Biosci. Biotechnol. Biochem., 70 (12), 3031–3035, 2006
The Composition of Intestinal Bacteria Aﬀects the Level of Luminal IgA Yuji O HASHI, Mari H IRAGUCHI, and Kazunari U SHIDAy Laboratory of Animal Science, Kyoto Prefectural University, Shimogamo, Kyoto 606-8522, Japan Received March 23, 2006; Accepted September 13, 2006; Online Publication, December 7, 2006 [doi:10.1271/bbb.60164]
An essential role of several speciﬁc intestinal bacteria in the intestinal IgA level is suggested. Fecal IgA concentration in mice from one breeder was signiﬁcantly higher than that in mice from two other breeders. The level of segmented ﬁlamentous bacteria and four particular clostridia in mice from the former breeder are of particular importance in developing the IgA production and secretion system. Key words:
segmented ﬁlamentous bacteria; clostridia; IgA; breeder
The epithelial layer in the gastrointestinal tract is exposed to various antigens, such as food antigens, commensal bacteria, and pathogenic bacteria. Secretory IgA (sIgA) released from the mucosal surface has a central role in mucosal immune system in inhibiting the adherence of pathogenic bacteria and neutralizing biologically active antigens (e.g., bacterial toxins, viruses, and enzymes).1) In germ-free rodents, in comparison with conventional animals, IgA production is limited.2) As shown by gnotobiotic studies, production of IgA is stimulated by the colonization of bacteria,3) particularly, segmented ﬁlamentous bacteria (SFB) and clostridia.4,5) However, the mucosal immune system is also modulated by various probiotic bacteria, such as lactobacilli and biﬁdobacteria, which are common members of the intestinal microﬂora.6) Therefore, a wide range of intestinal bacteria can stimulate the development of the mucosal immune system. Indeed, common bacterial cell components, such as the lipopolysaccharide of gram-negative bacteria, the lipoteichoic acid of gram-positive bacteria, and ﬂagellin, induce the immune response.7) It is still unclear whether several speciﬁc bacteria, previously suggested by gnotobiotic studies, are involved only in the development of the mucosal immune system, particularly IgA production and secretion. The objective of the present study was to demonstrate that several speciﬁc bacteria are largely responsible for the higher luminal sIgA concentration that results in proper development of IgA y
production and the secretion system. The diﬀerence in intestinal microﬂora might explain the diﬀerence in IgA production and intestinal secretion among experimental rodents provided by diﬀerent breeders. Five male Balb/c mice (7 weeks old) were purchased from three commercial suppliers: A, B, and C. The mice were divided into three groups according to the supplier. They were fed an MF diet (Oriental Yeast, Tokyo) and maintained under barrier-free conditions. All animals were handled with due regard for their welfare, as recommended by the Experimental Animal Care and Use Committee of Kyoto Prefectural University, Kyoto, Japan. After the mice underwent a 2-week adaptation period, their feces were collected to analyze IgA and microﬂora. To measure IgA concentration, feces were appropriately diluted with phosphate-buﬀered saline and centrifuged at 12;000 g for 5 min. The concentration of IgA in the supernatant was measured with a Mouse IgA ELISA Quantitation Kit (Bethyl, Montgomery, TX) according to the manufacturer’s instructions. Bacterial DNA was extracted from fresh feces according to Godon et al.8) Temperature-gradient gel electrophoresis (TGGE) on bacterial 16S rDNA was carried out to analyze the diversity of fecal bacteria. The variable region (from V6 to V8) of 16S rDNA of the fecal bacteria was ampliﬁed by PCR with primers 968GC and UnivR.9) Amplicons were analyzed by TGGE and ampliﬁed rDNA restriction analysis (ARDRA) according to Ohashi et al.9) After electrophoresis, the gel was silver-stained and analyzed using the Gel-Pro Analyzer (Media Cybernetics, Silver Spring, MD). Hierarchical clustering of band-pattern similarity without band density was used to create a dendrogram, as described previously.10) To obtain species information on bacteria that contributed to the particular proﬁle of fecal ﬂora of mice purchased from breeder B, DNA from four speciﬁc bands (indicated in Fig. 2) were sequenced. These bands were excised, and DNA was eluted overnight into sterile distilled water at 4 C. The eluted DNA was subjected to PCR using primers, 968-f and
To whom correspondence should be addressed. Tel/Fax: +81-75-703-5620; E-mail: k [email protected]
Abbreviations: ARDRA, ampliﬁed rDNA restriction analysis; SFB, segmented ﬁlamentous bacteria; TGGE, temperature-gradient gel electrophoresis; SPF, speciﬁc pathogen-free; FOS, fructo-oligosaccharide
Concentration of fecal IgA (log 10 µg/g feces)
Y. OHASHI et al. A B C (A) M 1 2 3 4 5 M1 2 3 4 5 M 1 2 3 4 5 M
3 1 2 3
Number of SFB (log 10 cell/g feces) Fig. 1. The Correlation between the Number of Segmented Filamentous Bacteria (SFB) and the Concentration of Fecal IgA. The value of SFB-monoassociatred mice was taken from Umesaki et al.4) and Umesaki and Setoyama.5) , breeder A; , breeder B; , breeder C; , SFB-monoassociated mice. The level of SFB in this study was estimated by real-time PCR. For details, see text.
UnivR, and the PCR product was transformed into competent E. coli JM107 cells (Toyobo, Tokyo) with the pGEM-T vector system (Promega, Tokyo) according to the manufacturer’s instructions. The eight colonies of ampicillin-resistant transformants generated from each band were subjected to colony PCR using vectorspeciﬁc primers T7 and SP6, and subsequent ARDRA grouping was done by a method described elsewhere.9) The plasmid of one clone from the predominant ARDRA group was sequenced by the Shimadzu Genomic Research Laboratory (Shimadzu, Kyoto, Japan). Obtained and reference sequences were aligned using the CLUSTAL X program.11) The reference sequences were obtained from the DDBJ. A phylogenic tree was constructed using the maximum parsimony method with SEQBOOT, DNAPARS, CONSENSE, and DRAWGRAM of the PHYLIP package.12) Quantiﬁcation of SFB-DNA in total fecal DNA was performed with the LightCycler system (Roche, Mannheim, Germany). The FastStart DNA Master SYBR Green I was used for PCR ampliﬁcation. Speciﬁc primers for the 16S rRNA gene of SFB (SFB-f, 50 -AGGAGGAGTCTGCGGCACATTAGC-30 and SFB-r, 50 CGCATCCTTTACGCCCAGTTATTC-30 ) were used. The reaction mixture (20 ml) contained 4 mM MgCl2 , 2 ml of the 10 Mastermix (including the FastStart enzyme, FastStart Taq DNA polymerase, a reaction buﬀer, a dNTP mixture, MgCl2 , and SYBR Green I dye), 50 ng of fecal DNA, and 0.5 mM of each of the SFB-speciﬁc primers. The thermal program consisted of initial denaturation at 95 C for 10 min, followed by 40 cycles of 95 C for 15 s and 65 C for 5 s, and the reaction was completed with a ﬁnal elongation at 72 C for 25 s. Melting curve analysis was performed on the product after completion of ampliﬁcation to determine
C Cluster 1 A Cluster 2
5 3 2 1 4 5 4 3 2 1 5 3 1 4 2
Fig. 2. Gel Image of Temperature-Gradient Gel Electrophoresis (TGGE) (A) and Dendrogram Created from the TGGE Band Proﬁle (B) of Fecal Microﬂora of Mice Purchased from Three Experimental Animal Suppliers. (A), 1, 2, 3, 4, and 5, the bacterial 16S rDNA of feces was obtained from ﬁve mice from each breeder (A, B, and C). M, marker. Four bacterial 16S rDNA sequences, Ruminococcus hydrogenotroplicus-like (91% identity to X95624), Clostridium sp.-like (85% identity to AF157053), Escherichia coli (100% identity to AF14565), and Lactobacillus gasseri strain KC5a-like (97% identity to AF243165), were used as standard marker of TGGE.10) Arrows indicate the positions of speciﬁc bands of mice from breeder B to be sequenced. These bands were excised, and DNA was eluted and sequenced. For details, see text. (B), The cluster analysis employed hierarchical clustering with Euclidean square distances.
the speciﬁcity of the PCR. Since SFB is a still uncultured bacterium, a mouse fecal sample previously determined for the number of SFB was used as a standard feces for real-time PCR. The number of SFB in that sample was microscopically determined depending upon SFB’s particular segmented and ﬁlamentous morphology. This particular sample contains relatively large number of SFB, which made it possible to estimate the level of this bacterium. DNA extracted from this fecal sample was serially diluted and ampliﬁed to establish the standard curve for SFB in real-time PCR. The concentration of fecal IgA and the number of SFB were statistically analyzed by Tukey-PLSD after one-way ANOVA with StatView Ver.5.0 (SAS Institute, Cary, NC). SFB has a particular function in the development of the IgA secretion system.4,5) SFB is closely associated with epithelial cells, particularly follicle-associated epithelial cells, in stimulating mucosal immune reactions.13) Luminal IgA was nearly zero in the germ-free
Composition of Intestinal Bacteria and Luminal IgA
Methanobacterium formicicum (M36508) C. coccoides (M59090) C. boltei (AJ508452) Band-1 Band-2 Clostridium sp. DSM 6877 (X76747) C. clostridiiformes (M59089) 97 C. xylanolyticum (X71855) C. aerotolerans (X76163) C. sphenoides (X73449) 87 90 80 C. celerecrescens (X71848) C. oroticum (M59109) Cluster XIVa C. aminophilum (L04165) C. symbiosum (M59112) C. polysaccharolyticum (X71858) C. populeti (X71853) C. aminovalericums (X73436) C. fimetarium (AF126687) C. fusiformis (AF028349) C. indolis (AF1028351) C. scindens (AF262238) C. hylemonae (AB117570) 100 Band-4 Clostridium sp. ASF502 (AF157053) C. nexile (X73443) 100 C. ramosum (AY699288) Cluster XVIII C. spiroforme (X73441) 92 SFB (X77814) SFB (D86305) 100 100 SFB (X87244) C. tetani (NC_004557) C. intestinale (AY781385) Cluster I C. butyricum (X68176)
C. cellulosi (L09177) C. sporosphaeroides (X66002) C. methylpentosum (Y18181) C. leptum (AJ305238) Band-3 C. aldrichii (X71846) C. straminisolvens (AB125279) 100 C. stercorarium subsp. stercorarium (AJ310082) C. thermolacticum (X72870) Clostridium sp. JC3 (AB093546) 80 C. cellulolyticum (X71847) 88 C.papyrosolvens (X71852) 100 C.termitidis (X71854) C. cellobioparum (X71856) C. thermocellum (L09173)
Fig. 3. Phylogenetic Tree Derived from the 16S rDNA Sequences of Four Speciﬁc TGGE-Bands in Mice from Breeder B and Related Species in Clostridium Clusters I, III, IV, XIVa, and XVIII. The tree was constructed using the maximum parsimony method. The sequence of Methanobacterium formicicum (M36508) was used as the outgroup to root the tree. Bootstrap values, based on 1,000 replications, at the nodes of the tree display conﬁdence levels. Bootstrap values above 80% are given in the tree. The database sequences show GenBank accession numbers in brackets.
animals, but mono-association with SFB restored luminal IgA to 2/3 of the conventional level.14) This previous report strongly suggests the particular importance of SFB in the development of the mucosal immune system expressed partly as luminal IgA level. Hence, the level of SFB was estimated by real-time PCR in this experiment. The cell number of SFB (log10 cell number/g feces) was signiﬁcantly larger in mice from breeder B (6:0 0:2) than in the other mice (A, 5:2 0:3; C, 4:9 0:4; p < 0:01). The IgA concentration in feces (mg/g feces) was also signiﬁcantly higher in mice from breeder B (2;138:2 666:5) than in the other mice (A,
127:5 54:5; C, 111:6 61:8; p < 0:01). Perhaps the concentration of fecal IgA depended on the cell number of SFB, but in the SFB-monoassociated mice,4,5) the cell number of SFB was approximately 100 times larger than in the mice from breeder B, although the fecal IgA level was same as in the mice from breeder A or breeder C (Fig. 1). This suggests that the degree of stimulation of the fecal IgA level by colonization of SFB is limited to approximately 200 mg/g feces. Elevation of the fecal IgA level more than 200 mg/g feces requires other factors. TGGE band proﬁles generated from mice from two
Y. OHASHI et al.
breeders (A and C) made one cluster (Fig. 2). TGGE was performed three times in this experiment, and the TGGE proﬁles showed the same cluster pattern. Hence those from breeder B were separated from this cluster. Therefore, the composition of the fecal microﬂora of mice from breeder B diﬀers from those of mice raised by the two other breeders. Experimental animals are generally maintained under speciﬁc pathogen-free (SPF) conditions.15) However, SPF animals are raised from gnotobiotic animals, which harbor deﬁned (or semi-deﬁned) intestinal microﬂora.15) Each breeder might have his own bacterial cocktail for the creation of gnotobiotic animals. Therefore, the composition of intestinal bacteria produced by diﬀerent breeders is expected to vary. Since the origin of Balb/c mice is diﬀerent for each breeder, diﬀerences in the genetic backgrounds of Balb/c mice can aﬀect the composition of intestinal bacteria, but there is no information available on how diﬀerent these strains are. Diﬀerences in the composition of intestinal bacteria might aﬀect fecal IgA levels. Interestingly, the traits shown by the concentration of fecal IgA (Fig. 1) were identical to those shown by the dendrogram of the fecal microﬂora (Fig. 2). The diﬀerence in the fecal microﬂora of mice from breeder B and the other mice (from breeders A and C) was mainly explained by the presence of four bacteria in the former as detected by TGGE analysis (Fig. 2). The 16S rDNA sequences of these bacteria perhaps belong to Firmicutes (clostridia) according to the phylogenic tree (Fig. 3). Three of these bacteria belong to Clostridium cluster XIVa, which is the predominant clostridium group in mice.16) The remaining sequence branches deeply into Clostridium cluster III. Clostridia have a speciﬁc role in developing the mucosal immune system.4,5) Hence these four clostridia might be of particular importance in the elevation of fecal IgA level. The metabolite and cell components of some clostridia have been thought to be mediators involved in the gut immune response.5) The importance of SFB and certain clostridia in stimulating the development of the mucosal immune system previously suggested by a gnotobiotic study was therefore demonstrated in a complex conventionalized system. The present study was concerned with diﬀerences in the intestinal microﬂora of the experimental rodents, Balb/c mice, among Japanese breeders. As shown in Fig. 1, the fecal sIgA level apparently depends on the intestinal microﬂora, determined by the breeder. This can bias experimental results, particularly in an immunological study. For instance, Umemoto et al.17) reported that fructo-oligosaccharide (FOS) suppressed colonic inﬂammation in a dextran sodium sulfate-induced colitis model using the Sprague-Dawley rat, while Moreau et al.18) reported the ineﬀectiveness of FOS in the same colitis model. Therefore, the construction of deﬁned intestinal microﬂora is necessary, and the information should be available to researchers.
Mestecky, J., and Russell, M. W., Intestinal immunoglobulin A: role in host defense. In ‘‘Microbial Pathogenesis and Intestinal Epithelial Cell,’’ ed. Hecht, G. A., ASM Press, Washington, DC, pp. 95–112 (2003). Shoﬀ, K. H., Meslin, K., and Cebra, J. J., Commensal enteric bacteria engender a self-limiting humoral mucosal immune response while permanently colonizing the gut. Infect. Immun., 63, 3904–3914 (1995). Macpherson, A. J., Hunziker, H., McCoy, K., and Lamarre, A., IgA responses in the intestinal mucosa against pathogenic and non-pathogenic microorganisms. Microbes Infect., 3, 1021–1035 (2001). Umesaki, Y., Okada, Y., Setoyama, H., Matsumoto, S., Imaoka, A., and Ito, K., Diﬀerential roles of segmented ﬁlamentous bacteria and clostridia in development of the intestinal immune system. Infect. Immun., 67, 3504– 3511 (1999). Umesaki, Y., and Setoyama, H., Structure of the intestinal ﬂora responsible for development of the gut immune system in a rodent model. Microbes Infect., 2, 1345–1351 (2000). Isolauri, E., Salminen, S., and Ouwehand, A. C., Microbial-gut interactions in health and disease: probiotics. Best Pract. Res. Clin. Gastroenterol., 18, 299– 313 (2004). Kaisho, T., and Akira, S., Toll-like receptors as adjuvant receptors. Biochim. Biophys. Acta, 1589, 1–13 (2002). Godon, J. J., Zumstein, E., Dabert, P., Habouzit, F., and Moletta, R., Molecular microbial diversity of an anaerobic digester as determined by small-subunit rDNA sequence analysis. Appl. Environ. Microbiol., 63, 2802– 2813 (1997). Ohashi, Y., Tokunaga, M., and Ushida, K., The eﬀects of Lactobacillus casei strain Shirota on the cecal fermentation pattern depend on the individual cecal microﬂora in pigs. J. Nutr. Sci. Vitaminol., 50, 399–403 (2004). Inoue, R., and Ushida, K., Development of the intestinal microbiota in rats and its possible interaction with the evolution of the luminal IgA in the intestine. FEMS Microbiol., 45, 147–153 (2003). Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F., and Higgins, D. G., The ClustalX Windows interface: ﬂexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res., 24, 4876–4882 (1997). Felsenstein, J., PHYLIP: phylogeny inference package (version 3.2). Cladistics, 5, 164–166 (1989). Meyerholz, D. K., Stabel, T. J., and Cheville, N. F., Segmented ﬁlamentous bacteria interact with intraepithelial mononuclear cells. Infect. Immun., 70, 3277–3280 (2002). Talham, G. L., Hiang, H., Bos, N. A., and Cebra, J. J., Segmented ﬁlamentous bacteria are potent stimuli of a physiologically normal state of the murine gut mucosal immune system. Infect. Immun., 67, 1992–2000 (1999). Yanabe, M., Shibuya, M., Gonda, T., Asai, H., Tanaka, T., Sudou, K., Narita, T., and Itoh, K., Establishment of speciﬁc pathogen-free (SPF) rat colonies using gnotobiotic techniques. Exp. Anim., 50, 293–298 (2001). Salzman, N. H., de Jong, H., Paterson, Y., Harmsen, H. J. M., Welling, G. W., and Bos, N. A., Analysis of 16S
Composition of Intestinal Bacteria and Luminal IgA
libraries of mouse gastrointestinal microﬂora reveals a large new group of mouse intestinal bacteria. Microbiology, 148, 3651–3660 (2002). Umemoto, Y., Tanimura, H., Ishimoto, K., Masaki, K., Maniwa, Y., Murakami, K., Sahara, M., Yamada, K., Takizawa, T., and Hidaka, H., Fructo-oligosaccharide improves the colonic environment in rats with experimental ulcerative colitis. Digestion and Absorption, 16,
84–87 (1993). Moreau, N. M., Martin, L. J., Toquet, C. S., Laboisse, C. L., Nguyen, P. G., Siliart, B. S., Dumon, H. J., and Champ, M. M., Restoration of the integrity of rat caecocolonic mucosa by resistant starch, but not by fructooligosaccharides, in dextran sulfate sodium-induced experimental colitis. Br. J. Nutr., 90, 75–85 (2003).