The gut-associated lymphoid system: nature and properties of the large dividing cells

Gut mucosal mast cells. Origin, traffic, and differentiation

Gut mucosal mast cells (MMC), which are nearly absent in normal mice are abundant during nematode infection. In normal mice, study of MMC precursors (MMC-P: cells giving rise to MMC colonies in the presence of IL-3) show that: (a) their frequency, judged by limiting dilution is very high in bone marrow (BM) and gut, and very low in most lymphoid organs and thoracic duct lymph (TDL); (b) gut MMC-P are Thy-1- Lyt-1-2- and are not rapidly replicating; (c) they are the progeny of less differentiated BM MMC-P which are attracted from the blood to the gut mucosa by local factor(s), other than antigen and T cell factors (since normal amounts of gut MMC-P are found in germ-free, nude, and newborn mice). In mice bearing the Wehi 3 tumor (which releases enough IL-3 to produce detectable blood levels) spleen and mesenteric lymph nodes (LN) show increased MMC-P frequency, the greatest increase being in the gut and BM, where numerous differentiated MMC are found. In Nippostrongylus brasiliensis (Nb)-infested mice (known to develop a large, T cell-dependent, gut MMC infiltration), gut MMC-P proliferation is induced by IL-3 released from gut mucosal Thy-1+ Lyt-2- cells, whose in vitro IL-3 release capability is much higher than that of similar cells from normal mice. Both Nb-stimulated T blasts and proliferating MMC-P undergo cyclic traffic, migrating into the TDL and then seeding the whole length of the gut (a process which allows a widespread immune defense after a local antigenic stimulus). Experiments using 2-d interruption of this traffic and fetal gut grafts, suggest that the continuous homing of T blasts back to the gut which leads to permanent Nb-stimulated IL-3 release, is essential for the full maturation of MMC. Transfer experiments in the rat show that TDL circulating MMC-P rapidly mature into MMC when they home back to the Nb-infested gut. It is proposed that gut MMC arise after several stages of progressive differentiation of MMC-P, influenced both by IL-3 and unidentified gut factor(s).

Lymphocyte migration into the lacrimal gland is random

The murine lacrimal gland demonstrates a high proportion of both IgA-producing cells and IgA-bearing cells, an observation consistent with tissues of the secretory immune system. This selective accumulation of IgA cells in the lacrimal gland was studied by observing the accumulation of various populations of fluorescein isothiocyanate-labeled lymphocytes in the gland. Labeled spleen cells were found to accumulate equally as well as mesenteric lymph node cells. T cells and T-cell-depleted populations of lymphocytes from spleen also showed migration into the gland. Surface staining of labeled cells found in the gland after 24 hr revealed that IgA-, IgG-, and IgM-bearing cells were all present in high proportions. These experiments demonstrate a random migration of lymphocytes into the lacrimal gland and suggest that the accumulation of IgA-bearing cells in the gland is regulated by the nonvascular microenvironment within the gland.

The interaction of high endothelial venules with T and B cells in peripheral lymph nodes after antigenic stimulation

The influence of an ongoing immune response on the specific interaction of high endothelial venules (HEV) in peripheral lymph nodes with T and B cells was determined using short-term homing experiments and an in vitro HEV adherence assay. It was demonstrated that stimulation with various antigens did not change the receptor specificity of the HEV. However, the retention and/or the efflux of immigrant cells in the stimulated lymph nodes had changed.

Repopulation of the gut lamina propria with IgA-containing cells by lymphoid cells isolated from the gut lamina propria

Migration and lodging properties of IgA cells and their precursors of the gut lamina propria (GLP) were studied by transfer of highly purified GLP lymphoid cells between immunoglobulin allotype-congenic mice (CB-20 to BALB/c). The donor IgA-containing ( cIgA ) cells appeared in the GLP of the recipients at day 1, peaked at days 12-15 (late repopulation peak) and persisted up to day 20 after cell transfer. An additional peak at day 3 (early repopulation peak) was seen only when large numbers of GLP B cells were transferred. Few cIgA cells appearing in spleen and mesenteric lymph nodes were seen mainly at days 12-15. The early repopulation peak was probably formed by the differentiation and accumulation of IgA blasts, recirculating IgA-bearing cells and some IgA precursors. The former 2 cell types homed directly back, while the later homed indirectly via spleen to the GLP. The late repopulation peak was possibly formed by the homing and differentiation of Peyer's patch IgA precursors arriving in the GLP with membrane immunoglobulins unchanged. These GLP IgA precursors migrated first to the spleen and later back to the GLP, where they differentiated into IgA plasma cells.

Induction and delivery of mucosal immune responses

The oral cavity is relatively deficient in lymphoid tissue capable of directing the inductive and effector events required for effective mucosal IgA responses to local antigen. Recent evidence regarding a common mucosal immune system has raised new possibilities in immunization of mucosal surfaces. This review describes the role of the intestine in contributing to responses at distant mucosal sites and in particular emphasizes the immune potential of the intestine in determining the success of immunization strategies to control dental caries.

Morphological study of antigen-sampling structures in the rat large intestine

Organized lymphoid tissue in the rat colon exists as clusters (colonic lymphoid patches) of intramucosal and submucosal follicles in the proximal, mid, and distal colon, interspersed by solitary follicles. The follicular lymphoid cells of colonic lymphoid patches are separated from the gut lumen by a highly specialized lymphoepithelium which lacks mature goblet cells. Cells of this epithelium are of two types: those characterized by an electron-dense cytoplasm, large numbers of apical vesicles and lysosomes, and prolonged extensions of the apical cytoplasm forming thin partitions between the gut lumen and underlying intercellular spaces; and cells with a less electron-dense cytoplasm, distorted mitochondria, and little endoplasmic reticulum. Both cell types bear normal microvilli and have numerous lateral membrane processes which penetrate large intercellular spaces. A ferritin-India ink label infused into the colonic lumen was preferentially adsorbed onto the surface of this follicle-associated epithelium. Indigenous colonic bacteria were observed penetrating the superficial cytoplasm of the electron-dense cells where they were enclosed in lysosomes and digested. An antigen-sampling role is proposed for the colonic lymphoid patch epithelium.

Distribution of immunoglobulin-containing cells in normal and neonatally thymectomized rats

The distribution of immunoglobulin-containing cells (ICC) of the immunoglobulin A (IgA), IgG, and IgM isotypes was examined in various lymphoid and secretory tissues of rats. The effect of neonatal thymectomy of rats on T cells, B cells, and ICC in these tissues was determined by immunofluorescence. The results showed that although T cells were severely depleted in both lymphoid and secretory tissues of the thymectomized (Tx) rats, Tx and normal rats showed comparable numbers of B cells staining for IgA, IgG, and IgM. After neonatal thymectomy, IgA ICC in both lymphoid and secretory tissues were significantly decreased. However, the Tx rats exhibited a compensatory increase in IgM ICC in the identical tissues. Local injection of normal and Tx rats with Streptococcus mutans 6715 resulted in an increase in all isotypes of ICC in the secretory tissues. Although the primary increase in normal rats was due to IgA ICC, Tx rats exhibited the greatest change in the number of IgM ICC.

Migration of B and T lymphocytes to M cells in Peyer's patch follicle epithelium: an autoradiographic and immunocytochemical study in mice

Intimate interaction occurs between M cells which take up antigens in Peyer's patch epithelium and underlying lymphocytes but the migratory pathway and subtypes of these lymphocytes have remained uncertain. Lymphocyte homing to Peyer's patches was investigated by separating mouse splenic B and T lymphocytes, labeling with [3H]adenosine, reinjecting into syngeneic mice, and localizing them by autoradiography after 17, 40, and 68 hr. Both B and T lymphocytes migrated across postcapillary venules through intercellular spaces. Most labeled cells of both types localized around postcapillary venules in the initial time period, then migrated through the lymphoid mass to follicle epithelium. T-cell number in follicle epithelium peaked at 17 hr but B cells peaked at 40 hr. At each time point more labeled B than T lymphocytes were found below M cells and in follicle epithelium but the differences were only significant (P less than 0.01) at 40 hr. Endogenous B lymphocytes, identified by antibody staining, were also associated with M cells, possibly reflecting a role for secretory antibody in modulating antigen uptake.

Isolation of murine small bowel intraepithelial lymphocytes

A method for the preparation of cellular suspensions of the epithelium of murine small bowel for the purpose of isolation and recovery of intraepithelial lymphocytes employing intraluminal hyaluronidase digestion is described. Discontinuous Percoll centrifugation of these suspensions yielded a 76-82% pure population of intraepithelial lymphocytes. Between 1.6 and 4.0 X 10(7) viable intraepithelial lymphocytes were obtained per gram of gut. The isolate contained approximately equal numbers of T, B and null lymphocytes.

Cellular determinants of mammary cell-mediated immunity in the rat: kinetics of lymphocyte subset accumulation in the rat mammary gland during pregnancy and lactation

The present study was undertaken to determine the pattern of T- and B-lymphocyte accumulation in the rat mammary gland during gestation, lactation, and weaning. Adoptive transfer experiments were then used to identify the likely source and extent of immigrating cells during this period. Our results indicate that the mammary gland accumulates levels of cIgA+, W3/13+, and W3/25+ cells that are approximately one third their densities in the small intestine. Cell migration studies suggest that the input of immigrating B-cells is not sufficient to maintain these levels in the gland, and we have suggested that substantial clonal expansion must occur among cIgA+ B-cells within the mammary tissue itself. The migratory behavior of T-cells, their pattern of appearance throughout gestation and lactation, and the predominance of the T-helper cell subset suggest that T-cells regulate B-cell development within the mammary gland.

Analysis of the effector functions of different populations of mucosal lymphocytes

Lymphocytes separated from the epithelial layer of mouse small intestine, IEL, were tested for their NK cytotoxicity against Yac-1 targets. There was little NK activity in a 4 hour assay, but high activity in an 18 hour assay, and the NK activity of IEL did not parallel that in the spleen in any of the mouse strains tested. Furthermore, IEL exerted a suppressor activity on mouse spleen NK activity. Specific T-cell cytotoxicity appeared in IEL in mice immunized with an intraperitoneal injection of P-815 tumor cells. By contrast with IEL, LPL had little NK or NK suppressor activity, but higher levels of specific T-cell cytotoxicity in tumor-immunized mice than intraepithelial lymphocytes. A high proportion of IEL had granules that stained with Giemsa and Astra blue. Furthermore many IEL carried Lyt-2+ phenotype and no other T-cell surface antigen. Intraepithelial lymphocytes appeared, therefore, to have staining and phenotype characteristics of both granular NK cells and suppressor cells. It was clear that the intestinal mucosa contained populations of immune effector cells that were heterogeneous in nature and function.

Studies on the local immune response of the lactating ewe infected with Ostertagia circumcincta

Changes in the flow and composition of gastric lymph were monitored in groups of lactating and non-lactating ewes which were repeatedly infected with Ostertagia circumcincta. As judged by faecal egg counts and worm burdens the lactating group was more susceptible than the non-lactating controls to the challenge infection. Increased amounts of pepsinogen as well as larger numbers of eosinophils and neutrophils entering the gastric lymph indicated considerable abomasal damage and inflammation in the lactating ewes. However, measurement of the flow of lymphocytes as well as the amount of IgA and IgA antibody in the lymph did not indicate that these aspects of the local immune response were impaired during lactation; in fact the output of IgA-containing lymphocytes as well as IgA itself was usually raised in the lactating sheep. Lymph flow was increased and lymph globulin concentrations proportionally decreased in lactating ewes, irrespectively of whether they were infected with worms. It is suggested that these changes were caused by the increase in voluntary food intake which occurs during lactation.

Germinal center B cells lack homing receptors necessary for normal lymphocyte recirculation

Germinal center B cells (GCLC) are a discrete population of antigen-activated lymphoblasts that lack surface IgD and express abundant cell surface binding sites for peanut agglutinin (PNA). These phenotypic features render GCLC easily distinguishable from nearly all plasma cells, T cells, and unstimulated B cells, and have enabled us to identify and isolate GCLC from antigen-stimulated murine lymphoid organs. We have examined the migratory properties of these lymphoblasts in (a) short-term in vivo homing studies, and (b) an in vitro assay of lymphocyte binding to post-capillary, high endothelial venules (HEV) in frozen sections of Peyer's patches and peripheral lymph nodes. In the in vivo experiments, intravenously injected GCLC failed to migrate in significant numbers to peripheral lymphoid organs in comparison with T cells or IgD+ B cells. In the in vitro binding assay, GCLC did not adhere to HEV in either Peyer's patch or peripheral node sections. A variety of factors, such as preferential sequestration in the liver, may operate in vivo to influence the localization of these cells. However, their nearly total failure to migrate into lymphoid organs can best be explained by their inability to recognize and adhere to the specialized HEV which normally mediate the emigration of recirculating lymphocytes from the blood into these sites. The concept that GCLC fail to express functional homing receptors for HEV has been further supported by studies using MEL-14, a monoclonal antibody that appears to recognize the lymphocyte surface receptor for peripheral node HEV: In contrast to most peripheral lymphocytes, GCLC fail to bind MEL-14. These migratory and endothelial-recognition properties of GCLC, when viewed in the context of the possible role of these cells as precursors of plasma cells and/or memory B cells, have led us to propose that the inability of GCLC to recognize HEV may be transient and related to a phase of sessile B cell differentiation.

Alpha-chain disease without qualitative serum IgA abnormality. Report of two cases, including a "nonsecretory" form

Two cases of alpha-chain disease (alpha-CD) without detectable amounts of alpha-CD protein in serum are reported. Both patients presented with the usual epidemiologic, clinical and pathologic features of alpha-CD, including a predominantly plasmacytic infiltration of the whole small intestine, alpha-CD Protein was found in the jejunal fluid in one case but not in the other, the latter presenting as a "nonsecretory" form of alpha-CD. In both cases, immunofluorescence study of the small bowel mucosa showed that most of the infiltrating cells were positive for alpha chains and negative for other isotypes, and kappa and lambda light chains. These findings warrant a reevaluation of the prevalence of alpha-CD protein synthesis in immunoproliferative small intestinal disease (IPSID) (including lymphomas previously described as Mediterranean lymphoma) by performing an adequate search for alpha-CD protein in the jejunal juice, and at the cellular level in patients without detectable amounts of this protein in serum.

Antigen-trapping cells in Peyer's patches of the rat

Peyer's patches (PP) of rats were studied with immunocytochemical, enzyme cytochemical, and electron microscopic techniques. Special attention was paid to the non-lymphoid antigen-trapping cells. Follicular dendritic cells, morphologically comparable to those present in spleen and lymph nodes, were found in the germinal centers of PP. Ia-positive cells with the morphological characteristics of interdigitating cells were seen in the interfollicular area. Thus, both types of antigen-presenting cells, characteristic for B- and T-cell areas, respectively, were present in PP.

Gut and bronchus associated lymphoid tissue: an overview

The mucosal lymphoid aggregates in lung and gut contain precursor cells destined to seed mucosal tissues with IgA containing cells. It is likely that both tissues also are sources of both IgE B cell precursors as well as the cells responsible for their regulation (38). Although the traffic of B cells is relatively well understood from the standpoint of IgA, the factors responsible for localization in mucosal tissue are at best unclear. Furthermore, little is known about the migration patterns of helper or suppressor cells of any kind and these may have, if derived from mucosal tissue, a preferential site of action in the mucosa. The mucosal mast cell may well be a cell of different lineage than the mast cell from the peritoneal cavity and may itself have a mucosal localization pattern. Much more work needs to be done to render some of these speculations onto a better factual base, and to harness these systems in provision of better approaches to vaccination and control of disease.

Natural killer activity of gut mucosal lymphoid cells in mice

Intraepithelial lymphoid cells (IEL) obtained from the mucosa of mouse small intestine were tested for natural killer (NK) activity in a 20-h 51Cr-release assay against YAC-1 and RLfemale1 tumor cells. It was found that IEL possess a strong NK activity, higher than spleen cells, whereas lymphocytes from Peyer's patches (PPL) did not show any NK activity. The cytotoxic activity of IEL could be boosted by interferon (IFN-beta ) treatment, but no NK activity could be induced in PPL by in vitro IFN-beta treatment. The characteristics of the NK effector cells present in the mouse intestine strictly resemble those of NK cells in the spleen. In fact, intestinal NK activity is not affected by depletion of adherent cells and only partially reduced by anti-Thy-1.2 antibodies plus complement. Moreover, the age dependency of NK cytotoxicity is identical for IEL and spleen. Finally, NK-insensitive target cells are not lysed by IEL.

Regional blood flow and the localisation of lymphoblasts in the small intestine of the mouse: effect of an elemental diet

To test the hypothesis that food antigens influence the in vivo migration of lymphoblasts to the small intestine, the effect of an elemental diet (Vivonex) on the distribution of lymphoblasts within the small intestine of mice has been examined. Viable lymphoblasts from the mesenteric nodes of conventionally fed animals were labelled in vitro and given intravenously to recipient mice fed either a standard diet or elemental diet. The localisation of these cells within the small intestine was altered in the animals fed the elemental diet but only in the distal half of the small intestine. The relationship of the localisation of blast cells to the delivery of cardiac output along the small intestine was examined by assessing cell localisation in conjunction with the distribution of an isotopic indicator (86RbC1). The results show that the pattern of localisation of lymphoblasts within the small intestine is related to the probability that they will be delivered to different regions by the blood stream. Therefore, the alterations in blast localisation in the small intestine of animals of the elemental diet can be viewed as a consequence of changes in the perfusion of the distal small intestine. These results do not support the concept that antigens directly influence the efficiency with which blast cells migrate into the intestinal mucosa.

Systemic tolerance and secretory immunity after oral immunization

Diminished systemic immune reaction after ingestion of antigen has been reported in several animal models. Conversely, it has been reported recently that oral immunization may lead to the production of secretory antibodies. To determine whether these events could occur concurrently, CBA/J mice were immunized intragastrically with varying doses of ovalbumin (OVA) and Streptococcus mutans. After 7 d, the animals were challenged systemically with antigen in complete adjuvant and 8 d later serum and saliva taken, and the draining lymph nodes assayed for a proliferative response. Intragastric doses of 1 mg OVA or 10(9) S. mutans led to significant suppression of the proliferative response, and intragastric doses of 10 mg OVA or 2.5 X 10(9) S. mutans led to the production of detectable salivary antibodies using hemagglutination. Serum antibodies were not detected after intragastric administration of OVA or S. mutans. Suppression of the proliferative response could be detected from 2-60 d after intragastric administration of OVA, and 2-21 d after S. mutans. Prior intragastric immunization with heterologous antigens did not suppress the response to OVA or S. mutans. Transfer of 40 X 10(6) mesenteric lymph node cells from mice given 20 mg OVA or 10(9) S. mutans led to suppression of the proliferative response in syngeneic recipients. Salivary antibodies wer removed by absorption with anti-IgA, but not anti-IgG or IgM, indicating that they were of the IgA class. It appears that intragastric administration of soluble or particulate antigens in mice may lead to the concurrent induction of salivary antibodies and systemic suppression.

Direct fluorescent labeling of cells with fluorescein or rhodamine isothiocyanate. II. Potential application to studies of lymphocyte migration and maturation

The effect of direct cell labeling with fluorescein or tetramethyl rhodamine isothiocyanate on lymphocyte migration is examined. In vitro conditions of labeling are defined which (1) do not significantly affect immediate or long term viability of lymphocytes (up to 2 weeks after transfer in vivo), (2) do not alter normal lymphocyte migration, (3) do not affect expression or detectability of surface antigens, and (4) permit direct visualization and counter-staining with fluorescent antibody reagents for days after intravenous injection. The potential application of this method to studies of lymphocyte migration and maturation is discussed.

Characterization of mouse thoracic duct B lymphocytes. I. Evidence of functional heterogeneity

Thoracic duct lymphocytes (TDL) from C3H/Tif and BALB/c mice were studied for their in vitro reactivity to the B cell mitogens lipopolysaccharide (LPS) and lipoprotein (LP). Roughly 4% and 10% of the surface immunoglobulin (Ig)-positive cells in these populations could be stimulated by LPS and LP, respectively, to generate clones of IgM-secreting cells. Among LPS-reactive B cells, roughly 30% developed into clones which also produced IgG3 or IgG2, while only a very small fraction (1-2%) of all precursors could give rise to clones secreting IgG1 and IgA. Freshly collected TDL from some batches of C3H/HeJ mice displayed a high proportion of Ig-containing B cell blasts (5-10%), which did not secrete enough Ig to be detected as plaque-forming cells (PFC). These cells, however, under appropriate culture conditions and stimulated by LP (but not by Nocardia mitogen), differentiated to PFC of the various Ig classes without dividing.

Organ specificity of lymphocyte migration: mediation by highly selective lymphocyte interaction with organ-specific determinants on high endothelial venules

Evidence is presented that the organ specificity of lymphocyte migration is determined by selective interaction of lymphocytes with specialized endothelial cells. Mouse Peyer's patch and lymph node lymphocytes bind preferentially to high endothelial venules (HEV) in frozen sections of Peyer's patches and peripheral nodes, respectively, and this in vitro binding preference accurately predicts their differential segregation in vivo 30 min after i.v. injection. Both in vivo and in vitro, about 1.4 times as many as many Peyer's patch as lymph node lymphocytes bind HEV in Peyer's patches, and, conversely, twice as many lymph node cells interact with HEV in nonmesenteric lymph nodes. Even greater specificity is shown by certain homogeneous lymphocyte populations, i.e. thymic lymphomas. Some lymphomas bind with remarkable selectivity to HEV in Peyer's patches, and others interact almost exclusively with those in lymph nodes indicating that the mechanisms mediating selective recognition of HEV are capable of nearly absolute discrimination. Mesenteric node HEV are unique in that they allow both Peyer's patch- and lymph node-specific cells to bind. It is proposed that lymphocyte surface receptors specific for organ-restricted endothelial cell determinants mediate the antigen-independent organ specificity of lymphocyte migration. According to this model, there are at least 2 sets of complementary lymphocyte and endothelial cell receptors, one mediating lymphocyte-HEV adherence in Peyer's patches, the other in lymph nodes.

Breast milk lymphocyte response to K1 antigen of Escherichia coli

Comparison milk and blood lymphocyte blastogenic responses to the K1 antigen of Escherichia coli and lipopolysaccharide (LPS) from E. coli O127,B8 were examined in 16 postpartum women by [3H]thymidine uptake. Rabbit hemolysincoated sheep erythrocyte monolayers were used to deplete macrophages from milk lymphocyte preparations and to enrich for T lymphocytes in order to make milk preparations more comparable to blood preparations. Response was defined as a stimulation index of greater than or equal to 2.0. There was no evidence of selective response to K1 antigen by milk lymphocytes, since both blood and milk lymphocytes responded in four women and neither blood nor milk lymphocytes responded in nine. Milk lymphocytes alone responded to K1 in one woman, whereas blood lymphocytes alone responded in two women. Additional nonpaired milk or blood cultures were available from three women. None of these responded to K1 antigen. Corresponding lymphocyte cultures were stimulated with LPS. A positive K1 response was always accompanied by an LPS response, and the LPS response correlated with the K1 response in 17 of 19 women. Stool cultures examined with an antiserum agar showed no correlation between the presence of K1 E. coli in the stool and milk or blood lymphocyte response to K1 antigen. In the system used here, no selectivity of response of breast milk lymphocytes to K1 antigen was noted.

Interepithelial cells of the oral mucosa. Light and electron microscopic observations in germfree, specific pathogen-free and conventionalized mice

Interepithelial cells are found in all epithelia of the internal and external surfaces of the mammalian body. The regional differences of these interepithelial cells and their function are not completely known so far. The quantitative and qualitative changes of the interepithelial cell population were investigated in germfree, specific pathogen-free and conventionalized mice by light and electron microscopy. Germfree and specific pathogen-free animals did not show significant differences in the number of interepithelial cells. In the epithelium of the tongue a mean of 7.4 cells per 1000 basal cells is found. After conventionalization a significant increase to 14.4 interepithelial cells per 1000 basal cells is observed. The number of cells in the buccal epithelium is constantly about 20% higher than in the epithelium of the tongue. In the oral mucosa lymphocytes, cerebriform cells and Langerhans cells are an integral component of the epithelium. In contrast to the monostratified intestinal mucosal epithelium, which is considered a secondary lymphatic tissue, the interepithelial lymphocytes of the oral mucosa are not significantly decreased in germfree animals. This could indicate that the oral mucosa functions partly as a primary lymphatic tissue. Interepithelial cerebriform cells and Langerhans cells increased after conventionalization with a maximum after 10 days in response to exogenous antigens. Both cells are immunologically important. The observations prove that the oral mucosa represents a local immunologic system in which the Langerhans cells plays an important part by formation a reticulo-epithelial tissue.