William Allan

Immunohistochemical Characterisation of Macrophages in Human Liver and Gastrointestinal Tract: Expression of CD4, HLA-DR, OKM1, and the Mature Macrophage Marker 25F9 in Normal and Diseased Tissue

This report describes the immunocytochemical characterisation of macrophages in sections of human liver, gastrointestinal tract, and associated lymphoid tissue and the inflammatory lesions of Crohns disease. 25F9 is an antigen reported to be induced during the maturation of blood monocytes in vitro. The antigen was concentrated in cytoplasmic vesicular structures of isolated gastrointestinal macrophages. Similar labelled cells were observed in the apical regions of lamina propria in both small and large intestine in vivo. Their numbers and size were greatly increased in specimens of colon from patients with melanosis coll. Mucosal inflammatory lesions in specimens from patients with Crohns disease did not contain 25F9‐positive cells. The antigen was absent from giant cells and epithelioid cells in granulomata but was expressed on histiocytes in submucosal microgranulomata. In lymphoid organs, 25F9‐positive cells were found in germinal centres, in the dome region of Peyers patch, and in the medulla, but were largely excluded from T cell areas. In reactive nodes from Crohns disease patients, the number of labelled cells in germinal centres and T cell areas was greatly increased. 25F9 was absent from the majority of typical liver Kupffer cells, but was expressed on cytoplasmic granules in a minor subpopulation of larger, more rounded cells in the liver. The results suggest that 25F9 is a marker for endocytosis rather than maturation. In parallel sections, resident macrophages of both liver and gastrointestinal tract labelled with Leu 3a/OKT4 (CD4) and with OKIa (HLA‐DR antigen) but did not express OKM1 (type III complement receptor). By contrast, OKM1 was present on inflammatory cells, epithelioid cells, and giant cells in mucosal lesions of Crohns disease.

Preparation and characterization of human bone marrow-derived macrophages.

Bone marrow–derived macrophages were prepared from human bone marrow mononuclear cells following cultivation in GCT‐conditioned medium (GCT‐CM) and purification by adherence to fibronectin‐coated flasks. The growth of bone marrow mononuclear cells in GCT‐CM was dependent on the shape of the culture vessels, being increased in round‐bottomed versus flat‐bottomed wells. Proliferation was confined to nonadherent cells; like blood monocytes, bone marrow‐derived macrophages did not incorporate [3H]thymidine in response to GCT‐CM or human serum. Purified macrophages from this source expressed nonspecific esterase and OKM1, OKIa, FMC 17, 32, and 34 and 25F9 antigens but lacked Mo2. They expressed high levels of an inactivator of plasminogen activator, minactivin, and gave a substantial metabolic burst in response to phorbol myristate acetate or opsonized (but not unopsonized) zymosan. Bone marrow–derived macrophages acted as accessory cells in the response of T lymphocytes to phytohemagglutinin. The results suggest that liquid bone marrow cultures are useful in the study of the differentiation of human mononuclear phagocytes.

Induction of class II major histocompatibility antigens on thyroid, adult pancreatic islet, and fetal proislet allografts.

Cultured BALB/c (H-2d) thyroid, adult pancreatic islet and fetal proislet tissues can be accepted permanently in CBA/H (H-2k) recipients until rejection is triggered by the transfer of antidonor strain activated immune T cells. Class II MHC antigens are not expressed on the accepted grafts and are induced following immune cell transfer, but expression is quantitatively and qualitatively different for the different tissues. Thyroid allografts show intense class II MHC antigen expression early after immune cell transfer and before extensive tissue destruction has occurred. In contrast, adult islet and fetal proislet allografts show patchy and cytoplasmic staining usually associated with areas of tissue destruction. Gamma-interferon treatment of tissues in vitro showed that proislet tissue has a greater capacity for class II MHC antigen induction than adult islet tissue. The differences in the capacity for class II MHC induction by fetal and adult islet tissue may be important in relation to the pathogenesis of autoimmune disease.

Antigenic Variation and Macrophage Infiltration of Human Bladder-Tumors Xenografted Into Nude-Mice

The presence of macrophages both within and around human bladder transitional cell carcinomas xenografted into nude mice has been examined using Immunocytochemical staining. Nine different xenograft lines derived from bladder tumors of eight patients were stained for F4/80, MHC II and PGP‐1 labelled cells. The results revealed considerable heterogeneity both within and between tumors. All of the tumors generated some macrophage response at the tumor periphery, and this was marked in two instances, UCRU‐BL‐13, passage 9 and UCRU‐BL‐17, passage 2. In only two tumors was there substantial penetration of the tumor epithelium by macrophages, though infiltrating la+, F4/80+, PGP‐1+ cells were sometimes seen spreading along the base of the epithelial sheets. Three of the tumors were characterized by the presence of L3T4+ T lymphocytes at the invasion front. The presence of macrophages or lymphocytes either at the invasion front or within the tumors did not correlate with the pathological grade of the tumors. PGP‐1 monoclonal antibodies also stained granulocytes, which were present within the tumor mass in UCRU‐BL‐15, passage 2, possibly reflecting the production by the tumor of G‐CSF. In addition, the PGP‐1 antibodies stained some of the bladder tumors cells themselves, and, in some cases, the interstitial structures within the tumors. The significance of this staining is not yet understood. The xenografted tumors will provide a useful model to examine the constraints on tumor therapy using macrophage activating stimuli.

Regulation of the production of granulocyte-macrophage colony-stimulating factor by macrophage-like tumour cell lines.

Macrophage tumour cell lines (PU5‐1.8, P388D1) produced detectable granulocyte macrophage colonystimulating (CSF‐2) activity measured using a factor‐dependent cell line FDC‐P1. The production ofCSF‐2 was enhanced by endotoxin and inhibited by serum, and correlated inversely with [3H]TdR incorporation. mRNA isolated from PU5‐1.8 or P388D1 cells initiated CSF‐2 production when injected into Xenopus laevis oocytes. The specific activity in this assay was unaltered in mRNA isolated from endotoxin‐treated cells. The results suggest that endotoxin acts at a post‐transcriptional level.

Influence of non-major histocompatibility complex differences on the severity of lymphocytic choriomeningitis.

The role of the non-major histocompatibility complex (MHC) genetic background in the development of lymphocytic choriomeningitis (LCM) was examined for a range of mouse strains of the H-2k haplotype. The onset of meningitis relative to the time of injection of LCM virus was delayed and the maximal level of cellular extravasation into cerebrospinal fluid was lower in C3H/HeJ and CBA/H compared with AKR/J, B10.Br and BALB/c.H-2k mice. Adoptive transfer experiments indicated that the C3H mice are genuine low responders, but immune spleen cells from the CBA/H were as potent on a cell-for-cell basis as those from the AKR/J. Further analysis with CBA/H, AKR/J and (CBA/H x AKR/J)F1 mice showed that the pattern of high response for the AKR/J was dominant, with the differential kinetics of the development of meningitis correlating with the cellularity of the cervical lymph nodes. Thus, the generation of the LCM inflammatory process is not dictated solely by the MHC phenotype.

Role of CD8+αβ T Cells in Respiratory Infections Caused by Sendai Virus and Influenza Virus

Current understanding of the part played by CD8+ αβ T cells in respiratory virus infections is based on findings from a spectrum of approaches. These include direct analysis of the patterns of lymphocyte involvement in the virus-infected lung (McDermott et al., 1987; Pena-Cruz et al., 1989; Allan et al., 1990; Openshaw, 1991; Eichelberger et al., 1991b), adoptive transfer experiments utilizing either bulk immune T-cell populations or cell lines (Leung and Ada, 1992; Lukacher et al., 1984, 1986; Ada and Jones, 1986; Taylor and Askonas, 1986; Kast et al., 1986; Cannon et al., 1987; Askonas et al., 1988; Mackenzie et al., 1989), in vivo depletion with monoclonal antibody (mAb) (Lightman et al., 1987; Waldmann, 1989; Allan et al., 1990; Eichelberger et al., 1991b), and the use of H-2 mutant or genetically manipulated mice (de Waal et al., 1983; Eichelberger et al., 1991b).