| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Laboratory for Immunohistochemistry and Immunopathology, Institute of Pathology, University of Oslo, Rikshospitalet, Oslo, Norway;
Department of Bacteriology, Osaka Dental University, Osaka, Japan;
Department of Pathology, Haukeland Hospital, Bergen, Norway; and
Department of Medical Microbial Ecology, Karolinska Institute, Stockholm, Sweden
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
GALT of the small intestine consists mainly of well-organized PPs, and
recent studies (8, 9, 10) have suggested that interactions
between B cells and enterocytes are important for the development of a
FAE with M cells. Hashi et al. (11) demonstrated that the
basic architecture of murine PPs was formed by
IL-7R
+ PP inducer cells,
VCAM-1+ mesenchymal cells, and
CD11c+ dendritic cells (DCs) at day 18 of
gestation; the PP anlagen was assembled without entry of mature
lymphocytes in scid/scid mice. In humans, PPs have been
shown to develop at 19 wk of gestation (12). Colonization
of germfree (GF) mice with Salmonella typhimurium aroA
caused a 2- to 3-fold increase in M cells accompanied by a 2-fold
increase in FAE-associated CD4+ T cells
(13). In rabbits, short-term exposure of the FAE to
nonintestinal bacteria (pneumococci) markedly increased the number of M
cells (14).
When M cell-transported Ags reach GALT, cognate interactions among APCs, naive T cells, and B cells are presumably initiated for subsequent establishment of potentially productive immunological memory or mucosal tolerance (1, 5). In human PPs from adolescents and young adults, we have previously shown that M cells harbor equal numbers of memory B and T cells (15). Moreover, we found that the M cell-associated memory B cells by far outnumber juxtaposed DCs and expressed the costimulatory molecules B7.1 (CD80) and B7.2 (CD86) at high levels, and that the M cell-associated memory T cells often expressed the CD40 ligand (CD154), suggesting B-T cell cognate interactions via CD40-CD154 ligation in the pockets (16). In adult humans, the M cell areas, identified by clustered B cells with strong CD80/CD86 expression, were always spatially related to germinal center (GC) extensions from the underlying B cell follicles. Notably, however, the M cell-associated memory B cell phenotype (surface (s) IgD-CD20lowCD80/CD86highCD10-CD27+Bcl-2+) differed from that of GC B cells (sIgD-CD20+CD80/86highCD10+CD27-Bcl-2-) (16). Therefore, we proposed that the M cell areas are important not only for Ag uptake, but also for cognate B-T cell interactions against recall Ags and subsequent B cell survival as well as differentiation. This notion was supported by mechanistic in vitro studies with a similarly induced memory B cell phenotype from peripheral blood (16). However, it has remained unclear how the memory B and T cells establish their dominance in the M cell pockets.
The present study was undertaken to visualize in situ the impact of
initial microbial colonization in the gut on 1) accumulation of
FAE-associated lymphocyte subsets and 2) modulation of CD80/CD86
expression on B cells. Rats taken from GF to conventionalized (CV)
animal house conditions were used for these experiments because this
model is well established in our laboratory and has been used to study
the effect of the normal microflora on the development of TCR
+ intraepithelial lymphocytes (IELs), both
in terms of subset expansion (17) and shaping of the
receptor V repertoire (18). Moreover, the size of PPs
is larger in rats than in mice, thus facilitating sample excision.
GF rats showed virtually no CD80/86 expression in their FAE or underlying lymphoid follicles. Although some scattered B and T cells were seen in the FAE under GF conditions, intraepithelial accumulation of CD86+ B cells and CD4+ T cells did not start until 1 wk after microbial colonization. Under GF conditions, CD4+CD86- DCs dominated markedly over B and T cells in the FAE, but they decreased in number after microbial colonization. In the mantle zone of underlying lymphoid follicles, B cells started to express Bcl-2 at a high level 12 wk after microbial colonization. These initial cellular events suggested that immature DCs surveying the FAE are crucial for the primary immune response against commensal bacteria. After migration of these DCs into the PPs, their interactions with Ag-specific naive T cells are most likely crucial for initiating immune responses that over time give rise to post-GC B cells subsequently residing within the M cell pockets ready to respond against recall Ags.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
GF AGUS rats were bred at the Department of Medical Microbial Ecology (Karolinska Institute, Stockholm, Sweden). The GF rats were exposed to a conventional indigenous microbial flora at 2 mo of age by receiving oral and rectal colonic fecal suspensions from rats housed under ordinary specific pathogen-free conditions. This protocol ensured the establishment of a commensal microbiota within 3 days (19). Such GF to CV (GF-CV) rats were sacrificed 1 wk (n = 3), 5 wk (n = 6) or 12 wk (n = 6) later. Control GF rats were killed at the age of 2 or 5 mo (six rats in each group).
Antibodies
Mouse mAbs against rat Ig 
L chain (OX12, IgG2a) and CD4
(W3/25, IgG1) were obtained from the Medical Research Council Cellular
Immunology Unit (Oxford, U.K.). Anti-rat CD3 (G4.18, IgG3),
anti-rat CD80 (3H5, IgG1), anti-rat CD86 (24F, IgG1), and
anti-rat DCs (OX62, IgG1) were purchased from BD PharMingen (San
Diego, CA). A mAb against rat follicular DCs (FDCs; OX2, IgG1) was
obtained from Cytotech (Copenhagen, Denmark), and anti-rat Bcl-2
(10C4, IgG1) from Zymed Laboratories (San Francisco, CA).
Immunohistochemistry
Terminal ileum segments with visible PPs were opened
longitudinally, embedded in OCT compound (Tissue-Tek; Miles
Laboratories, Elkhart, IN), and snap frozen in liquid nitrogen.
Cryosections cut at 4 µm were subjected to multicolor
immunofluorescence staining as previously detailed (17).
Briefly, the sections were first incubated with a mixture of mAbs
anti-CD3 and OX62 (2.5 µg/ml), anti-CD3 and anti-CD4
(1/4), or anti-Ig L chain and anti-CD80 or anti-CD86 (6.25
µg/ml) for 60 min at room temperature. The former two combinations of
primary mAbs were followed by a mixture of Cy3
(indocarbocyanine)-conjugated goat anti-mouse IgG1 (1.25 µg/ml;
Southern Biotechnology Associates, Birmingham, AL), biotinylated goat
anti-mouse IgG3 (5 µg/ml; Southern Biotechnology Associates), and
a rabbit antiserum specific for cytokeratin (1/100; our laboratory),
all reagents being preabsorbed with 20% rat serum (90 min). The latter
two combinations of primary mAbs were followed by a mixture of
Cy3-conjugated goat anti-mouse IgG1, biotinylated goat
anti-mouse IgG2a (25 µg/ml; Southern Biotechnology Associates),
and the rabbit antiserum against cytokeratin (to visualize the
epithelium). The sections were finally incubated with Cy2-conjugated
streptavidin (1/1000; Amersham, Aylesbury, U.K.) and amminomethyl
coumarin acetic acid-conjugated goat anti-rabbit IgG (20 µg/ml;
Vector Laboratories, Burlingame, CA) for 30 min. Anti-Bcl-2 (2.5
µg/ml) or anti-OX2 (1/50) was combined with anti-Ig -chain
followed by appropriate secondary and tertiary immunoreagents as above.
Tissue sections incubated with irrelevant isotype- and
concentration-matched primary mAbs served as negative controls.
Microscopy was performed with a Nikon E-800 fluorescence microscope (Nikon, Tokyo, Japan) equipped with a charge-coupled device video camera system (C5810; Hamamatsu Photonics, Hamamatsu, Japan). Digitalized images were captured with a computerized imaging system (Foto-station; Interfoto, Oslo, Norway). Recording of the immunophenotypes of infiltrating cells in the FAE was performed at x600 magnification. In each immunostained section, infiltrating cells were counted from all FAE areas (median number, 5; range, 3–8) and recorded as the average number of positive cells per FAE area. Putative M cell pockets were identified by intraepithelial clustering of B and T cells as described previously (15, 16).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
H&E-stained sections revealed well-developed PPs with B cell follicles in both GF and GF-CV rats, but no GCs occurred in the GF state (Fig. 1, A and B). Conversely, GF-CV rat PPs showed overt GC formation 1 wk after microbial colonization (Fig. 1C), and the GCs were maintained throughout the experimental period (Fig. 1, D and E).
|
|
|
-chain became down-regulated in GF-CV rats
over time (Fig. 3, D and E). The expression of
CD80 on B cells was hardly visible under both GF and GF-CV conditions
(data not shown). Only a small number of T cells had invaded the FAE under GF conditions, mainly being CD3+CD4- and therefore presumably CD8+ (Figs. 2B and 4, A and B). Accumulation of CD3+CD4+ T cells in the B cell follicles (Fig. 4, C–E) and the FAE (Figs. 2B and 5) clearly depended on microbial colonization. In GF rats, this subset was not seen in these two defined tissue compartments (Fig. 4, A and B).
|
The limited mAb combinations available for multicolor
immunofluorescence staining in rats rendered it difficult to
distinguish between different subsets of DCs. Therefore, we attempted
to trace numerical fluctuations of CD4+ DCs,
OX62+ DCs, and FDCs in the FAE. Under GF
conditions, the CD4+ cells appearing in the FAE
were virtually all CD3- DCs (Fig. 4, A and B). The number of this DC phenotype
remained unchanged in GF rats, but decreased strikingly after
conventionalization (Fig. 2C). OX62+
DCs increased slightly 1 wk after conventionalization and then declined
(Fig. 2D). The number of FDCs was less than five cells per
FAE area in GF-CV rats, with little difference from the GF state (Fig. 2E). Because mAb OX2 recognizes a 47-kDa glycoprotein on rat
thymocytes, brain tissue, endothelium, some smooth muscle cells, and B
cells, it is difficult to identify strictly the rat FDC lineage outside
of lymphoid follicles. Therefore, caution has to be exerted in an
attempt to identify the distribution of FDCs in rat PPs. However,
solitary polyhedral OX2+ cells, distinguished
from B cells by lack of Ig -chain expression and from endothelial
cells by morphology, were located in the subepithelial dome area, at
the periphery of lymphoid follicles, and in the B cell follicles quite
diffusely in PPs of GF rats (Fig. 6, A and B). Such putative FDCs gradually changed
their location, along with OX2+ B cells, toward
the serosal side and contributed to the formation of GCs after 5–12 wk
of conventionalization (Fig. 6, C–E).
|
Under GF conditions, B cells of lymphoid follicles or within the FAE did not express the apoptosis-preventing protein Bcl-2 (Fig. 7, A and B). After microbial colonization, however, B cells just beneath the FAE started to express Bcl-2, although the B cells of lymphoid follicles remained negative (Fig. 7, C, F, and G). Mantle zone B lymphocytes became faintly positive for Bcl-2 only at 5 wk after conventionalization (Fig. 7D) and more strongly positive at 12 wk (Fig. 7E). Throughout the experimental period, no substantial expression of Bcl-2 was observed on B cells within the FAE (Fig. 7, F–H). This was in contrast to the situation in M cell pockets of human adults in which B cells consistently expressed Bcl-2 at a high level along with the post-GC marker CD27 (16).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
+, but
not the TCR
+, subset of IELs is influenced
by microbial colonization, resulting in oligoclonal expansion and
phenotypic alterations both in rats and mice. GF mice are known to have
reduced numbers of PPs and their lymphoid follicles, as well as M
cells, show a numerical increase after transfer to a conventional
specific pathogen-free animal facility (23). Repopulating
GF mice either with a single nonpathogenic microorganism
(Clostridium indolis) or with a single pathogen (S.
typhimurium) is sufficient to establish the normal number of PPs
with a FAE containing M cells (1, 13). In humans, PPs start to develop at 19 wk of gestation (12), and the local B cells apparently have a crucial role in the induction of M cells in the FAE (8, 9, 10). We have previously reported that B cells present in the M cell pockets of human PPs by far outnumber other types of cells with potential Ag-presenting function (15) and often coexist with CD4+CD154+ memory T cells (16). Therefore, we have proposed that cognate B-T cell interactions may take place in the M cell pockets (5). However, it has remained elusive when the B cells accumulate in the FAE to establish an ideal microcompartment for encountering luminal Ags and to start their putative interaction with memory T cells. Because human neonatal PP specimens generally are unavailable, we performed our study in rats reared under GF conditions to observe the redistribution of immune cells in PPs during initial microbial colonization. Interestingly, neither CD86+ B cells nor CD4+ T cells were present in the FAE of GF rats in any significant numbers. Notably, before microbial colonization, mainly CD4+ DCs occurred in the FAE and clearly outnumbered other leukocytes. These DCs could be characterized as immature (or resting) because they did not express the CD80 or CD86 costimulatory molecules.
Liu et al. (24) reported that DCs present in rat intestinal lymph on their way to the mesenteric lymph nodes could be divided into two subsets by their CD4 expression; CD4+OX41+ DCs had short fine processes and low levels of nonspecific esterase, whereas CD4-OX41- DCs had long pseudopodia and high levels of nonspecific esterase. The CD4+ subset expressed MHC class II and CD11b/c and was better at stimulating an allogeneic MLR, presentation of Ag to sensitized T cells, and particularly Ag-specific activation of naive T cells than the CD4- subset (24). Recent experiments have shown that DCs are able to sample bacteria from the surface of a monolayered epithelium (25). Therefore, our study suggested that the CD4+ DC subset observed in the FAE of GF rats has a role in eliciting primary PP immune responses against commensal bacteria after their initial colonization. Interestingly, this subset disappeared from the FAE soon after the conventionalization, probably as a result of rapid DC migration to the interfollicular T cell zone where DC activation/maturation and stimulation of naive T cells are known to take place. Such a compartmentalized interaction between APCs and T cells in PPs has recently been demonstrated in a peroral infection model with S. typhimurium in mice (26). That this could be the case also in GF rats was suggested by the presence of CD86+ DCs in the interfollicular T cell zones of their PPs, although observations based on functional markers in tissue sections are only hypothesis generating and not conclusive in mechanistic terms. Nevertheless, our presumption is supported by a murine lymph node study showing that it is the CD11b+CD8- (presumably CD4+), and not the CD11b-CD8+, DC subset that undergoes physical interactions with Ag-specific CD4+ T cells just outside of the B cell follicles (27).
B cells are known to be semiprofessional APCs but are likely to tolerize naive T cells (28, 29, 30). We found that neither CD86+ B cells nor CD4+ T cells accumulated in the FAE without the presence of an indigenous microbiota, whereas such intraepithelial subsets showed a substantial elevation 5 wk after conventionalization. One might speculate that B cells accumulate in the FAE to capture recall Ags from colonizing commensal gut bacteria to avoid excessive primary immune responses against new bacterial Ags from the normal flora. Moreover, Alpan et al. (31) reported that T cells from OVA-fed µMT mice, which lack B cells, PPs, and M cells, responded to OVA in vitro like T cells from wild-type mice. They also suggested that B cell-induced GALT structures such as PPs are not necessary for immune responses against fed soluble Ag. Others, however, have proposed that B cells may have an important, although not obligatory, role as APCs for primary T cell responses. Thus, in vitro studies have shown that activated B cells are nearly as effective in this respect as DCs (32). Linton et al. (33) compared responses of CD4+ T cells from normal and B cell-deficient mice to keyhole limpet hemocyanin over 6 mo and reported diminished IL-2 production by T cells primed in the absence of B cells. The transfer of B cells restored immunological memory and Ag presentation was not essential to this end, because also B cells activated in vitro with irrelevant Ag restored the number of memory T cells. The cellular redistributions we observed in PPs suggested that DCs must be most important for the primary responses against initial microbial colonization, but it is possible that B cells can aid the expansion of the CD4+ T cell population by acting as APCs in the face of the limited number of DCs available early on in the ontogeny. Alternatively, B cells may aid T cell maturation without Ag presentation as alluded to above (32).
Moreover, it is known that B cells can regulate cytokine profiles of T cells by affecting the function of DCs (34). Thus, DCs from B cell-deficient µMT mice failed to sensitize T cells to produce IL-4. Also, DCs have been suggested to be able to interact with naive B cells and to initiate Ig class switching in a primary T cell-dependent response (35). Further studies are needed to learn whether direct DC-B cell interactions may skew the T cell cytokine profile in PPs or promote B cell isotype switching in primary immune responses to initial microbial colonization of the gut.
It is of note that the B cells in rat PPs did not express Bcl-2 without the presence of a commensal bacterial flora; the mantle zones of lymphoid follicles, which consist of naive but mature recirculating B cells, became strongly positive for Bcl-2 only 12 wk after microbial colonization, showing an expression level similar to mantle zone B cells in adult human PPs (16). Bcl-2 is involved in resistance against apoptosis and is a marker of cellular differentiation or longevity (36, 37, 38), but very little is known about factors modulating its level of expression in vivo. B cells from sheep ileal PPs have been shown to be rescued from apoptosis in vitro by induction of Bcl-2 expression after stimulation with phorbol ester (39).
In contrast to GC B cells, which mainly lack Bcl-2, the M cell pockets of PPs in adult humans contain numerous sIgD-Bcl-2+ post-GC (CD27+) memory/effector B cells, suggesting that these microcompartments are important in B cell differentiation and survival (16). During initial microbial colonization in rats, however, as observed in the present study, the activated (CD86+) sIgD-B cells accumulating in the M cell pockets did not express a substantial level of Bcl-2. This species disparity might be explained by the fact that the microbial load is more restricted in rats under specific pathogen-free conditions than in adult human beings. Fig. 8 presents a hypothetical model for modulation of Bcl-2 expression in B cell follicles. We suggest that microbial components are crucial, both for up-regulation of Bcl-2 in mature naive mantle zone B lymphocytes and to provide Ags for the positive selection process in GCs, which depends on B cell survival at least partly mediated by Bcl-2. The putative role of macrophage/DC activation could be secretion of IL-10, which is known to induce Bcl-2 in B cells (40). Importantly, gut bacteria such as Escherichia coli can directly stimulate macrophages to secrete IL-10 (41) and lactobacilli can do the same with T cells (42). Different levels of microbial stimulation might also explain that the PP B cell follicles of colonized rats did not make contact with the putative M cell pockets, in contrast to the situation in adult human PPs (16).
|
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Per Brandtzaeg, Institute of Pathology, Rikshospitalet, N-0027 Oslo, Norway. E-mail address: per.brandtzaeg{at}labmed.uio.no
3 Abbreviations used in this paper: FAE, follicle-associated epithelium; PP, Peyer’s patch; DC, dendritic cell; GALT, gut-associated lymphoid tissue; GC, germinal center; GF, germfree; CV, conventionalized; FDC, follicular DC; IEL, intraepithelial lymphocyte; s, surface.
Received for publication July 16, 2002. Accepted for publication November 5, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
repertoire of intraepithelial lymphocytes in rat intestine. Immunology 89:494.[Medline]
T-cell receptor-bearing intestinal intraepithelial lymphocytes after microbial colonization in germ-free mice and its independence from thymus. Immunology 79:32.[Medline]
/ T cells to the intestinal epithelium is independent of normal microbial colonization. J. Exp. Med. 172:239. and interleukin-10. Clin. Diagn. Lab. Immunol. 8:695.This article has been cited by other articles:
|
|
 |
|
  S. Velaga, H. Herbrand, M. Friedrichsen, T. Jiong, M. Dorsch, M. W. Hoffmann, R. Forster, and O. Pabst Chemokine Receptor CXCR5 Supports Solitary Intestinal Lymphoid Tissue Formation, B Cell Homing, and Induction of Intestinal IgA Responses J. Immunol., March 1, 2009; 182(5): 2610 - 2619. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. H. Niess, F. Leithauser, G. Adler, and J. Reimann Commensal Gut Flora Drives the Expansion of Proinflammatory CD4 T Cells in the Colonic Lamina Propria under Normal and Inflammatory Conditions J. Immunol., January 1, 2008; 180(1): 559 - 568. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Corthesy, H. R. Gaskins, and A. Mercenier Cross-Talk between Probiotic Bacteria and the Host Immune System J. Nutr., March 1, 2007; 137(3): 781S - 790S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C Mueller and A J Macpherson Layers of mutualism with commensal bacteria protect us from intestinal inflammation Gut, February 1, 2006; 55(2): 276 - 284. [Full Text] [PDF]
|
|
|
|
 |
|
 G. Henriksson, L. Helgeland, T. Midtvedt, P. Stierna, and P. Brandtzaeg Immune Response to Mycoplasma pulmonis in Nasal Mucosa Is Modulated by the Normal Microbiota Am. J. Respir. Cell Mol. Biol., December 1, 2004; 31(6): 657 - 662. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. Noverr, R. M. Noggle, G. B. Toews, and G. B. Huffnagle Role of Antibiotics and Fungal Microbiota in Driving Pulmonary Allergic Responses Infect. Immun., September 1, 2004; 72(9): 4996 - 5003. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, 
, and 
represent data from GF rats and •,
, and 
represent GF rats after conventionalization.
A, B cells identified by an anti-Ig 
s) or DCs
activated by microbial components. Induction of a high level of Bcl-2
in memory/effector B cells depends on cognate T cells and sustained Ag
stimulation through a B cell receptor (BCR) with high affinity after
somatic hypermutation of Ig variable region genes. Such positive B cell
selection normally takes place on Ag-retaining FDCs in GCs, but this
mechanism appears to be less robust in rats under clean, specific
pathogen-free conditions than in adult humans.