High-frequency spin torque oscillation in orthogonal magnetization disks with strong biquadratic magnetic coupling.

In this study, we numerically investigate the spin transfer torque oscillation (STO) in a magnetic orthogonal configuration by introducing a strong biquadratic magnetic coupling. The orthogonal configuration consists of top and bottom layers with in-plane and perpendicular magnetic anisotropy sandwiching a nonmagnetic spacer. The advantage of an orthogonal configuration is the high efficiency of spin transfer torque leading a high STO frequency; however, maintaining the STO in a wide range of electric current is challenging. By introducing biquadratic magnetic coupling into the orthogonal structure of FePt/spacer/Co90Fe10, Ni80Fe20 or Ni, we were able to expand the electric current region in which the stable STO is realized, resulting in a relatively high STO frequency. For example, approximately 50 GHz can be achieved in an Ni layer at a current density of 5.5 × 107 A/cm2. In addition, we investigated two types of initial magnetic state: out-of-plane and in-plane magnetic saturation; this leads to a vortex and an in-plane magnetic domain structure after relaxation, respectively. The transient time before the stable STO was reduced to between 0.5 and 1.8 ns by changing the initial state from out-of-plane to in-plane.

Biochemical characteristics and inhibitor selectivity of mouse indoleamine 2,3-dioxygenase-2.

The first step in the kynurenine pathway of tryptophan catabolism is the cleavage of the 2,3-double bond of the indole ring of tryptophan. In mammals, this reaction is performed independently by indoleamine 2,3-dioxygenase-1 (IDO1), tryptophan 2,3-dioxygenase (TDO) and the recently discovered indoleamine 2,3-dioxygenase-2 (IDO2). Here we describe characteristics of a purified recombinant mouse IDO2 enzyme, including its pH stability, thermal stability and structural features. An improved assay system for future studies of recombinant/isolated IDO2 has been developed using cytochrome b (5) as an electron donor. This, the first description of the interaction between IDO2 and cytochrome b (5), provides further evidence of the presence of a physiological electron carrier necessary for activity of enzymes in the "IDO family". Using this assay, the kinetic activity and substrate range of IDO2 were shown to be different to those of IDO1. 1-Methyl-D-tryptophan, a current lead IDO inhibitor used in clinical trials, was a poor inhibitor of both IDO1 and IDO2 activity. This suggests that its immunosuppressive effect may be independent of pharmacological inhibition of IDO enzymes, in the mouse at least. The different biochemical characteristics of the mouse IDO proteins suggest that they have evolved to have distinct biological roles.

Brainstem in Machado-Joseph disease: atrophy or small size?

Machado-Joseph disease (MJD), one of the most common types of hereditary spinocerebellar degeneration caused by abnormal expansion of the CAG repeat in the MJD1 gene, presents atrophy of the infratentorial structures neuropathologically and neuroradiologically. Although a significant positive correlation has been reported between infratentorial atrophy and the number of expanded CAG repeat units, the exact changing course of brainstem size in the individual case remains to be resolved. We investigated seven cases of genetically confirmed MJD longitudinally by magnetic resonance imaging with observation periods of 4.5-10.6 years. Measurement of the midsagittal areas of infratentorial structures disclosed progressive atrophy of the pontine base and cerebellum, which correlated significantly with age, whilst midbrain and pontine tegmentum showed atrophy with no significant progression, suggesting it was better identified as 'small size' and might have mostly been completed before the initial symptoms. Such differences between regions in atrophy progression must be caused by a difference in the neuropathological course.

Lipopolysaccharide-induced decreased protein S expression in liver cells is mediated by MEK/ERK signaling and NFkappaB activation: involvement of membrane-bound CD14 and toll-like receptor-4.

BACKGROUND: The vitamin K-dependent protein S (PS), mainly synthesized in hepatocytes and endothelial cells, plays a critical role in the anticoagulant activity of plasma. The decreased plasma level of PS in sepsis is associated with thrombotic tendency, but the mechanism is unclear. OBJECTIVES: In the present study, we examined the effect of lipopolysaccharide (LPS) on PS expression in vivo in rat liver, and in vitro in isolated hepatocytes and sinusoidal endothelial cells (SECs) from normal rats. RESULTS: LPS induced a progressive decrease of plasma PS antigen level up to 12 h with a slight recovery at 24 h, and a transient decrease of liver PS mRNA level at 4-8 h with a complete recovery at 24 h. In the in vitro studies, LPS decreased PS antigen and mRNA levels in both hepatocytes and SECs. After LPS treatment, tumor necrosis factor-alpha (TNF-alpha), interleukin-6 (IL-6) and interferon-gamma (IFN-gamma) transiently increased in plasma. IL-6 increased the protein expression of PS from hepatocytes, while TNF-alpha decreased it from SECs. LPS increased CD14 in hepatocytes and decreased it in SECs, but did not affect toll-like receptor-4 (TLR-4) expression in both cells. Antirat CD14 and antirat TLR-4 antibodies inhibited LPS-induced NFkappaB activation, and a NFkappaB inhibitor suppressed LPS-induced decreased PS expression in both cells. Furthermore, MEK inhibitor blocked LPS-induced decreased PS expression in both cells. CONCLUSIONS: These findings suggest that LPS-induced decreased PS expression in hepatocytes and SECs is mediated by MEK/ERK signaling and NFkappaB activation and that membrane-bound CD14 and TLR-4 are involved in this mechanism. These findings may explain in part the decreased level of plasma PS and thrombotic tendency in sepsis.

Convergent evolution. The gene structure of Sulculus 41 kDa myoglobin is homologous with that of human indoleamine dioxygenase.

The abalone Sulculus diversicolor contains abundant myoglobin in its buccal mass. The myoglobin consists of 377 amino acid residues and has a molecular mass of 41 000 Da, 2.5 times larger than that of other myoglobins. Sulculus myoglobin can bind oxygen reversibly, and the P50 was determined to be 3.8 mmHg at 20 degrees C and pH 7.4, showing that the oxygen affinity of Sulculus myoglobin is lower than those of vertebrate and invertebrate myoglobins. The cDNA-derived amino acid sequence showed no significant homology with those of any other invertebrate myoglobins and hemoglobins, but surprisingly showed 35% homology with a vertebrate tryptophan-degrading enzyme, indoleamine dioxygenase (IDO). The structure of the Sulculus myoglobin gene has been determined to consist of 14 exons and 13 introns (15.3 kbp). Compared with the gene of human IDO (10 exon-9 intron structure), the splice junctions of 7 introns were exactly conserved between the two genes, suggesting that these introns have been conserved for at least 600 million years. The Sulculus gene has 5 additional introns, one of which is located outside the coding region. From these results we conclude that Sulculus myoglobin evolved from an IDO gene and represents a typical case of functional convergence. Comparison of the amino acid sequence of each exon of Sulculus myoglobin with those of usual globin sequences showed that there is no significant evolutionary relationship between them. The IDO-like myoglobin is unexpectedly widely distributed among gastropodic molluscs, such as Sulculus, Nordotis, Battilus, Omphalius and Chlorostoma.

Comparative assessment of the resistance of the unstirred water layer to solute transport between two different intestinal perfusion systems.

The resistance of the unstirred water layer to solute transport was estimated in two different intestinal single-pass perfusion systems for a comparative study, using D-glucose as a model compound. One is a well established perfusion system in anesthetized rats as a standard (system A). The other is the one in unanesthetized rats for comparison (system B). It was demonstrated that in system B as well as in system A the resistance of the unstirred water layer to D-glucose transport should be taken into account and this resistance, accordingly, the effective thickness of the unstirred water layer (delta) which is assumed to be in proportion to its resistance, could be described as a function of the perfusion rate by using a film model. The delta decreased with increasing perfusion rate and was larger in system A than in system B at each perfusion rate; 785 microns in system A versus 319 microns in system B at the perfusion rate of 0.16 ml/min and 337 microns versus 184 micron at that of 2.95 ml/min. Thus in system B the effective thickness, accordingly, the resistance, of the unstirred water layer was reduced to about 50% of that in system A, but the resistance of the unstirred water layer could still account for 85% of the total resistance at the maximum as far as D-glucose absorption was concerned, while 93% in system A. These results suggest that, compared with perfusion experiments in anesthetized rats (system A), the resistance of the unstirred water layer is reduced but cannot be left out of consideration even if perfusion experiments are performed in unanesthetized rats (system B). And the lower resistance of the unstirred water layer in system B was attributed to a turbulent flow in contrary to a laminar flow in system A.

Determination of kinetic parameters of a carrier-mediated transport in the perfused intestine by two-dimensional laminar flow model: effects of the unstirred water layer.

The kinetic parameters of a carrier-mediated transport for D-glucose and for taurocholate were determined from rat in situ intestinal single perfusion experiments. The true parameters were obtained by the two-dimensional laminar flow model, in which the solute concentration at the aqueous-intestinal membrane interface can be calculated numerically without assuming the aqueous diffusion layer, discriminating the effects of the unstirred water layer. The true Michaelis constant was 4.5 mM for D-glucose and 1.5 mM for taurocholate. The true maximal transport velocity was 3.4 nmol/s per cm2 for D-glucose and 0.29 nmol/s per cm2 for taurocholate. The apparent Michaelis constant was raised by the factor of 6.6 for D-glucose and 3.6 for taurocholate due to the effects of the unstirred water layer. The maximal transport velocity was relatively unaffected by the unstirred water layer in both compounds. The values of the effective (operational) thickness of the unstirred water layer were compatible with those reported previously by employing various experimental methods. The kinetic parameters obtained in vitro everted sacs, for comparison, almost coincided with the true ones in situ. Therefore, the two-dimensional laminar flow model is shown to be valid not only for determining the kinetic parameters of a carrier-mediated transport in situ but also for predicting the absorption rate in situ from the uptake rate in vitro.

Determination of the membrane permeability coefficient and the reflection coefficient by the two-dimensional laminar flow model for intestinal perfusion experiments.

We performed single perfusion experiments in the small intestine of rats in order to prove that the two-dimensional laminar flow model is suitable to determine the membrane permeability coefficient and the reflection coefficient. We used progesterone as an aqueous-diffusion-limited drug, urea as a membrane transport-limited drug and the tritiated water as an intermediate substance. The membrane permeability coefficient for progesterone was calculated to be 3.6 X 10(-4) cm/s. This value did not change when the thickness of the aqueous diffusion layer was altered by increasing the perfusion rate 10-fold. It was directly demonstrated that the two-dimensional laminar flow model was suitable to analyze the data of intestinal perfusion experiments. Membrane permeability coefficients for urea and tritiated water were determined to be 3.4 X 10(-5) cm/s and 8.9 X 10(-5) cm/s, respectively. In the presence of water absorption with the hypotonic perfusion solution, the reflection coefficient for urea was 0.84. This value is thought to be theoretically reasonable, suggesting the usefullness of the two-dimensional laminar flow model to obtain the reflection coefficient in the intestinal membrane.

The giant extracellular hemoglobin from the polychaete Neanthes diversicolor. The cDNA-derived amino acid sequence of linker chain L2 and the exon/intron boundary conserved in linker genes.

The 4000 kDa extracellular hemoglobin from the polychaete Neanthes diversicolor consists of three types of subunits; three 15 kDa monomers (chains M1, M2 and M3), a 45 kDa disulfide-bonded trimer of chains T1, T2 and T3, and two 50-55 kDa disulfide-bonded homodimeric linkers (chains L1 and L2). The latter linker subunits are essential for the assembly of the other heme-containing subunits, monomers and a trimer. The cDNA encoding the linker chain L2 was amplified by polymerase chain reaction (PCR), and the cDNA-derived amino acid sequence of 235 residues has been determined. The sequence showed 22-75% identity with other linker chains. All of the linker sequences examined so far have a highly conserved cysteine-rich segment at positions 89-130: Xaa3-Cys-Xaa6-Cys-Xaa6-Cys-Xaa6-Cys-Asp-Gly-X aa2-Asp-Cys-Xaa4-Asp-Glu-Xaa4-Cys, and the motif corresponds exactly to the cysteine-rich repeats of the ligand-binding domains of vertebrate low-density lipoprotein (LDL) receptors (Suzuki, T. and Riggs, A.F. (1993) J. Biol. Chem. 268, 13548-13555). A 287 bp intron interrupts the coding sequence of Neanthes L2 gene just at the N-terminal boundary of this motif, and the position of the splice junction was exactly conserved in Neanthes and Lumbricus linker genes. This suggests that the intron has been conserved for at least 450 million years in annelid linker genes. The evolutionary origin of the remaining parts of linker chains is unclear, but it is noteworthy that the topology of the two intrachain disulfide bridges in the C-terminal segment of linker chains is homologous with that of the carbohydrate-recognition domain of animal C-type lectin.

Primary structure of Tetrahymena hemoglobins.

The primary structure of hemoglobins of Tetrahymena thermophila and T. pyriformis was determined by peptide and cDNA sequence analysis. Their sequences were 83.5% identical to each other and homologous to other protozoan and cyanobacteria hemoglobins, but not to proteins of the globin family. An intron was not present in the T. thermophila hemoglobin gene.

Electrospray ionization mass spectrometric composition of the 400 kDa hemoglobin from the pogonophoran Oligobrachia mashikoi and the primary structures of three major globin chains.

Maximum entropy analysis of the electrospray ionization mass spectra of the native, carbamidomethylated, reduced and reduced and carbamidomethylated forms of the extracellular ca. 400 kDa hemoglobin of the pogonophoran Oligobrachia mashikoi has shown it to consist of eight globin chains: (a1-a5), 14861.1, 14937.1, 15040.7, 15070.6 and 15310.6 Da and b-dl, 15173.2, 15605.1 and 14775.4 Da, respectively. Although chains a1-a5 are monomeric, chains b + c form a disulfide-bonded dimer of 30776.8 Da and chains b + c + d1 form a disulfide-bonded trimer of 45551.9 Da. The major chains a5, b and c were separated by reverse-phase chromatography, and their cDNA's amplified by PCR using redundant oligomers based on their N-terminal amino-acid sequences. The complete amino-acid sequences of chains a5 (142 residues), b (140 residues) and c (147 residues) were derived from protein and cDNA sequencing and represent the first pogonophoran globin sequences. They have a high percent identity (35-52%) with the globin chains of the approximately 3500 kDa hexagonal bilayer hemoglobins from the annelids Lumbricus and Tylorrhynchus and the vestimentiferan Lamellibrachia, suggesting a very close relationship among the phyla Annelida, Pogonophora and Vestimentifera. Two free cysteine residues (Cys-73 and Cys-83), which we proposed to be the most probable candidates for the sulfide-binding sites in the Lamellibrachia chains (Suzuki, T., Takagi, T. and Ohta, S. (1990) Biochem. J. 266, 221-225), are also conserved in three chains (Cys-73 for chains b and c, and Cys-83 for chain a5) of Oligobrachia hemoglobin, in agreement with the probable role of the hemoglobin in the binding and transport of sulfide to the symbiotic bacteria which provide the metabolic fuel in the two phyla.

Surgical treatment of primary lung cancer in patients less than 40 years of age.

PURPOSE AND METHODS: The major purpose of this study was to determine whether the survival rate in young lung cancer patients after surgical treatment differs from that in older patients. An analysis was performed for all patients with bronchogenic carcinoma who underwent surgery at Mie University Hospital from 1965 to 1990. RESULTS: Of 803 patients, 24 (2.99%) were 33 to 39 years old. At the time of surgery, the disease was diagnosed as stage I in seven patients (29%), stage II in four (17%), stage IIIa in seven (29%), stage IIIb in two (8%), and stage IV in four (17%), while 46.3% of the patients older than 40 years of age had either stage IIIa, IIIb, or IV disease. All of the 24 patients less than 40 years of age underwent thoracotomy: curative resection in 14 cases, palliative resection in sex, and probe-thoracotomy in four. The 5-year survival rate for all stages of disease was 31.4% in these 24 patients, and 41.9% in 603 patients greater than 40 years of age. The 5-year survival rate for stage I disease was 35.7% in the seven younger patients and 78.0% in the 207 older patients; for stage II, it was 25.5% in the four younger patients and 40.6% in the 98 older patients; for stage III, it was 33.3% in the nine younger patients and 15.6% in the 250 older patients; and for stage IV, it was 25% in the four younger patients and 6.6% in the 48 older patients. There were no significant differences in survival rate between the two age groups for all patients or for those with each stage of disease. CONCLUSION: Although younger patients tended to have more advanced disease, long-term survival in these patients did not differ from that of older patients.

Glu-47, which forms a salt bridge between neurophysin-II and arginine vasopressin, is deleted in patients with familial central diabetes insipidus.

The arginine vasopressin (AVP) gene was sequenced in a pedigree with familial central diabetes insipidus (DI). When polymerase chain reaction-amplified DNAs from affected subjects were subjected to polyacrylamide gel electrophoresis, fragments including exon 2 displayed two additional, slower migrating bands. These extra bands represented DNA heteroduplexes, indicating that there was a deletion or insertion mutation in exon 2. As the region with such a mutation was identified by direct sequence analysis, polymerase chain reaction-amplified fragments including the region were subcloned and sequenced. A 3-basepair deletion (AGG) out of two consecutive AGG sequences (nucleotides 1824-1829) was identified in one of two alleles. The cosegregation of the mutation with the DI phenotype in the family was confirmed by restriction enzyme analyses. This mutation should yield an abnormal AVP precursor lacking Glu47 in its neurophysin-II (NP) moiety. Since Glu47 is essential for NP molecules to form a salt bridge with AVP, it is very likely that the function of NP as a carrier protein for AVP would be impaired. We suggest that AVP would undergo accelerated proteolytic degradation, and this mechanism would be involved in the pathogenesis of DI in this pedigree.

Novel mutations in the V2 vasopressin receptor gene in two pedigrees with congenital nephrogenic diabetes insipidus.

Novel mutations in the V2 vasopressin receptor gene were identified in two Japanese pedigrees with X-linked congenital nephrogenic diabetes insipidus. The V2 receptor belongs to the family of G-protein-coupled receptors that contain seven distinct transmembrane domains, and the V2 receptor gene is encoded by three exons. The coding regions amplified by polymerase chain reaction were directly sequenced. In a pedigree, one of four consecutive guanine sequences (nucleotides 528-531) in the second exon was deleted (528delG). This deletion mutation results in a frame shift beginning at codon 154 in the second intracellular domain and a premature termination at codon 161. In another pedigree, a missense mutation (A-->G) was identified at nucleotide position 310 in the second exon. This point mutation, H80R, changes a histidine at codon 80 in the second transmembrane domain to an arginine that is more positively charged than histidine under the neutral environment. Each mutation cosegregated with the phenotype of diabetes insipidus and supposed to be a cause for resistance to arginine vasopressin.

Two novel mutations in the coding region for neurophysin-II associated with familial central diabetes insipidus.

Familial central diabetes insipidus is an autosomal dominant disease caused by a deficiency of arginine vasopressin (AVP). We previously reported three distinct mutations in the AVP gene in Japanese familial central diabetes insipidus pedigrees that result in a substitution of Ser for Gly57 in the neurophysin-II (NPII) moiety of the AVP precursor, a substitution of Thr for Ala at the COOH-terminus of the signal peptide, and a deletion of Glu47 in the NPII moiety. In this study, we analyzed the AVP gene in two pedigrees by direct sequencing of the polymerase chain reaction-amplified DNA and found two novel mutations in exon 2, which encodes the central part of the NPII moiety of the precursor. The mutation in one pedigree was a C to A transition at nucleotide position 1891, which replaces Cys67 (TGC) with stop codon (TGA). As the premature termination eliminates part of the COOH domain of the NPII moiety and the glycoprotein moiety, the conformation of the truncated protein is likely to be markedly different from that of normal precursor. In another pedigree, a G to T transversion was detected at nucleotide position 1874, which substitutes polar Trp (TGG) for hydrophobic Gly62 (GGG). It is possible that mutated NPII molecules, as a consequence of a conformational change, cannot bind AVP or self-associate to form higher oligomer complexes. Interestingly, all mutations we have identified to date, with the exception of the signal peptide mutation, are located in exon 2, suggesting the importance of the highly conserved central part of the NPII molecules and/or the NPII moiety in the precursor for AVP synthesis.

Genomic structure of the sandworm, Perinereis vancaurica tetradentata, troponin C.

Troponin C (TnC) superfamily genes essentially possess five introns, the positions of all but the fourth being highly conserved. The fourth intron is frequently absent from protostomian invertebrate genes, such as calmodulin or TnC. We previously proposed that the common ancestor of TnC superfamily genes never possessed an intron corresponding to today's fourth introns, and that members of the superfamily independently gained a fourth intron in the evolutionary pathway of each lineage. In the present study, we isolated the TnC cDNA from the sandworm, Perinereis vancaurica tetradentata and determined its genomic structure. Sandworm TnC appears to exist as a single copy gene consisting of six exons and five introns. The positions of the first, second, third and fifth introns are identical to other TnCs, but that of the fourth intron is unique. This is in good agreement with the above-mentioned scheme, i.e. the gain of the fourth intron of sandworm TnC might have occurred within the annelid lineage after annelida/mollusca divergence.

The genomic structure of the scallop, Patinopecten yessoensis, troponin C gene: a hypothesis for the evolution of troponin C.

Two cDNAs encoding troponin C (TnC) isoforms are isolated from the scallop, Patinopecten yessoensis, striated adductor muscle. The sequential differences between these isoforms, named TnC(long) and TnC(short), are restricted in several residues of the C-terminal region. TnC(long) is commonly expressed in both the striated and the smooth adductor muscle; however, TnC(short) is only in the striated adductor muscle. The TnC gene is a single copy gene in the scallop, thus they are expressed through the alternative splicing from the same gene. The scallop TnC gene is constructed from five exons and four introns, and positions of introns are identical with chordate TnC genes, although the scallop TnC possesses no corresponding intron to the fourth intron of chordates. The loss of this intron is also observed in Drosophila TnC; these may be remnants of their ancestor, namely the early metazoan TnC gene might be a five exons-four introns structure. In addition, the absence of the corresponding intron is also observed among protostomian calmodulins (CaMs), a molecule closely related to TnC. This suggests that the common ancestor gene of the TnC superfamily might also be a five exons-four introns structure. Assuming this to be true, the discordance of the fourth intron positions observed among members of the family is well explained by the evolutionary independent gain of the intron on each member's lineage.

Structural organization of lower marine nonvertebrate calmodulin genes.

The troponin C (TnC) superfamily genes generally possess five introns, and the positions where they are inserted are well conserved except for the fourth intron. Based on a structural comparison of TnC genes, we proposed that the common ancestor of TnC or TnC superfamily genes had no intron corresponding to the modern fourth intron, and therefore members of the superfamily have gained the fourth intron independently within each lineage. Here, we cloned calmodulin (CaM, one of the members of the TnC superfamily) cDNAs from two lower marine nonvertebrates, the sea anemone, Metridium senile, belonging to the Cnidaria, and the sponge, Halichondria okadai, belonging to the Porifera, and also determined their genomic organization. Chordate CaM genes generally possess five introns, but neither sea anemone nor sponge CaM has anything corresponding to the fourth intron of chordate CaMs, suggesting that the early metazoan CaM must have had only four introns. The modern fourth intron of chordate CaMs was acquired within the chordate lineage after nonvertebrate/chordate divergence. This notion concurs with our proposal explaining the evolution of the TnC superfamily genes.

The structural organization of the ascidian, Halocynthia roretzi, calmodulin genes. The vicissitude of introns during the evolution of calmodulin genes.

Two distinct calmodulin (CaM) genes are isolated from the ascidian, Halocynthia roretzi, (Hr-CaM A and Hr-CaM B) and those structures are determined. There are three nucleotide substitutions, producing two amino acid differences between Hr-CaM A and Hr-CaM B, and those are corresponding to two of the known eight variable residues among metazoan CaMs. Both Hr-CaM A and Hr-CaM B are constructed from six exons and five introns, and the positions of introns are identical. The positions of introns of Hr-CaMs are also identical with those of vertebrate CaMs, except third introns. The third introns of Hr-CaMs are inserted at 28bp upstream when compared with vertebrate CaMs. Thus, sliding of the third intron might have occurred in only the ascidian lineage prior to the gene duplication that also occurred only in that lineage. In addition, with the comparison of the intron positions, we attempt to investigate the vicissitude of introns during the evolution of metazoan CaMs.

Organization and chromosomal localization of the human endothelial protein C receptor gene.

Endothelial protein C receptor (EPCR), present on endothelial cells of relatively large veins and arteries, plays a role in the enhancement of protein C activation by the thrombin-thrombomodulin complex. In the present study, we determined the organization and the complete nucleotide sequence of the human EPCR gene using polymerase chain reaction-direct sequencing method. The transcription initiation site of the EPCR gene was also determined by the cap site hunting method, using a cap site cDNA prepared from human placenta. The human EPCR gene spanned approx. 6 kb and was composed of four exons and three introns. All exon-intron boundaries agreed with the GT-AG rule. The 5'-flanking region (300 bp) of the EPCR gene contained a putative AP1-binding site, two Sp1-binding sites and two AP2-binding sites, but not definite TATAA or CCAAT sequences. Fluorescence in situ hybridization analysis showed that the EPCR gene is located in chromosome 20q11.2.