Common cell surface antigen associated with mammalian C-type RNA viruses. Cell membrane-bound gs antigen.

The indirect membrane immunofluorescence test and the absorption analysis of rabbit anti-FeLV, rabbit anti-FeLVp 30, and rabbit anti-MuLVp 30 antisera yielded the following conclusions. An antigen shared by mammalian (murine and feline) C-type RNA leukemia and sarcoma viruses was detected on the surface of cells infected or transformed by C-type viruses. The antigen was characterized as membrane-bound gs antigen bearing two determinants, membrane-bound gs-1, intraspecies-specific antigenic determinant, and membrane-bound gs-3, interspecies-specific antigenic determinant. Membrane-bound gs antigen was located on the cell surface, frequently near the site of virus budding but not on the envelope of murine C-type RNA virus.

The viral envelope glycoprotein of murine leukemia virus and the pathogenesis of immune complex glomerulonephritis of New Zealand mice.

The use of monospecific antisera for the analysis by radioimmunoassay and immunofluorescence study of two major viral proteins, gp69/71 and p30 of murine leukemia virus, that could be of significance in the pathogenesis of immune complex glomerulonephritis of mice, particularly NZB and B/WF(1) hybrid mice, yielded the following conclusions. A remarkably high concentration of viral envelope glycoprotein, gp69/71, was detected in the spleen and serum of New Zealand mice (NZB, NZW, B/WF(1), and W/BF(1)); the concentration in the spleen was 10-fold greater than that found in AKR mice and 30-fold greater than that present in C57BL/6 mice. The gp69/71 was deposited along with bound immunoglobulins, apparently as an immune complex, in the diseased kidneys of mice, and the glomerular site and extent of deposition of gp69/71 was related to the severity of the glomerulonephritis. This study suggests that the pathogenesis of immune complex glomerulonephritis (and vasculitis) in mice is related to the expression of this specific viral envelope glycoprotein and to the host immune response to this protein.

Rat lymphoid cell lines with human T cell leukemia virus production. I. Biological and serological characterization.

Cocultivation of spleen cells, lymph node cells, and thymocytes of female Wistar-King-Aptekman rats with short-term cultured male adult T cell leukemia (ATL) cells in the presence of 5-bromo-2'-deoxyuridine (BrdUrd) resulted in the establishment of rat lymphoid cell lines, TARS-1, TARL-2, and TART-1. Cytogenetic analysis of the three cell lines showed a female rat karyotype with 42 chromosomes. The surface phenotypes of TARS-1 and TART-1 were those of rat T cells. TARL-2 was only positive for rat Ia and leukocyte common antigens. The cell lines continuously produced a type C retrovirus, human T cell leukemia virus (HTLV) and expressed ATL-associated antigens. TARS-1 and TART-1, but not TARL-2 were transplantable into newborn syngeneic rats and nude mice. These results strongly indicate that HTLV not only immortalizes, but also transforms rat T cells in vitro. Adult rats immunized with either TARS-1 or TARL-2 produced antibodies specific for HTLV. The biochemical analysis of the antigens that reacted with rat sera revealed that they are the two HTLV-specific polypeptides, p24 and p28.

A rat model of human T lymphocyte virus type I (HTLV-I) infection. 1. Humoral antibody response, provirus integration, and HTLV-I-associated myelopathy/tropical spastic paraparesis-like myelopathy in seronegative HTLV-I carrier rats.

Human T lymphocyte virus type I (HTLV-I) can be transmitted into several inbred strains of newborn and adult rats by inoculating newly established HTLV-I-immortalized rat T cell lines or the human T cell line MT-2. The transmission efficiency exceeds 80%, regardless of strain differences or the age at transmission. The production of anti-HTLV-I antibodies significantly differs among the strains and depends on the age at the time of transmission. Rats neonatally inoculated with HTLV-I-positive rat or human cells generally become seronegative HTLV-I carriers throughout their lives, whereas adult rats inoculated with HTLV-I-positive cells at 16 wk of age become seropositive HTLV-I carriers. The HTLV-I provirus genome is present in almost all organs, regardless of whether the carriers are seronegative or seropositive. According to antibody titers to HTLV-I, there are three groups of inbred rat strains: ACI, F344, and SDJ (high responders); WKA, BUF, and LEJ (intermediate responders); and LEW (low responder). Three of three 16-mo-old seronegative HTLV-I carrier rats of the WKA strain developed spastic paraparesis of the hind legs. Neuropathological examinations revealed that the lesions were confined primarily to the lateral and anterior funiculi of the spinal cord. Both myelin and axons were extensively damaged in a symmetrical fashion, and infiltration with massive foamy macrophages was evident. The most severe lesions were at levels of the thoracic cord and continued from the cervical to the lumbar area. These histopathological features as well as clinical symptoms largely parallel findings in humans with HTLV-I-associated myelopathy/tropical spastic paraparesis (HAM/TSP). These HTLV-I carrier rats, in particular the WKA rats described above, can serve as a useful animal model for investigating virus-host interactions in the etiopathogenesis of HTLV-I-related immunological diseases, particularly HAM/TSP.

Short communication: expression of human endogenous retrovirus-R gene links to differentiation of squamous cells.

To investigate the biological roles of human endogenous retrovirus-R (HERV-R) in vivo, we established transgenic rats carrying the full sequence of the viral genome with control of its own long terminal repeat promoter. The Env protein was expressed on the surface of the epidermis of fetal HERV-R transgenic rats on day 10 of gestation. The epidermal Env expression disappeared by day 18 of gestation. After day 18 of gestation, the Env protein was detected in the prickle layer of the esophageal epithelium of transgenic rats. Interestingly, it was not detected in the basal layer of the epithelium, and the expression in the granular layer was weaker than in the prickle layer. These findings suggest that expression of HERV-R is linked not only to the development but also to the differentiation of squamous cells. Next, we examined alterations in the expression of the HERV-R env gene in cultured human squamous cells after exposure to all-trans retinoic acids (ATRA). The env expression was increased by ATRA in a dose-dependent manner, while the expression of transglutaminase 1 (TGM1), a terminal marker for squamous differentiation, was decreased. TGM1 is expressed in the granular layer of the squamous epithelium, and ATRA suppresses the differentiation of cultured squamous cells. Thus, these in vitro data also suggest that HERV-R expression is regulated by a mechanism closely related to the differentiation of squamous cells. This study is the first to demonstrate the association of HERV-R expression and differentiation of squamous cells.

Articular tissues expressing the env-pX transgene are required for generation of arthritogenic T cells in human T cell leukemia virus type I transgenic rats.

OBJECTIVE: Human T cell leukemia virus type I env-pX transgenic rats (env-pX rats) were used to investigate the pathogenesis of arthritis. METHODS: Phenotype of cells infiltrated into arthritic joints in env-pX rats was analyzed using flow cytometry and cell-transfer experiments were done using env-pX and wild-type WKAH rats. RESULTS: The majority of T cells infiltrated into arthritic joints in env-pX rats exhibited a CD4 and activated phenotype. Transfer of these T cells into articular space in wild-type WKAH rats succeeded to induce arthritis similarly seen in env-pX rats. However, injection of the cells into sites other than joints did not induce inflammation. Transfer of in vitro-stimulated lymph node cells from disease-free env-pX rats into articular space did not induce arthritis in wild-type WKAH rats. CONCLUSION: These findings suggest that articular tissues carrying the env-pX transgene are required for generation of arthritogenic T cells in env-pX rats. However, the constitutive antigens other than the transgene products are recognized as immunological targets by the arthritogenic T cells in the advanced arthritic joints. Molecules expressed specifically in articular tissues may be needed to maintain the inflammatory cell infiltration.

Diagnosis of bone metastasis in men with prostate cancer by measurement of serum ICTP in combination with alkali phosphatase and prostate-specific antigen.

AIMS: Carboxy-terminal telopeptide of type I collagen (ICTP) is a parameter of bone absorption, and has recently been introduced to monitor bone metastases. The aim of this retrospective study was to investigate the potential of ICTP as a candidate serum marker of bone metastasis in prostate cancer. MATERIALS AND METHODS: Serum markers in 155 men pathologically diagnosed with prostate cancer were measured. The serum levels of ICTP, prostate-specific antigen (PSA), and alkali phosphatase (ALP) were compared to assess the extent of disease (EOD) scores from bone scans and then analysed statistically. RESULTS: The serum ICTP levels were not well correlated with the EOD scores in the total group of men, men newly diagnosed with prostate cancer, or men previously diagnosed with prostate cancer who were followed up. The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of ICTP (cut-off value, 5.0 ng/ ml) of the men newly diagnosed with prostate cancer were 78.6%, 88.0%, 78.6%, and 88.0%, respectively. In these men, the specificity and PPV of ALP (cut-off value, 335 IU/l) were 100%, whereas the sensitivity and NPV of PSA (cut-off value, 40 ng/ml) were 100% in this study. The serum levels of ICTP in the men with low ALP (< 335 IU/l) and high PSA (> or = 40 ng/ ml) clearly separated the men with or without bone metastasis, as judged by bone scans. CONCLUSION: We found that the ICTP is not a superior serum marker for bone metastases compared with ALP or PSA. Our study suggests, however, that the ICTP measurement is useful in a certain subset of men with the combination of PSA and ALP in distinguishing men with bone metastasis from those without.

A novel form of prostate-specific antigen transcript produced by alternative splicing.

Molecular characterization of prostate-specific antigen (PSA) has not been well elucidated, despite a great deal of clinical study. We examined the heterogeneity of PSA using reverse transcription-PCR and direct sequencing. A novel, alternatively spliced variant of the PSA transcript was found in prostate cancer (PC), as well as in benign prostatic tissue. This alternative splicing leads to the deletion of 44 amino acid residues (amino acids 45-88) from mature PSA, resulting in the loss of asparagine 45, which is a binding site for a carbohydrate chain. By these nested reverse transcription-PCR systems, this novel, alternatively spliced PSA gene was recognized in 13 of 18 (72.2%) cases with noncancerous prostate tissue, 4 of 5 (80.0%) PC cases, and 3 of 12 (25.0%) blood samples from PC patients (noncancerous prostate tissue group versus blood sample group, P = 0.011). At present, the biological significance of this alternative splicing remains to be established.

Cytokine-producing mammary carcinomas in transgenic rats carrying the pX gene of human T-lymphotropic virus type I.

In two lines of transgenic rats (pX rats) from WKAH and F344 strains and carrying the human T-lymphotropic virus type I pX gene, undifferentiated mammary carcinomas developed predominantly in females starting at about 5 months of age, and there was massive infiltration of granulocytes in the tumor tissue. The incidence of the tumor reached about 40% when the rats were 12 months old. mRNAs of both pX and host genes Gro and MIP-2, which are granulocyte chemoattractants of the interleukin 8 family, were highly expressed in the tumor tissue. Since expression and point mutation of several oncogenes and the antioncogene were not demonstrated, hitherto unidentified novel oncogenic pathways may be transactivated by the pX transgene in these pX rats.

Promotion of early osteoclastogenesis and B lymphopoiesis in the bone marrow of transgenic rats with the env-pX gene of human T-cell lymphotropic virus type I.

Human T-cell lymphotropic virus type I (HTLV-I) is associated with various clinical disorders including adult T cell leukemia, myelopathy, arthropathy. Hypercalcemia resulting from osteoclast activation and a variety of hematopoietic abnormalities have been also observed in HTLV-I infected patients, however, precise mechanism about initial trigger(s) prior to presenting symptoms is still unknown. In this study, to assess effects of HTLV-I on hematopoiesis, we analysed characteristics of early hematopoietic precursors in HTLV-I env-pX transgenic rats. Progenitor cells for osteoclasts were significantly increased even in the marrow of asymptomatic env-pX rats. Progenitors for B cells were also highly enriched, while colony forming cells (CFC) elicited by GM-CSF(CFU-GM) and M-CSF(CFU-M) were comparable to normal littermates. Following arthritis in env-pX transgenic rats, osteoclastogenesis was further augmented and the CFCs were increased. Bone marrow cells carrying adjuvant-induced arthritis retained a constant number of progenitors for osteoclast and B lymphocytes, whereas the number of CFU-GM and CFU-M increased. These results indicate that the env-pX transgene affect early stages of osteoclast and B-cell lineages prior to developing diseases, in contrast, an increase of the CFCs was caused indirectly by arthritis. This study provides a novel standpoint for the mechanisms of pathogenesis by HTLV-I.

HTLV-I induced myeloneuropathy in WKAH rats: apoptosis and local activation of the HTLV-I pX and TNF-alpha genes implicated in the pathogenesis.

To investigate the pathogenesis of HTLV-I associated diseases, we established a rat model for HTLV-I-associated myelopathy/tropical spastic paraparesis (HAM/TSP) in WKAH rats. In the spinal cords of WKAH rats carrying HTLV-I, chronological histopathology revealed the occurrence of apoptotic cell death starting at 9 months after the infection, followed by demyelination, macrophage infiltration, and the activation of astrocytes starting at 12, 15 and 20 months, respectively. Apoptosis of the Schwann cells was also observed in the peripheral nerves of these rats. By RT-PCR, pX mRNA of HTLV-I was selectively expressed in the diseased spinal cords and peripheral nerves, but not in the unaffected cerebra, cerebella, even though provirus DNAs were consistently identified in these tissues. Among several cytokines examined, mRNA expression and production of TNF-alpha were frequently detected in the spinal cord and the cerebrospinal fluid. The collective evidence suggests that the selective activation of HTLV-I, in particular Tax expression, and/or the production of TNF-alpha in target spinal cord and peripheral nerves are causally related to apoptotic death of the oligodendrocytes and Schwann cells, a major pathogenetic pathway of HTLV-I induced myeloneuropathy in the WKAH rat.

HTLV-I env-pX transgenic rats: prototype animal model for collagen vascular diseases.

To evaluate the function of HTLV-I env-pX gene in vivo, we developed two lines of transgenic rats (env-pX rats) that expressed env-pX gene products, under control of own LTR promotor. In various tissues of the rats, env and pX mRNAs were constitutively expressed, irrespective of age. At age 5 weeks, swelling of the bilateral ankle joints histologically showing synovial lining hyperplasia, severe chronic inflammation, erosion of the joint cartilage, and bone destruction with pannus formation began to develop in these env-pX rats. These histologic features resemble those of rheumatoid arthritis (RA) in man. High titered rheumatoid factors and low anti-dsDNA antibodies and hyper-gamma globulinemia were detected. Necrotizing arteritis resembling polyarteritis nodosa, polymyositis, myocarditis and Sjögren syndrome-like sialoadenitis developed, together with RA-like arthritis even in one individual animal. Thymic atrophy with low body weight was also observed. The evidence indicates that env-pX rats appear to be suitable animal models for elucidating pathogenetic mechanisms involved in not only HTLV-I related diseases but also various collegen vascular and autoimmune diseases of unknown etiology in man.

HTLV-I pX transgenic rats: development of cytokine-producing mammary carcinomas and establishment of the pX mammary carcinoma cell lines.

In two lines of transgenic rats (pX rats) from WKAH and F344 strains and carrying the HTLV-I pX gene under control of the mouse H-2Kd promoter, mammary carcinomas developed predominantly in females starting at about 5 months of age. The incidence of the tumor reached about 40% when the rats were 12 months old. Histology of the tumor was undifferentiated carcinoma with massive infiltration of granulocytes into the tumor tissue. Systemic granulocytosis and hepato-splenomegaly due to extramedullary granulocytopoiesis were seen in pX rats and nude mice bearing pX mammary tumor. mRNAs of both pX and host genes, Gro and MIP-2, which are granulocyte chemoattractants of the IL-8 family, were highly expressed in the tumor tissue. Since expression and point mutation of several oncogenes and anti-oncogene, related with mammary carcinomas, were not demonstrated, hitherto unidentified novel oncogenic pathways may be transactivated by the pX transgene in these pX rats. pX mammary carcinoma cell lines, which have similar characteristics to the primary tumor, were established and the cells underwent apoptosis under the serum deprived conditions. The pX rats and the pX mammary carcinomas appear to be suitable models for analyses of HTLV-I pX oncogenesis and immune pathogenesis in vivo and in vitro.

Models of HTLV-I-induced diseases. Infectious transmission of HTLV-I in inbred rats and HTVL-I env-pX transgenic rats.

To examine the pathogenic roles of HTLV-I in HTLV-I-induced diseases, we developed two models; namely HTLV-I carrier rats and HTLV-I env-pX transgenic rats. Among life long HTLV-I carriers in seven rat strains, only WKAH rats with the RT1k haplotype developed chronic progressive myeloneuropathy, resembling HAM/TSP clinically and histologically in humans, designated as HAM rat disease and after long incubation periods. Apoptosis of myelin forming cells, oligodendrocytes and Schwann cells associated with HTLV-I infection appears to be the primary cause of HAM rat disease. Local activation of the pX gene and TNF alpha gene was evident in these rats. WKAH rats transgenic for HTLV-I env-pX gene were established and at age 5 weeks, swelling of the bilateral ankle joints began to develop and histological features of the affected joints resembled findings in cases of rheumatoid arthritis (RA): high-titers of rheumatoid factors were present in these rats. A series of vascular collagen diseases such as polyarteritis nodosa-like angiitis, polymyositis, myocarditis, and Sjögren's syndrome-like sialodenitis together with RA were present, even in one individual animal. These transgenic rats as well as HAM rats appear to be suitable animal models for elucidating pathogenic mechanisms implicated in HTLV-I-induced diseases and also various demyelinating vascular collagen diseases of unknown etiology.

Detection of circulating cancer cells by reverse transcription-polymerase chain reaction for uroplakin II in peripheral blood of patients with urothelial cancer.

Few attempts have been made at the molecular detection of urothelial cancer cells in the blood or lymph nodes mainly because of an absence of good candidate molecular or genetic changes specific to urothelial cancer or urothelium. In 1990, however, genes that encode urothelium-specific transmembrane proteins, uroplakins (UPs), were cloned. We have established a method of detecting circulating cancer cells in peripheral blood of patients with transitional cell carcinoma by nested reverse transcription-PCR assay for UP II. UP II mRNA-positive cells were detected in 3 (10.3%) of 29 patients with superficial cancers (pTa-1N0M0), 4 (28.6%) of 14 patients with muscularly invasive cancers (pT2-4N0M0), 2 (40.0%) of 5 loco-regional node-positive patients (pN1-2M0), and 6 (75.0%) of 8 patients with distant metastases. Positive rates, therefore, increased with tumor extension (P = 0.0033, Kruskal-Wallis test). Furthermore, sequential blood sampling was performed in three patients with metastases during and after systemic chemotherapy, and UP-II-positive cells were found to have disappeared in two patients who responded well to the systemic chemotherapy. These results suggest that our nested reverse transcription-PCR assay for UP II is highly specific and might be used as a tumor marker for molecular staging of urothelial cancers, although the sensitivity is not so optimal.

Persistent and secondary adenovirus-mediated hepatic gene expression using adenovirus vector containing CTLA4IgG.

Adenovirus vectors can transfer recombinant genes efficiently into a wide variety of cells in vivo, but have serious limitations: gene expression is transient and secondary gene transfer is inefficient or impossible because of cellular and humoral immune responses against adenovirus-transduced cells. To solve these limitations, we have constructed an adenovirus vector, Adex1CACTLA4IgG, that expresses CTLA4IgG molecules. After in vivo administration of Adex1CACTLA4IgG (9.0 x 10(9) PFU), the peak level of serum CTLA4IgG was 29.8 mg/ml on day 4. The serum CTLA4IgG concentration gradually fell but was still 5.7 mg/ml on day 90. However, the serum concentration of CTLA4IgG was elevated after a second administration of Adex1CACTLA4IgG. The production of antibody against adenovirus was completely prevented after treatment with Adex1CACTLA4IgG. In addition, coadministration of Adex1CALacZ with Adex1CACTLA4IgG induced persistent hepatic expression of beta-Gal molecules, while administration of Adex1CALacZ alone induced transient expression of beta-Gal molecules. More importantly, on day 160 a secondary challenge with Adex1CALacZ was possible in mice treated with Adex1CALacZ plus Adex1CACTLA4IgG. Thus, we have demonstrated that (1) gene expression of a recombinant adenovirus, Adex1CACTLA4IgG, is persistent in liver and secondary administration of this adenovirus is possible, (2) coadministration of Adex1CACTLA4IgG virus with another adenovirus, AdexCALacZ, prolongs AdexCALacZ-mediated gene expression, and (3) Adex1CACTLA4IgG is useful for secondary challenge with Adex1CALacZ.

Lamina densa malformation involved in histogenesis of primary localized cutaneous amyloidosis.

Skin lesions of lichenoid amyloidosis and macular amyloidosis were immunohistochemically investigated using five monoclonal antibodies against basement membrane zone (BMZ) components. A hemidesmosomal component did not contribute to amyloid deposits, but components of the lamina densa and anchoring fibrils were associated with amyloid deposits in the uppermost dermis. Immunoelectron microscopy revealed that these BMZ components were not only aggregated in the BMZ and dermis, but were also involved in the individual amyloid islets. The lamina densa was disrupted in the interface areas just above the amyloid deposits, where cytoplasm of the basal cells directly faced the aggregate of amyloid filaments. Aggregates of some BMZ components were continuous to the amyloid islets from the lamina densa area. These findings suggest that a lamina densa malformation is involved in amyloid production in the interface of the BMZ, and support the secretion theory rather than the fibrillar body theory of amyloidogenesis in these types of primary localized cutaneous amyloidosis.

Mapping and transcriptional properties of RT1 class II region genes.

The class II region of major histocompatibility complex of the rat. Rattus norvegicus (RT1) consists of RT1.B, RT1.D, and RT1.H subregions. The gene order around the H subregion was determined as RT1.A--H beta-H alpha--B by RFLP analysis of naturally occurring intra-RT1 recombinant rats with HLA DP probes. A unique recombinant strain, LEJ, was found to have its recombinational site between H beta and H alpha (RT1.AuH beta uH alpha bBbDb). Northern analysis of class II mRNAs showed that transcripts of RT1.D alpha, RT1.D beta, and RT1.B alpha shared identical sizes among various strains of rats, but RT1.B beta mRNA showed allele-specific size heterogeneities. Northern hybridization with HLA DP alpha probes detected possible RT1.H alpha transcripts. On the other hand, no clear signal of H beta was observed. BDIX whose RT1.B products had not been identified was found to transcribe B alpha and B beta mRNAs.

Cytokine regulation of ICAM-1 expression on human renal tubular epithelial cells in vitro.

Regulation of the intercellular adhesion molecule-1 (ICAM-1) expression on human renal tubular epithelial cells in culture (hKEC-1) was investigated. A large proportion of hKEC-1 cells from the primary cultures expressed the ICAM-1 antigen. Supernatants from mixed lymphocyte reaction (MLR) of both specific and third-party combinations augmented the expression of the ICAM-1 antigen, in a dose-dependent manner. A kinetic study revealed maximal augmentation by MLR supernatant on the first day, with a gradual decrease thereafter. Among several recombinant human cytokines tested, i.e., interferon-gamma, tumor necrosis factor-alpha, interleukin 1 alpha and beta, and IL-4, IFN-gamma, TNF-alpha, and IL-1 alpha/beta were shown to augment the expression of ICAM-1. MLR supernatants and IFN-gamma were more effective in augmenting ICAM expression than TNF-alpha and IL-1 alpha/beta. IFN-gamma upregulated ICAM-1 expression in a dose-dependent manner, and maximal augmentation was achieved on the first day. The MLR supernatants were shown to contain IFN-gamma and TNF-alpha, and the activity of the MLR supernatant was partially inhibited by neutralizing antibody against IFN-gamma. These data suggest that cytokines, especially IFN-gamma, TNF-alpha, and IL-1 alpha/beta, released by T cells and antigen-presenting cells upon recognition of alloantigens upregulate ICAM-1 expression on renal tubular epithelial cells. This may result in an increase in the attachment of graft-infiltrating T cells to the renal tubular cells, by the ICAM-1-LFA-1 interaction.

Identification of ICAM-1-positive cells in the nongrafted and transplanted rat kidney--an immunohistochemical and ultrastructural study.

Cellular localization of intercellular adhesion molecule-1 (ICAM-1) in rat nongrafted intact kidneys and in transplanted kidneys was investigated using monoclonal anti-rat ICAM-1, 1A29. The major ICAM-1-positive cells in the nongrafted and isografted kidneys were endothelial cells in the large vessels and intertubular capillaries, as observed using light microscopy. A weak, but specific expression of ICAM-1 antigen was noted in the glomeruli, but the exact localization and cell type were not clearly discernible. In the allograft, the ICAM-1-positive cells found in the nongrafted and isografted kidneys also expressed ICAM-1 antigen. In addition, tubular epithelial cells at the luminal border and some infiltrating cells in the allograft expressed ICAM-1. In the allograft, some graft-infiltrating cells were shown to be lymphocyte function-associated antigen-1(LFA-1)-positive. As the nature of ICAM-1-positive cells in the infiltrates was unclear, we examined ICAM-1-positive cells using immunoelectron microscopy and the direct immunoperoxidase method. Glomerular endothelial cells, podocytes, and Bowman's capsular epithelial cells expressed ICAM-1 antigen in the nongrafted and transplanted kidneys. Among the infiltrating cells in the allograft, the major ICAM-1 positive cells were macrophagelike tissues, and some blastic lymphocytes also expressed ICAM-1. Only rarely did the proximal tubular cells express ICAM-1 antigen at the luminal surfaces in the intact kidney. In the allograft, the proximal, distal, and collecting ductular epithelial cells expressed ICAM-1 at the luminal surface, and in addition, the ICAM-1 antigen was also localized at the basal surfaces of some of the renal proximal tubular epithelial cells. The upregulated ICAM-1 expression in the allograft may accelerate graft rejection by augmenting adhesiveness of LFA-1-positive graft-infiltrating cells.