Downloaded from
`Downloaded from
`Downloaded from
`
`http://aem.asm.org/
`http://aem.asm.org/
`http://aem.asm.org/
`
` on November 15, 2016 by guest
` on November 15, 2016 by guest
` on November 15, 2016 by guest
`
`APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Feb. 2008, p. 660–666
`0099-2240/08/$08.00⫹0 doi:10.1128/AEM.01261-07
`Copyright © 2008, American Society for Microbiology. All Rights Reserved.
`
`Vol. 74, No. 3
`
`Gnotobiotic Mouse Immune Response Induced by Bifidobacterium sp.
`Strains Isolated from Infants䌤
`Odile Me´nard,1 Marie-Jose´ Butel,1 Vale´rie Gaboriau-Routhiau,2 and Anne-Judith Waligora-Dupriet1*
`EA4065, Faculte´ des Sciences Pharmaceutiques et Biologiques, Universite´ Paris Descartes, 4 Avenue de l’Observatoire,
`75270 Paris Cedex 06, France,1 and Unite´ d’Ecologie et de Physiologie du Syste`me Digestif, Institut National de
`la Recherche Agronomique, Domaine de Vilvert, 78352 Jouy en Josas Cedex, France2
`
`Received 7 June 2007/Accepted 28 November 2007
`
`Bifidobacterium, which is a dominant genus in infants’ fecal flora and can be used as a probiotic, has shown
`beneficial effects in various pathologies, including allergic diseases, but its role in immunity has so far been
`little known. Numerous studies have shown the crucial role of the initial intestinal colonization in the
`development of the intestinal immune system, and bifidobacteria could play a major role in this process. For
`a better understanding of the effect of Bifidobacterium on the immune system, we aimed at determining the
`impact of Bifidobacterium on the T-helper 1 (TH1)/TH2 balance by using gnotobiotic mice. Germfree mice were
`inoculated with Bifidobacterium longum NCC2705, whose genome is sequenced, and with nine Bifidobacterium
`strains isolated from infants’ fecal flora. Five days after inoculation, mice were killed. Transforming growth
`factor ␤1 (TGF-␤1), interleukin-4 (IL-4), IL-10, and gamma interferon (IFN-␥) gene expressions in the ileum
`and IFN-␥, tumor necrosis factor alpha (TNF-␣), IL-10, IL-4, and IL-5 secretions by splenocytes cultivated for
`48 h with concanavalin A were quantified. Two Bifidobacterium species had no effect (B. adolescentis) or little
`effect (B. breve) on the immune system. Bifidobacterium bifidum, Bifidobacterium dentium, and one B. longum
`strain induced TH1 and TH2 cytokines at the systemic and intestinal levels. One B. longum strain induced a TH2
`orientation with high levels of IL-4 and IL-10, both secreted by splenocytes, and of TGF-␤ gene expression in
`the ileum. The other two strains induced TH1 orientations with high levels of IFN-␥ and TNF-␣ splenocyte
`secretions. Bifidobacterium’s capacity to stimulate immunity is species specific, but its influence on the orien-
`tation of the immune system is strain specific.
`
`Bifidobacteria are gram-positive, anaerobic bacteria that
`represent up to 90% of the total gut microflora in breast-fed
`babies (7) and up to 15% in adults (22). The presence of such
`high levels of bifidobacteria in the human intestine is suggested
`to contribute to human health, leading to the use of bifidobac-
`teria as probiotics. Indeed, beneficial effects of bifidobacteria
`are shown for various diseases, e.g., diarrhea associated with
`rotavirus or antibiotics and some inflammatory intestinal dis-
`eases (27). Though some studies suggest a role for bifidobac-
`teria in prevention of allergic diseases, data are scarce and
`controversial (2, 5). Moreover, the role of bifidobacteria in
`immunity has so far been little known.
`For a few years, the incidences of allergic diseases have been
`increasing in developed countries. A disturbance in the bal-
`ance of T-helper 1 (TH1)/TH2 lymphocyte responses to exog-
`enous antigens toward a TH2 phenotype is considered a major
`event in the onset of allergic diseases. Though the mechanisms
`of these disturbances are still controversial, several studies
`suggest a prominent role for microorganisms from the envi-
`ronment and the normal commensal flora of the gastrointes-
`tinal tract in these disregulations or in the prevention of aller-
`gic sensitization (15, 28). In fact, intestinal colonization, in
`
`* Corresponding author. Mailing address: EA4065 Ecosyste`me
`Intestinal, Probiotiques, Antibiotiques, Faculte´ des Sciences Phar-
`maceutiques et Biologiques, Universite´ Paris Descartes, 4 avenue
`de l’Observatoire, 75270 Paris Cedex 06, France. Phone: 33 (1) 53
`73 99 20. Fax: 33 (1) 53 73 99 23. E-mail: anne-judith.waligora
`@univ-paris5.fr.
`䌤 Published ahead of print on 14 December 2007.
`
`660
`
`which the earliest contact with microbes occurs, has a marked
`impact on the maturation of the immune system (IS), which is
`underdeveloped at birth (9). The natural sequential coloniza-
`tion process of the digestive tract is complex and was shown to
`crucially influence immune response establishment during the
`neonatal period, especially TH1/TH2 balance (8, 34, 35). Sev-
`eral clinical studies showed a relationship between allergic
`diseases and the gut microbiota, pointing out quantitative and
`qualitative differences in bifidobacterial colonization (4, 12, 26,
`37). These studies showed that allergic patients had lower
`counts of Bifidobacterium than healthy control subjects. Fur-
`thermore, differences in species were observed. Bifidobacte-
`rium adolescentis and Bifidobacterium longum were isolated
`from allergic infants as
`the predominant bifidobacteria,
`whereas the predominant ones isolated from age-matched
`healthy infants were Bifidobacterium infantis, Bifidobacterium
`bifidum, and Bifidobacterium breve (26). Although these data
`suggested a link between bifidobacterial species and atopy or
`tolerance, no clear relation of causes and effects has been
`demonstrated yet. An in vivo study of newborn mice showed
`that B. infantis establishment was important for restoration of
`the usual oral tolerance process and especially for immuno-
`globulin E suppression (34). The impact of bifidobacteria on
`immunity has been specified with in vitro studies showing the
`roles of individual bifidobacterial species (11, 21, 29, 39).
`Nonetheless, no in vivo study has confirmed such observations.
`Thus, it is essential, at a time when dramatic increases in the
`prevalences of allergic diseases are a major concern in western
`countries, to understand the connections between allergy and
`intestinal flora and more particularly between allergy and bi-
`
`Genome & Co. v. Univ. of Chicago, PGR2019-00002
`UNIV. CHICAGO EX. 2004 - 1/8
`
`

`

`VOL. 74, 2008
`
`IMPACT OF BIFIDOBACTERIUM ON IMMUNITY
`
`661
`
`Downloaded from
`
`http://aem.asm.org/
`
` on November 15, 2016 by guest
`
`FIG. 1. Levels of pro-TH1 cytokine secretion by splenocytes stimulated for 48 h with ConA (3 ␮g/␮l). (A) IFN-␥; (B) TNF-␣. Data obtained
`by ELISA are presented as means plus SEMs. P was ⬍0.05 (ⴱ) and ⬍0.01 (ⴱⴱ) for comparison with GF mice. Reference groups are represented
`by open bars.
`
`fidobacteria, which may be used as a probiotic for preventing
`these diseases (3, 5).
`In this study, using gnotobiotic mice and quantification of
`pro- and anti-inflammatory-cytokine production and/or gene
`expression, we aimed to assess the roles of bifidobacterial
`strains belonging to different species and isolated from the
`fecal flora of infants on TH1/TH2 balance.
`
`MATERIALS AND METHODS
`
`Bacterial strains. Ten Bifidobacterium strains were studied. Nine strains were
`isolated from the dominant fecal flora of 3- to 24-month-old healthy infants,
`including three B. longum strains (BS64, BS49, and CUETM 89-215), three B.
`breve strains (BS94, BS52, BS56), one B. bifidum strain (BS42), one B. dentium
`strain (BS21), and one B. adolescentis strain (BS27). The tenth strain was B.
`longum NCC2705, whose genome is entirely sequenced (31). Bifidobacterial
`genera were identified according to their morphological characteristics, by the
`presence of fructose-6-phospho-phosphoketolase, and by PCR according to Kok
`et al. (18). Species identification was realized using a multiplex PCR targeting the
`16S-23S rRNA gene intergenic spacer according to Mullie´ et al. (23). All strains
`were grown on Wilkins Chalgren agar base containing D-glucose (10 g/liter),
`L-cysteine (0.5 g/liter), and Tween 80 (0.5%, vol/vol) (6) and in trypticase-
`glucose-yeast extract-hemin broth. They were incubated at 37°C in an anaerobic
`chamber (dw Scientific, AES laboratoires, Bruzz, France).
`Experimental animals. Six- to 7-week-old female germfree (GF) C3H/HeN
`and conventional C3H/HeN mice were purchased from the INRA (Jouy-en-
`
`Josas, France). GF and conventional mice were maintained in sterile isolators
`(JCE Biotechnology, Hauterive, France) and fed ad libitum with a commercial
`rodent diet sterilized by gamma irradiation (40 megarads). The animal experi-
`mentation was conducted in accordance with the rules of our institution and with
`Council of Europe Guidelines, with license for experimental studies on living
`animals and animal facility agreement no. A750602 (Direction of Veterinary
`Services, Prefecture de Police de Paris, France).
`Twelve groups of 6 to 12 mice were studied. GF and conventional mice as well
`as mice which had been mono-associated with one of the 10 Bifidobacterium
`strains were included.
`Colonization of GF mice. Bifidobacterial mono-associated mice were obtained
`by gastric intubation of GF mice with a 48-h culture of one of the bifidobacterial
`strains comprising 106 to 108 CFU/ml. This range of doses did not influence the
`colonization level. Both GF and conventional mice received the same volume
`(300 ␮l) of sterile water. Feces were collected 48 h after force feeding to check
`the sterility in GF mice and the bacterial establishment in monobiotic mice.
`Cecal contents were collected after sacrifice in order to measure the bifidobac-
`terial levels in monobiotic mice, to determine the fecal microbiota compositions
`in conventional mice, and to check the sterility in GF mice. Dilutions of feces and
`cecal contents in a prereduced peptone liquid medium were performed in anaer-
`obiosis and spread on Wilkins Chalgren agar base for monobiotic mice and on
`various media allowing the isolation and the quantification of the main bacterial
`genera for conventional mice, as previously described (6).
`Lymphocyte culture. Six days after their inoculation, mice were sacrificed with
`an intraperitoneal injection of pentobarbital sodium (CEVA sante´ animale,
`Libourne, France). Spleens were gently crushed, filtered through a 70-␮m nylon
`filter (Falcon, VWR, Val de Fontenay, France), and rinsed in RPMI 1640
`
`Genome & Co. v. Univ. of Chicago, PGR2019-00002
`UNIV. CHICAGO EX. 2004 - 2/8
`
`

`

`662
`
`ME´NARD ET AL.
`
`APPL. ENVIRON. MICROBIOL.
`
`Downloaded from
`
`http://aem.asm.org/
`
` on November 15, 2016 by guest
`
`FIG. 2. Levels of pro-TH2 cytokine secretion by splenocytes stimulated for 48 h with ConA (3 ␮g/␮l). (A) IL-10; (B) IL-4. Data obtained by
`ELISA are presented as means plus SEMs. P was ⬍0.05 (ⴱ) and ⬍0.01 (ⴱⴱ) for comparison with GF mice. Reference groups are represented by
`open bars.
`
`(Gibco, Fisher-Bioblock, Illkirch, France). Spleen cells (SC) were then purified
`for 5 minutes in ice with Tris-buffered NH4Cl according to Nicaise et al. (25). SC
`were rinsed and resuspended in 1 ml of RPMI 1640 containing 25 mM of HEPES
`buffer, 1% of L-glutamine, 1% of penicillin, streptomycin, 1% of fungizon, and
`10% of fetal calf serum. The number of viable cells was determined by the trypan
`blue dye (0.25%) exclusion method. SC were adjusted to 2.106 cells/well and
`cultured in 24-well plates with and without 3 ␮g/␮l of concanavalin A (ConA)
`(Sigma, France) at 37°C in a 5% CO2-95% air atmosphere. Supernatants were
`collected after 48 h of culture by centrifugation at 1,800 rpm, 20°C, for 5 minutes
`and conserved at ⫺20°C.
`Cytokine level measurement. Levels of gamma interferon (IFN-␥), tumor
`necrosis factor alpha (TNF-␣), interleukin 10 (IL-10), IL-4, and IL-5 in culture
`supernatants were quantified using enzyme-linked immunosorbent assay
`(ELISA) kits (eBiosciences, Montrouge, France) according to the manufactur-
`er’s instructions. Detection limits for ELISAs were as follows: IL-4 and IL-5, 4
`pg/ml; TNF-␣, 8 pg/ml; and IFN-␥ and IL-10, 15 pg/ml. Duplicate wells were run
`for each sample. Results are presented as mean values plus standard errors of the
`means (SEMs) for each group.
`Cytokine expression from the terminal ileum. A total of 2.5 cm of the entire
`terminal ileum, including the Peyer patches, was crushed with an Ultra-Turrax
`J25 instrument (Fisher-Bioblock) for 40 seconds, and total mRNA was extracted
`using the TRIzol reagent method (Invitrogen, Illkirch, France) according to the
`manufacturer’s instructions. mRNA was treated by DNase I (Invitrogen) and
`converted into cDNA by reverse transcription using the Superscript II and Oligo
`dT12-18 primers (Invitrogen). The cDNA obtained was subjected to real-time
`PCR using Smart Cycler (Cepheid, Sunnyvale, Canada). The Quantitect Probe
`PCR master mix and Quantitect gene expression assay (Qiagen, Courtaboeuf,
`
`France) kits were used to quantify the transforming growth factor ␤1 (TGF-␤1)-
`and IL-4 gene expressions directly on mouse ileum. IFN-␥- and IL-10 gene
`expressions were quantified by using a Smart kit for Sybr green 1 (Eurogentec,
`Angers, France) and 900 nM specific primers (IFN-␥ forward [5⬘-AGCAACAG
`CAAGGCGAAAA-3⬘] and reverse [5⬘-CTGGACCTGTGGGTTGTTGA-3⬘]
`and IL-10 forward [5⬘-TTTGAATTCCCTGGGTGAGAA-3⬘] and reverse [5⬘-A
`CAGGGGAGAAATCGATGACA-3⬘]) (38). Dosages were performed in dupli-
`cate. As the efficacy of amplification for each gene was confirmed to be similar
`to that of the ␤-actin-encoding gene, which was used as a reference (Qiagen),
`data analysis was performed with the 2⫺⌬⌬Ct method as described by Livak and
`Schmittgen (20). For each group and each cytokine, gene expression level is
`determined by comparison with that in the GF mouse group.
`Statistical analysis. The Kruskal-Wallis test was used to determine the signif-
`icance of the differences between the groups, and the Mann-Whitney U test was
`performed for pairwise comparison. P values of less than 0.05 were considered to
`be statistically significant. Data were analyzed using SPSS (version 12.0).
`
`RESULTS
`
`Bacterial colonization level. GF mice were sterile before
`inoculation. After inoculation, bifidobacteria were established
`at a high level, with a mean of 9.5 ⫾ 0.8 log10 CFU/g of cecal
`content, ranging from 8.0 ⫾ 0.8 log10 CFU/g of cecal content
`for B. adolescentis to 10.3 ⫾ 0.3 log10 CFU/g of cecal content
`for B. breve BS56. No exogenous bifidobacteria were found.
`
`Genome & Co. v. Univ. of Chicago, PGR2019-00002
`UNIV. CHICAGO EX. 2004 - 3/8
`
`

`

`Downloaded from
`
`http://aem.asm.org/
`
` on November 15, 2016 by guest
`
`VOL. 74, 2008
`
`IMPACT OF BIFIDOBACTERIUM ON IMMUNITY
`
`663
`
`TABLE 1. Orientation of the IS in GF mice mono-associated with different Bifidobacterium strains
`
`Bifidobacterium strain
`
`Bacterial
`colonization
`levela
`
`Splenocyte secretion (peripheral IS)b
`
`Gene expression in final
`ileum (intestinal IS)b
`
`IL-10
`
`IL-4
`
`IL-5
`
`TNF-␣
`
`IFN-␥
`
`IFN-␥
`
`TGF-␤
`
`IL-4
`
`Orientation of TH1/TH2 balance
`according to our in vivo study
`
`B. adolescentis BS27
`B. breve BS94
`
`8.0 ⫾ 0.8
`10.3 ⫾ 0.3
`
`B. breve BS52
`
`B. breve BS56
`B. longum BS64
`
`9.9 ⫾ 0.9
`
`9.1 ⫾ 0.1
`9.6 ⫾ 0.2
`
`B. longum NCC2705
`
`9.8 ⫾ 0.1
`
`B. longum BS49
`
`9.9 ⫾ 0.1
`
`0c
`/4d
`
`0
`
`0
`0
`
`0
`
`0
`⫻11.5d
`
`⫻7d
`⫻11e
`
`0
`
`0
`
`0
`
`0
`
`0
`/2
`
`0
`
`0
`0
`
`0
`
`0
`
`B. longum CUETM
`89–215
`B. bifidum BS42
`B. dentium BS21
`
`10.0 ⫾ 0.1
`
`⫻4.5e
`
`⫻6.5e ⫻1.5
`
`9.1 ⫾ 0.2
`8.7 ⫾ 0.3
`
`0
`0
`
`⫻2e
`⫻3.5e
`
`0
`0
`
`0
`/2
`
`0
`
`0
`0
`
`⫻2e
`
`⫻2d
`
`⫻2e
`
`⫻1.5
`0
`
`0
`0
`
`0
`
`0
`/4d
`
`⫻3e
`
`⫻2d
`
`⫻7.5e
`
`⫻3.5e
`⫻13e
`
`0
`0
`
`0
`
`0
`0
`
`0
`
`/7d
`
`/5
`
`0
`0
`
`0
`⫻2d
`
`⫻5.5d
`
`⫻2.5d
`⫻3.5e
`
`⫻4e
`
`0
`
`⫻3e
`
`⫻2
`⫻3e
`
`0
`/5d
`
`0
`0
`
`0
`
`Noneffector
`Suppressor (peripheral IS)/TH1/
`TH2 (intestinal IS)
`/12d Non inducer (peripheral IS)/
`TH1/TH2 (intestinal IS)
`Low TH2 inducer
`TH2 (peripheral IS)/suppressor
`(intestinal IS)
`TH1 (peripheral IS)/suppressor
`(intestinal IS)
`TH1 (peripheral IS)/TH2
`(intestinal IS)
`TH1/TH2
`
`0
`
`0
`
`/8d
`0
`
`TH1/TH2
`TH1/TH2
`
`a Expressed as log10 numbers of UFC/g cecal content.
`b Differences are expressed in comparison with the level for GF mice. ⫻n, upregulation by a factor of n; /n, downregulation by a factor of n.
`c No difference.
`d P ⬍ 0.05.
`e P ⬍ 0.01.
`
`The microbiota of conventional mice consisted of 8.2 log10
`CFU total aerobes/g of cecal content and 9.4 log10 CFU total
`anaerobes/g of cecal content, with numerous lactobacilli (8.7
`log10 CFU/g of cecal content) and no bifidobacteria.
`Cytokine production from splenocytes in vitro. Cytokine
`secretions by SC are shown in Fig. 1 and 2, and data are
`compiled in Table 1. Spontaneous secretions of IFN-␥, TNF-␣,
`IL-4, IL-5, and IL-10 were observed in the unstimulated
`splenocytes’ supernatants, but with rates below the detection/
`quantification threshold. Only the results obtained with SC
`stimulated with 3 ␮g/␮l of ConA are considered.
`TH1 cytokines. According to their capacities to influence
`IFN-␥ secretion, strains can be significantly separated into
`three groups in comparison with GF mice, i.e., inducers (three
`B. longum strains out of four, B. bifidum, and B. dentium),
`suppressors (the fourth B. longum strain), and strains with no
`effect (the three B. breve and B. adolescentis strains) (Fig. 1A).
`Various effects were observed for the B. longum strains: B.
`longum CUETM 89-215 was a higher inducer than B. longum
`NCC2705 and BS49, and B. longum BS64 was shown to be a
`suppressor.
`Three strains of B. longum significantly induced TNF-␣ se-
`cretion, and the other strains, regardless of species, had no
`effect in comparison with the level for GF mice (Fig. 1B).
`TH2 cytokines. According to their capacities to influence
`IL-4 secretion, two groups of strains can be observed: two B.
`longum strains, one B. breve strain, B. bifidum, and B. dentium
`significantly induced this secretion, and the other strains had
`no effect in comparison with the level for GF mice (Fig. 2B).
`No strains had a significant effect on IL-5 secretion.
`Regulatory cytokine: IL-10. Strains can be significantly sep-
`arated into three groups, i.e., inducers (two B. longum strains),
`a suppressor (BS94), and strains with no effect on IL-10 secre-
`tion (Fig. 2A).
`
`Cytokine gene expression in the terminal ileum. Cytokine
`gene expressions in the terminal ileum are shown in Fig. 3, and
`data are compiled in Table 1. IL-10 gene expression was below
`the threshold of detection. By contrast, TGF-␤1, IL-4, and
`IFN-␥ gene expressions were detected in conventional, GF,
`and mono-associated mice.
`TGF-␤1 gene expression. All the strains but three (the B.
`bifidum strain, the B. adolescentis strain, and B. longum BS49)
`induced TGF-␤1 gene expression compared with the level for
`GF mice (Fig. 3A).
`IFN-␥ gene expression. Only the establishment of a conven-
`tional flora induced a significant increase in IFN-␥ gene ex-
`pression compared with the level for GF mice (Fig. 3B). This
`expression was significantly inhibited only in mice mono-asso-
`ciated with B. longum BS49 (P ⬍ 0.05) and CUETM 89-215
`(P ⫽ 0.067) compared with the level for GF mice. By contrast,
`the two other B. longum strains (NCC2705 and BS64) did not
`exert such an effect.
`IL-4 gene expression. Two B. breve strains (BS94 and BS52)
`had suppressor effects on IL-4 gene expression, and the other
`strains, regardless of species, had no effect compared with the
`level for GF mice (Fig. 3C). Except for one mouse, B. bifidum
`significantly inhibited IL-4 gene expression compared with the
`level for GF mice.
`
`DISCUSSION
`
`This is the first time that the impacts of various Bifidobac-
`terium strains belonging to several species on immunity have
`been compared in an in vivo model, i.e., gnotobiotic mice. Our
`study has shown that Bifidobacterium’s capacity to stimulate
`immunity is species specific but its influence on the orientation
`of the IS operates in a strain-specific manner.
`In our study, we chose to include four strains of B. longum
`
`Genome & Co. v. Univ. of Chicago, PGR2019-00002
`UNIV. CHICAGO EX. 2004 - 4/8
`
`

`

`664
`
`ME´NARD ET AL.
`
`APPL. ENVIRON. MICROBIOL.
`
`Downloaded from
`
`http://aem.asm.org/
`
` on November 15, 2016 by guest
`
`FIG. 3. Increases (n-fold) in cytokine gene expression in the terminal ileum. (A) TGF-␤1; (B) IFN-␥; (C) IL-4. Results are expressed as
`individual data points, and bars represent means. The y axis represents the amount of the target gene relative to the level for GF mice, normalized
`with ␤-actin. P values were ⬍0.05 (ⴱ) and ⬍0.01 (ⴱⴱ) for comparison with GF mice.
`
`and three strains of B. breve among the 10 Bifidobacterium
`strains tested. Indeed, in a previous prospective study, these
`species were isolated as the predominant bifidobacteria in in-
`fants (personal data). In GF mice, Bifidobacterium coloniza-
`tion occurred at a high level, as high as that in the infants’ fecal
`flora from which the strains were isolated. To investigate the
`influence of bifidobacteria on TH1/TH2 orientation, IL-4 and
`
`IL-5 were studied as pro-TH2 cytokines, IFN-␥ and TNF-␣ as
`pro-TH1 cytokines, and IL-10 and TGF-␤ as regulatory cyto-
`kines (28). The last group consists of cytokines that can inhibit
`T-cell proliferation and differentiation but in some cases can
`be associated with a TH2 profile. Indeed, IL-10 and TGF-␤ can
`be produced by TH2 lymphocytes as well as regulatory T cells
`(16, 28).
`
`Genome & Co. v. Univ. of Chicago, PGR2019-00002
`UNIV. CHICAGO EX. 2004 - 5/8
`
`

`

`Downloaded from
`
`http://aem.asm.org/
`
` on November 15, 2016 by guest
`
`VOL. 74, 2008
`
`IMPACT OF BIFIDOBACTERIUM ON IMMUNITY
`
`665
`
`The Bifidobacterium strains can be separated into three
`groups according to their impacts on cytokine profiles (Table
`1). The first group includes strains which induced TH1 and/or
`TH2 cytokines, which are the B. bifidum and B. dentium strains
`as well as the four B. longum strains. The second group gathers
`the three B. breve strains which had few effects on immunity.
`The third group includes strains without any effect on immu-
`nity compared with the level for GF mice and comprises only
`one strain, namely, that of B. adolescentis.
`In the first group, B. longum strains direct systemic immunity
`toward a TH1 or TH2 profile, depending on the strains. B.
`longum BS64 induced a predominant TH2 profile. By contrast,
`B. longum BS49 and NCC2705 appeared to be high inducers of
`TH1 cytokines. B.
`longum CUETM 89-215 induced a high
`production level of every cytokine tested then, inducing TH1
`and TH2 cytokines at the same time. At the intestinal level, B.
`longum BS49 inhibited only IFN-␥ gene expression and then
`induced a TH2 profile. Three strains induced TGF-␤ gene
`expression; thus, they have a suppressor effect. Among them,
`B. longum CUETM 89-215 also inhibited IFN-␥ gene expres-
`sion. As TGF-␤ produced by TH3 cells could act toward a TH2
`profile (28), this strain was considered to act toward a TH2
`profile. These results, showing that B. longum appeared to
`orientate TH1/TH2 balance depending on the strain, tally with
`in vitro studies using heat-inactivated B. longum strains or
`bacterial extracts (11, 19, 30, 39). Our in vivo model allows us
`to consider interactions between bacteria and intestinal cells
`(whether immune or not, as with epithelial cells) and to take
`the IS’s complexity into account. However, the fact that we are
`working with entire tissues and not with isolated populations of
`immune cells could account for the differences between our
`results and those previously cited (11, 19, 30, 39).
`As with B. longum CUETM 89-215, B. bifidum and B. den-
`tium colonization in GF mice induced TH1 and TH2 cytokines
`at the systemic level at the same time. These data contrast with
`previous in vitro studies suggesting that B. bifidum induces only
`TH2 cytokines (11, 39). In our study, except for the inhibition
`of IL-4 gene expression, B. bifidum had no impact at the gut
`level. This corroborates previous data obtained in vivo and in
`vitro, showing that peritoneal macrophages from B. bifidum
`mono-associated GF mice produced neither TH1 cytokines
`(IL-1 and TNF-␣) nor TH2 ones (IL-6) (24) and showing that
`B. bifidum was unable to induce TNF-␣, IL-6, and IL-1␤ pro-
`duction (29). However, some B. bifidum strains isolated from
`an infant’s intestine could induce IL-8 secretion but not IL-6
`secretion (21). The multiplicity of strains and models used
`(macrophage-like cell line, dendritic cells from cord blood,
`Caco2, etc.) makes it difficult to draw a conclusion and could
`explain discrepancies between data. The last Bifidobacterium
`species of this group, B. dentium, induced TGF-␤ gene expres-
`sion, but that was the only influence of this strain on the
`terminal ileum. This is the first report of the impact on immu-
`nity of B. dentium, which is not among the usual species in the
`dominant gut microbiota.
`In the second group, B. breve strains appear to be poor
`inducers of immunity, suppressors, or strains ineffective at the
`systemic level. Only B. breve BS56 had an impact and induced
`a TH2-driven peripheral immune response. This TH2 immune
`orientation confirmed in vitro studies with living B. breve (11)
`and B. breve supernatant (14). B. breve BS94 was a suppressor,
`
`and B. breve BS52 was ineffective; their establishment induced
`an intestinal immune response activating TGF-␤1 gene expres-
`sion but inhibiting IL-4 gene expression. Similarly, a strain-
`specific effect was observed by Morita et al., who tested four B.
`breve strains (21).
`The establishment in GF mice of B. adolescentis, which is the
`constituent of the third group, had no ability to stimulate the
`IS, as observed by numerous authors (17, 21, 32, 39). However,
`He et al. showed that B. adolescentis induced a TH1-driven
`immune response, but they used a murine macrophage-like cell
`line (11), and Hessle et al. showed that one strain of B. ado-
`lescentis induced TNF-␣ secretion by human mononuclear
`cells, but this was with bacteria killed by UV (13). This killing
`process can change the capacity of bacteria to stimulate im-
`mune responses, and these authors used at least 10 times more
`bacteria/ml than the others and we did.
`Therefore, the present study demonstrates that strains more
`than species are linked to IS orientation. This strain-specific
`effect may be linked with specific secretion products (14) and
`structures, like DNA CpG unmethylated patterns (19, 33),
`exopolysaccharides (1), and peptidoglycans (30, 36, 40). The
`last varies in structure between bifidobacterial species but also
`between strains belonging to the same species (10).
`Our study offers a better understanding of the heterogeneity
`of Bifidobacterium species and of their potential role in IS
`stimulation. Interestingly, this study raises numerous questions
`concerning the use of Bifidobacterium as a probiotic in immune
`diseases such as allergies. The matters regarding the choice of
`one or several strains and the interaction between several bac-
`terial genera (synergy or antagonism) for probiotic supplemen-
`tation as a preventive treatment of allergic disease are still to
`be clarified.
`
`ACKNOWLEDGMENTS
`
`The work of O. Me´nard was supported by the Socie´te´ Franc¸aise de
`Nutrition Ente´rale et Parente´rale and Nutricia—Nutrition Clinique.
`Our thanks to M. C. Moreau for her useful advice.
`
`REFERENCES
`1. Amrouche, T., Y. Boutin, G. Prioult, and I. Fliss. 2006. Effects of bifidobac-
`terial cytoplasm, cell wall and exopolysaccharide on mouse lymphocyte pro-
`liferation and cytokine production. Int. Dairy J. 16:70–80.
`2. Bjorksten, B. 2004. Effects of intestinal microflora and the environment on
`the development of asthma and allergy. Springer Semin. Immunopathol.
`25:257–270.
`3. Bjorksten, B. 2005. Evidence of probiotics in prevention of allergy and
`asthma. Curr. Drug Targets Inflamm. Allergy 4:599–604.
`4. Bjorksten, B., P. Naaber, E. Sepp, and M. Mikelsaar. 1999. The intestinal
`microflora in allergic Estonian and Swedish 2-year-old children. Clin. Exp.
`Allergy 29:342–346.
`5. Boyle, R. J., and M. L. Tang. 2006. The role of probiotics in the management
`of allergic disease. Clin. Exp. Allergy 36:568–576.
`6. Butel, M. J., N. Roland, A. Hibert, F. Popot, A. Favre, A. C. Tesse`dre, M.
`Bensaada, A. Rimbault, and O. Szylit. 1998. Clostridial pathogenicity in
`experimental necrotising enterocolitis in gnotobiotic quails and protective
`role of bifidobacteria. J. Med. Microbiol. 47:391–399.
`7. Fanaro, S., R. Chierici, P. Guerrini, and V. Vigi. 2003. Intestinal microflora
`in early infancy: composition and development. Acta Paediatr. Suppl.
`91(441):48–55.
`8. Gaboriau-Routhiau, V., P. Raibaud, C. Dubuquoy, and M. C. Moreau. 2003.
`Colonization of gnotobiotic mice with human gut microflora at birth protects
`against Escherichia coli heat-labile enterotoxin-mediated abrogation of oral
`tolerance. Pediatr. Res. 54:739–746.
`9. Guy-Grand, D., C. Griscelli, and P. Vassalli. 1974. The gut-associated lym-
`phoid system: nature and properties of the large dividing cells. Eur. J. Im-
`munol. 4:435–443.
`10. Habu, Y., M. Nagaoka, T. Yokokura, and I. Azuma. 1987. Structural studies
`of cell wall polysaccharides from Bifidobacterium breve YIT 4010 and re-
`lated Bifidobacterium species. J. Biochem. (Tokyo) 102:1423–1432.
`
`Genome & Co. v. Univ. of Chicago, PGR2019-00002
`UNIV. CHICAGO EX. 2004 - 6/8
`
`

`

`Downloaded from
`
`http://aem.asm.org/
`
` on November 15, 2016 by guest
`
`666
`
`ME´NARD ET AL.
`
`APPL. ENVIRON. MICROBIOL.
`
`11. He, F., H. Morita, A. C. Ouwehand, M. Hosoda, M. Hiramatsu, J. Kurisaki,
`E. Isolauri, Y. Benno, and S. Salminen. 2002. Stimulation of the secretion of
`pro-inflammatory cytokines by Bifidobacterium strains. Microbiol. Immunol.
`46:781–785.
`12. He, F., A. C. Ouwehand, E. Isolauri, H. Hashimoto, Y. Benno, and S.
`Salminen. 2001. Comparison of mucosal adhesion and species identification
`of bifidobacteria isolated from healthy and allergic infants. FEMS Immunol.
`Med. Microbiol. 30:43–47.
`13. Hessle, C. C., B. Andersson, and A. E. Wold. 2005. Gram-positive and
`Gram-negative bacteria elicit different patterns of pro-inflammatory cyto-
`kines in human monocytes. Cytokine 30:311–318.
`14. Hoarau, C., C. Lagaraine, L. Martin, F. Velge-Roussel, and Y. Lebranchu.
`2006. Supernatant of Bifidobacterium breve induces dendritic cell matura-
`tion, activation, and survival through a Toll-like receptor 2 pathway. J.
`Allergy Clin. Immunol. 117:696–702.
`15. Holt, P. G., P. D. Sly, and B. Bjorksten. 1997. Atopic versus infectious
`diseases in childhood: a question of balance? Pediatr. Allergy Immunol.
`8:53–58.
`16. Janeway, C., P. Travers, M. Walport, and M. Shlomchik. 2007. Immunobi-
`ology: the immune system in health and disease, 8th ed. Garland Science
`Publishing, New York, NY.
`17. Karlsson, H., C. Hessle, and A. Rudin. 2002. Innate immune responses of
`human neonatal cells to bacteria from the normal gastrointestinal flora.
`Infect. Immun. 70:6688–6696.
`18. Kok, R. G., A. de Waal, F. Schut, G. W. Welling, G. Weenk, and K. J.
`Hellingwerf. 1996. Specific detection and analysis of a probiotic Bifidobac-
`terium strain in infant feces. Appl. Environ. Microbiol. 62:3668–3672.
`19. Lammers, K. M., P. Brigidi, B. Vitali, P. Gionchetti, F. Rizzello, E. Cara-
`melli, D. Matteuzzi, and M. Campieri. 2003. Immunomodulatory effects of
`probiotic bacteria DNA: IL-1 and IL-10 response in human peripheral blood
`mononuclear cells. FEMS Immunol. Med. Microbiol. 38:165–172.
`20. Livak, K. J., and T. D. Schmittgen. 2001. Analysis of relative gene expression
`data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method.
`Methods 25:402–408.
`21. Morita, H., F. He, T. Fuse, A. C. Ouwehand, H. Hashimoto, M. Hosoda, K.
`Mizumachi, and J. Kurisaki. 2002. Adhesion of lactic acid bacteria to caco-2
`cells and their effect on cytokine secretion. Microbiol. Immunol. 46:293–297.
`22. Mueller, S., K. Saunier, C. Hanisch, E. Norin, L. Alm, T. Midtvedt, A.
`Cresci, S. Silvi, C. Orpianesi, M. C. Verdenelli, T. Clavel, C. Koebnick, H. J.
`Zunft, J. Dore, and M. Blaut. 2006. Differences in fecal microbiota in dif-
`ferent European study populations in relation to age, gender, and country: a
`cross-sectional study. Appl. Environ. Microbiol. 72:1027–1033.
`23. Mullie´, C., M. F. Odou, E. Singer, M. B. Romond, and D. Izard. 2003.
`Multiplex PCR using 16S rRNA gene-targeted primers for the identification
`of bifidobacteria from human origin. FEMS Microbiol. Lett. 222:129–136.
`24. Nicaise, P., A. Gleizes, F. Forestier, A. M. Quero, and C. Labarre. 1993.
`Influence of intestinal bacterial flora on cytokine (IL-1, IL-6 and TNF-alpha)
`production by mouse peritoneal macrophages. Eur. Cytokine Netw. 4:133–
`138.
`25. Nicaise, P., A. Gleizes, C. Sandre, R. Kergot, H. Lebrec, F. Forestier, and C.
`Labarre. 1999. The intestinal microflora regulates cytokine production pos-
`itively in spleen-derived macrophages but negatively in bone marrow-derived
`macrophages. Eur. Cytokine Netw. 10:365–372.
`26. Ouwehand, A. C., E. Isolauri, F. He, H. Hashimoto, Y. Benno, and S.
`Salminen. 2001. Differences in Bifidobacterium flora composition in allergic
`and healthy infants. J. Alle