YY
Yasuhiko Yamamoto
Author with expertise in Biochemistry of Diabetic Complications
Achievements
Cited Author
Open Access Advocate
Key Stats
Upvotes received:
0
Publications:
9
(78% Open Access)
Cited by:
2,598
h-index:
53
/
i10-index:
143
Reputation
Biology
< 1%
Chemistry
< 1%
Economics
< 1%
Show more
How is this calculated?
Publications
0

Novel splice variants of the receptor for advanced glycation end-products expressed in human vascular endothelial cells and pericytes, and their putative roles in diabetes-induced vascular injury

Hideto Yonekura et al.Mar 12, 2003
The binding of advanced glycation end-products (AGE) to the receptor for AGE (RAGE) is known to deteriorate various cell functions and is implicated in the pathogenesis of diabetic vascular complications. In the present study, we show that the cellular constituents of small vessels, endothelial cells (EC) and pericytes express novel splice variants of RAGE mRNA coding for the isoforms that lack the N-terminal V-type immunoglobulin-like domain (N-truncated) or the C-terminal transmembrane domain (C-truncated), as well as the known full-length mRNA. The ratio of the expression of the three variants was different between EC and pericytes; the content of the C-truncated form was highest in EC, whereas the full-length form was the most abundant in pericytes. Transfection experiments with COS-7 cells demonstrated that those variant mRNAs were translated into proteins as deduced; C-truncated RAGE was efficiently secreted into the culture media, and N-truncated RAGE was located mainly on the plasma membrane. The three isoforms were also detected in primary cultured human EC and pericytes. Further, full-length and C-truncated forms of RAGE bound to an AGE-conjugated column, whereas N-truncated RAGE did not. The AGE induction of extracellular-signal-related kinase phosphorylation and vascular endothelial growth factor in EC and of the growth and cord-like structure formation of EC was abolished completely by C-truncated RAGE, indicating that this endogenous secretory receptor (endogenous secretory RAGE) is cytoprotective against AGE. The results may contribute to our understanding of the molecular basis for the diversity of cellular responses to AGE and for individual variations in the susceptibility to diabetic vascular complications.
0
Citation693
0
Save
0

The Receptor for Advanced Glycation End Products Is Induced by the Glycation Products Themselves and Tumor Necrosis Factor-α through Nuclear Factor-κB, and by 17β-Estradiol through Sp-1 in Human Vascular Endothelial Cells

Nobushige Tanaka et al.Aug 1, 2000
The binding of advanced glycation end products (AGE) to the receptor for AGE (RAGE) is known to deteriorate various cell functions and is implicated in the pathogenesis of diabetic vascular complications. Here we show that AGE, tumor necrosis factor-α (TNF-α), and 17β-estradiol (E2) up-regulated RAGE mRNA and protein levels in human microvascular endothelial cells and ECV304 cells, with the mRNA stability being essentially invariant. Transient transfection experiments with human RAGE promoter-luciferase chimeras revealed that the region from nucleotide number −751 to −629 and the region from −239 to −89 in the RAGE 5′-flanking sequence exhibited the AGE/TNF-α and E2 responsiveness, respectively. Site-directed mutation of an nuclear factor-κB (NF-κB) site at −671 or of Sp-1 sites at −189 and −172 residing in those regions resulted in an abrogation of the AGE/TNF-α- or E2-mediated transcriptional activation. Electrophoretic mobility shift assays revealed that ECV304 cell nuclear extracts contained factors which retarded the NF-κB and Sp-1 elements, and that the DNA-protein complexes were supershifted by anti-p65/p50 NF-κB and anti-Sp-1/estrogen receptor α antibodies, respectively. These results suggest that AGE, TNF-α, and E2 can activate the RAGE gene through NF-κB and Sp-1, causing enhanced AGE-RAGE interactions, which would lead to an exacerbation of diabetic microvasculopathy. The binding of advanced glycation end products (AGE) to the receptor for AGE (RAGE) is known to deteriorate various cell functions and is implicated in the pathogenesis of diabetic vascular complications. Here we show that AGE, tumor necrosis factor-α (TNF-α), and 17β-estradiol (E2) up-regulated RAGE mRNA and protein levels in human microvascular endothelial cells and ECV304 cells, with the mRNA stability being essentially invariant. Transient transfection experiments with human RAGE promoter-luciferase chimeras revealed that the region from nucleotide number −751 to −629 and the region from −239 to −89 in the RAGE 5′-flanking sequence exhibited the AGE/TNF-α and E2 responsiveness, respectively. Site-directed mutation of an nuclear factor-κB (NF-κB) site at −671 or of Sp-1 sites at −189 and −172 residing in those regions resulted in an abrogation of the AGE/TNF-α- or E2-mediated transcriptional activation. Electrophoretic mobility shift assays revealed that ECV304 cell nuclear extracts contained factors which retarded the NF-κB and Sp-1 elements, and that the DNA-protein complexes were supershifted by anti-p65/p50 NF-κB and anti-Sp-1/estrogen receptor α antibodies, respectively. These results suggest that AGE, TNF-α, and E2 can activate the RAGE gene through NF-κB and Sp-1, causing enhanced AGE-RAGE interactions, which would lead to an exacerbation of diabetic microvasculopathy. advanced glycation end products receptor for AGE vascular endothelial growth factor nuclear factor-κB tumor necrosis factor-α 17β-estradiol bovine serum albumin transforming growth factor-β1 interferon-γ human skin microvascular endothelial cells fetal bovine serum phosphate-buffered saline reverse transcription polymerase chain reaction polyacrylamide gel electrophoresis estrogen receptor analysis of variance Glucose and other reducing sugars can react nonenzymatically with the amino groups of proteins to form reversible Schiff bases and, then, Amadori compounds. These early glycation products undergo further complex reactions to become irreversibly cross-linked, heterogeneous fluorescent derivatives termed advanced glycation end products (AGE)1 (1Brownlee M. Cerami A. Vlassara H. N. Engl. J. Med. 1988; 318: 1315-1321Crossref PubMed Scopus (2424) Google Scholar). AGE are known to accumulate in various tissues at an extremely accelerated rate under a diabetic state, and are implicated in the development of diabetic vascular complications, e.g. retinopathy and nephropathy (1Brownlee M. Cerami A. Vlassara H. N. Engl. J. Med. 1988; 318: 1315-1321Crossref PubMed Scopus (2424) Google Scholar). We have shown previously that AGE exert their effects on endothelial cells and pericytes, the constituents of microvessels, through interactions with a cell-surface receptor for AGE (RAGE); AGE stimulate the growth of microvascular endothelial cells through an induction of vascular endothelial growth factor (VEGF) leading to angiogenesis on the one hand (2Yamagishi S. Yonekura H. Yamamoto Y. Katsuno K. Sato F. Mita I. Ooka H. Satozawa N. Kawakami T. Nomura M. Yamamoto H. J. Biol. Chem. 1997; 272: 8723-8730Abstract Full Text Full Text PDF PubMed Scopus (300) Google Scholar), and inhibit prostacyclin production and stimulated plasminogen activator inhibitor-1 synthesis by the endothelial cells on the other (3Yamagishi S. Fujimori H. Yonekura H. Yamamoto Y. Yamamoto H. Diabetologia. 1998; 41: 1435-1441Crossref PubMed Scopus (212) Google Scholar). AGE exhibit a growth inhibitory action on pericytes (4Yamagishi S. Hsu C.-C. Taniguchi M. Harada S. Yamamoto Y. Ohsawa K. Kobayashi K. Yamamoto H. Biochem. Biophys. Res. Commun. 1995; 213: 681-687Crossref PubMed Scopus (142) Google Scholar), which would lead to pericyte loss, the earliest histological hallmark in diabetic retinopathy (5Cogan D.G. Toussaint D. Kuwabara T. Arch. Ophthalmol. 1961; 66: 366-378Crossref PubMed Scopus (531) Google Scholar). RAGE belongs to the immunoglobulin superfamily of cell surface molecules (6Schmidt A.M. Vianna M. Gerlach M. Brett J. Ryan J. Kao J. Esposito C. Hegarty H. Hurley W. Clauss M. Wang F. Pan Y.-C.E. Tsang T.C. Stern D. J. Biol. Chem. 1992; 267: 14987-14997Abstract Full Text PDF PubMed Google Scholar, 7Neeper M. Schmidt A.M. Brett J. Yan S.D. Wang F. Pan Y.-C.E. Elliston K. Stern D. Shaw A. J. Biol. Chem. 1992; 267: 14998-15004Abstract Full Text PDF PubMed Google Scholar). It is expressed in multiple tissues (8Brett J. Schmidt A.M. Yan S.D. Zou Y.S. Weidman E. Pinsky D. Nowygrod R. Neeper M. Przysiecki C. Shaw A. Migheli A. Stern D. Am. J. Pathol. 1993; 143: 1699-1712PubMed Google Scholar) and interacts with various ligands including AGE (9Schmidt A.M. Hori O. Brett J. Yan S.D. Wautier J.L. Stern D. Arterioscler. Thromb. 1994; 14: 1521-1528Crossref PubMed Google Scholar, 10Schmidt A.M. Yan S.D. Wautier J.L. Stern D. Circ. Res. 1999; 84: 489-497Crossref PubMed Scopus (720) Google Scholar). The engagement of RAGE by AGE has been reported to induce cellular oxidant stress, activating the transcription factor nuclear factor-κB (NF-κB) (11Yan S.D. Schmidt A.M. Anderson G.M. Zhang J. Brett J. Zou Y.S. Pinsky D. Stern D. J. Biol. Chem. 1994; 269: 9889-9897Abstract Full Text PDF PubMed Google Scholar, 12Lander H.M. Tauras J.M. Ogiste J.O. Hori O. Moss R.A. Scmidt A.M. J. Biol. Chem. 1997; 272: 17810-17814Abstract Full Text Full Text PDF PubMed Scopus (692) Google Scholar), resulting in the perturbation of a variety of vascular homeostatic functions (9Schmidt A.M. Hori O. Brett J. Yan S.D. Wautier J.L. Stern D. Arterioscler. Thromb. 1994; 14: 1521-1528Crossref PubMed Google Scholar,10Schmidt A.M. Yan S.D. Wautier J.L. Stern D. Circ. Res. 1999; 84: 489-497Crossref PubMed Scopus (720) Google Scholar). AGE-RAGE interaction thus has been thought to play a central role in the development of diabetic vasculopathy. To determine how the RAGE gene is regulated under a diabetic state is, therefore, important for clarifying the pathogenesis of diabetic complications as well as for understanding the physiological roles of RAGE. It has been reported that AGE-rich blood vessels show enhanced RAGE immunoreactivity (13Soulis T. Thallas V. Youssef S. Gilbert R.E. McWilliam B.G. Murray-McIntosh R.P. Cooper M.E. Diabetologia. 1997; 40: 619-628Crossref PubMed Scopus (202) Google Scholar); this implies the possibility that AGE themselves may up-regulate the RAGE expression. Among cytokines, tumor necrosis factor-α (TNF-α) is thought to be involved in the development of diabetes (14Moller D.E. Flier J.S. N. Engl. J. Med. 1991; 325: 938-948Crossref PubMed Scopus (802) Google Scholar). Evidence has accumulated that serum TNF-α levels are increased in non-insulin-dependent diabetes mellitus (15Hotamisligil G.S. Shargill N.S. Spiegelman B.M. Science. 1993; 259: 87-91Crossref PubMed Scopus (6341) Google Scholar, 16Hotamisligil G.S. Spiegelman B.M. Diabetes. 1994; 43: 1271-1278Crossref PubMed Google Scholar, 17Hotamisligil G.S. Arner P. Caro J.F. Atkinson R.L. Spiegelman B.M. J. Clin. Invest. 1995; 95: 2409-2415Crossref PubMed Scopus (3048) Google Scholar) and that TNF-α can activate the NF-κB pathway (18Beg A.A. Baltimore D. Science. 1996; 274: 782-784Crossref PubMed Scopus (2949) Google Scholar, 19Wang C.Y. Mayo M.W. Baldwin A.S.J. Science. 1996; 274: 784-787Crossref PubMed Scopus (2522) Google Scholar, 20Antwerp D.J.V. Martin S.J. Kafri T. Green D.R. Verma I.M. Science. 1996; 274: 787-789Crossref PubMed Scopus (2463) Google Scholar). Recently, NF-κB has been reported to play a role in the basal and lipopolysaccharide-induced expression of the RAGE gene (21Li J.F. Schmidt A.M. J. Biol. Chem. 1997; 272: 16498-16506Abstract Full Text Full Text PDF PubMed Scopus (457) Google Scholar). It has also been reported that diabetic vasculopathy is often aggravated during pregnancy, probably due to the increased level of serum estrogen (22Moloney J.B.M. Drury M.I. Am. J. Ophthalmol. 1982; 93: 745-756Abstract Full Text PDF PubMed Scopus (141) Google Scholar, 23Klein B.E.K. Moss S.E. Klein R. Diabetes Care. 1990; 13: 34-40Crossref PubMed Scopus (280) Google Scholar, 24Axer-Siegel R. Hod M. Fink-Cohen S. Kramer M. Weinberger D. Schindel B. Yassur Y. Ophthalmology. 1996; 103: 1815-1819Abstract Full Text PDF PubMed Scopus (108) Google Scholar). From these observations, the possibility that TNF-α and estrogen worsen the diabetic complications through the induction of RAGE gene expression should also be considered. In the present study, we thus examined the effects of AGE, TNF-α, and 17β-estradiol (E2) on RAGE gene expression and found that the three agents are capable of up-regulating the RAGE mRNA and protein levels in human microvascular endothelial cells. On the other hand, non-glycated BSA, other cytokines, and an anti-estrogen did not affect the RAGE mRNA levels. mRNA stability and promoter assays demonstrated that the induction was at the transcriptional level, and that AGE and TNF-α induced the RAGE gene through an activation of NF-κB while E2 induced the gene through Sp-1. Bovine serum albumin (BSA) (fraction V, fatty acid-free, endotoxin-free) was purchased from Roche Molecular Biochemicals (Mannheim, Germany). TNF-α was purchased from Becton Dickinson Labware (Bedford, MA). Transforming growth factor-β1 (TGF-β1) and interferon-γ (IFN-γ) were from R&D Systems, Inc. (Minneapolis, MN). E2 and 4-hydroxytamoxifen (4-OH tamoxifen) were from Sigma. [γ-32P]ATP (6,000 Ci/mmol) was from NEN Life Science Products. Restriction enzymes and T4 polynucleotide kinase were from Takara (Kyoto, Japan). Rabbit polyclonal antibodies raised against recombinant human RAGE extracellular domain (amino acids 24–321) (7Neeper M. Schmidt A.M. Brett J. Yan S.D. Wang F. Pan Y.-C.E. Elliston K. Stern D. Shaw A. J. Biol. Chem. 1992; 267: 14998-15004Abstract Full Text PDF PubMed Google Scholar) were kindly provided by the Institute of Biological Science, Mitsui Pharmaceutical Inc (Mobara, Japan). Human skin microvascular endothelial cells (HMVEC) (Cascade Biologics, Inc., Portland, OR) were maintained in HuMedia-EB2 supplemented with 5% fetal bovine serum (FBS), gentamycin (50 μg/ml), amphotericin B (50 ng/ml), basic fibroblast growth factor (5 ng/ml), heparin (10 μg/ml), epidermal growth factor (10 ng/ml), and hydrocortisone (1 μg/ml) according to the supplier's instructions (Kurabo Corp., Osaka, Japan) in humidified incubators containing 5% CO2. The human umbilical vein endothelial cell-derived cell line ECV304 (25Takahashi K. Sawasaki Y. Hata J. Mukai K. Goto T. In Vitro Cell. Dev. Biol. 1990; 25: 265-274Crossref Scopus (421) Google Scholar) was kindly donated by Dr. Yoshio Sawasaki (National Defense Medical College, Tokorozawa, Japan), and was maintained in M199 (Nissui Pharmaceutical Co., Ltd., Tokyo, Japan) supplemented with 10% FBS, penicillin (100 IU/ml), and streptomycin (100 μg/ml). All experiments using HMVEC and ECV304 cells were carried out in media lacking epidermal growth factor and hydrocortisone, and FBS, respectively. These media are referred to as the experimental media. AGE-BSA was prepared by incubating BSA with 0.5 m glucose at 37 °C for 6 weeks under sterile conditions as described previously (2Yamagishi S. Yonekura H. Yamamoto Y. Katsuno K. Sato F. Mita I. Ooka H. Satozawa N. Kawakami T. Nomura M. Yamamoto H. J. Biol. Chem. 1997; 272: 8723-8730Abstract Full Text Full Text PDF PubMed Scopus (300) Google Scholar). After unincorporated sugars were removed by dialysis against phosphate-buffered saline (PBS), glucose-modified higher molecular weight materials were used as AGE-BSA. Control non-glycated BSA was incubated under the same conditions except for the absence of glucose. The concentration of AGE-BSA and control BSA were determined by the method of Bradford (26Bradford M.M. Anal. Biochem. 1972; 71: 248-254Google Scholar). Poly(A)+ RNAs were isolated from subconfluent cultures of HMVEC or ECV304 cells incubated under various conditions, using a Quick Prep Micro mRNA purification kit (Amersham Pharmacia Biotech, Buckinghamshire, United Kingdom), and analyzed by RT-PCR with a GeneAmp RNA PCR kit (Perkin-Elmer, Foster City, CA) as described previously (27Hatakeyama H. Miyamori I. Fujita T. Takeda Y. Takeda R. Yamamoto H. J. Biol. Chem. 1994; 269: 24316-24320Abstract Full Text PDF PubMed Google Scholar). Oligodeoxyribonucleotide primers and probes for human RAGE, VEGF, and β-actin mRNA were the same as described previously (4Yamagishi S. Hsu C.-C. Taniguchi M. Harada S. Yamamoto Y. Ohsawa K. Kobayashi K. Yamamoto H. Biochem. Biophys. Res. Commun. 1995; 213: 681-687Crossref PubMed Scopus (142) Google Scholar, 28Nomura M. Yamagishi S. Harada S. Hayashi Y. Yamashima T. Yamashita J. Yamamoto H. J. Biol. Chem. 1995; 270: 28316-28324Abstract Full Text Full Text PDF PubMed Scopus (305) Google Scholar). The amounts of poly(A)+ RNA templates and cycle numbers for amplification were chosen in quantitative ranges where reactions proceeded linearly as described previously (28Nomura M. Yamagishi S. Harada S. Hayashi Y. Yamashima T. Yamashita J. Yamamoto H. J. Biol. Chem. 1995; 270: 28316-28324Abstract Full Text Full Text PDF PubMed Scopus (305) Google Scholar, 29Yonekura H. Sakurai S. Liu X. Migita H. Wang H. Yamagishi S. Nomura M. Abedin Md.J. Unoki H. Yamamoto Y. Yamamoto H. J. Biol. Chem. 1999; 274: 35172-35178Abstract Full Text Full Text PDF PubMed Scopus (173) Google Scholar); 30 ng of templates and 30 cycles were chosen for amplifying human RAGE mRNA, 30 ng and 40 cycles for human VEGF mRNA, and 30 ng and 20 cycles for β-actin mRNA. The fragments amplified with PCR were sequence-verified on both strands by the chain termination method (30Sanger F. Nicklen S. Coulson A.R. Proc. Natl. Acad. Sci. U. S. A. 1977; 74: 5463-5467Crossref PubMed Scopus (58971) Google Scholar). Five-μl aliquots of each RT-PCR reaction mixture were electrophoresed on a 2% agarose gel and transferred to a Hybond-N+ nylon membrane (Amersham Pharmacia Biotech), and the membrane was hybridized with the respective probes, which had been 32P-end-labeled with polynucleotide kinase (27Hatakeyama H. Miyamori I. Fujita T. Takeda Y. Takeda R. Yamamoto H. J. Biol. Chem. 1994; 269: 24316-24320Abstract Full Text PDF PubMed Google Scholar). Signal intensities of hybridized bands were measured by a BAS 1000 BioImage analyzer (Fuji Photo Film Co. Ltd., Hamamatsu, Japan). HMVEC or ECV304 cells were treated with TNF-α, E2, or AGE-BSA for 4 h, and further cultured in the presence of 10 μg/ml of actinomycin D (Sigma) for various time periods. Total RNAs were isolated from the cultures with Isogen (Nippon Gene, Toyama, Japan) according to the method described by Chomczynski and Sacchi (31Chomczynski P. Sacchi N. Anal. Biochem. 1987; 162: 156-159Crossref PubMed Scopus (66740) Google Scholar), and analyzed by the quantitative RT-PCR method described above. The amounts of total RNA templates and cycle numbers for amplification were chosen in quantitative ranges; 300 ng of templates and 35 cycles were chosen for amplifying human RAGE mRNA, and 300 ng and 20 cycles for β-actin mRNA. Subconfluent cultures of HMVEC or ECV304 cells were incubated with TNF-α, E2, or AGE-BSA for 24 h. After the incubation, cells were washed with cold PBS, scraped off in cold PBS, and pelleted by centrifugation at 300 ×g at 4 °C for 5 min. The cells were lysed immediately by sonication in SDS-polyacrylamide gel electrophoresis (PAGE) sample buffer (62.5 mm Tris-HCl (pH 6.8), 2% SDS, 5% 2-mercaptoethanol, 10% glycerol, and 0.002% bromphenol blue) and boiled at 95 °C for 5 min. Twelve μg of the cell lysates were resolved by SDS-PAGE (12.5%), and then transferred onto a nitrocellulose membrane (Amersham Pharmacia Biotech). The membrane was treated with the anti-RAGE polyclonal antibodies, and the immunoreactive bands were visualized with an ECL detection system (Amarsham Pharmacia Biotech) as described previously (32Kaji M. Yonemura Y. Harada S. Liu X. Terada I. Yamamoto H. Cancer Gene Ther. 1996; 3: 393-404PubMed Google Scholar). Signal intensities of the resultant bands were determined by densitometry using BIO-PROFIL 1-D (version 5.08) software (Vilber Lourmat Biotechnology, Marne La Vallée, France). The chimeric genes for transfection experiments were constructed by ligating the 5′-flanking regions of differing lengths of the human RAGE gene upstream of the luciferase gene in a pGL3-basic vector (Promega Corp., Madison, WI) (see Fig. 5 A). The genomic DNA fragments of the human RAGE gene were amplified by PCR using a cosmid named KS71 (33Sugaya K. Fukagawa T. Matsumoto K. Mita K. Takahashi E. Ando A. Inoko H. Ikemura T. Genomics. 1994; 23: 408-419Crossref PubMed Scopus (211) Google Scholar) as a template, which was kindly provided by Professor Toshimichi Ikemura (National Institute of Genetics, Shizuoka, Japan). The PCR primers employed in the amplification reactions are shown in Table I. Furthermore, we constructed two additional chimeric genes. The DNA fragments containing exons 1–11 plus introns 1–10 or the 3′-flanking region of the human RAGE gene were amplified by PCR using the cosmid KS71 and specific primers shown in TableI, and were ligated to downstream of the luciferase gene in the pGL3-Basic vector that had carried the longest fragment of the 5′-flanking region (see Fig. 5 A). All the fragments obtained were sequence-verified.Table IPrimers used for the construction of RAGE promoter-luciferase fusion genesNames of the constructsNucleotide sequencesCorresponding nucleotide positions in the RAGE gene (33)Forward primers pGL-15′-GAACGCGT AGAGATGCCAAAAATGGGGA−1689 to −1669 pGL-25′-GAACGCGT CTTACCTCATACGCAGCTCA−1599 to −1579 pGL-35′-GAACGCGT TACCTCGGAGGGAGTTTCTG−1528 to −1508 pGL-45′-GAACGCGT GGGAATTATTGACTGGGCTT−1500 to −1480 pGL-55′-GAACGCGT ACTGTCCCATCCTCAGCCCT−751 to −731 pGL-65′-GAACGCGT TTGGAAGTGTGATGGGTGGG−629 to −609 pGL-75′-GAACGCGT GAGACCCTAGAGGGTACACT−239 to −219 pGL-85′-GAACGCGT CGACTGAAAGATGGGGGCTG−89 to −69Reverse primer pGL-1∼85′-GAAGATCT CAGGCTCCAACTGCTGTTCC+43 to +23Forward primers pGL-el-115′-GAGTCGAC TCCTCAGTCTGTGGGGTGAG+52 to +72 pGL-3′fl5′-GAGTCGAC CCCCACAATGATGATTAAAC+3253 to +3273Reverse primers pGL-el-115′-GAGTCGAC AGAGCAAGATGTGTCAGGTG+3272 to +3252 pGL-3′fl5′-GAGTCGAC AAGAGGTGGAACAAGAGAAG+6318 to +6298 Open table in a new tab Mutations were introduced into the RAGE promoter-luciferase chimeras using a GeneEditor™ in vitro site-directed mutagenesis system (Promega) according to the manufacturer's protocol. The nucleotide sequences of the mutagenic oligodeoxyribonucleotide primers were 5′-ACTGTCAGAGTGGGGTTCCCCTCCCATTAAAG-3′ (nucleotides −683 to −652), 5′-GTGACTGTACCCAGAAACTGGTAGTACCCAGG-3′ (−202 to −171), 5′-CTGGTAGTACCCAGGAATGGGGTGATAATTAT-3′ (−185 to −154), and 5′-ACTGTACCCAGAAACTGGTAGTACCCAGGAATGGGGTGATAAT-3′ (−199 to −157) for yielding pGL-5 NF-κB2m, pGL-7 Im, pGL-7 IIm, or pGL-7 ImIIm, respectively (mutated sites are indicated by underlines) (see schematic representations in Fig. 6 A). pGL-5 NF-κB2m contained the region −751 to +43, but with the mutation in an NF-κB site (−671 to −663). pGL-7 Im and pGL-7 IIm contained the region −239 to +43, but with the mutations in one of two Sp-1 sites (−189 to −181 and −172 to −166), respectively. pGL-7 ImIIm had the mutations in both Sp-1 sites. All the mutated constructs were sequence-verified. ECV304 cells or HMVEC (2 × 105 cells) were plated into the wells of six-well tissue culture plates (Becton Dickinson Labware) 1 day before transfection. For ECV304 cells, the DNA/cationic lipid mixture for transfection was composed of test plasmid (2 μg), pRL-SV40 vector (0.5 μg) (Promega) that served as an internal control to normalize luciferase activities, TransFastTM transfection reagent (7.5 μl) (Promega), and 1 ml of the experimental medium. Cells were exposed to the DNA/cationic lipid mixture for 2 h, then received 5 ml of the FBS-containing medium and were further incubated for 36 h at 37 °C. For transfection of HMVEC, test plasmid (2 μg) and pRL-CMV vector (1 μg) (Promega) were added to a tube containing 100 μl of Opti-MEM I (Life Technologies, Inc.). Lipofectin reagent (6 μl) (Life Technologies, Inc.) was added to another tube containing 100 μl of Opti-MEM I. The plasmid DNA and lipofectin reagent were then mixed together. After incubation at room temperature for 30 min, DNA-liposome complex was diluted with 800 μl of Opti-MEM I. Cells were exposed to the mixture for 8 h, then received 2 ml of the FBS-containing medium and were further incubated for 36 h at 37 °C. After the incubation, the cells were treated with TNF-α, E2, or AGE-BSA for 8 h in the experimental medium. Luciferase activities were measured using a Dual-Luciferase™ reporter assay system (Promega) according to the manufacturer's protocol with a luminometer (Fluoroskan Ascent FL version 2.2.4, Labsystems, Helsinki, Finland). Nuclear extracts were prepared essentially as described by Schreiber et al. (34Schreiber E. Matthias P. Müller M.M. Schaffner W. Nucleic Acids Res. 1989; 17: 6419Crossref PubMed Scopus (4074) Google Scholar). Briefly, ECV304 cells (2 × 106) were plated onto 75-cm2 tissue culture flasks (Becton Dickinson Labware) in the complete medium and left for 24 h at 37 °C. The cells were further incubated in the experimental medium at 37 °C for 12 h and then treated with TNF-α, E2, or AGE-BSA for 4 h. After the treatment, the cells were washed twice with ice-cold PBS, scraped off in PBS (10 ml), and pelleted by centrifugation at 3,000 rpm in a Beckman GH3.7 roter at 4 °C for 5 min. The pelleted cells were resuspended in 0.4 ml of ice-cold buffer A (10 mm HEPES (pH 7.9), 10 mm KCl, 0.1 mm EDTA, 1 mm dithiothreitol, and 0.5 mm phenylmethylsulfonyl fluoride) and placed on ice for 15 min. After the addition of 25 μl of 10% Nonidet P-40, the suspension was vortexed for 15 s and centrifuged at 13,000 rpm in a Hitachi T15S roter at 4 °C for 30 min. The resultant nuclear pellets were washed with buffer A and resuspended in 0.1 ml of a solution containing 20 mm HEPES (pH 7.9), 400 mm NaCl, 1 mm EDTA, 1 mm dithiothreitol, and 1 mm phenylmethylsulfonyl fluoride by constant agitation for 15 min at 4 °C. The nuclear lysates were centrifugated at 13,000 rpm at 4 °C for 10 min, and the supernatants were collected as nuclear extracts. The protein concentrations of the nuclear extracts were determined by the method of Bradford (26Bradford M.M. Anal. Biochem. 1972; 71: 248-254Google Scholar). The wild type and mutant double-stranded oligodeoxyribonucleotides encompassing the NF-κB site (nucleotide numbers −671 to −663) or the two Sp-1 sites (−189 to −181 and −172 to −166) were prepared. Their sequences were 5′-AGAGTGGGGAACCCCTCCCA-3′ and 5′-AGAGTGGGGTTCCCCTCCCA-3′ (−677 to −658) as the wild and mutated NF-κB elements, respectively, and 5′-CCCAGAGGCTGGTAGTACCCAGGGGTGGGGTGA-3′ and 5′-CCCAGAAACTGGTAGTACCCAGGAATGGGGTGA-3′ (−193 to −161) as the wild and mutated Sp-1 elements, respectively (mutated sites are indicated by underlines). Twenty-five fmol of wild-type oligodeoxyribonucleotides, which had been 32P-end-labeled with polynucleotide kinase (27Hatakeyama H. Miyamori I. Fujita T. Takeda Y. Takeda R. Yamamoto H. J. Biol. Chem. 1994; 269: 24316-24320Abstract Full Text PDF PubMed Google Scholar), were incubated with 5 μg of nuclear extracts at room temperature for 30 min. Samples were then loaded onto 6% polyacrylamide gels and run in 0.2× Tris borate/EDTA electrophoresis buffer at 10 V/cm for 2–3 h. The gels were dried and autoradiographed at −80 °C overnight. For competition assay, nuclear extracts were first incubated with a 50-fold excess of unlabeled wild-type or mutant oligodeoxyribonucleotides at room temperature for 15 min, and then incubated with the labeled wild-type probe for 30 min under the same conditions as described above. For supershift assays, antibodies to NF-κB p65, NF-κB p50, Sp-1, estrogen receptor (ER) α, or ERβ (Santa Cruz Biotechnology Inc., Santa Cruz, CA) were added to the nuclear extracts and incubated at 4 °C for 12 h. The antibody-treated nuclear extracts were subsequently incubated with the labeled oligodeoxyribonucleotides for 30 min under the same conditions as described above. Paired t tests and one-way analysis of variance (ANOVA) with Tukey's range tests were used to test for significant differences between groups. All experiments were carried out at least three times. To examine the effects of TNF-α, E2, and AGE-BSA on the RAGE mRNA level in HMVEC, poly(A)+ RNAs were isolated from cells that had been exposed to various concentrations of these agents for 4 h, and analyzed by the quantitative RT-PCR method. As shown in Fig.1 A, TNF-α, E2, and AGE-BSA increased the RAGE mRNA levels in dose-dependent manners. The extents of induction and the peak concentrations were about 3-fold at 100 ng/ml TNF-α, 10 nm E2, and 50 μg/ml AGE-BSA, respectively, when normalized by β-actin mRNA-derived signals used as an internal control. Next, we examined the time course of the RAGE mRNA induction. HMVEC were treated for various time periods with TNF-α, E2, and AGE-BSA at their most effective doses. As shown in Fig. 1 B, the mRNA levels began to increase at 2 h and reached a maximum at ∼8 h after the addition of either of the three agents. On the other hand, exposure of HMVEC to TGF-β1 (10 ng/ml), IFN-γ (165 ng/ml), and non-glycated BSA (50 μg/ml) for 4 h did not affect the RAGE mRNA levels (Fig. 1 C). 4-OH tamoxifen (10 nm), an anti-estrogen, abolished the E2-induced RAGE mRNA induction (Fig.1 C). We next determined the RAGE mRNA levels in ECV304 cells, an immortalized cell line derived from human umbilical vein endothelial cells (25Takahashi K. Sawasaki Y. Hata J. Mukai K. Goto T. In Vitro Cell. Dev. Biol. 1990; 25: 265-274Crossref Scopus (421) Google Scholar). This cell line exhibited the same responsiveness to AGE-BSA with respect to VEGF induction (Fig.2 A) as did HMVEC (2Yamagishi S. Yonekura H. Yamamoto Y. Katsuno K. Sato F. Mita I. Ooka H. Satozawa N. Kawakami T. Nomura M. Yamamoto H. J. Biol. Chem. 1997; 272: 8723-8730Abstract Full Text Full Text PDF PubMed Scopus (300) Google Scholar), the most effective dosage of AGE-BSA being 50 μg/ml in both cultures. The extent of induction was about 4-fold. When exposed to TNF-α (100 ng/ml), E2 (10 nm), or AGE-BSA (50 μg/ml) for 4 h, the RAGE mRNA levels in the ECV304 cells were also increased about 2-fold compared with those in the control unexposed cells (Fig. 2 B). We next examined whether the increase in RAGE mRNA was actually followed by an increase in RAGE proteins in HMVEC and ECV304 cells. The cells were treated with TNF-α, E2, or AGE-BSA for 24 h, and subjected to Western blot analysis with anti-RAGE polyclonal antibodies. As shown in Fig. 3, one major immunoreactive band was marked at 46 kDa in either HMVEC (Fig.3 A) or ECV304 cells (Fig. 3 B), being consistent with our previous report (3Yamagishi S. Fujimori H. Yonekura H. Yamamoto Y. Yamamoto H. Diabetologia. 1998; 41: 1435-1441Crossref PubMed Scopus (212) Google Scholar), and its intensity was increased by the treatment with the three agents. The RAGE protein levels in HMVEC and ECV304 cells treated with TNF-α, E2, or AGE-BSA were about 2-fold higher than the basal levels. The results indicated that ECV304 cells retained the ability to respond to those agents as did primary cultured endothelial cells, and subsequently we mainly used this cell line to examine the mechanisms of RAGE gene induction. We next determined the RAGE mRNA stability in HMVEC and ECV304 cells exposed or not exposed to the three agents to determine which step of gene expression accounted for the increase in RAGE mRNA levels. The cells were incubated in the presence or absence of TNF-α, E2, or AGE-BSA for 4 h, then incubated with actinomycin D for various time periods, and underwent quantitative RT-PCR analyses. As shown in Fig.4 (A and B), the half-lives of RAGE mRNA in TNF-α-, E2-, and AGE-BSA-treated or untreated HMVEC and ECV304 cells were calculated from the RAGE and β-actin mRNA-derived signals to be between 2.1 and 2.8 h, and there was no statistically significant difference among them. The results suggested that the TNF-α-, E2-, and AGE-BSA-induced increase in RAGE mRNA was achieved at the transcriptional level. To confirm whether TNF-α, E2, and AGE-BSA did induce the RAGE gene transcription and, if so, to delimit the regions involved in such transcriptional activations, a series of chimeric 5′-deletion promoter-luciferase reporter constructs were prepared.
0

CCR5 Plays a Critical Role in Obesity-Induced Adipose Tissue Inflammation and Insulin Resistance by Regulating Both Macrophage Recruitment and M1/M2 Status

Hironori Kitade et al.Apr 4, 2012
C-C motif chemokine receptor (CCR)2 and its ligand, monocyte chemoattractant protein (MCP)-1, are pivotal for adipose tissue macrophage (ATM) recruitment and the development of insulin resistance. However, other chemokine systems also may play a role in these processes. In this study, we investigated the role of CCR5 in obesity-induced adipose tissue inflammation and insulin resistance. We analyzed expression levels of CCR5 and its ligands in white adipose tissue (WAT) of genetically (ob/ob) and high-fat (HF) diet-induced obese (DIO) mice. Furthermore, we examined the metabolic phenotype of Ccr5(-/-) mice. CCR5 and its ligands were markedly upregulated in WAT of DIO and ob/ob mice. Fluorescence-activated cell sorter analysis also revealed that DIO mice had a robust increase in CCR5(+) cells within ATMs compared with chow-fed mice. Furthermore, Ccr5(-/-) mice were protected from insulin resistance, glucose intolerance, and hepatic steatosis induced by HF feeding. The effects of loss of CCR5 were related to both reduction of total ATM content and an M2-dominant shift in ATM polarization. It is noteworthy that transplantation of Ccr5(-/-) bone marrow was sufficient to protect against impaired glucose tolerance. CCR5 plays a critical role in ATM recruitment and polarization and subsequent development of insulin resistance.
0

Tumor-associated macrophages of the M2 phenotype contribute to progression in gastric cancer with peritoneal dissemination

Takahisa Yamaguchi et al.Nov 30, 2015
Tumor-associated macrophages (TAMs) of the M2 phenotype are known to promote tumor proliferation and to be associated with a poor prognosis in numerous cancers. Here, we investigated whether M2 macrophages participate in the development of peritoneal dissemination in gastric cancer.The characteristics of peritoneal macrophages in gastric cancer patients with or without peritoneal dissemination were examined by flow cytometry and the real-time quantitative polymerase chain reaction. The effects of M2 macrophages on phenotypic changes of the gastric cancer cell line MKN45 were assessed with a direct or indirect co-culture system in vitro and an in vivo mouse xenograft model.The number of peritoneal macrophages with the M2 phenotype (CD68(+)CD163(+) or CD68(+)CD204(+)) was significantly higher in gastric cancer patients with peritoneal dissemination than in those without peritoneal dissemination. Higher expression of the M2-related messenger RNAs (IL-10, vascular endothelial growth factor A, vascular endothelial growth factor C, matrix metalloproteinase 1, and amphiregulin) and lower expression of M1-related messenger RNAs (TNF-α, CD80, CD86, and IL-12p40) were also confirmed in the TAMs. Macrophage co-culture with gastric cancer cells converted M1 phenotype into M2 phenotype. Moreover, the coexistence of MKN45 cells with M2 macrophages resulted in cancer cell proliferation and an acceleration of tumor growth in the xenograft model.Intraperitoneal TAMs in gastric cancer patients with peritoneal dissemination were polarized to the M2 phenotype, and could contribute to tumor proliferation and progression. Therefore, intraperitoneal TAMs are expected to be a promising target in the treatment of peritoneal dissemination in gastric cancer.
0

Anti–high mobility group box‐1 antibody therapy for traumatic brain injury

Yu Okuma et al.Apr 4, 2012
Abstract Objective: High mobility group box‐1 (HMGB1) plays an important role in triggering inflammatory responses in many types of diseases. In this study, we examined the involvement of HMGB1 in traumatic brain injury (TBI) and evaluated the ability of intravenously administered neutralizing anti‐HMGB1 monoclonal antibody (mAb) to attenuate brain injury. Methods: Traumatic brain injury was induced in rats or mice by fluid percussion. Anti‐HMGB1 mAb or control mAb was administered intravenously after TBI. Results: Anti‐HMGB1 mAb remarkably inhibited fluid percussion‐induced brain edema in rats, as detected by T2‐weighted magnetic resonance imaging; this was associated with inhibition of HMGB1 translocation, protection of blood–brain barrier (BBB) integrity, suppression of inflammatory molecule expression, and improvement of motor function. In contrast, intravenous injection of recombinant HMGB1 dose‐dependently produced the opposite effects. Experiments using receptor for advanced glycation end product (RAGE) −/− , toll‐like receptor‐4 (TLR4) −/− , and TLR2 −/− mice suggested the involvement of RAGE as the predominant receptor for HMGB1. Interpretation: Anti‐HMGB1 mAb may provide a novel and effective therapy for TBI by protecting against BBB disruption and reducing the inflammatory responses induced by HMGB1. ANN NEUROL 2012;72:373–384
0
Citation220
0
Save
1

Astroglial CD38 regulates social memory and synapse formation through SPARCL1 in the medial prefrontal cortex

Tsuyoshi Hattori et al.Dec 24, 2021
Abstract Social behavior is essential for the health, survival and reproduction of animals, yet the role of astrocytes in social behavior is largely unknown. CD38 is critical for social behaviors by regulating oxytocin release from hypothalamic neurons. On the other hand, CD38 is most abundantly expressed in astrocytes especially in the postnatal cortex, and is important for astroglial development. Here, we demonstrate that astroglial CD38 plays a pivotal role in the social behavior. Selective deletion of CD38 in postnatal astrocytes, but not in adult astrocytes, specifically impaired social memory without any other behavioral abnormalities. Morphological analysis revealed reductions in spine numbers, mature spines and excitatory synapse numbers in the pyramidal neurons of the medial prefrontal cortex (mPFC) due to deletion of astroglial CD38 in the postnatal brain. Astrocyte-conditioned medium (ACM) of CD38 KO astrocytes reduced synaptogenesis of cortical neurons by reducing extracellular SPARCL1, a synaptogenic protein. Finally, the release of SPARCL1 from astrocytes is regulated by CD38/cADPR/calcium signaling. Our data indicate that astroglial CD38 developmentally regulates social memory and neural circuit formation in the developing brain by promoting synaptogenesis through SPARCL1.
0

Nuclear Expression and DNA Binding Capacity of Receptor for Advanced Glycation End Products in Renal Tissue

Brooke Harcourt et al.May 9, 2019
The AGER gene encodes for a number of RAGE isoforms, with the membrane bound signal transduction and decoy circulating soluble RAGE being the best characterised. Here we demonstrate a novel nuclear isoform of RAGE in mice and human kidney cortex which by cell and size fractionation we determined to be approximately 37kda. This nuclear RAGE isoform is functional and binds to DNA sequences within the upstream 5' promoter region of its own gene, AGER. This binding was shown to be abrogated by mutating the DNA consensus binding sequences during electromobility shift assay (EMSA) and was independent of NFκ-B or AP-1 binding. Cotransfection of expression constructs encoding various RAGE isoforms along with AGER gene promoter reporter-plasmids identified that the most likely source of the nuclear isoform of RAGE was a cleavage product of the nt-RAGE isoform. In obese mice with impaired kidney function, there was increased binding of nuclear RAGE within the A. Region of ager gene promoter with corresponding increases in membrane bound RAGE in renal cells. These findings were reproduced in vitro using proximal tubule cells. Hence, we postulate that RAGE expression is in part, self-regulated by the binding of a nuclear RAGE isoform to the promoter of the AGER gene (encoding RAGE) in the kidney. We also suggest that this RAGE self-regulation is altered under pathological conditions and this may have implications for chronic kidney disease.