vivaspin columns (Sartorius, Göttingen, Germany) down to a final
`volume of approximately 2.5 ml. Glycerol was added to a final
`concentration of 2% and the supernatants were quickly frozen in liquid
`nitrogen and stored at −80 °C until use. Protein concentration of the
`supernatants was determined with the BCA method. Bacterial pellets
`were washed three times in PBS buffer (Oxoid) and resuspended in 5 ml
`of the same buffer. Bacterial
`levels in the cell suspensions were
`determined by plate counting and cultures were killed by exposing
`them to UV light in a UV chamber (15 W, Selecta, Barcelona, Spain) for
`90 min. Plate counting was carried out after UV treatment to
`corroborate the absence of bacteria that are able to recover in the
`proper medium. UV treated bacterial suspensions were then distributed
`in single use aliquots, frozen in liquid N2 and stored at −80 °C until use.
`
`2.2. Isolation of PBMCs
`
`Human peripheral blood mononuclear cells (PBMCs) were obtained
`from standard buffy-coat preparations from routine blood donors
`(Asturian Blood Transfusion Center, Oviedo, Spain) by centrifugation
`over Ficoll-Hypaque gradients (Lymphoprep, Nycomed, Oslo, Norway).
`All blood donors (the number is specified in each figure legend) were
`healthy adult volunteers (58% male and 42% female) between 18–60 years
`(median age 43.34±12.04) without any pathology or treatment. Approval
`for this study was obtained from the Regional Ethics Committee for Clinical
`Investigation.
`
`The different bacterial strains used in this study are shown in Table 1.
`Bifidobacterial strains were grown in MRS medium (Difco, Becton,
`Dickinson and Company, Le Pont de Claix, France) supplemented with a
`
`2.3. Proliferation assays
`
`PBMCs proliferation was determined by quantifying [3H]thymi-
`dine incorporation to cultured cells. To quantify PBMC responses to
`
`Table 1
`Bacterial strains, origin and relevant phenotype.
`
`Strain
`
`B. animalis subsp. lactis BB-12
`B. animalis subsp. lactis IPLA 4549
`B. animalis subsp. lactis 4549dOx
`B. longum NCIMB8809
`B. longum BM 6/2
`B. longum IF 3/6
`B. bifidum IF 10/10
`B. bifidum A8
`B. bifidum L22
`B. bifidum LMG11041T
`B. breve B27
`B. breve LMG13208T
`L. lactis IL594
`E. coli LMG 2092T
`P. antarcticus DSM 14505
`
`Origin
`
`Intestine of adult
`Culture collection
`Laboratory strain
`Nursling stools
`Infant
`Infant
`Infant
`Dairy product
`Adult
`Breast-fed infant
`Infant
`Infant
`Cheese starter
`Urine
`Ponds in Antarctica
`
`Relevant characteristics
`
`Selected reference
`
`Known probiotic strain
`–
`Bile-resistant derivative
`Producer of antimicrobial compounds
`–
`–
`–
`–
`Mucin-degrading strain
`–
`–
`In vitro antimutagenic activity
`–
`–
`–
`
`Garrigues et al. (2005)
`Ruas-Madiedo et al. (2005)
`Ruas-Madiedo et al. (2005)
`O'Riordan and Fitzgerald (1998)
`This work
`This work
`This work
`This work
`Ruas-Madiedo et al. (2008)
`Ventura et al. (2006)
`This work
`Chalova et al. (2008)
`Chopin et al. (1984)
`–
`Reddy et al. (2002)
`
`NCIMB, The National Collection of Industrial and Marine Bacteria, Aberdeen, Scotland, United Kingdom; LMG/BCCM, Belgian co-ordinated collections of micro-organims; DSM,
`German Collection of Microorganisms and Cell Cultures; IPLA, IPLA Culture Collection; T, type strain.
`
`Genome Ex. 1038
`Page 2 of 9
`
`

`

`P. López et al. / International Journal of Food Microbiology 138 (2010) 157–165
`
`159
`
`different UV-killed bacterial strains, 2×104 PBMCs were incubated
`with bacteria at a PBMCs:bacteria ratio of 1:5. All cultures were
`performed in triplicate wells in 200 μl of complete medium in 96-well
`round-bottom microtiter plates (Costar, Cambridge, MA). After 4 days
`of culture, 100 µl of supernatants from these cultures was collected for
`cytokine determination and 100 µl of complete medium was added
`back to the wells plus 1 μCi/well [3H]thymidine. After 16 h cells were
`harvested onto glass fiber filters for counting cell-incorporated [3H]
`thymidine using standard scintillation techniques (Packard Instru-
`ments, Downers Grove, IL). Results were determined as a stimulation
`index (SI) which was calculated as the ratio between the mean counts
`per minute (cpm) values measured in stimulated triplicates and the
`mean cpm values measured in unstimulated triplicates.
`
`2.4. Generation of monocyte-derived DCs
`
`Monocytes were isolated from previously obtained PBMCs by
`negative selection using the Human Monocyte enrichment kit, accord-
`ing to the protocol provided by EasySep, StemCell Technologies
`(Canada). Purified monocytes were N95% CD14+. Independent experi-
`ments were performed with cells from different individuals. Immature
`DCs were obtained from isolated monocytes by standard procedures.
`Thus, monocytes were cultured for 7 days in the presence of
`recombinant human (rh) IL-4 (35 ng/ml) and rhGM-CSF (70 ng/ml)
`(R&D Systems, Abingdon, UK). At days 2 and 5, 0.5 ml of the medium
`was removed without disturbing the clusters of developing DC and
`0.5 ml of freshly made GM-CSF- and IL-4-containing medium was added
`to the wells, restoring the final volume in each well to 1 ml. To generate
`immature DCs, monocytes were cultured in 24-well plates at a
`concentration of 5×105 cells/ml at 37 °C and 5% carbon dioxide in
`complete RPMI medium (RPMI 1640 containing 2 mM L-glutamine and
`25 mM Hepes, Bio Whitaker, Verviers, Belgium, supplemented with 10%
`heat-inactivated fetal calf serum and the antibiotics streptomycin and
`ampicillin at 100 μg/ml). At day 7, immature DCs were recovered,
`washed and resuspended in complete RPMI medium at 5×105 cells/ml
`for subsequent maturation.
`
`2.5. Stimulation of monocyte-derived DCs with bifidobacteria
`
`To examine the effects of the different Bifidobacterium strains on DC
`maturation, different bifidobacterial strains, L. lactis IL594 or E. coli LMG
`2092T, killed by UV radiation, were added to immature DC cells at a DC:
`bacteria ratio of 1:10 in complete RPMI medium or their cell-free culture
`supernatants (10%). Parallel cultures were treated with either 1 µg/ml
`LPS from E. coli 0111:B4 (Sigma), as a positive control of maturation, or
`left untreated, as a negative control. After 48 h, supernatants from these
`cultures were collected, clarified by centrifugation and stored at −20 °C
`for cytokine analysis whereas DCs were harvested for phenotypic
`characterization.
`
`2.6. Cell surface phenotype expression
`
`Phenotypic studies of DCs were performed after two or three-color
`staining with the appropriate monoclonal antibody (mAb) using a
`FACScan flow cytometer (Becton Dickinson, BD Biosciences, San Diego,
`CA). Briefly, cells were stained directly with anti-CD86 fluorescein
`isothiocyanate (FITC, clone 2331), CD80 phycoerythrin (PE, clone
`L307.4) and HLA-DR (PE-Cy5, clone L243) mAb; anti-CD1a FITC (clone
`HI149) and CD40 PE (clone 5C3) mAb and with the corresponding
`isotype-matched conjugated irrelevant mAb as a negative control. All
`mAb were supplied by Becton Dickinson Pharmingen. Staining was
`performed for 30 min at 4 °C, and cells were washed twice in staining
`buffer and resuspended in PBS. A minimum of 10,000 cells were
`acquired and analyzed using the CellQuest software (BD Biosciences).
`The specific fluorescence intensity was quantified as the mean
`
`fluorescence intensity (MFI) calculated by subtracting the background
`of isotype-matched control staining from the total fluorescence.
`
`2.7. Cytokine determination
`
`Cytokine levels in cell culture supernatants were quantified by a
`multiplex immunoassay (cytometric bead array, CBA, BD) using
`FacsCalibur flow cytometer (BD). To determinate cytokine production
`of stimulated DCs the Human Inflamation kit for CBA was used (IL-8,
`IL-6, IL-1β, IL-10, TNFα and IL-12p70) whereas quantification of
`cytokine production by cultured PBMCs was performed using a Flex
`Set for CBA including IL-2, IL-4, IL-17, IL-10, TNFα and IFNγ. Analysis
`was carried out using CellQuest and CBA BD software. The detection
`limits were: IL-8: 3.6 pg/ml; IL-1β: 7.2 pg/ml; IL-6: 2.5 pg/ml; IL-10:
`3.3 pg/ml; TNFα: 3.7 pg/ml and IL-12p70: 1.9 pg/ml for the Human
`Inflamation kit and IL-2: 5.6 pg/ml; IL-4: 0.7 pg/ml; IL-17: 0.3 pg/ml;
`IL-10: 0.13 pg/ml; TNFα: 0.7 pg/ml and IFNγ: 0.8 pg/ml for the Flex
`Set.
`
`2.8. Statistical analysis
`
`The Kolmogorov–Smirnov test was used to assess the normal
`distribution of the data. Cytokine levels were not normally distributed
`and then differences between cell treatments were evaluated by a non
`parametric test. Due to the existence of genetically determined
`differences between individuals in cytokine production, Wilcoxon or
`Kendall W tests for related samples were used. Differences in
`proliferative responses and expression levels of maturation markers
`(MFI) among different bifidobacteria strains were evaluated by Kendall
`W test and t test for paired data. Results were represented by mean±
`standard deviation or by median and interquartil range. The SPSS 15.0
`statistical software package (SPSS Inc) was used for all determinations
`and a value of p b0.05 was considered significant.
`
`3. Results
`
`3.1. Differential activation of dendritic cells by Bifidobacterium sp
`
`The effect of bifidobacteria on in vitro DC function was evaluated by
`analyzing maturation and cytokine production of immature monocyte-
`derived DCs exposed during 48 h to twelve different Bifidobacterium
`strains, Escherichia coli LMG 2092T or Lactococcus lactis IL594, a widely
`used food microorganism isolated from a cheese starter. Results were
`compared with that of immature DCs cultured in medium alone or with
`LPS, habitually used to induce in vitro DC maturation. To determine the
`possible stimulating effect of molecules secreted by bifidobacteria,
`immature DCs were also treated with cell-free culture supernatants.
`The maturation pattern of DCs was assessed by flow cytometric
`analysis of surface marker expression (Fig. 1A), determining down-
`regulation of CD1a and upregulation of CD40, CD80, CD86 and HLA-DR
`upon stimulation with UV-killed bacteria (bacteria:cell ratio 10:1) or
`their cell-free culture supernatants (10%). Results obtained with LPS
`treatment were used as reference for maturation. Fig. 1B shows that all
`bacterial strains included in the study up-regulated the expression of
`HLA-DR and costimulatory molecules (p b0.05, t test for paired samples;
`n=6) at similar or even higher levels to those observed with LPS.
`However, significant differences can be observed on the expression
`levels of maturation markers among different bifidobacteria (CD1a:
`p=0.003; CD40: p=0.0003; CD80: p=1.30 10− 6; CD86: p=0.0004;
`HLA-DR: p=0.002, Kendall W test). Remarkably, two different strains
`of Bifidobacterium bifidum, IF 10/10 and LMG11041, showed significant
`differences in their ability to induce DCs maturation (p b0.05 for CD40,
`CD80, CD86 and HLA-DR, t test for paired samples; n=6). On the other
`hand, cell-free culture supernatants were, in general, poor inducers of
`DC maturation and did not show significant differences with respect to
`MRSc values (Fig. 1C).
`
`Genome Ex. 1038
`Page 3 of 9
`
`

`

`160
`
`P. López et al. / International Journal of Food Microbiology 138 (2010) 157–165
`
`Fig. 1. Phenotype and maturation status of monocyte-derived DC stimulated by Bifidobacterium strains or their supernatants. Immature DCs were cultured with lipopolysaccharide (LPS) as
`maturation control and E. coli, L. lactis IL594 or different bifidobacteria strains at a bacteria:cell ratio of 10:1. After 48 h expression of CD14, CD40, CD80, CD86 and HLA-DR was analyzed by
`flow cytometry. (A) Representative profile of DC before (immature DCs) and after 48 h of culture with a bifidobacteria strain (B. bifidum A8). Empty histograms show background staining
`with isotype control mAbs, and solid histograms represent the specific staining of the indicated cell surface marker in immature and Bifidobacterium-stimulated DC. Mean fluorescence
`intensities (MFI) obtained after subtracting background staining are provided in each histogram. (B) Median and interquartil range of MFI expression of each cell surface marker in DC
`stimulated with diverse strains of UV-killed bacteria or (C) with their cell-free culture supernatants during 48 h. MFI levels after LPS stimulation was used as positive control of maturation
`(discontinuous line). Shown are results of six independent experiments performed with different blood donors. Differences on the expression levels of maturation markers between
`Bifidobacterium-stimulated DC and RPMI-stimulated DC was tested by t test for paired samples, whereas differences among different bifidobacteria were performed by Kendall W test.
`
`In addition to costimulatory molecule expression, DC competence to
`generate specific immune responses depends on the cytokine produc-
`tion profile. Thus, we further quantified the amount of IL-8, IL-6, IL-1β,
`IL-10, TNFα and IL-12 present in culture supernatants of DCs exposed to
`these bacteria. Cytokine quantification was carried out after 48 h of
`culture by a multiplex immunoassay. Due to the weak effect of bacterial
`supernatants on DC maturation, cytokine production in these cultures
`was not determined. As can be observed in Fig. 2A, no cytokine
`production was detected in unexposed DCs, whereas LPS (from E. coli)
`
`treatment only produced significant amounts of IL-8 and IL-6. However,
`E. coli and the whole set of Bifidobacterium strains tested induced the
`production of all studied cytokines, but to different degrees. No relevant
`differences were observed in the production of IL-8 (p=0.207, Kendall
`test) and IL-6 (p=0.192), but, interestingly, independently of the
`donor, the variety of bifidobacterial strains differed substantially in their
`capacities to induce IL-1β (p = 0.041),
`IL-10 (p = 0.001), TNFα
`(p=0.008) and IL-12 (p=0.008). Given that several bacteria are potent
`inducers of both proinflammatory and immunosuppressor cytokines,
`
`Genome Ex. 1038
`Page 4 of 9
`
`

`

`P. López et al. / International Journal of Food Microbiology 138 (2010) 157–165
`
`161
`
`Fig. 2. DC cytokine profile induced by exposure to Bifidobacterium strains. Immature DCs were cultured with 12 bifidobacterial strains, E. coli, L. lactis or LPS (bacteria:cell ratio 5:1)
`and the supernatants were recollected after 48 h for measuring the production of IL-8, IL-6, IL-1β, IL-10, TNFα and IL-12 by multiplex assay. Independent ratios were calculated from
`each donor. Data represent median and interquartil range of five independent experiments performed with different blood donors. Differences among different bacterial strains
`were evaluated by Kendall W test: IL-8, p =0.207; IL-6, p=0.192; IL-1β, p =0.041; IL-10, p =0.001; TNFα, p =0.008; IL-12, p=0.008; TNFα/IL-10, p=0.444; IL-10/IL-12,
`p =0.006; IL-1β/IL-12, p=0.022.
`
`in an attempt to predict
`i.e. E. coli and B. bifidum LMG11041,
`subsequently DC-mediated Th cell responses, we calculated ratios
`between cytokines that are relevant for T cell differentiation, for instance,
`TNFα/IL-10, IL-10/IL-12 and IL-1β/IL-12 (Fig. 2B). Furthermore, analysis
`of these cytokine relations showed important differences among bacterial
`strains. Thus, Bifidobacterium animalis subsp. lactis BB-12 and IPLA 4549,
`Bifidobacterium longum IF 3/6 and BM 6/2 and, to a lesser extent, B.
`bifidum LMG11041 exhibit high TNFα/IL-10 and low IL-10/IL-12 ratios in
`addition to an elevated IL-12 production. Therefore, it is reasonable to
`expect that DCs exposed to these strains may promote the induction of
`Th1 responses. An opposite cytokine profile was observed in the
`supernatant of DCs stimulated with Bifidobacterium breve LMG13208, in
`which the elevated IL-10/IL-12 ratio was suggestive of promoting a Th2
`polarization. Nevertheless, its low IL-12 production also results in an
`elevated IL-1β/IL-12 relation, indicative of a possible Th17 induction.
`Notably, L22, A8 and IF 10/10 strains of B. bifidum, in addition to E. coli,
`generated an elevated IL-1β/IL-12 ratio, suggestive of favouring a Th17
`polarization. Of note, DC cultured with B. bifidum IF 10/10 consistently
`exhibited a weak ability for cytokine production, in accordance with their
`low capacity for inducing DC maturation, whereas the LMG11041 strain of
`the same species stimulated DCs to produce high amounts of a wide
`spectrum of cytokines, also in line with DC maturation data. All these
`results indicate that Bifidobacterium are able to maturate and fully activate
`DCs, inducing a particular cytokine synthesis profile that is strain-
`dependent.
`
`3.2. Proliferation and cytokine production of peripheral blood mononuclear
`cells stimulated with bifidobacteria were strain-specific
`
`Most Bifidobacterium sp. are commensal microorganisms usually
`present in the gut of adult individuals and thus probably interact with
`
`immune cells. Therefore, to evaluate in vitro the possible consequences
`of these interactions, PBMCs were used as a responder population to
`determine cellular proliferation and cytokine production after bifido-
`bacteria stimulation. Thus, PBMCs isolated from healthy individuals
`were stimulated with UV-killed Bifidobacterium strains, E. coli, L. lactis or
`the non commensal non pathogenic psychrophilic bacteria P. antarc-
`ticus, used as a negative control for memory T cell responses (bacteria:
`cell ratio 5:1), as well as with their cell-free culture supernatants (10%).
`Proliferation and cytokine production were quantified in these cultures.
`Results showed that all bacterial strains were low inducers of
`lymphocyte proliferation (see Supplementary file 1), as was expected
`for an Ag-induced assay, and only a significant proliferative response
`was obtained with B. animalis subsp. lactis IPLA 4549 (p=0.029, t test
`for paired samples) and 4549dox (p=0.033) and with B. longum BM 6/2
`(p=0.048). PBMCs stimulated with cell-free bacterial culture super-
`natants did not show significant proliferative responses in any case
`(data not shown).
`Subsequently, to determine the pattern of cytokine expression
`induced by different Bifidobacterium strains, the amount of Th1 (IFNγ,
`TNFα and IL-2), Th2 (IL-4 and IL-10) and Th17 (IL-17) cytokines was
`determined in the supernatants of bacteria-stimulated PBMCs recov-
`ered after 4 days of culture, prior to 3[H]T addition to determine cellular
`proliferation. Due to the weak effect of bacterial supernatants, cytokines
`were not measured in these cultures. We found that, although the levels
`of produced cytokines varied among donors, the overall cytokine
`pattern for each bacterial strain seemed to be donor independent
`(Fig. 3). Undetectable or very low levels of IL-4 were produced by most
`cultures whereas all of them produced IL-2. All bacteria tested induced
`significant IL-10 secretion compared with the medium (p b0.05,
`Wilcoxon tests), but to varying degrees. However, the most notable
`differences were observed in the production of IL-17, IFNγ and TNFα.
`
`Genome Ex. 1038
`Page 5 of 9
`
`

`

`162
`
`P. López et al. / International Journal of Food Microbiology 138 (2010) 157–165
`
`Fig. 3. Profile of cytokine production by PBMCs stimulated with bifidobacteria. PBMCs were cultured in the absence or presence of LPS, and UV-killed E. coli, L. lactis or twelve
`Bifidobacterium strains at a bacteria:cell ratio of 5:1, and supernatants were recovered at day 4. IL-4, IL-2, IL-10, IL-17, IFNγ and TNFα cytokine levels were measured by multiplex
`assay (Flex set CBA). Bars represent median and interquartil range of five independent experiments performed with different blood donors. Differences were evaluated by Wilcoxon
`or Kendall W tests for related samples. Differences between medium and Bifidobacterium-stimulated PBMCs were evaluated by Wilcoxon test for related samples.
`
`Thus, compared with the medium, B. animalis subsp. lactis BB-12, IPLA
`4549 and 4549dOx and B. longum BM 6/2 and NCIMB8809 strains
`induced the secretion of large amounts of IFNγ (p=0.019, p=0.030,
`p=0.019, p=0.012, p=0.090, respectively) and TNFα (p=0.003,
`p = 0.009, p = 0.002, p = 0.0004, p b0.0001), respectively, but not of
`IL-17, suggesting a Th1 profile that was consistent with the cytokine
`pattern exhibited by DCs treated with the same bacteria. On the
`contrary, the strains of B. bifidum IF 10/10, A8 and L22 induced poor
`secretion of these cytokines but exhibited the highest capacities for
`IL-17 production (p = 0.020, p = 0.089, and p = 0.118, respectively),
`once again in accordance with data from DCs cultures (high IL-1β/IL-
`12 ratio).
`
`4. Discussion
`
`It is widely accepted that intestinal microbiota possesses immuno-
`modulatory capacity and plays an important role for the health of the
`host. Elucidation of the mechanisms by which intestinal microorgan-
`isms, including potential probiotics, modulate the immune system may
`facilitate implementation of probiotic supplements that are individually
`tailored for their immunoregulatory properties. In this regard, DC
`activation after bacterial stimulation is a pivotal factor for the generation
`of immune responses. These cells are found in immature status at
`different sites in the gastrointestinal tract. Contact with antigens or
`inflammatory stimuli
`induces their maturation, accompanied by
`functional and phenotypic changes such as upregulation of costimula-
`tory molecules and increase in cytokine production (Joffre et al., 2009).
`Moreover, the nature of the maturation agent may determine the
`pattern of cytokine production that, subsequently, will direct the
`polarization of naive T cells toward different subtypes of effector
`lymphocytes (Zhu and Paul, 2008). Thus, it appears reasonable to
`assume that intestinal probiotics could exert immunoregulatory effects
`by affecting the Th1, Th2, Th17 or regulatory T cell-promoting capacity
`of DCs in the gut.
`Bifidobacterium strains are among the most relevant probiotic and
`intestinal microorganisms. In the present study we examined the
`effect of 12 Bifidobacterium strains belonging to 4 different species on
`the activation pattern of human monocyte-derived DCs, demonstrat-
`ing their ability to upregulate phenotypic maturation markers and to
`induce cytokine production in a species and strain-specific manner.
`Moreover, study of cytokine production by PBMCs exposed to the
`different Bifidobacterium strains confirmed the specific profile of these
`probiotics in the polarization of the immune responses. A few works
`on the effects of bifidobacteria on human immunocompetent cells
`have been carried out previously (Boyle et al., 2006; Young et al.,
`2004) but, to the best of our knowledge, this is the first study
`
`analyzing immune responses on human DCs and PBMCs comparing a
`relatively large number of bifidobacterial strains from different
`species.
`In concordance with previous works (Baba et al., 2008; Hart et al.,
`2004; Young et al., 2004; Zeuthen et al., 2006), our results show that all
`bifidobacterial strains are able to induce full DC maturation, although
`significant differences can be observed on the expression levels of
`maturation markers among different strains. In contrast, cell-free
`culture supernatants were poor inducers of DC maturation, suggesting
`the absence or scarcity of relevant amounts of secreted immune
`activating molecules in the culture conditions used in this work. On the
`other hand, although different bifidobacteria elicited a common DC
`mature phenotype, important differences were observed in the pattern
`of cytokine production. In fact, bacterial strains could be categorized into
`different groups according to the amount of cytokines they triggered in
`DCs. We found that a common feature of all tested DC-stimuli, including
`the LPS used as a positive control for maturation, was their ability to
`significantly increase IL-6 and IL-8 production compared with non
`stimulated immature DCs. LPS treatment, h