Vascular smooth muscle RhoA counteracts abdominal aortic aneurysm formation by modulating MAP4K4 activity.

Whether a small GTPase RhoA plays a role in the pathology of abdominal aortic aneurysm (AAA) has not been determined. We show here that RhoA expression is reduced in human AAA lesions, compared with normal areas. Furthermore, incidence of AAA formation is increased in vascular smooth muscle cell (VSMC)-specific RhoA conditional knockout (cKO) mice. The contractility of the aortic rings and VSMCs from RhoA cKO mice is reduced, and expression of genes related to the VSMC contractility is attenuated by loss of RhoA. RhoA depletion activates the mitogen-activated protein (MAP) kinase signaling, including MAP4K4, in the aorta and VSMCs. Inhibition of MAP4K4 activity by DMX-5804 decreases AAA formation. Set, a binding protein to active RhoA, functions as an activator of MAP4K4 by sequestering PP2A, an inhibitor of MAP4K4, in the absence of RhoA. In conclusion, RhoA counteracts AAA formation through inhibition of MAP4K4 in cooperation with Set.

Lewis glycosphingolipids as critical determinants of TRAIL sensitivity in cancer cells.

Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) induces cancer cell death and contributes to tumor rejection by cytotoxic lymphocytes in cancer immunosurveillance and immunotherapy. TRAIL and TRAIL receptor agonists have garnered wide popularity as promising agents for cancer therapy. We previously demonstrated that the loss of fucosylation in cancer cells impairs TRAIL sensitivity; however, the precise structures of the fucosylated glycans that regulate TRAIL sensitivity and their carrier molecules remain elusive. Herein, we observed that Lewis glycans among various fucosylated glycans positively regulate TRAIL-induced cell death. Specifically, Lewis glycans on lacto/neolacto glycosphingolipids, but not glycoproteins including TRAIL receptors, enhanced TRAIL-induced formation of the cytosolic caspase 8 complex, without affecting the formation of the membranous receptor complex. Furthermore, type I Lewis glycan expression in colon cancer cell lines and patient-derived cancer organoids was positively correlated with TRAIL sensitivity. These findings provide novel insights into the regulatory mechanism of TRAIL-induced cell death and facilitate the identification of novel predictive biomarkers for TRAIL-related cancer therapies in future.

Increased levels of acidic free-N-glycans, including multi-antennary and fucosylated structures, in the urine of cancer patients.

We recently reported increased levels of urinary free-glycans in some cancer patients. Here, we focused on cancer related alterations in the levels of high molecular weight free-glycans. The rationale for this study was that branching, elongation, fucosylation and sialylation, which lead to increases in the molecular weight of glycans, are known to be up-regulated in cancer. Urine samples from patients with gastric cancer, pancreatic cancer, cholangiocarcinoma and colorectal cancer and normal controls were analyzed. The extracted free-glycans were fluorescently labeled with 2-aminopyridine and analyzed by multi-step liquid chromatography. Comparison of the glycan profiles revealed increased levels of glycans in some cancer patients. Structural analysis of the glycans was carried out by performing chromatography and mass spectrometry together with enzymatic or chemical treatments. To compare glycan levels between samples with high sensitivity and selectivity, simultaneous measurements by reversed-phase liquid chromatography-selected ion monitoring of mass spectrometry were also performed. As a result, three lactose-core glycans and 78 free-N-glycans (one phosphorylated oligomannose-type, four sialylated hybrid-type and 73 bi-, tri- and tetra-antennary complex-type structures) were identified. Among them, glycans with α1,3-fucosylation ((+/- sialyl) Lewis X), triply α2,6-sialylated tri-antennary structures and/or a (Man3)GlcNAc1-core displayed elevated levels in cancer patients. However, simple α2,3-sialylation and α1,6-core-fucosylation did not appear to contribute to the observed increase in the level of glycans. Interestingly, one tri-antennary free-N-glycan that showed remarkable elevation in some cancer patients contained a unique Glcß1-4GlcNAc-core instead of the common GlcNAc2-core at the reducing end. This study provides further insights into free-glycans as potential tumor markers and their processing pathways in cancer.

Occurrence of a D-arabinose-containing complex-type free-N-glycan in the urine of cancer patients.

Urinary free-glycans are promising markers of disease. In this study, we attempted to identify novel tumor markers by focusing on neutral free-glycans in urine. Free-glycans extracted from the urine of normal subjects and cancer patients with gastric, colorectal, pancreatic and bile duct were fluorescently labeled with 2-aminopyridine. Profiles of these neutral free-glycans constructed using multidimensional high performance liquid chromatography separation were compared between normal controls and cancer patients. The analysis identified one glycan in the urine of cancer patients with a unique structure, which included a pentose residue. To reveal the glycan structure, the linkage fashion, monosaccharide species and enantiomer of the pentose were analyzed by high performance liquid chromatography and mass spectrometry combined with several chemical treatments. The backbone of the glycan was a monoantennary complex-type free-N-glycan containing ß1,4-branch. The pentose residue was attached to the antennal GlcNAc and released by α1,3/4-L-fucosidase. Intriguingly, the pentose residue was consistent with D-arabinose. Collectively, this glycan structure was determined to be Galß1-4(D-Araß1-3)GlcNAcß1-4Manα1-3Manß1-4GlcNAc-PA. Elevation of D-arabinose-containing free-glycans in the urine of cancer patients was confirmed by selected reaction monitoring. This is the first study to unequivocally show the occurrence of a D-arabinose-containing oligosaccharide in human together with its detailed structure.

Identification of ß1-3 galactosylglucose-core free-glycans in human urine.

We previously reported the precise structure of acidic free-glycans in human urine. In the present study, structural analysis of neutral free-glycans in urine was performed in fine detail. Urine samples were collected from 21 healthy volunteers and free-glycans extracted from the creatinine-adjusted urine and then fluorescently labeled with 2-aminopyridine. Neutral glycan profiling was achieved by a combination of high-performance liquid chromatography, mass spectrometry, enzymatic digestion, and periodate cleavage. A total of 79 glycans were identified. Because the ABO-blood group antigen containing urinary neutral glycans are major components, profiling patterns were similar between individuals of the same ABO-group. The neutral glycans were composed of lactose-core (Galß1-4Glc) glycans, type-II N-acetyllactosamine-core (GlcNAcß1-4Glc) glycans, hexose oligomers, N-glycans and to our surprise ß1-3 galactosylglucose-core (Galß1-3Glc) glycans. Although glycans with a ß1-3 galactosylglucose-core were identified as major components in urine, comprising structurally simple isomers of a lactose-core, the core structure has not previously been reported. The major ß1-3 galactosylglucose-core glycans were Fucα1-2Galß1-3(Fucα1-4)Glc, GalNAcα1-3(Fucα1-2)Galß1-3(Fucα1-4)Glc and Galα1-3(Fucα1-2)Galß1-3(Fucα1-4)Glc, corresponding to H-, A-, and B-blood group antigens, respectively. Three lactosamine extended ß1-3 galactosylglucose-core glycans were also detected as minor components. Elucidating the biosynthesis of ß1-3 galactosylglucose will be crucial for understanding the in vivo function of these glycans.

Investigation of acidic free-glycans in urine and their alteration in cancer.

Alterations to glycans in cancer patients have been used to identify novel tumor biomarkers. Most of these studies have focused on protein glycosylation but less attention has been paid to free-glycans. Here, we analyzed acidic free-glycans in the urine of cancer patients to identify novel tumor marker candidates. Specifically, urine samples were collected from patients with gastric cancer, pancreatic cancer and cholangiocarcinoma as well as normal controls. The free-glycans were extracted from creatinine-adjusted urine and fluorescently labeled with 2-aminopyridine. Initially, we performed profiling of urinary free-glycans by high-performance liquid chromatography and mass spectrometry with enzymatic and chemical degradation. More than 100 glycans, including novel structures, were identified. The chromatographic peaks suggested some of these glycans were present at elevated levels in cancer patients. To verify cancer-associated alterations, we compared the glycan levels between cancer patients and normal controls by selected reaction monitoring. Representative structures of glycans with elevated levels in cancer patients included the following: small glycans related to sialyllactose; sialyl Lewis X; lactose- and N-acetyllactosamine (LacNAc) type-II-core glycans with LacNAc (type-I or II)-extensions and modifications of α1,3/4-fucose and/or 6-sulfate on the Glc/GlcNAc; free-N-glycans containing sialylation or ß1,6-branch of 6-sulfo Lewis X; novel NeuAcα2-3Galß1-4(+/-Fucα1-3) Xylα1-3Glc glycans. Our results provide further insight into urinary free-glycans and suggest the potential utility of these compounds as tumor markers.

Identification of internally sialylated carbohydrate tumor marker candidates, including Sda/CAD antigens, by focused glycomic analyses utilizing the substrate specificity of neuraminidase.

In our previous study, 14 sulfated carbohydrate tumor marker candidates were identified by focused glycomic analyses. Here, glycomic analyses focused on internally sialylated glycans to identify novel marker candidates. Internally sialylated glycans were enriched by digestion of pyridylaminated glycans prepared from sera with α-neuraminidase from Salmonella typhimurium, which did not cleave sialic acids linked to internal residues, followed by anion-exchange chromatography. Next, internally sialylated O-glycan profiles were constructed using two types of high performance liquid chromatography, which were compared between 20 healthy controls and 11 patients with gastric cancer and 9 patients with pancreatic cancer. In all, 17 marker candidates were identified. The structures of glycan candidates were precisely analyzed using enzymatic digestion, glycan synthesis, 2D mapping and mass spectrometry. Among 17 candidates, one was STn, and the other 16 comprised 10 core1, 1 core2 and 5 core3 glycans. The various structures included a α2,6-sialylated reducing terminal GalNAc and α2,6-sialylated type1 N-acetyl-lactosamine. Eight candidates possessed the Sda/CAD antigen. The levels of these candidate glycans in sera from all 40 subjects were quantified using a selected reaction monitoring assay and found to be elevated in at least one or more patients. Although the serum levels of each candidate glycan varied between patients, those candidates having the same backbone or determinant, such as core3 backbone and core1 structures with extended type1 N-acetyl-lactosamine, displayed similar patterns of elevation. These results suggest that analysis of multiple markers may be an effective means of diagnosing various cancers.

Correlation of serum sialyl Tn antigen values determined by immunoassay and SRM based method.

We previously identified four glycan tumor marker candidates using a HPLC-based method. One candidate was sialyl Tn (STN), NeuAcα2-6-GalNAc. In this study, glycans were prepared from sera by hydrazine treatment followed by fluorescent labeling with aminopyridine. Pyridylaminated-STN levels of 147 gastric cancer, 85 pancreatic cancer and 10 cholangiocarcinoma patients together with 102 normal controls were accurately quantified using HPLC separation followed by selected reaction monitoring (SRM) assay, which used a stable isotope, tetradeuterium-labeled pyridylamino glycan as an internal standard. Additionally, STN values were also quantified using conventional competitive inhibition radioimmunoassay (RIA). The two STN levels determined by RIA and SRM gave a similar distribution pattern in sera. STN levels were increased in sera from cancer patients compared to those from normal controls. Moreover, the STN levels in sera of cancer patients determined by the two different assay procedures showed a good correlation (i.e., correlation coefficient >0.9). Our results suggest it may be better to determine STN levels using SRM instead of RIA.

Protective effects of intercalated disk protein afadin on chronic pressure overload-induced myocardial damage.

Adhesive intercellular connections at cardiomyocyte intercalated disks (IDs) support contractile force and maintain structural integrity of the heart muscle. Disturbances of the proteins at IDs deteriorate cardiac function and morphology. An adaptor protein afadin, one of the components of adherens junctions, is expressed ubiquitously including IDs. At present, the precise role of afadin in cardiac physiology or disease is unknown. To explore this, we generated conditional knockout (cKO) mice with cardiomyocyte-targeted deletion of afadin. Afadin cKO mice were born according to the expected Mendelian ratio and have no detectable changes in cardiac phenotype. On the other hand, chronic pressure overload induced by transverse aortic constriction (TAC) caused systolic dysfunction, enhanced fibrogenesis and apoptosis in afadin cKO mice. Afadin deletion increased macrophage infiltration and monocyte chemoattractant protein-1 expression, and suppressed transforming growth factor (TGF) ß receptor signaling early after TAC procedure. Afadin also associated with TGFß receptor I at IDs. Pharmacological antagonist of TGFß receptor I (SB431542) augmented mononuclear infiltration and fibrosis in the hearts of TAC-operated control mice. In conclusion, afadin is a critical molecule for cardiac protection against chronic pressure overload. The beneficial effects are likely to be a result from modulation of TGFß receptor signaling pathways by afadin.

Various sulfated carbohydrate tumor marker candidates identified by focused glycomic analyses.

Glycomic analysis focused on sulfated O-glycans was performed to identify novel serum carbohydrate tumor markers. Sulfated glycans were enriched by α-neuraminidase digestion of pyridylaminated glycans prepared from sera, followed by anion exchange chromatography. Sulfated O-glycan profiles were constructed by two types of high performance liquid chromatography separation. Comparison of the profiles from 20 healthy controls with those of 11 gastric and 9 pancreatic cancer patients identified 14 marker candidates. The structures of these candidates were precisely analyzed using various methods including enzymatic digestion and mass spectrometry. The candidates comprised 9 core1 and 5 core2 glycans. All these candidates were monosulfated, and 11 were also mono- or difucosylated, and included various determinants such as 6-sulfo type2 lactosamine, 6-sulfo Lewis X, 6-sulfo Lewis Y, 3'-sulfo type1 lactosamine and 3'-sulfo Lewis A. Furthermore, among the core1 glycans, five candidates displayed a type1 and type2 lactosamine hybrid backbone. The levels of these candidate glycans in the sera from all 40 subjects were quantified using a selected reaction monitoring assay. These analyses revealed: (i) the levels of all candidates were elevated in sera of at least one or more patients; (ii) core1 candidates having type1-type2 hybrid backbones with 6-sulfo Lewis X, 6-sulfo type2 lactosamine or 3'-sulfo Lewis A were elevated in sera of variety of patients; and (iii) levels of the candidates varied widely among patients, suggesting analysis of multiple candidates will be an effective means of screening various cancers. To fully evaluate the clinical utility of these candidates, a further verification study is required.

The active zone protein CAST regulates synaptic vesicle recycling and quantal size in the mouse hippocampus.

Synaptic efficacy is determined by various factors, including the quantal size, which is dependent on the amount of neurotransmitters in synaptic vesicles at the presynaptic terminal. It is essential for stable synaptic transmission that the quantal size is kept within a constant range and that synaptic efficacy during and after repetitive synaptic activation is maintained by replenishing release sites with synaptic vesicles. However, the mechanisms for these fundamental properties have still been undetermined. We found that the active zone protein CAST (cytomatrix at the active zone structural protein) played pivotal roles in both presynaptic regulation of quantal size and recycling of endocytosed synaptic vesicles. In the CA1 region of hippocampal slices of the CAST knockout mice, miniature excitatory synaptic responses were increased in size, and synaptic depression after prolonged synaptic activation was larger, which was attributable to selective impairment of synaptic vesicle trafficking via the endosome in the presynaptic terminal likely mediated by Rab6. Therefore, CAST serves as a key molecule that regulates dynamics and neurotransmitter contents of synaptic vesicles in the excitatory presynaptic terminal in the central nervous system.

Elevation of CA19-9-Related Novel Marker, Core 1 Sialyl Lewis A, in Sera of Adenocarcinoma Patients Verified by a SRM-Based Method.

We have attempted to identify a novel glycan tumor marker. Pyridylaminated (PA) O-glycans were prepared from sera, and the corresponding O-glycan profiles were constructed by HPLC separation. By comparing the serum O-glycan profiles from healthy controls with those of cancer patients, we identified a marker candidate, core 1 sialyl Lewis A (NeuAcα2-3Galß1-3(Fucα1-4)GlcNAcß1-3Gal) (abbreviated C1SLA), whose concentration appeared to be weakly correlated with CA19-9 values. To quantify this glycan, we developed a selected reaction monitoring (SRM) assay that used a stable isotope, tetradeuterium-labeled pyridylamino (d4-PA) glycan, as an internal standard. The analyte (d0-PA-C1SLA) and the internal standard (d4-PA-C1SLA) were subjected to SRM analyses after two types of HPLC separation. Serum levels of C1SLA, determined as the relative ratio to total O-glycans, were then measured. These analyses revealed that (i) C1SLA is a CA19-9-related glycan, (ii) the mean value of C1SLA in normal controls is 3.41 ppm, (iii) the level of C1SLA was significantly higher in samples of stages II-IV stomach cancers (P = 0.0036) as well as pancreatic cancers (P < 0.0001) compared to that of normal controls, (iv) the relationship between C1SLA and CA19-9 varies from poor to weak depending on the cancer, and (v) C1SLA could be valuable as a diagnostic adjunct for cancer.

Genetic ablation of afadin causes mislocalization and deformation of Paneth cells in the mouse small intestinal epithelium.

Afadin is an actin filament-binding protein that acts cooperatively in cell adhesion with the cell adhesion molecule nectin, and in directional cell movement with the small G protein Rap1 in a nectin-independent manner. We studied the role of afadin in the organization of the small intestinal epithelium using afadin conditional gene knockout (cKO) mice. Afadin was localized at adherens junctions of all types of epithelial cells throughout the crypt-villus axis. Paneth cells were localized at the base of the crypt in control mice, but not confined there, and migrated into the villi in afadin-cKO mice. The distribution of other types of epithelial cells did not change significantly in the mutant mice. The Paneth cells remaining in the crypt exhibited abnormal shapes, were buried between adjacent cells, and did not face the lumen. In these cells, the formation of adherens junctions and tight junctions was impaired. Rap1 and EphB3 were highly expressed in control Paneth cells but markedly down-regulated in the afadin-deficient Paneth cells. Taken together, the results indicate that afadin plays a role in the restricted localization of Paneth cells at the base of the crypt by maintaining their adhesion to adjacent crypt cells and inhibiting their movement toward the top of villi.

Genetic deletion of afadin causes hydrocephalus by destruction of adherens junctions in radial glial and ependymal cells in the midbrain.

Adherens junctions (AJs) play a role in mechanically connecting adjacent cells to maintain tissue structure, particularly in epithelial cells. The major cell-cell adhesion molecules at AJs are cadherins and nectins. Afadin binds to both nectins and α-catenin and recruits the cadherin-ß-catenin complex to the nectin-based cell-cell adhesion site to form AJs. To explore the role of afadin in radial glial and ependymal cells in the brain, we generated mice carrying a nestin-Cre-mediated conditional knockout (cKO) of the afadin gene. Newborn afadin-cKO mice developed hydrocephalus and died neonatally. The afadin-cKO brain displayed enlarged lateral ventricles and cerebral aqueduct, resulting from stenosis of the caudal end of the cerebral aqueduct and obliteration of the ventral part of the third ventricle. Afadin deficiency further caused the loss of ependymal cells from the ventricular and aqueductal surfaces. During development, radial glial cells, which terminally differentiate into ependymal cells, scattered from the ventricular zone and were replaced by neurons that eventually covered the ventricular and aqueductal surfaces of the afadin-cKO midbrain. Moreover, the denuded ependymal cells were only occasionally observed in the third ventricle and the cerebral aqueduct of the afadin-cKO midbrain. Afadin was co-localized with nectin-1 and N-cadherin at AJs of radial glial and ependymal cells in the control midbrain, but these proteins were not concentrated at AJs in the afadin-cKO midbrain. Thus, the defects in the afadin-cKO midbrain most likely resulted from the destruction of AJs, because AJs in the midbrain were already established before afadin was genetically deleted. These results indicate that afadin is essential for the maintenance of AJs in radial glial and ependymal cells in the midbrain and is required for normal morphogenesis of the cerebral aqueduct and ventral third ventricle in the midbrain.

An Adaptor Molecule Afadin Regulates Lymphangiogenesis by Modulating RhoA Activity in the Developing Mouse Embryo.

Afadin is an intracellular binding partner of nectins, cell-cell adhesion molecules, and plays important roles in the formation of cell-cell junctions. Afadin-knockout mice show early embryonic lethality, therefore little is known about the function of afadin during organ development. In this study, we generated mice lacking afadin expression in endothelial cells, and found that the majority of these mice were embryonically lethal as a result of severe subcutaneous edema. Defects in the lymphatic vessels of the skin were observed, although the morphology in the blood vessels was almost normal. Severe disruption of VE-cadherin-mediated cell-cell junctions occurred only in lymphatic endothelial cells, but not in blood endothelial cells. Knockout of afadin did not affect the differentiation and proliferation of lymphatic endothelial cells. Using in vitro assays with blood and lymphatic microvascular endothelial cells (BMVECs and LMVECs, respectively), knockdown of afadin caused elongated cell shapes and disruption of cell-cell junctions among LMVECs, but not BMVECs. In afadin-knockdown LMVECs, enhanced F-actin bundles at the cell periphery and reduced VE-cadherin immunostaining were found, and activation of RhoA was strongly increased compared with that in afadin-knockdown BMVECs. Conversely, inhibition of RhoA activation in afadin-knockdown LMVECs restored the cell morphology. These results indicate that afadin has different effects on blood and lymphatic endothelial cells by controlling the levels of RhoA activation, which may critically regulate the lymphangiogenesis of mouse embryos.

Scmh1 has E3 ubiquitin ligase activity for geminin and histone H2A and regulates geminin stability directly or indirectly via transcriptional repression of Hoxa9 and Hoxb4.

Polycomb-group (PcG) complex 1 acts as an E3 ubiquitin ligase both for histone H2A to silence transcription and for geminin to regulate its stability. Scmh1 is a substoichiometric component of PcG complex 1 that provides the complex with an interaction domain for geminin. Scmh1 is unstable and regulated through the ubiquitin-proteasome system, but its molecular roles are unknown, so we generated Scmh1-deficient mice to elucidate its function. Loss of Scmh1 caused derepression of Hoxb4 and Hoxa9, direct targets of PcG complex 1-mediated transcriptional silencing in hematopoietic cells. Double knockdown of Hoxb4 and Hoxa9 or transduction of a dominant-negative Hoxb4N→A mutant caused geminin accumulation. Age-related transcriptional downregulation of derepressed Hoxa9 also leads to geminin accumulation. Transduction of Scmh1 lacking a geminin-binding domain restored derepressed expression of Hoxb4 and Hoxa9 but did not downregulate geminin like full-length Scmh1. Each of Hoxb4 and Hoxa9 can form a complex with Roc1-Ddb1-Cul4a to act as an E3 ubiquitin ligase for geminin. We suggest that geminin dysregulation may be restored by derepressed Hoxb4 and Hoxa9 in Scmh1-deficient mice. These findings suggest that PcG and a subset of Hox genes compose a homeostatic regulatory system for determining expression level of geminin.

Deletion of the presynaptic scaffold CAST reduces active zone size in rod photoreceptors and impairs visual processing.

How size and shape of presynaptic active zones are regulated at the molecular level has remained elusive. Here we provide insight from studying rod photoreceptor ribbon-type active zones after disruption of CAST/ERC2, one of the cytomatrix of the active zone (CAZ) proteins. Rod photoreceptors were present in normal numbers, and the a-wave of the electroretinogram (ERG)--reflecting their physiological population response--was unchanged in CAST knock-out (CAST(-/-)) mice. Using immunofluorescence and electron microscopy, we found that the size of the rod presynaptic active zones, their Ca(2+) channel complement, and the extension of the outer plexiform layer were diminished. Moreover, we observed sprouting of horizontal and bipolar cells toward the outer nuclear layer indicating impaired rod transmitter release. However, rod synapses of CAST(-/-) mice, unlike in mouse mutants for the CAZ protein Bassoon, displayed anchored ribbons, normal vesicle densities, clustered Ca(2+) channels, and essentially normal molecular organization. The reduction of the rod active zone size went along with diminished amplitudes of the b-wave in scotopic ERGs. Assuming, based on the otherwise intact synaptic structure, an unaltered function of the remaining release apparatus, we take our finding to suggest a scaling of release rate with the size of the active zone. Multielectrode-array recordings of retinal ganglion cells showed decreased contrast sensitivity. This was also observed by optometry, which, moreover, revealed reduced visual acuity. We conclude that CAST supports large active zone size and high rates of transmission at rod ribbon synapses, which are required for normal vision.

Systems analysis of ATF3 in stress response and cancer reveals opposing effects on pro-apoptotic genes in p53 pathway.

Stress-inducible transcription factors play a pivotal role in cellular adaptation to environment to maintain homeostasis and integrity of the genome. Activating transcription factor 3 (ATF3) is induced by a variety of stress and inflammatory conditions and is over-expressed in many kinds of cancer cells. However, molecular mechanisms underlying pleiotropic functions of ATF3 have remained elusive. Here we employed systems analysis to identify genome-wide targets of ATF3 that is either induced by an alkylating agent methyl methanesulfonate (MMS) or over-expressed in a prostate tumour cell line LNCaP. We show that stress-induced and cancer-associated ATF3 is recruited to 5,984 and 1,423 targets, respectively, in the human genome, 89% of which are common. Notably, ATF3 targets are highly enriched for not only ATF/CRE motifs but also binding sites of several other stress-inducible transcription factors indicating an extensive network of stress response factors in transcriptional regulation of target genes. Further analysis of effects of ATF3 knockdown on these targets revealed that stress-induced ATF3 regulates genes in metabolic pathways, cell cycle, apoptosis, cell adhesion, and signalling including insulin, p53, Wnt, and VEGF pathways. Cancer-associated ATF3 is involved in regulation of distinct sets of genes in processes such as calcium signalling, Wnt, p53 and diabetes pathways. Notably, stress-induced ATF3 binds to 40% of p53 targets and activates pro-apoptotic genes such as TNFRSF10B/DR5 and BBC3/PUMA. Cancer-associated ATF3, by contrast, represses these pro-apoptotic genes in addition to CDKN1A/p21. Taken together, our data reveal an extensive network of stress-inducible transcription factors and demonstrate that ATF3 has opposing, cell context-dependent effects on p53 target genes in DNA damage response and cancer development.

Involvement of afadin in barrier function and homeostasis of mouse intestinal epithelia.

Afadin interacts with the cytoplasmic region of nectins, which are immunoglobulin-like cell adhesion molecules at adherens junctions, and links them to the actin cytoskeleton. Afadin regulates activities of cells in culture such as directional motility, proliferation and survival. We used Cre-loxP technology to generate mice conditionally lacking afadin specifically in the intestinal epithelia after birth. The loss of afadin caused increased paracellular permeability in the intestinal mucosa and enhanced susceptibility to the tissue destruction induced by dextran sulfate sodium. The junctional architecture of the intestinal epithelia appeared to be preserved, whereas the deficiency of afadin caused the mislocalization of nectin-2 and nectin-3 from adherens junctions to basolateral membrane domains but not that of other components of apical junctions. By contrast, such phenotypic changes were undetected in mice lacking nectin-2, nectin-3 or both. These findings suggest that afadin plays crucial roles, independently of the role as the nectin-afadin module, in barrier function and homeostasis of the intestinal epithelia once the epithelial structure has been established.

Decreased expression of LMO7 and its clinicopathological significance in human lung adenocarcinoma.

LIM-domain only protein 7 (LMO7) has been suggested to act as a tumor suppressor for murine lung adenocarcinoma, while its splice variant p100 LMO7/#16 is associated with invasion and metastasis of rat AH130W1 cells. However, the importance of LMO7 in human lung cancer is unknown. We investigated LMO7 protein expression by immunohistochemistry in tumor tissues obtained from 57 patients with adenocarcinoma of the lung using a rabbit anti-LMO7 antibody. Signals for LMO7 were localized to the apical surface of the bronchial epithelium and to the cell membranes of pneumocytes in non-cancerous pulmonary tissues, but were noted circumferentially around the plasma membrane of cancer cells in all 57 patients with adenocarcinoma. The LMO7-positive group (24 patients, 42%) showed equivocal to strong expression of LMO7 in more than 50% cancer cells, while the remaining 33 patients (58%) showed LMO7 expression in less than 50% of their cancer cells. The latter group had significantly more advanced disease than the LMO7-positive group with regard to T factor (p=0.011), nodal involvement (p=0.026) and p-stage (p=0.010; χ(2) test). Multivariate analysis using a logistic regression model showed that LMO7 expression was independently associated with the T factor (p=0.041). Kaplan-Meier analysis showed that a poor prognosis was associated with low expression of LMO7 (p=0.036; log-rank test). Our findings are consistent with earlier observations and demonstrate that LMO7 is inversely correlated with the development and prognosis of human lung adenocarcinoma.