Liposomes bearing polyethyleneglycol-coupled transferrin with intracellular targeting property to the solid tumors in vivo.

PURPOSE: The purpose of this study was to determine the usefulness of transferrin (TF)-pendant-type polyethyleneglycol (PEG)-liposomes (TF-PEG-liposomes), in which TF was covalently linked to the distal terminal of PEG chains on the external surface of PEG-liposomes as a carrier for in vivo cytoplasmic targeting to tumor cells. METHODS: Small unilamellar TF-PEG-liposomes (100-140 nm in diameter) were prepared from DSPC, CH, DSPE-PEG, and DSPE-PEG-COOH (2:1:0.11:0.021, molar ratio), and were conjugated to TF via the carboxyl residue of DSPE-PEG-COOH. The intracellular targeting ability of TF-PEG-liposomes to tumor cells was examined in vitro and in Colon 26 tumor-bearing mice. RESULTS: TF-PEG-liposomes, bearing approximately 25 TF molecules per liposome, readily bound to mouse Colon 26 cells in vitro and were internalized by receptor-mediated endocytosis. TF-PEG-liposomes showed a prolonged residence time in the circulation and low RES uptake in Colon 26 tumor-bearing mice, resulting in enhanced extravasation of the liposomes into the solid tumor tissue. Electron microscopic studies in Colon 26 tumor-bearing mice revealed that the extravasated TF-PEG-liposomes were internalized into tumor cells by receptor-mediated endocytosis. CONCLUSION: TF-PEG-liposomes had the capabilities of specific receptor binding and receptor-mediated endocytosis to target cells after extravasation into solid tumors in vivo. Such liposomes should be useful for in vivo cytoplasmic targeting of chemotherapeutic agents or plasmid DNAs to target cells.

Three novel alternatively spliced isoforms of the human beta-site amyloid precursor protein cleaving enzyme (BACE) and their effect on amyloid beta-peptide production.

Three novel alternatively spliced transcripts of the beta-site amyloid precursor protein cleaving enzyme (BACE) were cloned from human brain. Alternative splicing of the RNA occurs at an internal donor in exon 3 and/or an internal acceptor in exon 4. The splicing events lead to a deletion of 25 (BACE-I-476), 44 (BACE-I-457) and 69 (BACE-I-432) amino acids and the latter two caused the loss of two of four N-linked glycosylation sites. Although the mature form of BACE-501 was resistant to endoglycosidase H treatment, glycosylated forms of BACE-I-457 and BACE-I-476 were sensitive. This result suggests that BACE-I-457 and BACE-I-476 underwent different post-translational modifications. Moreover, the beta-secretase activity of BACE-I-457 and BACE-I-476 was significantly weaker than that of BACE-501. Thus, these isoforms may contribute to a physiological function of BACE.

Alzheimer's disease-associated presenilin 2 interacts with DRAL, an LIM-domain protein.

Using the yeast two-hybrid system, we screened for proteins interacting with presenilin 2 (PS2) and cloned DRAL. DRAL is an LIM-only protein containing four LIM domains and an N-terminal half LIM domain. Previously DRAL has been cloned as a co-activator of the androgen receptor and as a protein interacting with a DNA replication regulatory protein, hCDC47. Our yeast two-hybrid assay showed that DRAL interacted with a hydrophilic loop region (amino acids 269-298) in the endoproteolytic N-terminal fragment of PS2, but not that of PS1, although the region 269-298 of PS2 and the corresponding PS1 sequence differ by only three amino acids. Each point mutation within this region, R275A, T280A, Q282A, R284A, N285A, P287T, I288L, F289A and S296A, in PS2 abolished the binding. This suggests that DRAL recognizes the PS2 structure specifically. The in vitro interaction was confirmed by affinity column assay and the physiological interactions between endogenous PS2 and DRAL by co-immunoprecipitation from human lung fibroblast MRC5 cells. Furthermore, in PS2-overexpressing HEK293 cells, we found an increase in the amount of DRAL in the membrane fraction and an increase in the amount of DRAL that was co-immunoprecipitated with PS2. The potential role of DRAL in the cellular signaling suggests that DRAL functions as an adaptor protein that links PS2 to an intracellular signaling.

Time to recover from atrial hormonal, mechanical, and electrical dysfunction after successful electrical cardioversion of persistent atrial fibrillation.

Although transient atrial dysfunction has been reported after electrical cardioversion of atrial fibrillation (AF), the difference in the time to recover from the atrial hormonal, mechanical, and electrical dysfunction has not been described. Thus, we evaluated the time course of recovery from atrial hormonal, mechanical, and electrical dysfunction after cardioversion in patients with nonvalvular AF. We attempted electrical cardioversion in 87 consecutive patients with nonvalvular AF that had persisted for > or =6 months, and in 24 patients (28%) with maintained sinus rhythm for > or =6 months. To evaluate atrial hormonal, mechanical, and electrical dysfunction in these 24 patients, we measured plasma concentration of atrial natriuretic peptide, the atrial peak velocity in transmitral flow, and the ratio of peak systolic-to-diastolic pulmonary venous flow (S/D ratio) using echocardiography, and the duration and the root mean voltage for the terminal 20 ms (LP20) of the filtered P wave using P-wave signal-averaged electrocardiography. Atrial natriuretic peptide rapidly returned to baseline within 1 day after cardioversion, and maintained these levels for 6 months. Atrial peak velocity in transmitral flow and S/D ratio were significantly increased at 2 weeks, and continued to increase until 1 month, and then reached a plateau. The duration and LP20 began to recover only 6 months after cardioversion. One to 3 years after conversion, the duration and LP20 had nearly reached a plateau, but the latter value remained below normal. In patients with nonvalvular AF of prolonged duration, recovery from atrial electrical dysfunction after sinus conversion took much longer than that from either atrial hormonal or mechanical dysfunction.

Serial analysis of gene expression in a microglial cell line.

We used the serial analysis of gene expression (SAGE) method to systematically analyze transcripts present in a microglial cell line. Over 10,000 SAGE tags were sequenced, and shown to represent 6,013 unique transcripts. Among the diverse transcripts that had not been previously detected in microglia were those for cytokines such as endothelial monocyte-activating polypeptide I (EMAP I), and for cell surface antigens, including adhesion molecules such as CD9, CD53, CD107a, CD147, CD162 and mast cell high affinity IgE receptor. In addition, we detected transcripts that were characteristic of hematopoietic cells or mesodermal structures, such as E3 protein, A1, EN-7, B94, and ufo. Furthermore, the profile contained a transcript, Hn1, that is important in hematopoietic cells and neurological development (Tang et al. Mamm Genome 8:695-696, 1997), suggesting the probable neural differentiation of microglia from the hematopoietic system in development. Messenger RNA expression of these genes was confirmed by RT-PCR in primary cultures of microglia. Significantly, this is the first systematic profiling of the genes expressed in a microglial cell line. The identification and further characterization of the genes described here should provide potential new targets for the study of microglial biology.

Genomic organization of the human X11L2 gene (APBA3), a third member of the X11 protein family interacting with Alzheimer's beta-amyloid precursor protein.

Recently we cloned a new member of the X11 protein family, X11L2 (gene symbol APBA3) which interacts with Alzheimer's beta-amyloid precursor protein (APP) and has three protein-protein interaction domains, a phosphotyrosine interacting domain (PID) and two PDZ. Here, we report the genomic structure and mapping of the APBA3. The gene spans about 11 kb and is composed of ten coding exons and one untranslated exon. The transcription start site of APBA3 was found in a CpG island. About 1.2 kb of the 5'-flanking region was also sequenced, and its functional promoter activity was confirmed by transient transfection experiments. The APBA3 was localized by the radiation hybrid mapping to chromosome 19. Determination of the APBA3 genome structure will facilitate the linkage analysis and search for mutations in the APBA3 in patients with Alzheimer's disease.

Genome structure and chromosomal mapping of the gene for Fe65L2 interacting with Alzheimer's beta-amyloid precursor protein.

Recently we cloned the cDNA encoding human Fe65L2, which interacts with Alzheimer's beta-amyloid precursor protein (APP). The protein has one WW domain and two PID elements (Neurosci. Lett. (1999) 261, 143-146.). Here, we report the entire genome structure and the chromosomal mapping of the Fe65L2 gene. The gene is composed of thirteen coding exons distributed over 6 kb and the genomic organization is similar to another Fe65 member, Fe65 gene. Two transcription start sites of the gene were found in a CpG island by primer extension analysis. Radiation hybrid mapping revealed that the Fe65L2 gene is on chromosome 5 between markers SHGC-9824 (D5S2374) and SHGC-8489 (D5S2569). Characterization of the Fe65L2 gene structure will be useful in the linkage analysis and search for mutations in the Fe65L2 gene in patients with Alzheimer's disease.

Transgenic mice with Alzheimer presenilin 1 mutations show accelerated neurodegeneration without amyloid plaque formation.

Familial Alzheimer disease mutations of presenilin 1 (PS-1) enhance the generation of A beta1-42, indicating that PS-1 is involved in amyloidogenesis. However, PS-1 transgenic mice have failed to show amyloid plaques in their brains. Because PS-1 mutations facilitate apoptotic neuronal death in vitro, we did careful quantitative studies in PS-1 transgenic mice and found that neurodegeneration was significantly accelerated in mice older than 13 months (aged mice) with familial Alzheimer disease mutant PS-1, without amyloid plaque formation. However, there were significantly more neurons containing intracellularly deposited A beta42 in aged mutant transgenic mice. Our data indicate that the pathogenic role of the PS-1 mutation is upstream of the amyloid cascade.

Isolation of human delta-catenin and its binding specificity with presenilin 1.

We screened proteins for interaction with presenilin (PS) 1, and cloned the full-length cDNA of human delta-catenin, which encoded 1225 amino acids. Yeast two-hybrid assay, GST binding assay and immunoprecipitation demonstrated that delta-catenin interacted with a hydrophilic loop region in the endoproteolytic C-terminal fragment of PS1, but not with that of PS-2. These results suggest that PS1 and PS2 partly differ in function. PS1 loop fragment containing the pathogenic mutation retained the binding ability. We also found another armadillo-protein, p0071, interacted with PS1.

X11L2, a new member of the X11 protein family, interacts with Alzheimer's beta-amyloid precursor protein.

We screened proteins for interaction with Alzheimer's beta-amyloid precursor protein (APP) and cloned a new member of the X11 protein family, X11L2. The PID/PTB element of X11L2 protein interacted with the intracellular domain of APP by GST binding assay, and in vivo interaction was confirmed by coimmunoprecipitation from cell extracts overexpressing APP and HA-tagged X11L2. This gene encoded 575 amino acids and the deduced amino acid sequence was highly homologous to rat Mint3. Three protein-protein interaction domains, a PID/PTB and two PDZ elements, were conserved among the X11 protein family, and the N-terminal region of X11L2 protein had several putative SH3 binding motifs, PXXP. Unlike other members of the X11 protein family, X11L2 mRNA was expressed in various tissues.

Dobutamine stress echocardiography at 7.5 mg/kg/min using color tissue Doppler imaging M-mode safely predicts reversible dysfunction early after reperfusion in patients with acute myocardial infarction.

Dobutamine stress echocardiography (DSE) is widely used to predict reversible left ventricular dysfunction, but evaluation by this method is subjective. The recently developed color tissue Doppler imaging (TDI) M-mode may permit objective and quantitative assessment of changes in wall motion induced by DSE. We tested the hypothesis that this new method can detect sensitively reversible dysfunction in the post-myocardial infarction setting. DSE with color TDI M-mode and conventional DSE were performed to predict reversible dysfunction in 53 patients at a mean of 3 days after infarction using 7.5 and 10 microg/kg/min of dobutamine. Follow-up regular echocardiography (4 weeks later) was used as the reference technique to define reversible dysfunction segments. To predict reversible dysfunction segments, the standard segmental wall motion score change on conventional DSE and the ratio of the segmental wall velocity difference at rest versus stress (7.5 and 10 microg/kg/ min) on DSE with color TDI M-mode (7.5-TDI-M and 10-TDI-M, respectively) were used. With 7.5 microg/kg/min of dobutamine, the sensitivity for predicting reversible dysfunction using color TDI M-mode (7.5-TDI-M) was significantly higher than that of conventional DSE (89% vs 73%, p <0.05) whereas specificities and predictive values were almost identical. With a 10-microg/kg/min dose, color TDI-M mode (10-TDI-M) and conventional DSE were not significantly different in predicting reversible dysfunction. With use of color TDI-M mode, regional wall motion during DSE was analyzed objectively and quantitatively. Moreover, combined TDI-M and conventional data were slightly superior to either mode alone. There were no arrhythmias during 7.5 microg/kg/min of dobutamine, but 9 arrhythmias occurred during the 10-microg/kg/min dose in patients with acute myocardial infarction. In conclusion, color TDI M-mode permits objective and quantitative assessment of regional ventricular wall motion and gives additional information for detecting reversible dysfunction in DSE. Improvement of sensitivity at a lower dose of dobutamine with color TDI-M mode may increase the safety of DSE in the post-myocardial infarction setting.

Molecular cloning of human Fe65L2 and its interaction with the Alzheimer's beta-amyloid precursor protein.

We report the cDNA sequence of human Fe65L2. The human Fe65L2 encoded 486 amino acids; the deduced amino acid sequence was shorter by 18 amino acids than the rat protein and had 86% identity to the rat protein Three protein-protein interaction domains, a WW and two PID/PTB elements, were conserved among the Fe65 protein family. Human Fe65L2 mRNA was expressed in various tissues; a transcript of about 2.2 kb was mainly expressed in the brain. A splicing variant lacking two amino acids in the first PID/PTB element was detected. We also confirmed that the carboxyl-terminal region of PID/PTB of the Fe65L2 interacted with the intracellular domain of the Alzheimer's beta-amyloid precursor protein (APP) and APP-like proteins.

Familial Alzheimer's disease genes in Japanese.

More than 40 missense mutations and a splice-site mutation in the presenilin 1 (PS-1) gene, two missense mutations of presenilin 2 (PS-2), and more than three missense mutations of amyloid precursor protein (APP) cosegregate with early onset familial Alzheimer's disease (FAD). In order to determine the incidence of mutations of these genes in Japanese patients, we screened 25 early onset FAD families, one late-onset FAD case, 33 early onset AD cases and five late-onset AD cases for mutations in the coding regions of the genes using SSCP analysis. Four different missense mutations of the PS-1 gene, including a novel mutation, Glu273Ala, were identified in five early onset FAD families and one missense mutation of PS-1 in one isolated AD patient. While no missense mutations of PS-2 were detected, four silent nucleotide substitutions were observed. Our data indicate that PS-1 mutations account for 20.0% of early onset FAD cases in Japan. Since mutations in PS-2 and APP genes were not found in the remaining cases, which could be explained only partially by apolipoprotein E epsilon4, important FAD genes or risk-factor genes remain to be identified.

Both N-terminal and C-terminal fragments of presenilin 1 colocalize with neurofibrillary tangles in neurons and dystrophic neurites of senile plaques in Alzheimer's disease.

Presenilin 1 (PS1) is a causative gene for chromosome 14-linked familial Alzheimer's disease. The gene product is known to be cleaved into N-terminal fragments (PS1-N) and C-terminal fragments (PS1-C). To understand the pathophysiological role of PS1, we conducted immunohistochemical studies using antibodies specific for PS1-N and PS1-C in sporadic Alzheimer's disease (AD). Both antibodies showed punctuate staining exclusively in neurons and their processes in both control and AD brains. PS1-N immunolabeling colocalized with neurofibrillary tangles (NFTs) in 36% of NFT-bearing neurons and with dystrophic neurites in 28% of senile plaques (SPs). PS1-C immunolabeling colocalized with dystrophic neurites in 70% of NFT-bearing SPs and with intraneuronal NFTs in 32% of NFT-bearing neurons. Both antibodies did not detect PHF-tau-positive neuropil threads and Abeta amyloid fibrils. The colocalization was also found in 33-38 % of NFT-bearing neurons in progressive supranuclear palsy. These results indicate that both PS1-N and PS1-C fragments are deposited in part of NFT-bearing neurons and dystrophic neurites in SPs; both are the pathologic hallmarks of AD.

Genetic risk factors in Japanese Alzheimer's disease patients: alpha1-ACT, VLDLR, and ApoE.

We studied the polymorphism of alpha1-antichymotrypsin (ACT), very low density lipoprotein receptor (VLDLR) and apolipoprotein E (ApoE) genes in 200 control subjects and 65 patients with Alzheimer's disease (AD) in Japanese. The subjects consisted of 30 patients with early onset familial Alzheimer's disease (FAD), a patient with late onset FAD, 29 patients with an early onset isolated form of AD, and 5 patients with late onset AD. ApoE genotypes were significantly different between controls and FAD (p < 0.0005) or AD (p < 0.05), and patients carrying at least one ApoE epsilon4 allele were found in 44% of FAD and 34.3% of AD; both were significantly different (p < 0.001) from the controls (12.5%). ACT genotypes and allele frequencies were not different among these groups except for genotypes between ApoE epsilon4 FAD and ApoE epsilon4 controls (p = 0.019). There was a slight but significant increase of the 5 repeat allele of VLDLR in AD (p = 0.014), but the difference was rather diminished in the presence of an ApoE epsilon4 allele. None of combinations of ACT and VLDLR genotypes in the presence or absence of an ApoE epsilon4 allele gave significant difference. Thus, we conclude that among the reported genetic risk factors, ApoE epsilon4 is the only definite risk factor for both FAD and AD, and the VLDLR polymorphism might be associated with AD cases in Japanese.

Cloning of the cDNA encoding rat presenilin-2.

We report here the cDNA sequence of rat homologue of presenilin-2 (PS-2). The rat PS-2 cDNA encoded 448 amino acids, and the deduced amino acid sequence was highly homologous to those of the human (94.9%), mouse (96.4%) and Xenopus (70.8%). A minor splicing variant lacking a single glutamate was detected, while the product corresponding to the exon 9 deleted splicing variant observed in human was not detected.

Prethrombotic state due to hypercoagulability in patients with permanent transvenous pacemakers.

Venous thrombosis is a relatively usual but serious complication of permanent transvenous pacing. However, the pathogenesis has not been defined. To clarify underlying abnormalities in the coagulation-fibrinolysis system in patients with permanent transvenous pacemakers, we measured serum levels of fibrinopeptide A (FPA), thrombin-antithrombin III complexes (TATs), plasmin-alpha 2 plasmin inhibitor complexes (PICs), D-dimer (D-D), beta-thromboglobulin (beta-TG), and platelet factor 4 (PF4) in 53 patients with permanent transvenous pacemakers and 10 control subjects. The patients were divided into two groups, as follows, according to the presence of mural thrombus documented along the pacing lead(s) by digital subtraction angiography and transesophageal echocardiography: Group Th (-), patients without venous route thrombus; and Group Th (+), patients with venous route thrombus. FPA and TAT levels increased significantly even in Group Th (-), and further increased in Group Th (+) compared with control subjects (FPA: 7.5 +/- 4.9, 15.3 +/- 8.8 vs 3.0 +/- 1.4 ng/mL, respectively, P < 0.05; TAT: 2.9 +/- 1.3, 4.8 +/- 2.3 vs 1.7 +/- 0.6 ng/mL, respectively, P < 0.05). There were no differences in levels of D-D, PIG, beta-TG, and PF4 among control subjects, Group Th (-), and Group Th (+). These findings suggest that the hypercoagulable state appears in patients with permanent transvenous pacemakers, even without apparent venous thrombosis. The patients with permanent transvenous pacemakers are thought to be in the prethrombotic state even if they have no venous route thrombosis.

A novel method for making nested deletions and its application for sequencing of a 300 kb region of human APP locus.

We developed a novel in vitro method for making nested deletions and applied it to a large-scale DNA sequencing. A DNA fragment to be sequenced (up to 15 kb long) was cloned with a new vector possessing two unique Sfi I sites, digested by Sfi I and ligated to generate a large head-to-tail concatemer. The large concatemer was randomly fragmented by sonication and then redigested by Sfi I to separate insert and vector DNAs. The fragments of various length were then cloned into the other vector(s) specifically designed for selective cloning of insert-derived DNA fragments to generate a library of nested deletions. This method allowed a single person to generate >20 nested deletion libraries sufficient to cover 100 kb in a few days. We applied the method for sequencing of P1 clones and successfully determined the complete sequence of approximately 300 kb of the human amyloid precursor protein (APP) locus on chromosome 21 with a redundancy of 3.8, reasonably low cost and very few gaps remaining to be closed. Development of some new instruments and software is also described which makes this method more applicable for large-scale sequencing.