Existence of Pink1 antisense RNAs in mouse and their localization.

PTEN-induced kinase 1 (PINK1), which is identified as the gene transactivated by the tumor suppressor PTEN, has been found to be one of the causative genes in Parkinson's disease (PD). In order to understand PD, rodent models containing affected Pink1 such as loss-of-function mutations have been exploited. Recently, natural antisense RNA of PINK1 has been demonstrated to be involved in the regulation of the PINK1 locus. However, no antisense RNAs of Pink1 except for human have been reported so far. Therefore, in the present study, while searching for the Pink1 antisense RNAs in mouse, we found that the antisense RNAs are transcribed from a mouse genomic region corresponding to the human region from which the antisense RNAs are produced. Further, we investigated the localization of the antisense RNAs in mouse brain using in situ hybridization; this demonstrated that the antisense RNAs were localized in the regions of brain where the Pink1 mRNA was found. In addition, the mRNA and antisense RNAs were found more densely in the hippocampus than in the other brain regions in newborn and 1-week-old mice, while those RNAs were found uniformly in the mouse brain regions of embryo day (E) 14, E17, and 8-weeks-old.

Genomic organization, expression and evolution of porcine CRSP1, 2, and 3.

Recently we identified and characterized porcine calcitonin receptor-stimulating peptide (CRSP) 1, CRSP2 and CRSP3 as members of the calcitonin/calcitonin gene-related peptide (CT/CGRP) family. In the present study, the genomic sequences and organization of CRSP1, 2, and 3 were determined, and the expression of the genes in the porcine brain was investigated using in situ hybridization. Analysis of 5'-upstream regions of the three CRSPs demonstrated that CRSP1 and CRSP2 have almost identical sequences (>98% similarity) and high sequence similarities including functional transcription binding sites with the corresponding region of human CALCA (CT/alpha CGRP), whereas CRSP3 retains less similarity with the above genes. RH mapping of CRSPs demonstrated that they resided in a region of swine chromosome 2 (SSC2). The arrangement of the genes in the region was found to be conserved in corresponding human and mouse regions. In situ hybridization demonstrated sense transcripts of the three genes in cerebrum, hippocampus, hypothalamus, pons/midbrain, and thalamus of 3-month-old pigs, and CRSP2 sense transcripts additionally in tractus opticus. The sense transcripts of alpha CGRP and CALCB (beta CGRP) were detected in cerebrum, hippocampus, and pons/midbrain of newborn mice, and to a lesser extent in pons/midbrain of 8-week-old mice. These results taken together with the chromosomal conservation and phylogenetic clustering of CT/CGRP family indicate that CRSP1, 2, and 3 may be functionally different from alpha CGRP and beta CGRP, though they are indicated to have a common progenitor gene.

Modulation of MUC1 mucin as an escape mechanism of breast cancer cells from autologous cytotoxic T-lymphocytes.

MUC1 mucin is known to serve as a target molecule in the killing of breast cancer cells by cytotoxic T-lymphocytes (CTLs). We searched for a possible mechanism allowing tumour cells to escape from autologous CTLs. When the killing of breast cancer cells by autologous lymphocytes was examined in 26 patients with breast cancer, significant tumour cell lysis was observed in 8 patients, whereas virtually no autologous tumour cell lysis was detected in as many as 18 patients. In the patients who showed negligible tumour cell lysis, the autologous tumour cells expressed MUC1-related antigenic epitopes much more weakly than the tumour cells in the patients who exhibited strong cytotoxicity (significant statistically at P< 0.0005-0.0045), suggesting that the unresponsiveness of cancer cells to CTLs observed in these patients was mainly due to loss of MUC1 expression or modulation of its antigenicity. A breast cancer cell line, NZK-1, established from one of the cytotoxicity-negative patients, did not express MUC1 and was resistant to killing by CTLs, while control breast cancer cell lines expressing MUC-1 were readily killed by CTLs. Transfection of NZK-1 cells with MUC1 cDNA induced significant lysis by autologous T-lymphocytes. These results supported the importance of MUC1 mucin in autologous anti-tumour immunity, but suggested that the major escape mechanism of tumour cells from autologous T-lymphocytes is the loss and/or modulation of MUC1 antigenicity on tumour cells, which would limit the effectiveness of possible immunotherapy designed to target the MUC1 mucin.

Loss of cortical and thalamic neuronal tenascin-C expression in a transgenic mouse expressing exon 1 of the human Huntington disease gene.

A transgenic mouse containing the first exon of the human Huntington's disease (HD) gene has revealed a variety of behavioral and pathophysiological anomalies reminiscent of certain aspects of human Huntington's disease (HD). The present study has found that expression of the extracellular matrix glycoprotein tenascin-C appears to be unaffected in astroglial cells in wild-type and R6/2 transgenic mice that express the mutant huntingtin protein but that it is conspicuously absent in two neuronal populations within the cerebral cortex and thalamus of the R6/2 mice. Loss of tenascin-C expression begins between the fourth and eighth postnatal weeks, coincidental with the onset of abnormal behavioral phenotype and the appearance of intranuclear inclusion bodies and neuropil aggregates. By 12 weeks, R6/2 mice exhibit a complete absence of tenascin-C neuronal immunolabeling, a disappearance of cRNA probe-positive neurons across discrete cytoarchitectonic regions of the dorsal thalamus (e.g., the ventromedial, parafascicular, lateral posterior, and posterior thalamic groups) and frontal cortex, and an accompanying thalamic astrogliosis. The loss of neuronal tenascin-C expression includes structures that are known to send converging excitatory axonal projections to the caudate-putamen, the structure that is most at risk for neurodegeneration in HD. Altered neuronal expression of tenascin-C in R6/2 mice implicates altered transcriptional activities of the mutant huntingtin protein. The abnormal biochemistry and possibly abnormal activity of thalamostriate and corticostriate projection neurons may also affect abnormal neuronal activities in their primary connectional target, the neostriatum, which is severely compromised in HD.

Tissue-specific carcinogenesis in transgenic mice expressing the RET proto-oncogene with a multiple endocrine neoplasia type 2A mutation.

Germ line mutations of the RET proto-oncogene are responsible for the development of multiple endocrine neoplasia type 2A (MEN 2A), an inherited cancer syndrome characterized by medullary thyroid carcinoma, pheochromocytoma, and parathyroid hyperplasia. To study the mechanism of tissue-specific tumor development by RET with a MEN2A (cysteine 634-->arginine) mutation, we generated transgenic mice by introducing the RET-MEN2A gene fused to Moloney murine leukemia virus long terminal repeat. Expression of the transgene and its product was detected at variable levels in a variety of tissues including thyroid, heart, liver, colon, parotid gland, and brain. All of 29 mice analyzed developed thyroid C-cell hyperplasia or medullary carcinoma, accompanying high levels of serum calcitonin. In addition, development of mammary or parotid gland adenocarcinoma was observed in one-half of the transgenic mice. RET dimerization and its complex formation with Shc and Grb2 adaptor proteins were detected in tumor tissues. Unexpectedly, no tumor formation was found in other tissues despite RET-MEN2A expression where RET dimerization was undetectable. Because these tissues but not tumors expressed glial cell line-derived neurotrophic factor family receptor alpha (GFR alpha) at high levels, this suggested that GFR alpha expression may interfere in the dimerization of the RET-MEN2A mutant proteins, leading to tissue-specific tumor development in vivo.

Autologous dendritic cells or cells expressing both B7-1 and MUC1 can rescue tumor-specific cytotoxic T lymphocytes from MUC1-mediated apoptotic cell death.

We attempted to induce MUC1-specific cytotoxic T lymphocytes (CTLs) by mixed-lymphocyte tumor cell culture (MLTC) using two allogeneic MUC1-positive cancer cell lines, T-47D and MCF7. The induced CTLs exhibited MUC1-specific cytotoxicity 16 days after the initial stimulation. However, these CTLs underwent apoptotic death within 16 days. To examine whether the B7-1 molecule is required for the expansion of the responder cells, a B7-1(+)/MUC1(-) cell line was transfected with MUC1 cDNA, and the resulting transfectant was employed as a stimulator in an autologous MLTC. The CTLs exhibited MUC1 specificity but also continued to propagate. In parallel, autologous dendritic cells (DCs) were added to an MLTC containing peripheral blood lymphocytes (PBLs) and the allogeneic MUC1-positive stimulators. The CTLs demonstrated MUC1 specificity and their number increased. This suggests that the B7-1 molecule is required for rescuing CTLs from MUC1-mediated apoptotic death, but not for the induction of MUC1-specific responsiveness. This strategy to obtain the CTLs efficiently may be useful for adoptive immunotherapy against cancer.

Vacuolar processing enzyme is up-regulated in the lytic vacuoles of vegetative tissues during senescence and under various stressed conditions.

Vacuolar processing enzyme (VPE) has been shown to be responsible for maturation of various seed proteins in protein-storage vacuoles. Arabidopsis has three VPE homologues; betaVPE is specific to seeds and alphaVPE and gammaVPE are specific to vegetative organs. To investigate the activity of the vegetative VPE, we expressed the gammaVPE in a pep4 strain of the yeast Saccharomyces cerevisiae and found that gammaVPE has the ability to cleave the peptide bond at the carbonyl side of asparagine residues. An immunocytochemical analysis revealed the specific localization of the gammaVPE in the lytic vacuoles of Arabidopsis leaves that had been treated with wounding. These findings indicate that gammaVPE functions in the lytic vacuoles as the betaVPE does in the protein-storage vacuoles. The betaVPE promoter was found to direct the expression of the beta-glucuronidase reporter gene in seeds and the root tip of transgenic Arabidopsis plants. On the other hand, both the alphaVPE and gammaVPE promoters directed the expression in senescent tissues, but not in young intact tissues. The mRNA levels of both alphaVPE and gammaVPE were increased in the primary leaves during senescence in parallel with the increase of the mRNA level of a senescence-associated gene (SAG2). Treatment with wounding, ethylene and salicylic acid up-regulated the expression of alphaVPE and gammaVPE, while jasmonate slightly up-regulated the expression of gammaVPE. These gene expression patterns of the VPEs were associated with the accumulation of vacuolar proteins that are known to respond to these treatments. Taken together, the results suggest that vegetative VPE might regulate the activation of some functional proteins in the lytic vacuoles.

Vacuolar processing enzyme is self-catalytically activated by sequential removal of the C-terminal and N-terminal propeptides.

A vacuolar processing enzyme (VPE) responsible for maturation of various vacuolar proteins is synthesized as an inactive precursor. To clarify how to convert the VPE precursor into the active enzyme, we expressed point mutated VPE precursors of castor bean in the pep4 strain of Saccharomyces cerevisiae. A VPE with a substitution of the active site Cys with Gly showed no ability to convert itself into the mature form, although a wild VPE had the ability. The mutated VPE was converted by the action of the VPE that had been purified from castor bean. Substitution of the conserved Asp-Asp at the putative cleavage site of the C-terminal propeptide with Gly-Gly abolished both the conversion into the mature form and the activation of the mutated VPE. In vitro assay with synthetic peptides demonstrated that a VPE exhibited activity towards Asp residues and that a VPE cleaved an Asp-Gln bond to remove the N-terminal propeptide. Taken together, the results indicate that the VPE is self-catalytically maturated to be converted into the active enzyme by removal of the C-terminal propeptide and subsequent removal of the N-terminal one.

Reconstitution of functional L-selectin ligands on a cultured human endothelial cell line by cotransfection of alpha1-->3 fucosyltransferase VII and newly cloned GlcNAcbeta:6-sulfotransferase cDNA.

Recently, we proposed sialyl 6-sulfo Lewis X as a major carbohydrate-capping group of the L-selectin ligands on high endothelial venules in human lymph nodes. In this study we succeeded in reconstituting functional L-selectin ligands on a cultured human endothelial cell line, ECV304, by transfecting the alpha1-->3fucosyltranseferase VII (Fuc-T VII) and newly cloned GlcNAcbeta:6-sulfotransferase (6-Sul-T) cDNAs. The ECV304 cells transfected with Fuc-T VII cDNA expressed conventional sialyl Lewis X detected with specific antibodies including 2H5, whereas the cells transfected with 6-Sul-T cDNA expressed sialyl 6-sulfo lactosamine as well as MECA-79-defined carbohydrate determinants, but these singly transfected cells failed to express sialyl 6-sulfo Lewis X, as detected with the antisialyl 6-sulfo Lewis X mAb G152. Sialyl 6-sulfo Lewis X appeared only on the cells that were cotransfected with both 6-Sul-T and Fuc-T VII cDNAs. Significant adhesion of L-selectin-expressing cells was seen only to the doubly transfected ECV304 cells and was inhibited by G152. No adhesion was observed to the cells transfected either with 6-Sul-T or with Fuc-T VII cDNA alone. The mRNAs of the two enzymes were expressed or were inducible upon interleukin 1 stimulation in human endothelial cells. These results indicate that a set of carbohydrate determinants synthesized by the concerted action of the two enzymes, as typically represented by the sialyl 6-sulfo Lewis X-capping group, serves as an essential component of the ligand for L-selectin and that the reagents 2H5 and MECA-79, utilized in earlier studies to detect L-selectin ligand on high endothelial venules, recognize two different aspects of the same set of synthetic products.

Effect of tenascin-C deficiency on chemically induced dermatitis in the mouse.

Tenascin-C is a large extracellular matrix glycoprotein characterized by its spatiotemporal expression during embryogenesis, carcinogenesis, and wound healing. Many in vitro studies on tenascin-C have revealed its multifunctional properties; however, disruption of the tenascin-C gene did not reveal any obvious abnormalities during development, and its function in vivo remains unclear. Here, we investigated whether tenascin-C is involved in inflammatory dermatitis in adults by studying chemically induced dermatitis in tenascin-C knockout mice. An epicutaneous application of a hapten, 2,4-dinitrofluorobenzene, to the ear skin of BALB/CA mice resulted in inflammation and induced the expression of tenascin-C. In congenic tenascin-C knockout mice, the dermatitis occurred more severely than in wild-type mice; infiltration of polymorphonuclear cells in knockout mice persisted longer than in wild-type mice, and the elastosis-like disorganized extracellular matrix was also seen in the ear. These results suggest that tenascin-C plays a role in vivo in inflammatory responses in the skin, and that the genetic background has profound effects on the function of tenascin-C in mouse dermatitis.

Gene expression of fucosyl- and sialyl-transferases which synthesize sialyl Lewisx, the carbohydrate ligands for E-selectin, in human breast cancer.

The adhesion of circulating cancer cells to vascular endothelium is an important step in the hematogenous metastasis of cancer. Until recently, it has been believed that carbohydrate antigens are expressed on cancer cells, and E-selectin is expressed on endothelial cells to effect this adhesion. We investigated the gene expression of fucosyl-transferase (Fuc-T) and sialyltransferase (ST), which are involved in the synthesis of sialyl Lewisx (s-Lex) in breast cancer by using Northern blot analysis. The concentration of s-Lex in the cancerous portion was increased, compared to that in the adjacent non-cancerous portion. A correlation was found between the concentration of s-Lex and the amount of Fuc-T VI message in 9 cases of breast cancer tissue. Expression of the Fuc-T III message was found in only one case who expressed s-Lea. No expression of the Fuc-T V or VII message was observed. There was no relationship between the concentration of s-Lex and the amount of ST3N and ST4 transcripts. Similar findings were obtained from an analysis using cell lines derived from human breast cancer. When Fuc-T VI gene was transfected to MCF-7 cells, the expression of s-Lex was markedly induced on MCF-7 cells, and the attachment of cancer cells to endothelial cells was enhanced. These findings suggest that Fuc-T VI is chiefly involved in the synthesis of s-Lex on breast cancer cells.

A methylation imprint mark in the mouse imprinted gene Grf1/Cdc25Mm locus shares a common feature with the U2afbp-rs gene: an association with a short tandem repeat and a hypermethylated region.

We identified a sperm-specific methylation imprint mark (Site II) associated with a short tandem repeat sequence and a site/region methylated in both gametes (Site I) in the Grf1 locus on mouse chromosome 9, which shared a common feature with the U2afbp-rs gene. Sites or regions of gamete-specific methylation in imprinted genes are strong candidates for carrying information regarding the parental origin of alleles. The gamete-specific methylation pattern of Sites I and II was conserved after fertilization, but attained the somatic cell pattern by the blastocyst stage. In primordial germ cells, Site I was methylated, but Site II was unmethylated in both male and female embryos, suggesting that the sperm-specific methylation imprint mark in Site II was established during spermatogenesis. These common features in methylation imprint regions may be a clue to identifying regions carrying primary information for the imprinting regulation.

Tenascin-C promotes healing of Habu-snake venom-induced glomerulonephritis: studies in knockout congenic mice and in culture.

Mice without the gene for tenascin-C, a multifunctional extracellular matrix protein expressed in many important biological events, including wound healing, did not show any phenotype. However, it is now obvious that the phenotype of deletion of one gene frequently depends on the genetic background. Therefore, we have newly generated tenascin-C knockout mice (KO) by backcrossing original KO into three congenic lines: C57BL/6N, BALB/cA, and GRS/A (GR). And we investigated the disease course of reversible kidney injury, Habu-snake venom-induced proliferative glomerulonephritis. In all strains, the disease was more severe in KO, but the severity varied with the strain. The KO-GR showed irreversibility; all treated KO-GR died by the 4th month due to renal failure. The diseased KO-GR showed abnormal regenerative reactions, including reduced proliferation of mesangial cells, key players in glomerulonephritis, and reduced production of some kinds of cytokines and matrices, leading to poor formation of granulation tissue. In culture, the mesangial cells from the KO-GR had the same potential for proliferation and response to cytokines as controls, but interestingly, to achieve this potential, they required contact with tenascin-C. These reactions were blocked by an anti-tenascin monoclonal antibody. The results of the present study, the first report showing the most dramatic phenotype so far discovered, have strongly suggested the importance of tenascin-C in the resolution of the renal inflammation and that of the genetic background on which the KO was developed.

Molecular cloning, chromosomal assignment and tissue-specific expression of a murine alpha(1,2)fucosyltransferase expressed in thymic and epididymal epithelial cells.

Terminal Fucalpha(1-2)Galbeta epitopes have been proposed to play significant roles in cell-cell interactions in development, cell adhesion, and malignant transformation. To begin to investigate the regulation and function of alpha(1-2)fucosylated epitopes in an animal model, we have isolated and characterized a mouse genomic DNA segment encoding a protein orthologous to the human H blood group locus alpha(1,2)fucosyltransferase (FUT1). This segment maintains an open reading frame encoding 376 amino acids sharing 75% sequence identity with the enzyme encoded by human FUT1, and 55% sequence identity with the enzyme encoded by the human Secretor blood group locus (FUT2). Expression of the open reading frame in COS-7 cells yields an alpha(1,2)fucosyltransferase activity with a Km of 7.6 mM for phenyl-beta-d-galactoside. Southern blotting and interspecific backcross analyses indicate that this murine locus represents a single copy sequence mapping to a novel locus 2.1 centimorgans from the Klk1 locus, in a region of homology between mouse chromosome 7 and the human FUT1 locus on the long arm of chromosome 19. Mouse FUT1 yields a 2.8 kb mRNA transcript identifiable in many organs, including thymus, lung, stomach, pancreas, small intestine, colon, uterus and epidiymis. Hybridization analyses in situ localize expression of FUT1 transcripts to thymic medullary and epididymal epithelial cells, implying that this gene determines the expression of cell surface Fucalpha(1-2)Galbeta epitopes in these tissues.

An aspartic endopeptidase is involved in the breakdown of propeptides of storage proteins in protein-storage vacuoles of plants.

To understand the mechanism of the maturation of various proteins in protein-storage vacuoles, we purified a 48-kDa aspartic endopeptidase composed of 32-kDa and 16-kDa subunits from castor bean. Immunocytochemical and cell fractionation analyses of the endosperm of maturing castor bean seed showed that the aspartic endopeptidase was localized in the matrix of the protein-storage vacuoles, where a variety of seed storage proteins were also present. The amount of the aspartic endopeptidase increased at the mid-maturation stage of the seeds before accumulation of the storage proteins. To determine how the aspartic endopeptidase is responsible for maturation of seed proteins in concert with the vacuolar processing enzyme, we prepared 35S-labeled proproteins of seed proteins from the endoplasmic reticulum fraction of pulse-labeled maturing endosperm and used the authentic proproteins as substrates for in vitro processing experiments. The purified aspartic endopeptidase was unable to convert any of three endosperm proproteins, pro2S albumin, proglobulin, and proricin, into their mature sizes, while the purified vacuolar processing enzyme could convert all three proproteins. We further examined the activity of aspartic endopeptidase on the cleavage of an internal propeptide of Arabidopsis pro2S albumin, which is known to be removed post-translationally. The aspartic endopeptidase cleaved the propeptide at three sites under acidic conditions. These results suggest that aspartic endopeptidase cannot directly convert pro2S albumin into the mature form, but it may play a role in trimming the C-terminal propeptides from the subunits that are produced by the action of the vacuolar processing enzyme.

Altered mRNA expression of specific molecular species of fucosyl- and sialyl-transferases in human colorectal cancer tissues.

Human colorectal cancers express various cancer-associated carbohydrate determinants such as Lewis Y or sialyl Lewis A, suggesting a considerable alteration in glycosyltransferase activities occurring upon malignant transformation. We investigated the mRNA amounts of fucosyltransferase (Fuc-T) and sialyltransferase (ST) isoenzymes, including Fuc-T III, IV, V, VI and VII and ST-3N, ST-30 and ST-4, in human colorectal cancer tissues by Northern blotting and RT-PCR. Regarding fucosyltransferases, mRNA of Fuc-T III and VI was not significantly altered, and only Fuc-T IV mRNA showed a moderate increase in cancer tissues when compared with adjacent non-malignant colonic epithelia taken from the same patient (273 +/- 96%; p < 0.001). The moderate increase of Fuc-T IV message may be related to an enhanced expression of Lewis Y in colon cancer tissues. In the ST isoenzymes, mRNA for ST-3N remained unchanged, whereas that for ST-4 decreased significantly in cancer tissues, to 32 +/- 29%, (p < 0.005). The most remarkable finding was that the message of ST-30 was prominently increased in cancer tissues compared with non-malignant colorectal mucosa. When further investigated by quantitative RT-PCR assays on a larger series of patients with colorectal cancers, the average increase in mRNA for ST-30 was 459 +/- 200% compared with that in adjacent non-malignant epithelium (significant at p < 0.0001). The increase of ST-30 message was more prominent in the cancer tissues strongly expressing sialyl Lewis A than in the cancer tissues expressing sialyl Lewis A only weakly or moderately (significant at p < 0.05). The marked increase in the message of ST-30 is suggested to be related to an enhanced expression of sialylated carbohydrate determinants in colon cancer tissues including sialyl Lewis A, since the enzyme exhibited a significant activity against the type 1 chain carbohydrate substrate and produced the precursors for sialyl Lewis A synthesis, when its cDNA was expressed in Cos-7 cells.

A rapid increase in the level of binding protein (BiP) is accompanied by synthesis and degradation of storage proteins in pumpkin cotyledons.

The binding protein (BiP) has been implicated in cotranslational folding of nascent polypeptides, and in the recognition and disposal of aberrant polypeptides. To elucidate the involvement of BiP in the biosynthesis of vacuolar proteins, we have characterized the protein in pumpkin cotyledons during seed maturation and seedling growth. Isolated microsomes from maturing pumpkin cotyledons contained a significant amount of BiP, protein-disulfide isomerase and calreticulin. We have purified a 70-kDa protein; sequences of the N-terminus and internal fragments of this protein exhibited a high identity to the sequence of soybean. Immunoblot analysis with specific antibodies raised against the purified BiP showed that the amount of BiP in a cotyledon increased markedly at the middle stages and then decreased. The increase was accompanied by the synthesis of storage proteins and the development of the endoplasmic reticulum in the cotyledons at the middle stage of seed maturation. Most of these storage proteins degraded dramatically between 2 and 5 days after seed germination, and the degradation was also accompanied by a rapid increase in the level of BiP. Subcellular fractionation of the 4-day-old cotyledons showed a high accumulation of BiP in the endoplasmic reticulum. It is possible that BiP might be involved in the synthesis of seed storage proteins during maturation and in the synthesis of hydrolytic enzymes responsible for the degradation of the storage proteins during seed germination.

Human T-cell leukemia virus-1 encoded Tax protein transactivates alpha 1-->3 fucosyltransferase Fuc-T VII, which synthesizes sialyl Lewis X, a selectin ligand expressed on adult T-cell leukemia cells.

Leukemia cells in patients with adult T-cell leukemia (ATL) and related cell lines strongly express the carbohydrate determinant sialyl Lewis X, a ligand for selectins. Its expression is thought to be related closely to the extravascular infiltration of the leukemia cells. Human T-cell leukemia virus type 1 (HTLV-1), the etiological agent of ATL, produces Tax protein, which is implicated in leukemogenesis through its transactivating effect on various cellular genes. In this study we investigated the transactivating effect of HTLV-1 Tax on the alpha 1-->3 fucosyltransferase Fuc-T VII, the putative rate-limiting enzyme in the synthesis of sialyl Lewis X in human leukocytes using JPX-9 cells. JPX-9 is a subclone of a non-ATL human lymphocytic leukemia cell line, Jurkat, and was established by introducing a metallothionein promoter-driven Tax expression plasmid. The JPX-9 cells as well as parental Jurkat cells did not express Fuc-T VII mRNA under normal culture conditions. When cultured in the presence of 10 microM CdCl2, Tax was induced and a significant amount of the Fuc-T VII message was ascertained by Northern blotting. The amount of the message was 24.5 times as much as was detected in non-treated cells, and was comparable to that which appeared by TPA stimulation of the cells, which is supposed to simulate the sequence of events occurring in normal activation of T lymphocytes activated by more physiological stimuli. Sialyl Lewis X determinant was expressed at the surface of CdCl2-treated cells, while the determinant was not detectable on either unstimulated JPX-9 or parental Jurkat cells. These results indicate that expression of sialyl Lewis X on leukemic cells in patients with ATL is at least partly due to the transactivation of the Fuc-T VII gene induced by the HTLV-1 Tax, and suggest that this leads to the accelerated extravascular infiltration of ATL cells.

Sulfated sialyl Lewis X, the putative L-selectin ligand, detected on endothelial cells of high endothelial venules by a distinct set of anti-sialyl Lewis X antibodies.

Endothelial cells of high endothelial venules (HEV) in human peripheral lymph nodes expressed a distinct type of sialyl Lewis X antigen, which was detected preferentially with a set of anti-sialyl Lewis X antibodies, 2F3, 2H5 and HECA-452 in immunohistochemistry, while another set of anti-sialyl Lewis X antibodies, FH-6 and CSLEX-1, failed to detect it. The adhesion of cells expressing L-selectin to HEV was inhibited by members of the former set of antibodies in Stamper-Woodruff assays performed on frozen sections of human peripheral lymph nodes. Transfection of a cultured endothelial cell line with a human alpha1-->3 fucosyltransferase, Fuc-T VII, resulted in the expression of a distinct type of sialyl Lewis X antigen having the reactivity similar to that of HEV; i.e., the antigen appearing on the transfectant clone was detectable only with the set of 2F3, 2H5 and HECA-452, but not with the set of FH-6 and CSLEX-1. Treatment of transfectant cells with sodium chlorate, a metabolic inhibitor of sulfation, resulted in reactivity to the members of the latter set of antibodies, suggesting that sulfation of sialyl Lewis X moiety was the cause of the discrepancy in the reactivity of the anti-sialyl Lewis X antibodies. When tested against various authentic sulfated sialyl Lewis X determinants, 6-sulfo sialyl Lewis X and 6,6'-bis-sulfo sialyl Lewis X were found to be reactive to the antibodies, 2F3, 2H5 and HECA-452, but not with antibodies FH-6 and CSLEX-1, suggesting that the distinct type of sialyl Lewis X determinant on the HEV endothelial cells and Fuc-T VII-transfected endothelial cell clone are mainly 6-sulfo and/or 6,6'-bis-sulfo sialyl Lewis X determinants.