Altered levels of urokinase on monocytes and in serum of children with AIDS; effects on lymphocyte activation and surface marker expression.

Urokinase (UK) type plasminogen activator is a serine protease produced by activated human monocytes. Despite the well-documented roles played by UK in cell-mediated immunity in healthy humans, the roles played by UK in the derangements of cell-mediated immune responses observed in HIV disease remain largely undefined. In these studies the numbers of peripheral blood lymphocytes and monocytes bearing surface UK (UK+) as well as serum levels of UK (flow microfluorimetry and ELISA, respectively) were determined in children with AIDS and in healthy HIV-negative children. The effects of exogenous UK on lymphocyte activation (cell cycle analysis using living cells) and surface marker (CD3, CD4, CD8, and CD19) expression (flow microfluorimetry using fixed cells) were also studied. Data are expressed as percent total cells. Numbers of UK+ lymphocytes in children with AIDS were similar to those observed in healthy children. In contrast, numbers of UK+ peripheral blood monocytes were dramatically decreased (> 70%) in the children with AIDS. However, serum levels of UK were increased (nearly threefold) in these children. When lymphocytes from these children were cultured with soluble UK, numbers of cells in S phase of cell cycle appeared suppressed. Incubation of fixed lymphocytes from either a child with AIDS or from a healthy child with exogenous UK appeared to increase numbers of cells expressing CD3. Incubation with UK had no effect on expression of any other surface marker (CD4, CD8, or CD19) using cells from the child with AIDS. In contrast, incubation with UK appeared to decrease (fivefold) numbers of cells expressing CD19 and increase numbers of cells expressing CD4 and CD8 only when fixed lymphocytes from a healthy HIV-negative child were used. The results suggest important roles for UK in regulation of lymphocyte surface markers in general and in CD3- and CD19-dependent lymphocyte activation pathways specifically. Furthermore, these studies add to a widening body of evidence implicating UK dysregulation in the pathogenesis of HIV disease and may point to pharmacological opportunities involving UK to delay or prevent progression of HIV infection into full-blown AIDS.

Control of IgE responses. V. Oral administration of a synthetic derivative of the inner bacterial cell wall (SDZ 280.636) to sensitized mice induces isotype specific suppression of peak hapten specific IgE antibody forming cell responses, serum IgE levels and immediate hypersensitivity responses.

SDZ 280.636, a nontoxic diacyl glycerol derivative of muramyl dipeptide (MDP), a component of the inner bacterial cell wall, which is suitable for use in man, suppressed hapten specific IgE antibody forming cell (AFC) responses in spleen, serum levels of hapten specific IgE and hapten specific immediate hypersensitivity (i.h.) responses in skin, when fed to mice at the peak of a hapten specific IgE AFC response. In addition, serum levels of IL-6 appeared increased while IFN gamma was decreased. To induce these IgE responses, BALB/c mice were injected i.p. with BPO-KLH (benzylpenicilloyl-keyhole limpet hemocyanin) (10 micrograms) in aluminum hydroxide gel (alum) on days 0, 21 and 42. Mice were fed (gavage) with either MDP or SDZ 280.636 (1.0 or 10 mg/kg) on day 44, or on days 44, 46 and 48, and killed on days 46 or 50. Numbers of BPO specific AFC in spleen, and serum levels of BPO specific immunoglobulins (IgG1, IgE and IgA) were determined (ELISPOT assay, ELISA). In addition, BPO specific IH responses were measured in these animals. Mice were injected in the right pinna with BPO-BSA (0.1 microgram) and in the left pinna with an equal volume of saline (0.05 ml). At 2 hr, pinnae were measured using a micrometer caliper. We found that 1 feeding with either MDP or SDZ 280.636 abrogated IgE AFC responses and dramatically suppressed serum levels of IgE, both in isotype specific fashion, and suppressed IH responses (> 50%). 3 feedings with SDZ 280.636 also abrogated IgE AFC responses and further decreased serum levels of IgE. In contrast to SDZ 280.636, 3 treatments with MDP had opposite effects in that IgE AFC responses and serum levels of IgE dramatically increased. A single treatment with SDZ 280.636 appeared to increase serum levels of IL-6 up to three fold, while IFN gamma levels decreased. Our data suggest that SDZ 280.636 may be useful in the therapeutic and prophylactic management of human atopic disease such as allergic rhinitis, asthma, and other atopic diseases.

IgE-bearing cells and epsilon-specific mRNA in lymphoid organs of two children with AIDS.

To characterize the cellular basis of IgE responses in HIV-positive (HIV+) children, we obtained central (bone marrow [BM], thymus) and peripheral (Peyer's patches [PP], mesenteric [MLN], and other lymph nodes [OLN], spleen), lymphoid organs from two children with AIDS (females, 2 and 8 years old), and from a non-HIV-infected trauma victim (female, 5 years old) at autopsy. PP were obtained from one of the HIV+ children (2 yr old) and from the non-infected child, but no PP were detected in small intestine of the 8-yr-old HIV+ child. Numbers of lymphocytes bearing surface IgE, CD19, CD3, CD4, and CD8 in lymphoid organs were determined (flow cytometry) and evaluated for expression of epsilon-specific (E) mRNA (RT-PCR). Thymus and MLN of the HIV+ child without PP contained high numbers of IgE+ (34% and 41%, respectively) and CD19+ (32% and 28%, respectively) cells; IgE+ cells were not found in any other organ. In contrast, in the HIV+ child with PP, IgE+ cells were detected in all organs, except BM. The thymus of this child contained fewer CD19+ cells (7%). However, in both HIV+ children, all lymphoid organs, including thymus, contained E mRNA. Because numbers of IgE+ cells often far exceeded numbers of CD19+ B cells, and because CD8+ T cells predominated in all organs, some of the IgE+ cells were probably CD8+ T cells with cytophilic IgE and may include IgE-specific regulatory and/or memory T cells. IgE responses were not detected in the healthy trauma victim nor were B cells found in thymus. The data suggest that during HIV infection, IgE+ B cells may be found in thymus and that synthesis of IgE may occur in all lymphoid organs except BM.

IFN-alpha-mediated suppression of low-affinity FC(epsilon) receptors on Peyer's patch lymphocytes and augmentation of soluble CD23: implications for IgE responses.

The ability of interleukin (IL)-6 or interferon-alpha (IFN-alpha) to regulate expression of low-affinity Fc(epsilon) receptor (CD23) and serum levels of CD23 was studied in benzylpenicilloyl-keyhole limpet hemocyanin-sensitized BALB/c mice at the peak of a hapten-specific immunoglobulin E (IgE) antibody-forming cell (AFC) response. These responses are analogous to those observed in human atopic disease. To induce peak IgE responses, mice were injected intraperitoneally with BPO-KLH (10 micrograms) in aluminum hydroxide gel (alum) on days 0, 21, and 42. On day 44, mice were injected subcutaneously with IL-6 (100-1000 U) or IFN-alpha (1000-10,000 U). On day 46, numbers of CD23+ lymphocytes in Peyer's patches (PP), mesenteric lymph nodes (MLN), and spleen and levels of soluble CD23 in serum were determined (flow microfluorimetry and enzyme-linked immunosorbent assay, confirmed by competition assay). Data are expressed as percent total cells or as optical density at 490 nm. IFN-alpha treatment strongly suppressed (up to 100%) numbers of CD23+ cells exclusively in PP (i.e., numbers of CD23+ cells in MLN and spleen were unchanged) whereas serum levels of soluble CD23 were dramatically increased (60%). IL-6 treatment had no effect on either numbers of CD23+ lymphocytes or on serum levels of soluble CD23. The data suggest that the mechanism(s) by which IFN-alpha, but not IL-6, regulates IgE responses involves suppression of CD23 expression on lymphocytes in PPs and supports a central role for these organs in regulation of IgE responses in vivo.

IL-6 mediated isotype specific suppression of hapten specific IgE in serum of BPO-KLH sensitized mice: role of IFN alpha in maintainance of hapten specific IgE responses.

The ability of IL-6 or IFN alpha or antibodies to these cytokines to regulate serum levels of hapten specific immunoglobulins (IgM, IgG1, IgE, IgA) was studied in BPO-KLH (benzylpenicilloyl-keyhole limpet hemocyanin) sensitized BALB/c mice at the peak of a hapten specific IgE antibody forming cell (AFC) response. To induce peak IgE responses, mice were injected intraperitonealy (i.p.) with BPO-KLH (10 micrograms) in aluminum hydroxide gel (alum) on days 0, 21, and 42. On day 44, mice were injected s.c. with IL-6 (100-1000 U), IFN alpha (1000-10,000 U), anti-IL-6 (100-1000 neutralizing units [NU]), or anti-IFN alpha (1000-10,000 NU). On day 46, levels of BPO specific IgM, IgG1, IgE and IgA in serum were determined (ELISA). Data are expressed as micrograms/ml. IL-6 suppressed BPO specific IgE in serum in isotype specific fashion (to > 90%), increasing IgA (approximately 3 fold), and leaving IgM and IgG1 unchanged. Since removal of endogenous IL-6 with anti-IL-6 increased serum IgE, and suppressed IgG1 (approximately 50%), with IgM and IgA unchanged, this suggests that IL-6 is an isotype specific suppressor of peak IgE responses and as such may be useful in the therapeutic management of atopic disease. IFN alpha treatment increased serum IgE levels (60%), and potentiated IgA responses (> 30 fold), with IgM and IgG1 unchanged. Since removal of endogenous IFN alpha with anti-IFN alpha decreased IgE levels (approximately 50%), increasing IgA, with IgM and IgG1 unchanged, this suggests a role for IFN alpha as an isotype specific helper of peak IgE responses and in maintenance of IgA responses.

Control of IgE responses. II. Isotype specific suppression of peak hapten specific IgE antibody forming cell responses in BPO-KLH sensitized mice after oral administration of muramyldipeptide or murabutide.

Muramyldipeptide (MDP) and murabutide (MB), a pyrogen free derivative of MDP, suppressed BPO specific IgE antibody forming cell (AFC) responses in vivo. To induce IgE responses, BALB/c mice were injected intraperitoneally (i.p.) with BPO-KLH (10 micrograms) in alum on days 0 and 21, or on days 0, 21 and 42. On day 44, mice were fed (gavage) or injected subcutaneously (s.c.) with MDP or MB (0.1-500 mg/kg). Mice were killed on days 45-70, and the numbers of BPO specific IgM, IgG1, IgE, and IgA antibody forming cells (AFC) in lymphoid organs determined in ELISPOT assay. With either immunization schedule, oral treatment with MDP or MB on day 44 suppressed BPO specific IgE AFC responses within 48 h (65-100%). With both molecules, the suppression was IgE isotype specific, dose dependent and transient. The suppression was also route specific since it was obtained only when MDP or MB was given by gavage, and not when injected s.c. These results show that peak antigen specific IgE responses can be suppressed in vivo, in isotype specific fashion, by a clearly defined class of molecules, one of which, MB, is a candidate for clinical studies in man. Pharmacologic agents of this type may be suitable for use in the therapeutic or prophylactic suppression of IgE and, hence, in the therapy of IgE mediated diseases such as allergic rhinitis, asthma, and other atopic diseases.

Suppression of human IgE antibody forming cell responses by IL-6.

To study the effects of cytokines on human IgE antibody forming cells (AFCs), log phase U266 myeloma cells (3 x 10(3)/ml), which secrete immunoglobulin E (IgE), were cultured for 0-24 h with and without cytokine or with or without antibodies against various cytokines. The numbers of IgE AFCs were determined in ELISPOT assay. We found that interleukin-6 (IL-6) suppressed (to 95%) whereas anti-IL-6 increased (to 148%) the numbers of IgE AFCs and that both worked in a dose-dependent fashion. IL-4 and interferon-gamma (IFN-gamma) also suppressed IgE AFC responses in a dose-dependent fashion. However, antibodies to these cytokines had no effect. In contrast, IFN-alpha increased (to fourfold) the numbers of IgE AFCs in a dose-dependent fashion. The data are the first to show a suppressive effect of IL-6 on human IgE responses and may also suggest a role for IL-6 in the treatment of atopic disease.

Control of IgE responses. III. IL-6 and IFN-alpha are isotype-specific regulators of peak BPO-specific IgE antibody-forming cell responses in mice.

The ability of cytokines (IL-4, IL-5, IL-6, IFN-alpha, IFN-gamma, TNF-alpha, GmCSF) to regulate peak benzylpenicilloyl (BPO)-specific IgE antibody-forming cell (AFC) responses was investigated. These responses were induced in BALB/c mice by ip injection of BPO-keyhole limpet hemocyanin (BPO-KLH; 10 micrograms) in aluminum hydroxide gel on Days 0, 21, and 42. On Day 44, or on Days 43, 44, and 45, mice were injected sc with varying doses of cytokine or anti-cytokine antibody. On Day 46, the numbers of BPO-specific AFC (IgM, IgG1, IgE and IgA) in spleen were determined ex vivo in enzyme-linked immunosorbent spot assay. Among the cytokines tested, only IL-6 suppressed BPO-specific IgE AFC responses in an isotype-specific fashion (60-90%). However, treatment of mice with anti-IL-6 also suppressed these responses, suggesting that IL-6 can either suppress or increase peak antigen specific IgE responses, depending upon its concentration. Among the cytokines tested, only IFN-alpha increased BPO-specific IgE AFC responses in an isotype-specific fashion. Since treatment with anti-IFN-alpha suppressed these responses, it appears that IFN-alpha is required to maintain peak antigen-specific IgE AFC responses. IL-4 or IFN-gamma nonspecifically suppressed responses of all isotypes. Treatment with anti-IL-4 also suppressed IgE responses, suggesting that this cytokine is required to maintain peak antigen specific IgE responses. Treatment with anti-IFN-gamma increased IgE responses, indicating that IFN-gamma suppresses peak antigen-specific IgE responses.

Control of IgE responses. 4. Isotype-specific suppression of peak BPO-specific IgE antibody-forming cell responses and of BPO-specific IgE in serum by muramyldipeptide or murabutide after administration to mice by gavage.

Muramyldipeptide (MDP) and murabutide (MB) suppressed hapten-specific IgE antibody-forming cell (AFC) responses in vivo. IgE responses were induced in BALB/c mice by intraperitoneal injection with benzylpenicilloyl-keyhole limpet hemocyanin (BPO-KLH) (10 micrograms) in aluminum hydroxide gel (Alum) on days 0, 21 and 42. On day 44, mice were fed (gavage) or injected subcutaneously with varying concentrations of MDP or MB (0.1-500 mg/kg). The mice were killed on days 45-70, and the numbers of BPO-specific IgM, IgG1, IgE, and IgA AFC in various lymphoid organs were determined in an enzyme-linked immunosorbent spot (ELISPOT) assay. In addition, levels of BPO-specific IgE in serum were determined by ELISA. Data are expressed as AFC/10(7) cells or as micrograms/ml. Feeding with MDP or MB on day 44 suppressed BPO-specific IgE AFC responses and serum levels of BPO-specific IgE within 48 h (day 46) (65-100% and approximately 50% decrease, respectively). With both molecules, the suppression was IgE isotype-specific, dose-dependent and transient. The suppression was also route-specific since it was obtained only when MDP or MB were given by gavage, and not when injected subcutaneously. These results show that peak antigen-specific IgE responses can be downregulated in vivo, in isotype-specific fashion, by a clearly defined class of molecules, MDP and MB, one of which, MB, is a candidate for clinical studies in man. The mechanism of suppression probably involves the modulation of gut-associated lymphoid tissue and mucosal immunity. The clinical implications are that pharmacologic agents of this type may be suitable for use in the therapeutic or prophylactic downregulation of IgE and, hence, in the therapy of IgE-mediated diseases in man such as allergic rhinitis, asthma, and other atopic diseases.

Origin and fate of IgE-bearing lymphocytes. II. Gut-associated lymphoid tissue as sites of first appearance of IgE-bearing B lymphocytes and hapten-specific IgE antibody-forming cells in mice immunized with benzylpenicilloyl-keyhole limpet hemocyanin by various routes: relation to asialo GM1 ganglioside+ cells and IgE/CD23 immune complexes.

The organs in which B cells bearing membrane-bound IgE (sIgE+) and benzylpenicilloyl (BPO)-specific IgE antibody-forming cells (AFC) first appeared were determined in BALB/c mice given BPO-keyhole limpet hemocyanin (10 micrograms) in aluminum hydroxide by various routes (i.p, gavage, s.c., i.v., or i.m.). In mice immunized by the i.p. route, the numbers and location of sIgE+ B cells and asialo GM1 ganglioside (AsGm1+) cells, the location of IgE/CD23 immune complexes, and the numbers of BPO-specific IgE AFC in lymphoid organs were determined. With all routes of immunization, no sIgE+ B cells or BPO-specific IgE AFC were ever detected in any organ before day 8. On day 8, with the s.c., i.v., or i.m. routes, sIgE+ B cells and IgE AFC appeared exclusively in Peyer's patches (PP); with the i.p. or gavage routes, sIgE+ B cells simultaneously appeared in both PP and mesenteric lymph nodes, whereas IgE AFC appeared only in PP. In mice immunized by the i.p. route, IgE/CD23 immune complexes and strikingly increased numbers of AsGm1+ cells transiently appeared only in PP after the appearance and preceding the "disappearance" of the sIgE+ B cells and IgE AFC. The data suggest that specific IgE responses originate in gut-associated lymphoid tissue and appear later in spleen. The data also associate the appearance of IgE/CD23 immune complexes and AsGm1+ cells with the "disappearance" of sIgE+ B cells and IgE AFC from PP.

Constitutive production of PAI-II and increased surface expression of GM1 ganglioside by peripheral blood monocytes from patients with AIDS: evidence of monocyte activation in vivo.

To characterize the activation state of monocytes during human immunodeficiency virus (HIV) infection, peripheral blood monocytes (PBMs) from patients with acquired immunodeficiency syndrome (n = 10) and from healthy controls (n = 10) were cultured for 4 days. Monocyte culture supernatant (MCS) was collected daily, and levels of urokinase (UK) inhibitor PAI-II, a product of activated monocytes, released into MCS were determined (fibrin plate assay). To examine the activation state of PBMs independently, expression of GM1 ganglioside on PBMs from patients with AIDS (n = 9), patients with AIDS-related complex (ARC) (n = 8), HIV+ asymptomatic patients (n = 6), and HIV- healthy controls (n = 11) was determined (flow cytometry; living cells in suspension). Data are expressed as percent inhibition of UK, or as percent total cells. Patients' MCS collected on days 1-4 of culture contained similar levels of PAI-II because it inhibited UK in similar fashion (70-90%). In contrast, MCS from healthy controls, collected after 2 days, had decreased ability to inhibit UK (15-50%) and thus contained lower levels of PAI-II. Monocyte activation, measured by increased expression of GM1 ganglioside on PBM surfaces, directly correlated with the progression of HIV infection into the development of AIDS, since the order of magnitude of GM1 ganglioside expression on PBMs was AIDS greater than ARC greater than HIV+ asymptomatic = healthy controls. Our data indicate that PBMs from patients with AIDS are constitutively activated and suggest that activation directly correlates with disease progression.

Peyer's patches and mesenteric lymph nodes are the sites of first appearance of IgE bearing B lymphocytes and hapten specific IgE antibody forming cells in BPO-KLH sensitized mice.

Antigen specific IgE responses originate in gut associated lymphoid tissue (GALT) of mice sensitized with benzylpenicilloyl-keyhole limpet hemocyanin (BPO-KLH) in alum, regardless of the route (intraperitoneal [i.p.], oral [gavage], subcutaneous [s.c.], intramuscular [i.m.] or intravenous [i.v.]) used for immunization. When BALB/c mice were injected i.p. with BPO-KLH (10 micrograms) in alum, B lymphocytes bearing membrane bound IgE (sIgE+B cells) first appeared simultaneously in Peyer's patches (PP) and mesenteric lymph node (MLN) on day 8. BPO specific IgE antibody forming cells (AFC) also appeared in PP on day 8, but were not found in MLN until day 10. On day 8, no sIgE+B cells or IgE AFC were found in bone marrow (BM) or other lymphoid organs. The appearance of sIgE+B cells and IgE AFC in PP and MLN was transient; these cells were no longer detected in PP on days 14 and 24, respectively, or in MLN on days 14 and 36, respectively. sIgE+B cells and IgE AFC did not appear in spleen until day 12, where they were detected through day 70. Although sIgE+B cells were never found in BM, IgE AFC appeared in BM on day 18, where they were detected through day 70. No sIgE+B cells or IgE AFC were found in other lymph nodes (OLN) on days 0-70. Boosting did not induce the reappearance of sIgE+B cells or IgE AFC in PP, the reappearance of sIgE+B cells in MLN, the appearance of sIgE+B cells in BM, or the appearance of sIgE+B cells or IgE AFC in OLN.(ABSTRACT TRUNCATED AT 250 WORDS)

Origin and fate of IgE-bearing lymphocytes. II. Modulation of IgE isotype expression on Peyer's patch cells by feeding with certain bacteria and bacterial cell wall components or by thymectomy.

Mechanisms regulating the appearance of sIgE+ B lymphocytes appear to be lacking in adult germfree (GF) rats in that their Peyer's patches (PP) contain high numbers of cells with sIgE (approximately 15% of total cells), one-half of which simultaneously express sIgA, whereas sIgE+ cells are absent from PP of conventional rats (less than 1%). GF rat PP also contain elevated numbers of sIgA+ cells and decreased numbers of sIgM+ cells, with elevated numbers of sThy-1+ RT 7.1+ Ig- T cells, and reduced numbers of sThy-1- RT 7.1+ Ig- T cells. The cellular composition of PP of GF rats was converted to that resembling a conventional rat within 18 h after either 1) use of standard (unautoclaved) food; 2) feeding with certain bacteria (Clostridium difficile, Corynebacterium pseudodiphtheriticum, Mycobacterium tuberculosis, and Klebsiella pneumoniae), in either live or heat-killed, but not autoclaved form; or with certain bacterial cell wall components: murein (peptidoglycan), and its synthetic derivatives, muramyltripeptide phosphatidylethanolamine and desmethyl-muramyldipeptide, but not with LPS, core lipid A or lipoprotein; there was no effect if any bacterial cell wall component was injected i.v.; or 3) thymectomy. Each procedure resulted in elimination of sIgE+ B cells and normalization of the other surface isotypes, and loss of sThy-1+ RT 7.1+ Ig- T cells and normalization of sThy-1- RT 7.1+ Ig- T cells. Irrespective of treatment, no sIgE+ cells were detected in bone marrow, thymus, other lymphoid organs or blood, excluding the possibility that the elimination of these cells from PP was associated with their redistribution to other sites. Thus, exposure to gut flora and bacterial peptidoglycan components may have resulted in IgE isotype switching, either directly or through the mediation of accessory and/or sThy-1+ RT 7.1+ regulatory T cells. The sites in which sIgE+ B cells are down-regulated appear to be PP.

Control of IgE responses.

Peyer's patches (PP) in germ-free rats (GF) and in the hyper-IgE syndrome patient (HIES) differ from their conventional rat (C) and healthy human (HH) counterparts in that GF rats contained fewer (two-fold) PP and none was detected in HIES. Existing PP in GF rats had reduced cellularity (three-fold) and different B and T cell subsets: high numbers of IgE-bearing (sIgE+) B cells (approximately 15% of total cells), one-half of which also expressed sIgA, were present in GF rat PP while none was detected in C rat PP (less than 1%). GF rat PP also contained elevated numbers of sIgA+ cells and decreased sIgM+ cells, with elevated numbers of sThy 1+ RT 7.1+ Ig- T cells (suppressor phenotype) and reduced sThy 1- RT 7.1+ Ig- T cells (helper phenotype). The cellular composition of GF rat PP was converted to that resembling a C rat within 18 hr after (a) use of standard (unautoclaved) chow; (b) feeding with certain bacteria or "working" bacterial cell wall components (BCWC) and synthetic derivatives, murein, MTP-PE, and norMDP, but not with LPS, core lipid A, or lipoprotein; BCWC had no effect if injected intravenously; or (c) thymectomy. Each procedure resulted in (i) elimination of sIgE+ B cells and normalization of the other isotypes, and (ii) loss of T suppressor cells and normalization of T helper cells. After treatments, no sIgE+ cells were detected in bone marrow (BM), thymus, other lymphoid organs, or blood. PP were not detected in HIES, although they were present in HH (approximately 10/individual). P blood contained two distinct sIgE+ B cell subpopulations, the apparent source of which was mesenteric lymph node (MLN), the only organ in which high numbers of these cells (35%) (five nodes examined) were detected; far fewer IgE+ cells were found in spleen (less than 5%), and none was detected in BM, thymus, other LN, or appendix, which was virtually acellular. Virtually no IgE secreting plasma cells were detected in MLN, spleen, appendix, other lymphoid organs, or in gut lamina propria. IgE+ B cells in MLN were not detected in follicles (classical B cell areas); instead, they were found in high numbers in the thymus-dependent area and in medulla. Most follicles (greater than 98%) in MLN and spleen contained intercellular IgE complexed to bacterial antigen and/or CD23 (IgE-binding factor? antigen?), but contained no germinal centers.(ABSTRACT TRUNCATED AT 400 WORDS)