High and pointed type of femoral localized reaction frequently extends to complete and incomplete atypical femoral fracture in patients with autoimmune diseases on long-term glucocorticoids and bisphosphonates.

Once a localized reaction (beaking) was detected, discontinuation of bisphosphonates (BPs) and switching to vitamin D supplementation or teriparatide therapy effectively improved its shape. When the localized reaction was high, of the pointed type, and/or accompanied by prodromal pain, the risks of complete and incomplete atypical femoral fracture increased and consideration of prophylactic fixation for such patients was required. INTRODUCTION: Femoral localized reaction (localized periosteal thickening of the lateral cortex, beaking) is reported to precede atypical femoral fractures (AFFs) and to develop in 8-10% of patients with autoimmune diseases taking BPs and glucocorticoids. The aims of the present study were to retrospectively investigate the shapes of localized reaction to consider how to manage the condition. METHODS: Twenty femora of 12 patients with autoimmune diseases who were on BPs and glucocorticoids exhibited femoral localized reaction. The heights of localized reaction were measured and the shapes classified as pointed, arched, and other. Localized reaction changes were divided into three categories: deterioration, no change, and improvement. A severe form of localized reaction was defined; this was associated with prodromal pain, de novo complete AFF, or incomplete AFF with a fracture line at the localized reaction. RESULTS: The mean height of localized reaction was 2.3 ± 0.8 mm (range, 1.0-3.7 mm) and the pointed type was 35%. Localized reaction was significantly higher (3.3 ± 0.8 vs. 2.1 ± 0.7 mm; p = 0.003) and the pointed type more common (78 vs. 27%; p = 0.035) in those with the severe form of localized reaction. Seven patients with localized reactions discontinued BPs just after localized reaction was detected, but five continued on BPs for 2 years. Localized reaction deterioration was more common in patients who continued than discontinued BPs (100 vs. 29%; p = 0.027). After 2 years, all patients had discontinued BPs and localized reaction did not deteriorate further in any patient. CONCLUSIONS: Once a localized reaction was detected, discontinuation of BPs and switching to vitamin D supplementation or teriparatide therapy effectively improved it. When the localized reaction was high, of the pointed type, and/or accompanied by prodromal pain, the risks of complete and incomplete AFF increased and consideration of prophylactic fixation for such patients was required.

The cumulative incidence of and risk factors for latent beaking in patients with autoimmune diseases taking long-term glucocorticoids and bisphosphonates.

SUMMARY: The incidence of beaking, which has been reported to precede atypical femoral fracture, was high and increased over 2 years in patients with autoimmune diseases who were taking bisphosphonates and glucocorticoids. Regular femoral X-rays are strongly recommended to screen for beaking, and bisphosphonate drug holidays should be considered. INTRODUCTION: Atypical femoral fractures (AFFs) have been recently recognized as complications associated with bisphosphonate (BP) use. AFFs are considered to be stress fractures; localized periosteal thickening of the lateral cortex is often present at the fracture site; this thickening is termed "beaking." Beaking has been reported to precede AFF. The aims of the present study were to evaluate the incidence of latent beaking in patients with autoimmune diseases taking BPs and glucocorticoids and to identify risk factors for beaking. METHODS: A total of 125 patients with autoimmune diseases who were taking BPs and glucocorticoids was included; 116 patients underwent X-rays and analysis of serum and urine bone metabolic markers annually for 2 years. Mean patient age was 54.5 years; there were 105 (90.5%) females and the mean duration of disease was 13.2 years. Focal lateral cortical thickening in femoral X-rays was defined as beaking. RESULTS: Beaking was detected in 15 femora of 10 patients (8.0%) at the time of recruitment. Over the 2-year observation period, the incidence of beaking increased to 21 femora of 12 patients (10.3%), and a complete AFF at the location of beaking occurred in one patient. Beaking was associated with a longer duration of BP treatment (6.1 ± 1.0 years vs. 5.0 ± 2.9 years, p = 0.01). Age 40-60 years, BP therapy ≥4 years, and diabetes mellitus were significantly associated with beaking. CONCLUSIONS: The incidence of beaking was high, and increased over 2 years, in patients with autoimmune diseases who were taking BPs and glucocorticoids. Regular femoral X-rays are strongly recommended to screen for beaking. Long-term BP/glucocorticoid use was a risk factor for beaking in patients with autoimmune diseases; BP drug holidays should be considered.

Detection of human coronavirus RNA in surgical smoke generated by surgical devices.

BACKGROUND: Gaseous by-products generated by surgical devices - collectively referred to as 'surgical smoke' - present the hazard of transmitting infective viruses from patients to surgical teams. However, insufficient evidence exists to evaluate and mitigate the risks of SARS-CoV-2 transmission via surgical smoke. AIM: To demonstrate the existence and infectivity of human coronavirus RNA in surgical smoke using a model experiment and to evaluate the possibility of lowering transmission risk by filtration through a surgical mask. METHODS: Pelleted HeLa-ACE2-TMPRSS2 cells infected with human coronavirus were incised by electric scalpel and ultrasonic scalpel, separately. A vacuum system was used to obtain surgical smoke in the form of hydrosol. Reverse transcription-quantitative polymerase chain reaction was used to analyse samples for the presence of viral RNA, and infectivity was determined through plaque assay. Furthermore, a surgical mask was placed centrally in the vacuum line to evaluate its ability to filter viral RNA present in the surgical smoke. FINDINGS: In this model, 1/106 to 1/105 of the viral RNA contained in the incision target was detected in the collected surgical smoke. The virus present in the smoke was unable to induce plaque formation in cultured cells. In addition, filtration of surgical smoke through a surgical mask effectively reduced the amount of viral RNA by at least 99.80%. CONCLUSION: This study demonstrated that surgical smoke may carry human coronavirus, though viral infectivity was considerably reduced. In clinical settings, surgical mask filtration should provide sufficient additional protection against potential coronavirus, including SARS-CoV-2, infection facilitated by surgical smoke.

Characterization of the promoter region of the src family gene lyn and its trans activation by human T-cell leukemia virus type I-encoded p40tax.

The src family gene lyn is expressed preferentially in B lymphocytes but very little in normal T lymphocytes. Transcription of the lyn gene in T lymphocytes was shown to be induced by the p40tax protein encoded by human T-cell lymphotropic virus type I. For determination of the mechanism of p40tax-mediated trans activation, the transcriptional promoter region of the lyn gene was characterized. By endonuclease S1 mapping, the transcriptional initiation sites were identified within the 770-bp EcoRI-SacI fragment of the 5'-terminal portion of the human lyn gene. This fragment showed promoter activity when placed upstream of the bacterial chloramphenicol acetyltransferase gene and transfected into various cell lines. Nucleotide sequence analysis revealed that the lyn promoter region contained four GC box-like sequences but not a TATA or CCAAT box. In addition, it contained sequences characteristic of a cyclic AMP-responsive element, octamer-binding motif, PEA3-like motifs, and NF kappa B-binding motif-like sequence. Mutational analysis suggested that the octamer-binding motif sequence is of primary importance for the lyn promoter activity but that the other elements are not. Cotransfection of various chloramphenicol acetyltransferase constructs containing different length of the lyn promoter together with p40tax expression plasmids into Jurkat T cells showed that the sequence responsible for p40tax-induced transcription is present around the transcription initiation sites.

SR alpha promoter: an efficient and versatile mammalian cDNA expression system composed of the simian virus 40 early promoter and the R-U5 segment of human T-cell leukemia virus type 1 long terminal repeat.

We developed a novel promoter system, designated SR alpha, which is composed of the simian virus 40 (SV40) early promoter and the R segment and part of the U5 sequence (R-U5') of the long terminal repeat of human T-cell leukemia virus type 1. The R-U5' sequence stimulated chloramphenicol acetyltransferase (CAT) gene expression only when placed immediately downstream of the SV40 early promoter in the sense orientation. The SR alpha expression system was 1 or 2 orders of magnitude more active than the SV40 early promoter in a wide variety of cell types, including fibroblasts and lymphoid cells, and was capable of promoting a high level of expression of various lymphokine cDNAs. These features of the SR alpha promoter were incorporated into the pcD-cDNA expression cloning vector originally developed by Okayama and Berg.

Autocrine stimulatory mechanism by transforming growth factor beta in human hepatocellular carcinoma.

The serum concentration of transforming growth factor beta (TGF-beta) is elevated as tumors progress in hepatocellular carcinoma (HCC) patients. In this study, we examined whether modulation of tumor-derived TGF-beta signal transduction contributes to malignant progression. We investigated the production of TGF-beta1, the biological effects of TGF-beta and neutralizing antibody on HCC cells, activation of Smad 2, Smad 3, and Smad 4, induction of antagonistic Smads (Smad 6 and Smad 7), and promoter activities of two target genes, plasminogen activator inhibitor type 1 (PAI-1) and p15INK4B. In human cell lines HCC-M and HCC-T, TGF-beta accelerates their proliferation. Smad 2 was activated constitutively by an autocrine mechanism, because in the absence of exogenous TGF-beta, a high level of Smad 2 phosphorylation, induction of PAI-1 transcripts, and nuclear localization of Smad 2 were observed. This constitutive activation of Smad 2 was, at least in part, attributable to the lack of induction of antagonistic Smads by TGF-beta. However, Smads activated by tumor-derived TGF-beta constantly suppressed p151NK4B expression. In addition, 3 of 10 human HCC tissues showed nuclear localization of Smad 2 and low mRNA levels of p15INK4B and antagonistic Smads but a high level of PAI-1. Our observations suggest that this constant suppression of the p15INK4B gene could be involved in the malignant progression of HCC.

A trans-activator Tax of human T-cell leukemia virus type 1 binds to NF-kappa B p50 and serum response factor (SRF) and associates with enhancer DNAs of the NF-kappa B site and CArG box.

A transcriptional activator of human T-cell leukemia virus type 1 (HTLV-1) activates at least three distinct enhancers: the viral 21-bp enhancer, the NF-kappa B binding site of the IL-2R alpha gene and the CArG box of the c-fos gene. To understand the mechanisms of Tax transactivations of the NF-kappa B enhancer and CArG box, the interactions of Tax protein with their binding factors were analysed. Using a DNA affinity precipitation (DNAP) assay, we found here that Tax associates with the DNA sequences of the NF-kappa B site and CArG box. These Tax associations with enhancers were observed only in the presence of a nuclear factor(s) and were equal to the activating capacities of Tax mutants. To identify the nuclear factor(s), we defined conditions under which no Tax binding to the NF-kappa B binding site and CArG box was detected with a nuclear extract of 293T cells. Under these conditions, transfections with cDNAs of the NF-kappa B p50 and serum response factor (SRF) produced a factor(s) that mediated Tax binding to the NF-kappa B site and the CArG box respectively. Furthermore, purified Tax protein interacted with purified NF-kappa B p50 and purified SRF, indicating their direct bindings. These observations indicate that Tax protein associates with enhancer sequences of the NF-kappa B site and CArG box through NF-kappa B p50 and SRF respectively. Previously we demonstrated that Tax interacts with CREB and CREM proteins that bind to the 21-bp enhancer DNA. These results together suggest that indirect binding of Tax to DNA through each enhancer binding protein is a general mechanism for Tax transactivation of transcription.

Cyclic AMP response element-binding protein, CRE-BP1, mediates the E1A-induced but not the Tax-induced trans-activation.

The adenovirus E1A protein and tax protein (Tax) of human T-cell leukemia virus-1 (HTLV-1) are transcriptional regulators that do not bind to DNA directly. The ATF sites/CRE (cyclic AMP response element) of the adenovirus E4 promoter and the long terminal repeat of HTLV-1 have been shown to be required for E1A and Tax inducibility, respectively. Using the c-Myb-CRE-BP1 fusion protein, it was shown that CRE-BP1, which could bind to the ATF sites/CRE, mediated the E1A-induced trans-activation. For this activation, the N-terminal portion of CRE-BP1, which contained the putative metal finger structure, was essential but not sufficient. In contrast, the trans-activation induced by HTLV-1 Tax was not mediated by CRE-BP1. These results strongly suggested that E1A activates transcription through interaction with CRE-BP1, but another CRE-binding protein participates in the Tax-induced trans-activation.

Transcriptional activator Tax of HTLV-1 binds to the NF-kappa B precursor p105.

Human T-cell leukemia virus type 1 (HTLV-1) induces adult T-cell leukemia and also a neurological disease, tropical spastic paraparesis. Tax protein (p40tax) of HTLV-1 activates in trans its own transcriptional enhancer in the long terminal repeat and also those in some cellular genes such as interleukin 2 receptor alpha, granulocyte-macrophage colony-stimulating factor, Fos, Jun and MHC class I. Thus, Tax has been proposed to play a critical role in the pathogenesis induced by HTLV-1 infection. Here, we report formation of a complex of Tax protein with the precursor protein p105 of the NF-kappa B p50 subunit. p105 was co-immunoprecipitated with Tax protein from cells infected with HTLV-1 from cells transfected with the Tax expression plasmid, but not from cells transfected with inactive mutants of Tax. Furthermore, a GST-p105 fusion protein produced in Escherichia coli bound to Tax protein. These results strongly suggest that the trans-activator Tax protein forms a complex with precursor NF-kappa B p105 and plays a role in trans-activation of transcriptional initiation.

HTLV-I encoded Tax in association with NF-kappa B precursor p105 enhances nuclear localization of NF-kappa B p50 and p65 in transfected cells.

NF-kappa B is a heterodimeric protein composed of two subunits, p50 and p65, and is sequestered in the cytoplasm as an inactive form through the association with an inhibitory protein, I kappa B. In the present study, the effect of HTLV-I encoded Tax on the exogenously expressed NF-kappa B p105, which encodes a precursor of p50, was investigated using a COS-7 expression system. When p105 was expressed in COS-7 cells, the precursor p105 and the processed p50 were retained in the cytosolic fraction by associating each other, and p50 was not detected in the nuclear fraction. In the cells co-expressing p105 and Tax, the p50/p105 ratio in the cytosolic fraction reduced with an induction of p50 in the nuclear fraction, which gave rise to a significant increase in NF-kappa B binding activity. Enhancement of NF-kappa B binding activity was not observed by adenovirus encoded E1A and bovine papilloma virus encoded E2, and correlated well with the ability of Tax to associate with p105. When p105 delta X which lacks the repeats of the ankyrin motif was expressed in this system, p105 delta X and processed p50 were detected in the nuclear fraction and p50/p105 delta X ratio was not affected by co-expression with Tax. In the same expression system, exogenously expressed NF-kappa B p65 was retained in the cytoplasm by p105 and further expression of Tax allowed entry of p65 in the nucleus. These results suggest that, in this model system, p105 acts as an I kappa B to sequester p50 and p65 in the cytoplasm and that Tax by inhibiting I kappa B activity of p105, enhances nuclear localization of p50 and p65. These findings raise a possibility for a novel mechanism for the induction of NF-kappa B in the nucleus by Tax.

Effect of tunicamycin on functions of PGE1 receptors from mouse mastocytoma P-815 cells.

Prostaglandin E1 (PGE1) receptors from mouse mastocytoma P-815 cells were found to bind to a wheat germ agglutinin (WGA)-Agarose column, suggesting that the receptors are glycoproteins. To further elucidate the role of carbohydrate moieties in the PGE1 receptors for their binding activity to ligand, the P-815 cells were treated with tunicamycin, swainsonine or monensin. Tunicamycin, an inhibitor of N-glycosylation, dose- and time-dependently inhibited the binding of PGE1 to mastocytoma P-815 cells. Neither swainsonine, an inhibitor of Golgi mannosidase II, nor monensin, an inhibitor of processing beyond the high mannose stage, altered PGE1 binding properties of the cells. The inhibition of PGE1 binding by tunicamycin was observed when incorporation of [3H]glucosamine into macromolecules was inhibited. The inhibitory effect was not on their affinity but on their number of binding sites. Subcellular distributions of [3H]PGE1-binding activity showed that decreases in the binding activity by tunicamycin were highest in plasma membrane fractions. Treatment of membranes with various endo- and exoglycosidases did not affect PGE1 binding. PGE1-stimulated cyclic AMP accumulation in the cells was also inhibited by tunicamycin. These results suggest that PGE1 receptors of mastocytoma P-815 cells are glycoproteins and that inhibition of N-glycosylation of PGE1 receptors by tunicamycin results in the arrest of the translocation of newly synthesized receptors to the surface of mastocytoma P-815 cells.

Prostaglandin D2 receptor of mastocytoma P-815 cells--possible regulation by phosphorylation and dephosphorylation.

The 3H-labeled prostaglandin D2 [( 3H]PGD2) binding protein in the membrane fraction of mastocytoma P-815 cells was characterized. The specific binding of [3H]PGD2 to the cells or the membranes reached a maximum at pH 5.6, and was saturable, displaceable and of high affinity when incubated at 0 or 37 degrees C. The Bmax values for [3H]PGD2 binding in the two preparations at pH 5.6 were much higher at 0 degrees C than at 37 degrees C, whereas the Kd values were almost equal (85.3 nM for the cells and 80.5 nM for the membranes, respectively). High specific [3H]PGD2 binding activity in the mildly acid-treated cells was still observed when the external pH was raised from 5.6 to 7.2. Furthermore, specific [3H]PGD2 binding to the membranes (at 0 degrees C, pH 5.6) increased on addition of phosphatase inhibitors (NaF and molybdate) in the presence of 10 microM ATP, but practically disappeared on pretreatment of the membranes with phosphatase. On incubation of the membrane with [gamma-32P]ATP and molybdate, the stimulated incorporation of the [32P]phosphate into several peptides, including ones having an Mr of around 100,000-120,000, was observed. These results suggest that [3H]PGD2 binding in the mastocytoma P-815 cell membrane is controlled through phosphorylation-dephosphorylation of the receptor itself.

Structural relationship of human interferon alpha genes and pseudogenes.

We have isolated and characterized DNA segments containing IFN-alpha-related sequences from human lambda and cosmid clone banks. We describe six linkage groups comprising 18 distinct IFN-alpha-related loci, and report the nucleotide sequences of nine chromosomal IFN-alpha-genes with intact reading frames, as well as of five pseudogenes. Taking into account as yet unsequenced genes as well as clones described by others, there are now seven linkage groups and 23 loci, of which 15 correspond to potentially functional genes and six to non-functional genes; two loci remain unsequenced. Eighteen additional sequences are likely to be allelic to the above. The finding that at least two IFN-alpha genes appear to be natural hybrids of other IFN-alpha genes, and that two distinct IFN-alpha loci have completely identical coding sequences, although their flanking regions are different, is evidence for information exchange between the individual genes.

A new approach to high sensitivity differential hybridization.

We describe a new approach to differential hybridization, designed to identify cDNA clones representing rare mRNA species. Duplicate filters carrying a library of cDNA from phorbolmyristate acetate (PMA)-induced EL-4 cells in lambda gt11 were hybridized with high concentrations of unlabeled, cloned, single-stranded cDNA from induced and control EL-4 cells, respectively. Plaques binding single-stranded cDNA were revealed by a second round of hybridization with 35S-labeled DNA complementary to the vector moiety of the single-stranded cDNA. Plaques corresponding to PMA-induced mRNAs occurring at a level of about 1 part in 15,000 were isolated. We believe the method is at least ten times more sensitive than conventional differential hybridization.

Uncleaved env gp160 of human immunodeficiency virus type 1 is degraded within the Golgi apparatus but not lysosomes in COS-1 cells.

The fate of newly synthesized human immunodeficiency virus type 1 env gp160 was examined in COS-1 cells. The results of morphological chase experiments involving cycloheximide demonstrated that gp160 was retained in the Golgi apparatus for longer than the half-life of the molecule. The degradation of gp160 was insensitive to both bafilomycin A1 and leupeptin (< 0.2 mM), which block lysosomal proteolysis. However, degradation was effectively suppressed by leupeptin at higher concentrations, maximally at 1.7 mM. Furthermore, undegraded gp160 was accumulated in the Golgi apparatus, but was not detected in lysosomes. These results indicate that in COS-1 cells gp160 is not degraded in lysosomes, but rather that degradation takes place in the Golgi apparatus.

A role for Rev in the association of HIV-1 gag mRNA with cytoskeletal beta-actin and viral protein expression.

Human immunodeficiency virus type-1 (HIV-1) Rev acts by inducing the specific nucleocytoplasmic transport of a class of incompletely spliced RNAs that encodes the viral structural proteins. The transfection of HeLa cells with a rev-defective HIV-1 expression plasmid, however, resulted in the export of overexpressed, intron-containing species of viral RNAs, possibly through a default process of nuclear retention. Thus, this system enabled us to directly compare Rev+ and Rev+ cells as to the usage of RRE-containing mRNAs by the cellular translational machinery. Biochemical examination of the transfected cells revealed that although significant levels of gag and env mRNAs were detected in both the presence and absence of Rev, efficient production of viral proteins was strictly dependent on the presence of Rev. A fluorescence in situ hybridisation assay confirmed these findings and provided further evidence that even in the presence of Rev, not all of the viral mRNA was equally translated. At the early phase of RNA export in Rev+ cells, gag mRNA was observed throughout both the cytoplasm and nucleoplasm as uniform fine stippling. In addition, the mRNA formed clusters mainly in the perinuclear region, which were not observed in Rev+ cells. In the presence of Rev, expression of the gag protein was limited to these perinuclear sites where the mRNA accumulated. Subsequent staining of the cytoskeletal proteins demonstrated that in Rev+ cells gag mRNA is colocalized with beta-actin in the sites where the RNA formed clusters. In the absence of Rev, in contrast, the gag mRNA failed to associate with the cytoskeletal proteins. These results suggest that in addition to promoting the emergence of intron-containing RNA from the nucleus, Rev plays an important role in the compartmentation of translation by directing RRE-containing mRNAs to the beta-actin to form the perinuclear clusters at which the synthesis of viral structural proteins begins.

Limited sequence divergence of HTLV-I of Indian HAM/TSP patients from a prototype Japanese isolate.

Nucleotide sequences of two HTLV-I proviruses isolated from Indian patients with HAM/TSP were analyzed. The sequence data of the env, pX, and LTR regions showed 98-99% homologies with the prototype HTLV-I, ATK-1, isolated from a Japanese ATL patient, indicating that HTLV-I isolates in India and Japan are similar, with minor variations. However, certain small sequences of noncoding regions in the pX and LTR showed differences of 6.1 and 7.2%, respectively, thus the conclusion could vary depending on the regions and length of the sequences used for comparison.

Antigenic cross-reaction between the alpha types of human and mouse interferon.

Cross-neutralization of the alpha and beta types of human and mouse interferons was tested using antibodies directed against the heterologous types of interferons. The alpha type of mouse interferon (MuIFN-alpha) prepared from L cells was found to be completely neutralizable by high-titered antibody against human alpha-type interferon (HuIFN-alpha), although the titers were much lower than those obtained in neutralization reactions with the homologous interferon. MuIFN-alpha from virus-induced lymphocytes, as well as non-glycosylated MuIFN-alpha from L cells, reacted similarly. The cross-reaction was also observed by the binding of anti-HuIFN-alpha antibody to a column of immobilized MuIFN-alpha. The binding experiment indicated that only a small fraction of the anti-HuIFN-alpha antibody population is heterologous reactive. Reciprocally, HuIFN-alpha from L cells, again with relatively low antibody titers, and the antibody responsible for the heterologous reaction was shown to be the anti-MuIFN-alpha and not anti-MuIFN-beta type. It is concluded that the alpha types of human and mouse interferon bear an antigenic homology. On the other hand, no significant antigenic cross-reactivity has been detected between the beta types of human and mouse interferons.

Novel isoforms of human cyclic AMP-responsive element modulator (hCREM) mRNA.

The cyclic AMP-response element (CRE), a transcriptional enhancer, is regulated by CREB (CRE-binding protein) which is the leucine zipper protein phosphorylated by protein kinase A in response to cAMP signal. The highly homologous protein CREM (CRE-modulator) is thought to modulate CREB-stimulated transcription, and is also involved in transcriptional control during spermatogenesis. In this paper, we report two types of cDNAs of human CREM (hCREM), type 1 and type 2; type 1 is a group of human counterparts of the mouse CREM alpha and type 2 is a novel form having a distinct 5' exon which is unrelated to any species of the CREB and CREM isoforms so far described. This unique 5' region of type 2 hCREM may suggest its independent expression from type 1 CREM. The specific 5' region of type 2 hCREM consisted of 88 bp, containing an initiation codon for translation, but no possible phosphorylation site, suggesting different roles from type 1 CREM. Both type 1 and 2 hCREMs are expressed in lymphoid and non-lymphoid cell lines. Their excess expression by transfection induced suppression of cAMP-mediated activation of transcription, suggesting their negative regulation of CRE-mediated transcription.

Identification of transcriptional activation domain of TREB5, a CREB/ATF family protein that binds to HTLV-1 enhancer.

TREB5 (hXBP-1) protein is a transcription factor that recognizes the CRE-like element in enhancers of human T-cell leukemia virus and MHC class II gene and activates their transcription. TREB5 is a member of the CREB/ATF family, containing a basic amino acid region and leucine zipper structure (b-Zip structure). To characterize the key domain of TREB5 for transcriptional activation, mutational analysis was carried out. The C-terminal region of 148-221 amino acids was identified as an activation domain and was also active when fused to Gal4 DNA binding domain. This domain contains three unique regions rich in glutamic acid, glutamine, or serine/threonine and is active in both osteosarcoma (HOS) and T (Jurkat) cell lines. All of these three regions are essential; however, a part of the serine/threonine region was dispensable in Jurkat, but not in HOS cells. In addition to the activation domain, the N-terminal region showed activity in conjunction with the b-Zip structure, but not with the Gal4 DNA binding domain. Furthermore, this region showed activity in Jurkat cells, but not in HOS cells. These results suggest that TREB5 has two activational functions in transcription and may provide diversity in cell-type-specific transcriptional activation, possibly through dimerization with other b-Zip proteins and phosphorylation.