The regulation of gut mucosal IgA B-cell responses: recent developments.

The majority of activated B cells differentiate into IgA plasma cells, with the gut being the largest producer of immunoglobulin in the body. Secretory IgA antibodies have numerous critical functions of which protection against infections and the role for establishing a healthy microbiota appear most important. Expanding our knowledge of the regulation of IgA B-cell responses and how effective mucosal vaccines can be designed are of critical importance. Here we discuss recent developments in the field that shed light on the uniqueness and complexity of mucosal IgA responses and the control of protective IgA responses in the gut, specifically.

Cholera toxin adjuvant promotes a balanced Th1/Th2/Th17 response independently of IL-12 and IL-17 by acting on Gsα in CD11b⁺ DCs.

Despite an extensive literature on the mechanism of action of cholera toxin (CT), we still lack critical information about how the toxin acts as an adjuvant and, especially, which dendritic cells (DCs) are the target cells. Although a T helper type 2 (Th2)-skewing effect of CT is most commonly reported, effective priming of Th17 cells as well as suppression of Th1 responses are well documented. However, the ability of CT to block interferon regulatory factor 8 (IRF8) function and interleukin (IL)-12 production in DCs, which blocks CD8α DC and Th1 cell development, is inconsistent with priming of Th1 and CD8 T cells in many other reports. This prompted us to investigate the adjuvant effect of CT in wild-type, IL-12p40-/-, Batf3-/-, and IL-17A-/- mice and in mice that selectively lack the Gsα target protein for CT adenosine diphosphate (ADP)-ribosylation in DCs. We found that CT promoted Th1 priming independently of IL-12, and whereas Th2 and also Th17 responses were augmented, the gut IgA responses did not require IL-17A. Adjuvanticity was intact in Batf3-/- mice, lacking CD8α(+) DCs, but completely lost in mice with Gsα-deficient CD11c cells. Thus, our data demonstrate that the adjuvant effect requires Gsα expression in CD11b(+) DCs, and that priming of mucosal IgA and CD4 T cells appears unbiased and is independent of IL-12 and IL-17A.

Re-utilization of germinal centers in multiple Peyer's patches results in highly synchronized, oligoclonal, and affinity-matured gut IgA responses.

Whereas gut IgA responses to the microbiota may be multi-centered and diverse, little is known about IgA responses to T-cell-dependent antigens following oral immunizations. Using a novel approach, gut IgA responses to oral hapten (4-hydroxy-3-nitrophenyl)acetyl-cholera toxin (NP-CT) conjugates were followed at the cellular and molecular level. Surprisingly, these responses were highly synchronized, strongly oligoclonal, and dominated by affinity matured cells. Extensive lineage trees revealed clonal relationships between NP-specific IgA cells in gut inductive and effector sites, suggesting expansion of the same B-cell clone in multiple Peyer's patches (PPs). Adoptive transfer experiments showed that this was achieved through re-utilization of already existing germinal centers (GCs) in multiple PPs by previously activated GC GL7(+) B cells, provided oral NP-CT was given before cell transfer. Taken together, these results explain why repeated oral immunizations are mandatory for an effective oral vaccine.

Gene expression profiling identifies STAT3 as a novel pathway for immunomodulation by cholera toxin adjuvant.

Earlier studies have reported on both proinflammatory and anti-inflammatory activities of cholera toxin (CT). As CT is a powerful adjuvant, we were interested in identifying genes with a possible involvement in these functions. A global gene expression analysis in mouse B cells showed that CT regulated <100 annotated genes, which encoded transcription factors, G proteins, cell-cycle regulators, and immunoregulating molecules. Interestingly, CT regulated the expression of the signal transducer and activator of transcription (STAT)3 gene and influenced the level and activation of both isoforms STAT3 alpha and STAT3 beta, in vitro in a B-cell line and in Peyer's patch (PP) B cells and in vivo in freshly isolated splenic B cells from CT-treated mice. This effect was cAMP dependent and was not seen with CTB. B cells pre-exposed to CT were significantly more susceptible to the activation of STAT3 by interleukin (IL)-6 and IL-10. This exerted a stronger inhibitory effect of IL-10 on lipopolysaccharide (LPS)-stimulated B-cell proliferation and cytokine production (IL-6). Moreover, IgG1 and IgA production induced by LPS and IL-10 were enhanced by the addition of CT to cultures of PP or splenic B cells. This is the first study to provide a molecular mechanism that can reconcile previous findings of proinflammatory and anti-inflammatory effects by CT adjuvant.

Impaired CD40-signalling in CD19-deficient mice selectively affects Th2-dependent isotype switching.

Activation of B lymphocytes involves binding of antigen to the specific receptor and signalling through several membrane coreceptors, of which CD19 has been found to play a pivotal role as a response regulator. Although previous studies in CD19 gene knockout mice have demonstrated that antibody responses to T-cell-dependent antigens are strongly impaired in the absence of this coreceptor, little is known about the consequences of CD19 deficiency for the interaction between T and B cells. Here we report that Th2 co-ordinated B-cell differentiation is selectively impaired in CD19-deficient mice in response to mucosal or systemic immunizations or following an intestinal infection with Nippostrongylus brasiliensis. Whereas immunoglobulin (Ig)G1 or IgE antibody responses were low or absent, IgG2a responses were normal. The selective defect was not caused by a poor Th2-development or interleukin (IL)-4 responsiveness in CD19-deficient mice. Rather, it was the result of an impaired Th2-B cell interaction, owing to a substantially reduced ability to signal via CD40 in CD19-deficient B cells. Thus, our study in CD19-deficient mice suggests that CD40L-CD40-interactions are more important for Th2 than for Th1 co-ordinated B-cell differentiation.

Measurement of immunoglobulin synthesis using the ELISPOT assay.

The enzyme-linked immunospot (ELISPOT) assay for detection of single antibody-secreting cells has become the best alternative method to the conventional plaque-forming cell (PFC) assays. Among its several advantages are better antigen stability and specificity as well as fewer limitations in the diversity of antigens that can be used in the assay. In addition, the ELISPOT assay can be used to detect two antigenically different secreted antibodies simultaneously by two-color analysis and offers the unique possibility of quantifying the number of antibody molecules secreted per cell. Finally, the assay can be used to detect single antibody-secreting cells in tissues that usually confront the immunologist with difficulties, e.g., gut lamina propria from humans or mice. This unit presents the ELISPOT assay in three steps: coating of antigen to a solid phase, incubation of antibody-producing cells in appropriate dilution, and detection of the antigen-antibody complex formed at the site of the active antibody-secreting cell. The assay can be performed using polystyrene plates or nitrocellulose membrane in microtiter plates or using nitrocellulose membrane in a blotting manifold.

CD19-deficient mice exhibit poor responsiveness to oral immunization despite evidence of unaltered total IgA levels, germinal centers and IgA-isotype switching in Peyer's patches.

CD19 exhibits a critical role as a response regulator in B cells, influencing activation, differentiation and survival. Accordingly, CD19-deficient mice largely lack B-1 cells, and their conventional B-2 cells are poor responders to thymus-dependent antigen. Since both B-1 and B-2 cells may contribute to the total intestinal IgA production, we investigated whether lack of CD19 negatively affected mucosal immunity. We found that CD19(-/-) mice have near normal total IgA levels in serum and gut mucosa and, contrary to systemic lymphoid tissues, Peyer's patches (PP) had germinal centers to which also IgA+ B cells localized. However, the mice demonstrated severely impaired responses to oral immunization with keyhole limpet hemocyanin plus cholera toxin adjuvant. Mucosal responses to oral immunization were significantly more impaired than systemic responses. Despite normal specific IL-4 production, a selective defect in Th2-regulated B cell isotypes was observed, with poor or no mucosal IgA, low serum IgG1 and no IgE, but intact serum IgA and IgG2a production. Ex vivo experiments revealed strongly inhibited CD40-stimulated proliferation and IgA differentiation in CD19-deficient PP B cells. Taken together, an impaired CD40 responsiveness selectively affected Th2, but not Th1, coordinated B cell responses. The CD19(-/-) mice provide compelling evidence for the differential regulation of serum and mucosal IgA immunity.

Adjuvanticity of the cholera toxin A1-based gene fusion protein, CTA1-DD, is critically dependent on the ADP-ribosyltransferase and Ig-binding activity.

The ADP-ribosylating enterotoxins, cholera toxin (CT) and Escherichia coli heat-labile toxin, are among the most powerful immunogens and adjuvants yet described. An innate problem, however, is their strong toxic effects, largely due to their promiscuous binding to all nucleated cells via their B subunits. Notwithstanding this, their exceptional immunomodulating ability is attracting increasing attention for use in systemic and mucosal vaccines. Whereas others have separated adjuvanticity from toxicity by disrupting the enzymatic activity of the A1 subunit by site-directed mutagenesis, we have constructed a nontoxic molecule that combines the full enzymatic activity of the A1 subunit with a B cell targeting moiety in a gene fusion protein, the CTA1-DD adjuvant. Despite its more selective binding properties, we found comparable adjuvant effects of the novel CTA1-DD adjuvant to that of CT. Here we unequivocally demonstrate, using a panel of mutant CTA1-DD molecules, that the immunomodulating ability of CTA1-DD is dependent on both an intact enzymatic activity and the Ig-binding ability of the DD dimer. Both agents, CT and CTA1-DD, ADP-ribosylate intact B cells. However, contrary to CT, no increase in intracellular cyclic AMP in the targeted cells was detected, suggesting that cyclic AMP may not be important for adjuvanticity. Most remarkably, CTA1-DD achieves similar immunomodulating effects to CT using a ganglioside-GM1 receptor-independent pathway for internalization.

Requirements for B7-CD28 costimulation in mucosal IgA responses: paradoxes observed in CTLA4-H gamma 1 transgenic mice.

The block in the CD80/CD86-CD28/CTLA-4 pathway in CTLA4-H gamma 1 transgenic (Tg) mice results in strongly impaired systemic IgG immunity and failure to develop germinal center reactions. By contrast, here we report that mucosal immunity and IgA B cell differentiation are not affected by this block. We found abundant germinal centers and evidence of IgA switch differentiation in Peyer's patches, normal total IgA levels, and normal numbers of IgA-labeling cells in the gut mucosa. The distribution of B-1 and B-2 cells and the relative contribution of B-1 cells to the total IgA B cells were similar in Tg and wild-type mice. Despite this, oral immunizations with keyhole limpet hemocyanin plus cholera toxin adjuvant failed to stimulate Ag-specific mucosal IgA responses in CTLA4-H gamma 1 Tg mice. This was not due to a lack of adjuvant activity of cholera toxin in Tg mice, nor was this secondary to an inability to take up Ag from the gut lumen. Rather, CD4+ T cells stimulated by oral immunization in Tg mice appeared to be inappropriately primed, as evidenced by a significantly reduced level of CD40 ligand and CD44 expression and an increased expression of CD95 compared to those in wild-type mice. This study reveals a paradox in the regulation of mucosal IgA responses.

Lack of local suppression in orally tolerant CD8-deficient mice reveals a critical regulatory role of CD8+ T cells in the normal gut mucosa.

We found that feeding keyhole limpet hemocyanin (KLH) to CD8-deficient (CD8-/-) mice induced oral tolerance that was comparable in both magnitude and quality to that induced in wild-type (wt) mice. The tolerance was dose dependent, and only higher doses of KLH caused significant reduction in specific Ab and T cell responses. Both Th1 and Th2 CD4+ T cell functions were affected. Feeding KLH together with cholera toxin (CT) adjuvant, however, abrogated the induction of oral tolerance equally well in CD8-/- and wt mice. On the contrary, CT adjuvant was unable to abrogate already established oral tolerance in both CD8-/- and wt mice. Most importantly, whereas Ag feeding induced hyporesponsiveness in systemic as well as in local gut IgA responses in wt mice, a lack of local suppression was evident in orally tolerant CD8-/- mice following oral immunizations. Thus, contrary to the situation in wt mice, Ag feeding induces systemic, but not local, gut IgA hyporesponsiveness in CD8-/- mice, suggesting that CD8+ T cells in the normal gut mucosa exert an important down-regulatory function. In wt mice the local suppression extended to an unrelated Ag, OVA, given together with KLH and CT adjuvant, i.e., bystander suppression. Based on these results we propose that tolerance induced by feeding Ag is highly compartmentalized, requiring CD8+ T cells for local suppression of IgA responses, whereas systemic tolerance may affect CD4+ T cells of both Th1 and Th2 types independently of CD8+ T cells. Finally, the adjuvant effect of CT abrogates induction, but not established, oral tolerance through a mechanism that does not require CD8+ T cells.

Interferon-gamma receptor-deficient mice exhibit impaired gut mucosal immune responses but intact oral tolerance.

Interferon-gamma (IFN-gamma) receptor knock-out (IFN-gamma R -/-) mice were used to analyse the role of IFN-gamma in mucosal immune responses following oral immunization. We found that the IFN-gamma R -/- mice demonstrated 50% reduced spot-forming cell (SFC) responses in the gut lamina propria and spleen after oral immunization with keyhold limpet haemocyanin (KLH) plus cholera toxin (CT) adjuvant. The IFN-gamma R -/- mice exhibited 10-fold reduced total serum KLH-specific antibody levels compared with wild-type mice after oral immunization, while after intravenous immunization, no such difference was seen, suggesting a selective impairment of mucosal immune responses. Moreover, oral immunizations resulted in impaired interleukin-4 (IL-4), IL-10 and IFN-gamma production by spleen T cells from IFN-gamma R -/- mice, indicating that no reciprocal up-regulation of Th2-activities had occurred despite the lack of IFN-gamma R function. No reduction in Th1 or Th2 cytokines was observed following systemic immunizations. Despite potentially strong modulating effects of IFN-gamma on epithelial cell IgA transcytosis and electrolyte barrier functions, CT-immunized IFN-gamma R -/- mice demonstrated unaltered protection against CT in ligated intestinal loops together with normal anti-CT IgA and total IgA levels in gut lavage. Oral feeding with KLH followed by parenteral immunization resulted in strongly suppressed SFC numbers and reduced cell-mediated immunity in both wild-type and IFN-gamma R -/- mice. CT-adjuvant abrogated induction of oral tolerance in both IFN-gamma R -/- and wild-type mice. Collectively, our data argue that the two major response patterns induced by oral administration of protein antigen, i.e. active IgA immunity and oral tolerance, are differently regulated. Thus, IFN-gamma R -/- mice have impaired mucosal immune responses while induction of oral tolerance appears to be unaffected by the lack of IFN-gamma functions.

Genetically engineered nontoxic vaccine adjuvant that combines B cell targeting with immunomodulation by cholera toxin A1 subunit.

Cholera toxin (CT) is an exceptionally potent adjuvant but, unfortunately, also very toxic. Here we present a powerful new approach to separate toxicity from adjuvanticity by constructing a fusion protein that combines the enzymatically active cholera toxin A1 subunit (CTA1) with targeting to B cells. The CTA1 was genetically linked at its C-terminal end to two Ig-binding domains, DD, of staphylococcal protein A and produced in Escherichia coli. The highly purified, monomeric CTA1-DD fusion protein, with a molecular mass of 37 kDa, was found to exhibit strong ADP-ribosyltransferase activity and bound, via the DD moiety, to both Fc and Fab fragments and to all IgG subclasses--IgE, IgA, and IgM. After i.v. injection of the fusion protein, FACS analysis revealed binding of CTA1-DD to splenic IgM+ B cells, but not CD3+ T cells, indicating cell-specific targeting in vivo. Strikingly, we found that the adjuvant ability of CTA1-DD to enhance systemic IgG as well as mucosal IgA responses to the unrelated Ags, OVA, or keyhole limpet hemocyanin, administered i.v or intranasally, was comparable to that of intact CT. In addition, the enhancing effect on specific IgG1, IgG2a, and IgG2b responses mimicked that of CT and suggested involvement of both Th1 and Th2 CD4+ T cell activity. The CTA1-DD, as well as CT, up-regulated expression of the CD80 and CD86 molecules on the targeted B cells, indicating that enhanced T cell costimulation may be responsible for the adjuvant effect. Contrary to CT, however, CTA1-DD was completely nontoxic. Thus, the CTA1-DD adjuvant should find general applicability in systemic and mucosal vaccines, and the strategy used may also be explored for other regimens requiring targeted immunomodulation.

IL-6-deficient mice exhibit normal mucosal IgA responses to local immunizations and Helicobacter felis infection.

Using IL-6-deficient (IL-6 -/-) or wild-type mice, we investigated whether IL-6 is involved in the intestinal adjuvant activity of cholera toxin (CT) and to what extent IL-6 is required for mucosal IgA responses against soluble protein Ags or live Helicobacter felis infection. In naive IL-6 -/- mice we found normal total IgA levels in serum, bronchial and intestinal lavage and unaltered frequencies of IgA plasma cells in intestinal lamina propria. In Peyer's patches (PP) and mesenteric lymph nodes (MLN) IgA-producing cells were as frequent in IL-6 -/- as in wild-type mice. Immunohistochemical analysis of PP revealed germinal centers that co-localized IgA+ cells, indicating B cell activation and isotype switching in situ in the intestinal immune inductive site. Phenotypic analysis of the distribution of conventional B-2 cells (B220+CD5-/Mac-1-) and B-1 cells (B220+, CD5+/Mac-1+) in intestine-associated tissues gave comparable results in IL-6 -/- and wild-type mice. The ability to respond with mucosal IgA following oral and intranasal immunization with specific Ag, KLH or OVA, in the presence of CT adjuvant or to live H. felis infection was similar in IL-6 -/- and wild-type mice. CT exerted strong and comparable adjuvant functions in IL-6 -/- and wild-type mice. Repeated oral immunizations with CT alone stimulated immune protection against CT-induced diarrhea in ligated loops that was of similar magnitude in IL-6 -/- and wild-type mice. We conclude that, although IL-6 has been ascribed a crucial role in terminal differentiation of IgA B cells in vitro, we found no evidence to support the notion that IL-6 is critically required for IgA B cell development or specific mucosal IgA responses in vivo.

Paradoxical IgA immunity in CD4-deficient mice. Lack of cholera toxin-specific protective immunity despite normal gut mucosal IgA differentiation.

Using normal and CD4 gene-targeted (CD4-/-) mice, we asked whether mucosal immune responses and IgA B cell differentiation require the presence of CD4+ T helper cells. We found that CD4-/- mice had numerous B cell germinal centers in Peyer's patches and other gut-associated lymphoid tissues. Membrane IgA+ B cells were found to co-localize to germinal center areas and CD4-CD8- double negative CD3+ T cells had replaced CD4+ T cells in the follicular areas of the Peyer's patches. CD4-/- mice had normal levels of IgA-producing cells in gut-associated lymphoid tissues, and gut lavage contained unaltered levels of total IgA. However, despite T cell help for IgA B cell differentiation, CD4-/- mice did not respond with Ag-specific intestinal IgA following oral immunization with the powerful mucosal immunogen cholera toxin (CT). By contrast, these mice demonstrated serum alpha-CT IgG following oral immunization, suggesting that double negative CD3+ T cells provided some help for systemic immune responses after oral immunization. Perorally immunized CD4-/- mice were completely unprotected against CT-induced diarrhea while both normal and CD8-/- mice were well protected and also demonstrated high levels of gut mucosal alpha-CT IgA. After reconstitution of the CD4-/- mice by adoptive transfer of naive mesenteric lymph node CD4+ T cells, the mice acquired the ability to respond with specific mucosal immune responses following oral immunization and also developed resistance against CT-induced diarrhea. Thus, paradoxically, although IgA B cell differentiation appears to proceed normally in CD4-/- mice, specific gut mucosal immune responses are grossly impaired in the absence of CD4+ T cells.

Cholera toxin promotes B cell isotype switching by two different mechanisms. cAMP induction augments germ-line Ig H-chain RNA transcripts whereas membrane ganglioside GM1-receptor binding enhances later events in differentiation.

In a recent study we provided evidence that isotype switching in LPS-stimulated murine B cells was greatly enhanced by cholera toxin (CT). We found that CT acted synergistically with IL-4 to promote IgG1 differentiation at the gene transcriptional level by strongly enhancing the expression of germ-line gamma 1-RNA transcripts. In this study we ask which mechanisms are responsible for the isotype-switching effect of CT on B cells and which second messenger systems are involved in this process. We found that at least two different mechanisms are involved: 1) increased intracellular cAMP levels stimulated by the A subunit potentiates isotype switching early in differentiation by augmenting the formation of sterile germ-line gamma 1-RNA transcripts and 2) the binding of the nontoxic B subunit to the membrane GM1-ganglioside receptor promotes later stages of differentiation. Although the whole toxin gave up to a ninefold increase in IgG1 differentiation the cAMP-independent effect of rCTB gave at most a fivefold increase in IgG1 differentiation as compared to that seen with IL-4 alone. However, on a molar basis whole CT was at least a 1000-fold more efficient at promoting B cell switch-differentiation as compared to rCTB. Moreover, IL-4 did not stimulate cAMP in murine B cells and its effect on LPS-stimulated B cell differentiation was not decreased by inhibitors of cAMP-dependent protein kinases. However, CT's effect on B cell switch differentiation was blocked by inhibitors of protein kinases and could be partially mimicked by dibutyryl cAMP. In contrast to CT, the enhancing effect of rCTB on IgG1-differentiation was not affected by blocking of the protein kinases and the combination of rCTB and dBcAMP was as potent as the intact CT in promoting IL-4-stimulated IgG1 differentiation. Finally, the IL-4 pathway, but not the CT pathway, was sensitive to phorbol esters: In IL-4 plus LPS-stimulated B cell cultures IgG1 production was almost completely blocked by PMA. This inhibition was not associated with a decreased B cell proliferation or expression of germ-line gamma 1-RNA transcripts. The addition of CT, and to a significantly lesser extent rCTB, to these cultures enhanced IgG1-differentiation and expression of germ-line gamma 1-RNA transcripts to the same extent as in cultures without PMA. The existence of dual mechanisms operating together on B cell differentiation help to explain the strong adjuvant function by CT on IgG and IgA antibody responses after oral and parenteral immunizations.

Cholera toxin adjuvant promotes long-term immunological memory in the gut mucosa to unrelated immunogens after oral immunization.

This study was conducted to investigate whether cholera toxin (CT), used as a mucosal adjuvant, would promote the development in mice of immunological memory to unrelated antigens administered by the oral route. We found that oral priming immunizations with keyhole limpet haemocyanin (KLH) in combination with CT adjuvant induced long-term, at least 22 months and perhaps lifelong, immunological memory in the intestinal lamina propria (LP). In contrast, oral priming immunizations with KLH alone failed to stimulate immunological memory. Moreover, memory responses in the KLH plus CT-immunized mice were elicited by antigen alone, i.e. without CT adjuvant, suggesting that once immunological memory is established in the intestinal mucosa, e.g. by oral vaccination, elicitation of secondary-type responses does not require the presence of CT and thus could result from re-encounter with specific bacterial or viral antigens in the intestine. We also found that a single priming-dose of KLH plus CT adjuvant was sufficient to stimulate long-term, antigen-specific memory in the intestinal mucosa. Finally, the ability of CT to induce immunological memory in the gut mucosa required the whole toxin and could not be achieved by using the toxoid, the cholera toxin B subunit (CTB), which lacks the adenylate cyclase/cAMP-activating property of the whole molecule. The results support the view that mucosal adjuvants, incorporated into oral vaccines, might be an effective means to achieve long-term immunological memory and protection against pathogenic micro-organisms at mucosal surfaces.

Antigen-specific and polyclonal CD4+ lamina propria T-cell lines: phenotypic and functional characterization.

Surface phenotype and function of lamina propria CD4+ T cells have been evaluated. In addition, long-term, antigen-specific and polyclonal lamina propria CD4+ T-cell lines have been generated and characterized. Lamina propria CD4+ T cells represent approximately 30% of lamina propria lymphocytes and are responsive to a variety of T-cell mitogens, including anti-CD3, concanavalin A, phytohaemagglutinin and pokeweed mitogen. In each case, however, lamina propria T cells are less responsive to these mitogens than spleen T cells. Freshly isolated lamina propria T cells produce substantial amounts of interleukin-2 (IL-2), interleukin-4 (IL-4), gamma interferon and to a lesser extent interleukin-5 (IL-5). Antigen-specific lamina propria CD4+ T-cell lines were generated by orally immunizing animals with antigen (KLH) in conjunction with cholera toxin as an oral adjuvant. Polyclonal lamina propria CD4+ T-cell lines were generated from unimmunized animals using anti-CD3 as a polyclonal stimulus. Both antigen-specific and polyclonal CD4+ T-cell lines were Thy-1+, alpha beta TCR+ and CD8-. The antigen-specific CD4+ T-cell line when stimulated by anti-CD3 and PMA produces predominantly IL-2, IL-4 and gamma interferon, with very little IL-5. In contrast, the polyclonal CD4+ T-cell line when similarly stimulated produces predominantly IL-4 and IL-5, with very little IL-2 and no detectable gamma interferon. In summary, lamina propria CD4+ T cells have been evaluated and in vitro conditions have been determined for successful generation of lamina propria CD4+ T-cell lines.

T lymphocytes that express CD4 and the alpha beta-T cell receptor but lack Thy-1. Preferential localization in Peyer's patches.

Thy-1- T cells expressing CD4 and the alpha beta-TCR have been identified in murine lymphoid tissues. These cells are particularly prevalent in Peyer's patches (PP), representing 17 +/- 3% of PP CD4 T cells, whereas they are much less prevalent in spleen, lymph nodes, lamina propria, or peritoneum. Phenotypic studies of fresh-isolated PP T cells demonstrate that all PP CD4 T cells (both Thy-1- and Thy-1+) express CD3, alpha beta-TCR, and CD5 (Lyt-1), whereas none coexpress CD8 (Lyt-2). Thy-1- and Thy-1+ CD4 T cell lines generated from PP also coexpress CD3 and alpha beta-TCR, but are heterogeneous in expression of CD5 and again do not coexpress CD8. Further studies revealed that Thy-1- CD4+ T cells were not present in nude mice. Short term stimulation of Thy-1+ CD4+ PP T cells with anti-CD3 resulted in loss of Thy-1 in a substantial fraction of these cells. Functional studies of Thy-1- and Thy-1+ CD4+ PP T cells indicate that fresh-isolated Thy-1- CD4+ cells do not proliferate in response to insoluble anti-CD3 but do proliferate when stimulated with soluble anti-CD3 in the presence of feeder cells. In contrast, Thy-1+ CD4+ cells proliferate well to both stimuli. However, Thy-1- CD4+ PP T cells adapted to in vitro culture exhibit vigorous proliferative responses when stimulated with either form of anti-CD3. Evaluation of lymphokine secretion by fresh-isolated Thy-1- and Thy-1+ CD4+ PP T cells revealed that both make substantial amounts of IL-2; however, Thy-1- T cells made less IL-4 than their Thy-1+ counterparts. Neither population made IL-5 or IFN-gamma. Similarly, Thy-1- and Thy-1+ CD4 T cell lines made similar amounts of IL-2; again Thy-1- T cells made less IL-4; and in this case Thy-1- T cells made IL-5 albeit significantly less than the Thy-1+ cells. Finally, immunohistochemical studies suggested that many of the CD4+ T cells in PP germinal centers were Thy-1-, indicating that Thy-1- and Thy-1+ CD4 T cells differ in their distribution within the PP. These studies thus define a phenotypically and functionally distinct T cell population which is most prevalent in murine Peyer's patches.