2003 ASCI/AAP Joint Meeting

Purpose: To identify changes in gene expression in normal adult human dermal endothelial cells (HDEC) after exposure to systemic sclerosis (SSc) sera containing either anti-centromere (ACA) or anti-topoisomerase-I (anti-Scl70) autoantibodies. Methods: RNA from subconfluent adult HDEC (that were incubated with pooled fractions of either 10% normal human sera or SSc sera over multiple times points) were hybridized to cDNA microarrays composed of 2,635 genes. Data were analyzed with Cluster/TreeView software and Self-Organizing Maps to demonstrate temporal differences in gene expression >3-fold. Unique patterns of gene expression in HDEC exposed to SSc sera were targeted for further investigation in the context of gene family maps and relevant biological pathways using GenMAPP software. Results: 478 genes were uniquely expressed (> 3-fold change) in HDEC exposed to SSc sera compared to153 genes in HDEC exposed to normal human sera. Autoantigen gene family maps demonstrated >3-fold upregulation of fibrillin-1 gene expression only in HDEC exposed to SSc sera. Transcriptional changes were verified at the protein level by immunohistochemistry with monoclonal antibody to fibrillin-1. Fibrillin-1 protein localized to the cytoplasm of HDEC undergoing apoptosis in response to SSc sera. Characteristic morphological changes of apoptosis (chromatin condensation, cell shrinkage, membrane blebbing, and apoptotic bodies) were confirmed by differential interference contrast microscopy to occur only in HDEC exposed to SSc sera. These microscopic changes correlated well with apoptosis gene expression as measured by microarrays. Proapoptotic gene expression was demonstrated as early as 1 hour and persisted over the multiple time points studied (1, 6, 12, and 24 hours). Pathway analysis demonstrated the upregulation of caspase-3 and requiem and the downregulation of survivin, NF-kappa-B (p100), and BCL2 genes. Analysis of the anti-apoptotic (angiogenic) pathway mediated by the endothelial differentiation gene 1 (EDG1) also correlated well with apoptosis as demonstrated by persistent downregulation of pathway members including cadherin 5, VEGF, phosphatidylinositol 3-kinase, and RAC serine/threonine protein kinase in HDEC exposed to SSc sera. Conclusion: HDEC exposed to SSc sera aberrantly express the microfibrillar protein, fibrillin-1, while undergoing apoptosis. Potential modification of fibrillin-1 during the process of apoptosis may uncover cryptic epitopes and thus provoke an autoimmune response to this antigen. The preliminary model of fibrillin-1 autoantigen generation during endothelial cell apoptosis proposed here may explain the presence of fibrillin-1 autoantibodies seen in the majority of patients with SSc and in the tight skin (tsk) murine model of scleroderma.

VIP--a Novel RhoGAP Protein with Rac-1 GTPase Activity, Modulates the Regulation of Endothelin-1 Promoter Expression by Vascular Endothelial Zinc Finger-1

Julius Aitsebaomo¹ and Cam Patterson²

Division of Cardiology, University of North Carolina, Chapel Hill¹; Director, Carolina Cardiovascular Biology Center, University of North Carolina, Chapel Hill².

Gene regulation in the vascular endothelium is required for maintenance of vascular integrity and homeostasis during vascular development and in adulthood. However, little is known about the coordinated interplay between transcription factors and signaling molecules that regulate endothelial cell-dependent transcriptional events. Vascular Endothelial Zinc Finger-1 (Vezf1/DB1) is a zinc finger-containing transcription factor that is specifically expressed within the vascular endothelium during vascular development. Vezf1/DB1 potently and specifically activates transcription of the human endothelin-1 promoter in transient transfection assays by binding to the Vezf-1/DB1-responsive element, ACCCCC, located 47 bp upstream of the human endothelin-1 transcriptional start site. VIP (Vezf-1 Interacting Protein) is a novel 68-kda N-terminal RhoGAP domain-containing protein we identified in a yeast two-hybrid screen of a mouse day 10.5 cDNA embryo library using the zinc finger- containing domain of Vezf-1/DB1 as bait. VIP mRNA is predominantly expressed in vascular endothelial cells by northern blot analysis. Recombinant Rac-1 and Vezf-1/DB1 both bind avidly to VIP, suggesting that VIP is not only a GTPase-activating protein for Rac-1, but that VIP may also be part of the protein complex that binds to and modulates Vezf-1/DB1 transcriptional activity. Functionally, VIP specifically activates the GTPase activity of Rac-1 in-vivo, and potently inhibits Vezf-1/DB1-mediated transcriptional activation of the human endothelin-1 promoter in a dose dependent fashion. Taken together, these data suggest that Vezf-1/DB1 is a multifunctional protein that regulate transcriptional activity of the endothelin-1 promoter and is involved in the Rac-1-dependent, VIP modulated, signal transduction pathway that regulates vasomotor tone and vascular development.

Mechanisms of Autoinhibition and STI-571/Imatinib Resistance Revealed by Mutagenesis of BCR-ABL

Mohammad Azam, Robert R. Latek, and George Q. Daley

1 Whitehead Institute, 9 Cambridge Center, Cambridge, MA.

The Bcr-Abl fusion protein kinase causes chronic myeloid leukemia and is targeted by the signal transduction inhibitor STI-571/Gleevec/imatinib (STI-571). Sequencing of the BCR-ABL gene in patients who have relapsed after STI-571 chemotherapy has revealed a limited set of kinase domain mutations that mediate drug resistance. To obtain a more comprehensive survey of the amino acid substitutions that confer STI-571 resistance, we performed an in vitro screen of randomly mutagenized BCR-ABL and recovered all of the major mutations previously identified in patients and numerous others that illuminate novel mechanisms of acquired drug resistance. Structural modeling implies that a novel class of variants acts allosterically to destabilize the autoinhibited conformation of the ABL kinase to which STI-571 preferentially binds. This screening strategy is a paradigm applicable to a growing list of target-directed anti-cancer agents and provides a means of anticipating the drug-resistant amino acid substitutions that are likely to be clinically problematic.

Phosphatidylcholine Transfer Protein Promotes Cellular Lipid Efflux Specifically through the Apolipoprotein A-I/ATP Binding Cassette A1 Pathway

Juan M. Baez and David E. Cohen

Marion Bessin Liver Research Center, Departments of Medicine and Biochemistry, Albert Einstein College of Medicine, Bronx, NY.

The phosphatidylcholine transfer protein (PC-TP) is a soluble lipid transporter that catalyzes intermembrane transfer of phosphatidylcholines exclusively. To explore the hypothesis that PC-TP supplies phosphatidylcholines to the plasma membrane for ATP Binding Cassette A1 (ABCA1)-dependent lipid efflux, we stably transfected Chinese hamster ovary (CHO) cells with a mammalian expression vector encoding a human PC-TP cDNA. Cellular cholesterol and phosphatidylcholines were radiolabeled by 48-h incubation with [³H]cholesterol or [³H]choline respectively and appearance of cholesterol and phosphatidylcholines in the medium assessed during a 6-h period following addition of extracellular acceptors at 37^oC. In experiments employing apolipoprotein A-I, PC-TP overexpression was associated with a >40% enhancement in apparent V_max for both cholesterol and phospholipid efflux, while a 50% reduction in apparent K_m was observed only for phosphatidylcholine efflux. Characterization of newly formed particles by fast protein liquid chromatography and agarose gel electrophoresis showed that PC-TP specifically enhanced the formation of discoidal nascent pre-beta high density lipoproteins (HDL). When HDL was used as the lipid acceptor, overexpression of PC-TP was associated with only a modest increase in phosphatidylcholine efflux (10% increase in V_max; 30% decrease in K_m) and no change in cholesterol efflux. Our findings provide strong evidence of a central role for PC-TP in the cellular ABCA1-controlled lipid efflux pathway and suggest physiological roles in HDL formation and cellular efflux of cholesterol. Because PC-TP is abundantly expressed in macrophages, we speculate that PC-TP may play a key role in protection against atherosclerosis.

Winner, Poster Session Award

Disruption of Leptin Signaling Contributes to Cardiac Hypertrophy Independently of Body Weight in Mice

Lili A. Barouch¹, Dan E. Berkowitz³, Robert W. Harrison¹, Christopher P. O�Donnell², Joshua M. Hare¹

Department of Medicine, Divisions of Cardiology¹ and Pulmonary Medicine², and the Department of Anesthesia and Critical Care Medicine³, Johns Hopkins Medical Institutions, Baltimore, MD.

Background: Whether left ventricular hypertrophy (LVH) in obesity results from increased hemodynamic load or altered neurohormonal signaling remains controversial. Dysregulation of leptin, a neurohormone essential to energy homeostasis, is implicated in the pathogenesis of obesity. Because leptin has cardiovascular bioactivity, we hypothesized that disruption of leptin signaling mediates the development of obesity-associated LVH. Methods and Results: We measured left ventricular (LV) wall thickness and LV mass with echocardiography in mice lacking leptin (ob/ob, n=15) or functional receptor (db/db, n=15), and controls at 2, 4, and 6 months of age. None of the mice had LVH at 2 months. Progressive obesity developed in ob/ob and db/db mice. At 6 months, LVH occurred in ob/ob and db/db compared to controls. We observed corresponding myocyte hypertrophy by light microscopy. To separate the direct contribution of leptin deficiency from mechanical effects of obesity, we induced weight loss in 6-8 month old ob/ob mice either by leptin infusion or caloric restriction. Mice in both groups lost similar weight compared to placebo-treated controls. Leptin infusion completely reversed the increase in wall thickness with partial resolution of myocyte hypertrophy, whereas calorie-restricted mice had no decrease in wall thickness and a lesser change in myocyte size. Conclusions: Together these data show that the effect of leptin on LV remodeling is not due to weight loss alone, indicating that leptin has anti-hypertrophic effects on the heart, either directly or through a leptin-regulated neurohumoral pathway. Disruption of leptin signaling may represent a novel mechanism in LVH and related cardiovascular disorders.

Molecular Profiling of Transcriptional Targeting Dynamics in Activated T-cells

Irene Collins, Wendy J. Freebern, James Smith, Cynthia M. Haggerty, Mark McNutt, Paul Hoover, Idalia Montano, and Kevin Gardner

Laboratory of Receptor Biology and Gene Expression, Centers for Cancer Research, National Cancer Institute, National Institutes of Health.

During immune response, extracellular signals trigger a series of biochemical events that result in T cell activation. How these signals are integrated by signal transduction pathways, to influence gene regulatory events in the nucleus, is unclear. Our laboratory is using three approaches to investigate this process. In the first approach we generate and analyze data from high-throughput transcriptional reporter assays. We challenge multiple gene regulatory elements targeted during T-cell activation with more than 100 different combinations of T-cell mitogens and drugs that modulate T-cell function. The resulting wealth of data is then analyzed using several statistical approaches, including principal component analysis, hierarchical clustering, and unsupervised neural networks, revealing trends and classifications in signal pathways and control points. By a second approach we combine Chromatin immunoprecipitation (ChIP), with micro-array technology ("ChIP to chip"), to assess the kinetic profiles of the in vivo assembly of higher order transcriptional complexes at multiple promoters in the nucleus. Results from this approach provide theoretical evidence for an "embedded language" or "kinetic code" that could explain why certain genes share common or convergent patterns of expression. In the third method we are using nuclear run on assays to profile the transcription activation of endogenous genes by direct measurement. We designed a protocol that employs the miniaturization of micro-array technology with the molecular specificity of the nuclear to run-on assay to gain powerful insights into the dynamic changes in the population of engaged RNA polymerases at multiple targets during T-cell activation.

Follicle-stimulating hormone (FSH) regulates the transcription factor FoxO1a (FKHR) in ovarian cells

Melissa A. Cunningham and James M. Hammond

Departments of Medicine and Cellular and Molecular Physiology, The Pennsylvania State University College of Medicine, Hershey, PA 17033

Follicle-stimulating hormone (FSH) stimulates the growth and differentiation of ovarian granulosa cells, and mediates cyclic recruitment of follicles leading to the emergence of the preovulatory follicle. To reach the preovulatory stage, the follicle must escape apoptosis, the fate of most follicles. While the differentiative effects of FSH in granulosa cells are accomplished largely through a cAMP/PKA-dependent pathway, the signaling pathways that mediate the survival and growth effects of FSH on granulosa cells have not been established. The forkhead (FoxO) family of transcription factors is known to play a critical role in the regulation of cellular differentiation and proliferation in many systems. For example, the negative regulation of FoxO1a (FKHR) by Akt/PKB promotes cell survival by effectively inhibiting FKHR�s role as a transactivator of apoptosis-inducing genes. Forkhead family members have also been shown to activate cell-cycle arrest genes such as p27kip, thereby implicating this protein as a potential regulator of proliferation. We have demonstrated for the first time that FKHR is expressed in porcine granulosa cells and that FSH is able to regulate FKHR phosphorylation and subcellular localization. RT-PCR and Western blot studies demonstrated that FKHR is expressed in this cell system and indicated that the acute regulation by FSH was a post-translational mechanism. Specifically, while we found no increase in message levels in response to FSH (0-6h), [32p]orthophosphate labeling of cultured granulosa cells revealed robust phosphorylation after FSH treatment for 30min. In addition, we find that the PI 3-kinase signal transduction pathway mediates the FSH-dependent effect on FKHR nuclear exclusion in these cells. The cytosolic accumulation of FKHR protein that was observed in FSH-treated cells both by Western blot and immunohistochemistry was blocked when the cells were pre-incubated with the PI 3-kinase inhibitor LY294002. These data suggest that FSH can regulate the transcription factor FKHR and that FKHR may play a role in FSH-stimulated growth and survival of granulosa cells.

Cerebral vascular dysfunction mediated by superoxide in hyperhomocysteinemic mice

Sanjana Dayal, Erland Arning, Teodoro Bottiglieri, Curt D. Sigmund, Frank M. Faraci, and Steven R. Lentz

Veterans Affairs Medical Center and University of Iowa, Iowa City, IA, USA; Baylor Institute of Metabolic Disease, Dallas, TX, USA.

Hyperhomocysteinemia is a risk factor for stroke and cardiovascular disease, but little is known about effects of hyperhomocysteinemia on cerebral vascular function in vivo. We tested the hypothesis that moderate hyperhomocysteinemia causes impairment of endothelium-dependent dilatation of cerebral arterioles through a mechanism that involves superoxide. Heterozygous cystathionine beta-synthase-deficient (Cbs +/-) or wild-type (Cbs +/+) mice were fed either a control diet or high-methionine (HM) diet for 11 months. Plasma total homocysteine was higher on the HM diet compared with the control diet for both Cbs +/+ (7.1+1.6 vs. 3.9+0.6 micromol/L; p<0.05) and Cbs +/- (12.3+2.1 vs. 7.3+1.3 micromol/L; p<0.05) mice. Dilatation of cerebral arterioles (~30 microm diameter) was measured in response to superfusion with the endothelium-dependent dilator acetylcholine. Dilatation to 10 microM acetylcholine was impaired on the HM diet compared with the control diet for both Cbs +/+ (12+2 vs. 27+3%; p<0.01) and Cbs +/- mice (12+2 vs. 22+2%; p<0.01). Dilatation to acetylcholine improved toward normal after addition of the superoxide scavenger Tiron (p<0.05). Dihydroethidine staining for superoxide in the carotid artery was elevated in Cbs +/- mice fed the HM diet compared with Cbs +/+ mice fed the control diet (p<0.05), and was inhibited by Tiron or pegylated superoxide dismutase. We conclude that superoxide is a key mediator of endothelial dysfunction in the cerebral circulation during hyperhomocysteinemia in vivo.

Tec family kinases multifaceted roles in T cell activation

Derek Dombroski, Pamela L. Schwartzberg

National Human Genome Research Institute, NIH Bethesda, MD.

T lymphocyte activation is initiated when a T cell antigen receptor (TCR) engages a non-self peptide ligand in the context on MHC molecules presented by other cells. Full activation requires signaling from the TCR complex in conjunction with rearrangement of the actin cytoskeleton, which allows signaling complexes to be positioned in an appropriate spatial and temporal manner. A key signaling complex in T cell activation includes the adaptor proteins LAT and SLP-76 along with the Tec family kinases (TFK) and phospholipase C-gamma (PLC-gamma). TFK first came to the forefront as key players in antigen receptor signaling when mutations in Bruton�s tyrosine kinase were found to be responsible for the B cell defect in the disease X-linked agammaglobulinemia. We have shown through gene targeting that TFK also are critical in T cells and required for optimum PLC-gamma activation and subsequent downstream effects such as calcium mobilization and NFAT activation.

One caveat to studies in gene targeted animals is that T lymphocytes from these mice undergo altered development. To address the function of TFK in T lymphocytes independent of altered development, we employed short interfering RNA (siRNA) in Jurkat cells, a well-characterized human T cell line. Using synthetic oligonucleotides specific for IL-2 inducible T cell kinase (Itk), the predominant TFK expressed in Jurkats, we have recapitulated defects in calcium mobilization and NFAT activation demonstrating these phenotypes are independent of altered maturation of mutant cells in the mice. We have extended the use of siRNA to demonstrate novel roles of TFK in T cell activation. We have found decreased Itk expression reduces TCR-induced adhesion and actin polymerization. Furthermore, we have found that TFK are important in chemokine responses by demonstrating impaired calcium flux to chemokine stimulation. Collectively these data point to TFK possessing crucial roles in many facets of T cell activation effector functions including gene expression, cytoskeletal reorganization, chemotaxis and cell adhesion. Importantly, we now have expanded our model to examine the specific domains within Itk responsible for these discrete functions by expressing mouse Itk in our human cell line. Thus we can knockdown endogenous protein levels and concomitantly express desired domain mutants to dissect the signaling functions of Itk. In summary we have used siRNA to set up a genetic system to evaluate novel roles for proteins in T cell activation while bypassing the question of altered T cell maturation.

CD8+ T Cell Phenotypes in HIV Infection

Brinda Emu, Elizabeth Sinclair, Walter Moretto, Steve Deeks, JM McCune, Douglas F Nixon

Gladstone Institute of Virology and Immunology, University of California, San Francisco

Background: There is increasing evidence of dysfunctional CD8+ T cell activity during HIV infection. One proposed hypothesis of CD8+ T cell dysfunction is that there might be abnormal maturation of HIV-specific cells. It is difficult to relate prior studies on this topic, however, because a variety of phenotypic markers for T cell differentiation stages have been used in each. Furthermore, we have previously observed that some antiretroviral-treated patients with multidrug-resistant (MDR) HIV maintain low levels of viremia and that these MDR variants are associated with preserved thymic function in vivo. Given these observations, we hypothesized that patients doing well with MDR HIV (�partial controllers�) may be immunologically controlling viral replication in a manner analogous to that observed with long-term non-progressors (LTNP). Methods: We developed a seven-parameter flow cytometry panel to define memory and effector CD8+ T cell phenotypes, using a panel of T cell differentiation markers that overlap those used in prior studies. PBMCs were stimulated with HIV peptide pools. Cytokine secretion as well as cell surface expression of CD45RA, CCR7, CD27, and CD28 was quantified using flow cytometry. We performed a cross-sectional analysis of long term non-progressors (LTNP, n=3) and patients on HAART with viral load detectable between 200-10,000 copies with stable CD4+ T cell counts and clinical status (�partial controllers�, n=13). Results: In LTNP and in partial controllers, most HIV-specific CD8+ T cells that produce interferon-gamma (IFN-gamma) have the phenotype CCR7-C45RA-. Irrespective of the expression of CCR7 and CD45RA, the majority of cells in both patient groups tended to be CD28- and evenly divided in the expression of CD27. Conclusions: LTNP and partial controllers demonstrated strong responses to HIV peptide pools from an effector population with the phenotype, CCR7-CD45RA-. The expression of CD27 and CD28 was similar in both groups of patients. Given the similarities between LTNP and partial controllers, our data suggest that the long-term control of MDR HIV is in part mediated by effective anti-HIV cellular immune responses. Longitudinal studies are in place to determine whether the presence of HIV-specific cells with the CCR7- CD45RA- is associated with durable virologic control.

Male infertility and thiamine-dependent erythroid hypoplasia in mice lacking thiamine transporter Slc19a2

Judith C. Fleming¹, Elena Tartaglini¹, Ryosuke Kawatsuji¹, David Yao¹,Yuko Fujiwara^1,2, Jeff Bednarski¹, Mark D. Fleming³, and Ellis J. Neufeld¹

¹Division of Hematology, Children�s Hospital, Dana-Farber Cancer Institute, and Harvard Medical School, Boston, Massachusetts 02115; ²Howard Hughes Medical Institute, Boston, Massachusetts 02115; ³Department of Pathology, Children's Hospital, Boston, Massachusetts 02115.

Thiamine-responsive megaloblastic anemia with deafness and diabetes (TRMA) is an autosomal recessive disease caused by mutations in high-affinity thiamine transporter gene SLC19A2. We hypothesize that thiamine transport via SLC19A2 plays a role in the development and maintenance of several organ systems, in particular the erythropoietic, auditory, and glucose homeostasis systems. To study the role of thiamine transport in the pathophysiology of TRMA syndrome and of each of the component disorders, we created a targeted disruption of the Slc19a2 gene in mice. Slc19a2 -/- mice are viable. Male -/- mice on a pure 129/Sv background are infertile with small testes (testis/body weight =0.13+0.04 knockout vs 0.35+0.05 wild type). Two independent lines were assessed. Histologic examination of the testes revealed the presence of interstitial Leydig cells and Sertoli cells in the tubules. The lack of developing germ cells beyond primary spermatocytes suggested an apparent arrest in spermatogenesis prior to the first meiotic division. Nuclear chromatin changes suggestive of apoptosis were detected. Mature sperm were found neither in the tubules nor in the epididymis. This phenotype suggests a previously unknown role for thiamine transport in spermatogenesis and male fertility. Slc19a2 -/- mice on a pure 129/Sv background develop reticulocytopenia after two weeks on thiamine-depleted chow with a virtual absence of reticulocytes in the peripheral blood (0.12% knockout vs 2.64% wild type). Examination of the bone marrow, revealed a complete lack of erythroid precursors. Contrary to human TRMA syndrome, we saw no evidence of megaloblastosis or ringed sideroblasts in the bone marrow of Slc19a2 -/- mice in thiamine-replete or thiamine-deficient dietary states. Systemic thiamine deficiency developed more rapidly in the knockout mice than in wild type mice (2-3 weeks vs 4-5 weeks), but knockout mice on regular mouse chow (22 mg/kg) had whole blood thiamine levels twice as high as normal human levels (12 ug/dL vs 6 ug/dL). Phenotypic differences between TRMA patients and Slc19a2 -/- mice might be explained by dissimilar tissue expression patterns of the transporter, as well as differing metabolic needs and possible different species-specific contributions of the related thiamine transporter SLC19A3.

Gene Expression Profiles in Polypoid Dysplastic Adenoma and Distinct Adenocarcinomas Associated with Chronic Ulcerative Colitis

Christopher Flynn¹, Hongyi Cui¹, Kishore Guda¹, Richard Muller², Paul V. Vignati³, Sanjeev Garg⁴, Luke E.K. Achenie⁴, and Daniel W. Rosenberg¹

¹Center for Molecular Medicine, University of Connecticut Health Center, Farmington, CT; ²Hartford Pathology Associates, Hartford, CT; ³CT Surgical Group, P.C. Hartford, CT; ⁴Department of Chemical Engineering, University of Connecticut, Storrs, CT.

We report a case of chronic ulcerative colitis (CUC) with two polypoid dysplastic adenomas (PDA), an exophytic adenocarcinoma (EAC) and a flat adenocarcinoma (FAC) present within the inflamed mucosa. Using a human 4.6K array slide containing 4,608 human cDNAs spotted in duplicate combined with a novel adaptive centroid algorithm for data analysis, we examined differentially expressed genes as well as overall transcription profiles within each of the lesions. A panel of genes involved in diverse cellular functions (e.g. K-ras, TGFbeta2, CEA family proteins and Wnt-signaling mediators) was differentially expressed in PDA in comparison with flat, inflamed mucosa. EAC, however, exhibited similar transcription profiles to PDA, suggesting that these lesions might be pathogenically related. Interestingly, several genes (including intestinal and liver tetraspan membrane protein, glutathione S-transferase pi, tissue inhibitor of metalloproteinase 3, and mitotic arrest deficient-like 1) in FAC displayed an expression pattern that was different from both PDA and EAC, indicating the involvement of distinct molecular pathways in lesion-specific tumor development. Further immunohistochemical analysis of both glutathione S-transferase pi and tissue inhibitor of metalloproteinase 3 was undertaken using additional patient samples to establish a consistent pattern of molecular alterations. Delineation of these molecular alterations may provide insights into the underlying mechanisms of CUC-related tumorigenesis and identify novel targets for more effective surveillance programs in the future.

A novel role for CD80 expressed in podocytes in proteinuric kidney diseases

Gero von Gersdorff¹, Jochen Reiser¹, Karin Schwarz¹, Martin Loos¹, Hikaru Sugimoto³, Jordan Kreidberg², Raghu Kalluri³, Maria-Pia Rastaldi⁵, Peter Mundel¹

¹Departments of Medicine and Cell Biology, Albert Einstein College of Medicine, Bronx, New York; ²Departments of Medicine and Pediatrics, Children�s Hospital and Harvard Medical School, Boston, Massachusetts; ³Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts; ⁴Renal Research Association, Renal Immunopathology Centre, San Carlo Borromeo Hospital, Milan, Italy

CD80 is a member of the immunoglobulin- superfamily normally found on B-cells and antigen presenting cells where it functions as a costimulatory molecule for T-cell activation. Here we show that CD80 was induced in renal glomerular podocytes in a variety of genetic and inducible models of podocyte foot process effacement. Podocyte CD80 expression correlated with the development of proteinuria in a mouse model of lupus nephritis, and with WHO class in human biopsy specimen. Similarly, LPS induced proteinuria and CD80 expression in wild-type BALB/c mice. In alpha3- integrin deficient mice, which cannot form normal foot processes, CD80 was strongly upregulated. In cultured alpha3-/- podocytes B7-1 could form an adhesion complex with beta1-integrin that bound to collagen IV and laminin alpha2. Outside-in signaling through CD80 lead to the relocation of beta1 integrin from focal contacts to sites of CD80 ligand binding. Activation of B7-1 by ligand also resulted in podocyte actin cytoskeleton reorganization, and in redistribution of slit diaphragm (SD) proteins (nephrin, CD2AP and ZO-1) to sites of B7-1 ligation. These data suggest that pathologically expressed CD80 can modulate focal contacts and cell spreading in podocytes. CD80 in podocytes may contribute to the pathogenesis of proteinuria by modulating podocyte-matrix adhesion and sequestration of vital SD molecules away from their normal sites of action, thereby altering glomerular permselectivity.

beta-Arrestin Expression Determines the Mechanism of ERK1/2 Activation and the Transcriptional Response to Lysophosphatidic Acid in Murine Embryo Fibroblasts

Diane Gesty-Palmer¹, Hesham El-Shewy¹, Moahad Dar¹, Trudy A. Kohout², and Louis M. Luttrell^1,3

Departments of Medicine¹ and Biochemistry², Duke University Medical Center, Durham, NC 27710, and The Geriatrics Research, Education and Clinical Center³, Durham Veterans Affairs Medical Center, Durham, NC 27705.

Over the past decade, it has become apparent that heptahelical, or G protein-coupled, receptors (GPCRs) play important roles in the regulation of cellular growth and differentiation in many pathophysiologic states. In many cases, these effects are mediated through the activation of the extracellular signal-regulated kinases 1 and 2 (ERK1/2). Previous work has shown that ERK1/2 activation by GPCRs often results from cross talk between GPCRs and the Epidermal Growth Factor (EGF) receptor. We have found that beta-arrestins, proteins that bind to agonist-occupied GPCRs and uncouple them from their cognate heterotrimeric G proteins, can act as scaffolds to recruit and activate ERK1/2 in a multiprotein complex that remains associated with the receptor as it internalizes. Since these two fundamentally distinct mechanisms of ERK1/2 activation may lead to the formation of spatially discrete pools of active kinase, we hypothesized that the expression of beta-arrestin would play a significant role in determining both the mechanism ERK1/2 activation and the transcriptional response to GPCR stimulation. In this study, we have investigated the role of beta-arrestins in ERK1/2 activation and function using murine embryo fibroblasts (MEFs) derived from beta-arrestin 1, beta-arrestin 2 and beta-arrestin 1/2 knockout mice. In wild type and beta-arrestin 1 null MEFs, ERK1/2 activation following stimulation of endogenous lysophosphatidic acid (LPA) receptors was insensitive to treatment with tyrphostin AG1478, a specific inhibitor of the EGF receptor tyrosine kinase. In contrast, LPA-stimulated ERK1/2 activation in beta-arrestin 2 null, and beta-arrestin 1/2 null MEFs was >90% AG1478-senstive. Stable expression of beta-arrestin 2 in the beta-arrestin 1/2 null background recapitulated the wild type phenotype, indicating that cross talk with EGF receptors was the predominant mechanism of ERK1/2 activation in the absence of endogenous beta-arrestin 2, while expression of beta-arrestin 2 conferred ERK1/2 activation through an EGF receptor-independent mechanism. We next examined the effect of beta-arrestin 2 expression on the transcriptional response to LPA stimulation using gene expression arrays containing gene specific cDNA markers for GPCR-mediated signaling. The contribution of EGF receptor cross talk and ERK1/2 to the transcriptional response was determined by testing sensitivity to AG1478 and the MEK1/2 inhibitor PD98059. In the beta-arrestin 1/2 null background, 1 hr exposure to LPA induced expression of several marker genes, including Egr1, Ptgs2, Myc, Junb, Serpine 1, F3, and Scya2. Six of the seven responses were sensitive to AG1478, and two were also sensitive to PD98059, indicating that EGF receptor transactivation is a significant mediator of the early transcriptional response to LPA. In stable beta-arrestin 2 expressing cells, the LPA stimulated transcriptional responses were either blunted or absent, and the residual responses were insensitive to AG1478 and PD98059. Thus, ERK1/2 activation in beta-arrestin 2 replete MEFs, while robust, did not contribute to the transcriptional response to LPA. These data indicate that beta-arrestins, through their dual role as terminators of receptor-G protein coupling and as scaffolds that direct the activation and spatial localization of ERK1/2, play a key role in determining the transcriptional response to GPCR stimulation.

Truncated DNMT3B Proteins Mediate the Hypomethylation of Repetitive Sequences in Cancer Cells

Lucy A. Godley¹, Kelly R. Ostler¹, Stacey L. Payne^1,2, Elizabeth M. Davis¹, and Michelle M. Le Beau¹

¹Section of Hematology/Oncology, University of Chicago, 5841 S. Maryland Ave., Chicago, IL 60637; ²Present address: Department of Anatomy and Cell Biology, Roy J. and Lucille A. Carver College of Medicine, Holden Comprehensive Cancer Center, The University of Iowa, 375 Newton Rd., Iowa City, IA, 52242.

DNA methylation plays an important role in the transcriptional regulation of cellular genes. In cancer cells, repetitive sequences are hypomethylated, transcriptionally active, and facilitate the formation of complex chromosomal rearrangements, whereas the promoters of many tumor suppressor genes are hypermethylated and transcriptionally silenced. The mechanism(s) leading to aberrant methylation in cancer cells is unknown. We have shown that there is aberrant splicing of the DNMT3B gene in numerous cancer cell lines and in primary leukemia cells. These abnormal transcripts are predicted to result in premature termination of translation, such that truncated forms of DNMT3B proteins lacking the C-terminus and the catalytically active domain are produced. Although the truncated proteins lack classical nuclear localization signals, they are concentrated in the nucleus. Overexpression of the most common version of these proteins in 293 cells results in hypomethylation of repetitive sequences, suggesting that the presence of catalytically-inactive DNMT3B proteins in cancer cells signals the abnormal methylation of their satellite DNA sequences. The predicted truncated forms of DNMT3B are reminiscent of those found in cells from patients with Immunodeficiency; Centromere instability; Facial anomalies (ICF) Syndrome, a rare disease in which germline mutations of the DNMT3B gene result in the hypomethylation of repetitive sequences and the formation of chromosomal rearrangements. An emerging paradigm in cancer biology is the production of altered proteins as a result of aberrant RNA splicing. Our results suggest that deregulation of DNMT3B function as a result of truncated DNMT3B proteins contributes to the aberrant DNA methyation characteristic of tumor cells.

Winner, Poster Session Award

Glucose Tolerance is Impaired with Increased Slow-Twitch Muscle: An Animal Model of Aging Skeletal Muscle

Laura Greenlund and K. Sreekumaran Nair

Mayo Clinic Department of Endocrinology, Rochester, MN.

Aging is associated with a decrease in muscle mass and a shift toward a slow twitch muscle fiber predominance. Since skeletal muscle is a major site of glucose uptake after a meal, these age-related changes may play an important role in the impairment of glucose tolerance that occurs with age. To test the hypothesis that changes muscle fiber type cause impairment of glucose tolerance, we generated transgenic mice that over-express myosin heavy chain 1 in skeletal muscle.

Using a mini-gene construct in which transgene expression is driven by a skeletal muscle-specific promoter (myosin light chain 2 promoter), we have generated mice that have a 53% + 10% (p<0.05) increase in slow-twitch skeletal muscle fibers in the quadriceps (vastus medialis) and a 13% + 5% (p<0.05) increase in the soleus. This shift in muscle fiber type results in animals that are normal in size and weight in comparison with control littermates. These mice show metabolic changes similar to those observed in aging humans. They demonstrate impaired post-prandial glucose tolerance with a peak glucose that is 40% + 3% higher than control mice, and serum insulin levels are 41% + 9.5% (p<0.01) higher than control animals. This impaired post-prandial glucose tolerance is associated with a reduction in the capacity of muscle to store glucose in the form of glycogen as evidenced by a 25.3% + 7.8% reduction in quadriceps muscle glycogen concentration. In addition, these mice have more body fat with a 60% + 14% (p<0.024) increase in the accumulation of intra-abdominal fat and a 6% + 3% (p<0.002) reduction in lean body tissue. Measurements of 24-hour spontaneous physical activity and measurement of food consumption showed no differences compared to control mice. The results strongly support that muscle fiber type is a key factor in determining glucose disposal. This mouse presents an animal model for age-related muscle changes and may in part explain the cause of impaired glucose tolerance and fat accumulation with age.

Regulation of Stat5 Function by c-Myb

Lisa N. Haubein, Michael A. Rycyzyn, and Charles V. Clevenger

Dept. of Pathology and Lab Medicine, University of Pennsylvania, Philadelphia, PA 19104

Our lab has recently discovered that Stat5 activity in mammary cells is regulated by intranuclear prolactin (PRL) in complex with the peptidyl-prolyl isomerase cyclophilin B (CypB; PNAS 99:6790, 2002). Following endocytosis and nuclear retrotransport of this complex, PRL/CypB interacts with Stat5 inducing the release of a repressor of Stat5 DNA-binding, namely PIAS3, thereby enhancing Stat5 DNA-binding activity and induced gene expression. To further delineate the mechanisms of Stat5 regulation, our laboratory has utilized Transcription Factor-Transcription Factor Interaction Array Analysis^TM (TF-TF array, Panomics). This approach permits the analysis of the association of 108 different transcription factors/co-activators/-repressors within anti-Stat5 immunoprecipitates obtained from resting, PRL, and PRL + CypB stimulated T47D breast cancer cell lysates. This method has enabled us to identify seven entirely novel transcription factor-Stat5 interactions. We report here the confirmation and analysis of one of these interactions, namely the interaction between Stat5 and c-Myb. Co-immunoprecipitation analysis has demonstrated a PRL-inducible interaction between Stat5 and c-Myb, an interaction that is diminished by the inclusion of CypB. Additional analysis has also revealed an interaction between CypB and Stat5; however, no interaction between PIAS3 and c-Myb were noted. Gene-reporter studies using a beta-casein-luciferase construct in CHO cells have revealed a five-fold potentiation of the expression of this reporter by co-transfection with c-Myb. Given the recognized overexpression of c-Myb known to occur in favorable histology breast cancers, the discovery of a functional interaction between Stat5 and c-Myb may represent a new site for targeted pharmacologic intervention in human breast cancer.

PPAR-gamma activation inhibits Helicobacter pylori-induced host inflammatory responses in vivo and in vitro

Dawn A. Israel, Rajnish Gupta, Kristin Gaus, Scott Shappell, Raymond N. DuBois, and Richard M. Peek, Jr.

Vanderbilt University School of Medicine

Chronic gastritis induced by Helicobacter pylori significantly increases the risk for peptic ulceration and gastric cancer, yet only a fraction of colonized persons develop pathologic outcomes and levels of inflammation vary considerably among infected subjects. H. pylori activates NF-kappaB and induces secretion of the pro-inflammatory cytokine IL-8 from gastric epithelial cells in vitro and in vivo. A specific host target that may attenuate such inflammatory responses is PPAR-gamma, since activation of this transcription factor suppresses NF-kappaB-dependent IL-8 secretion by colonic epithelial cells in vitro and inflammation in experimental models of colitis. We investigated the molecular pathways through which PPAR-gamma activation affects H. pylori-induced IL-8 secretion and inflammation and then used microarray to identify PPAR-gamma target genes that may regulate these events. Methods: Mongolian gerbils and C57/BL6 mice fed either control or BRL 49653-supplemented chow were infected with H. pylori or broth alone and sacrificed 4-8 weeks post-challenge. PPAR-gamma activity was measured in AGS gastric epithelial cells in vitro using transient reporter gene assays. AGS cells were co-cultured with H. pylori in the presence or absence of PPAR-gamma ligands 15d-PGJ₂ or BRL 49653, and IL-8 was quantified by ELISA. NF-kappaB activation was determined by immuno-fluorescence, Western analysis, and in vitro kinase assays. RNA isolated from AGS cells treated with or without BRL 49653 was hybridized to a cDNA microarray, and results for selected genes were confirmed by Northern analysis. PPAR-gamma in uninfected and H. pylori-infected human gastric tissue was quantified by real-time RT-PCR. Results: BRL 49653 significantly (p<0.05) decreased H. pylori-induced inflammation in gerbils (mean 3.1 + 0.3 vs. 5.0 + 0.4; BRL vs. control) and mice (0.2 + 0.2 vs. 2.0 + 0.5; BRL vs. control). BRL 49653 and 15d-PGJ₂ transcriptionally activated a PPAR luciferase reporter in AGS cells, and significantly (p<0.03) inhibited H. pylori-induced IL-8 release. PPAR-gamma activation also inhibited H. pylori-stimulated NF-kappaB signaling through prevention of IkappaB phosphorylation. PPAR-gamma-regulated transcripts included those involved in cell adhesion, complement activation, and response to oxidative stress. PPAR-gamma was expressed within uninfected and infected human gastric mucosa. Conclusions: PPAR-gamma activation inhibits NF-kappaB-mediated cytokine release induced by H. pylori in vitro and attenuates mucosal injury in two rodent models of H. pylori-induced gastritis, and candidate PPAR-gamma-regulated genes that may be required for these events were identified by microarray and confirmed by Northern analysis. Since the inflammatory response to H. pylori likely is dependent upon NF-kappaB signaling, PPAR-gamma may regulate bacterial:epithelial cell interactions related to pathogenesis and based upon its constitutive presence within gastric mucosa, ligands for PPAR-gamma have promise as novel therapeutic agents for the treatment of H. pylori-colonized patients and elimination of pathologic outcomes that arise within the context of microbially-initiated inflammatory states.

A mutation in the murine Galgt2 gene resulting in a tissue-specific switch in gene expression program

Jill Johnsen, Gallia Levy, David Siemieniak, David Ginsburg

University of Michigan, Ann Arbor, Michigan

von Willebrand factor (VWF) is a central protein in hemostasis. Abnormalities in VWF result in von Willebrand disease (VWD), the most common inherited bleeding disorder in humans. Plasma VWF levels vary widely among both normal individuals and VWD patients due, in part, to the action of one or more �modifier� gene(s). In an effort to characterize genetic factors altering VWF levels in the mouse as candidates for human modifier genes, we identified a dominant mutation responsible for low levels of VWF in the inbred mouse strain RIIIS/J as a regulatory mutation in the gene encoding an N-acetylgalactosaminyltransferase, GALGT2 (Mohlke et al., 1999). This mutation results in a tissue-specific switch from the highly gut epithelial specific expression pattern observed in most mouse strains (as well as in humans) to the RIIIS/J vascular endothelial specific expression pattern with loss of gut expression. This new Galgt2 expression program overlaps with VWF endothelial expression, resulting in altered glycosylation of VWF and accelerated clearance from plasma. The precise cis-regulatory DNA sequences responsible for this transcriptional switch have not yet been identified. These sequences must lie within the 300kb candidate region on chromosome 11 defined by our previous genetic mapping studies. A transgene derived from a bacterial artificial chromosome (BAC) within this region is sufficient to restore the wild-type GALGT2 intestinal epithelial expression pattern to deficient mice, suggesting that the critical elements lie within the 200kb encoded by this BAC. We are using a phylogenetic footprinting approach in an effort to identify the critical regulatory elements within this interval. Comparison of the publicly available Galgt2 genomic sequence from the C57BL6/J mouse to the sequence of its human ortholog, B4GALT, identified a single, 62% conserved, 270bp noncoding element in intron 4. However, sequence analysis of this segment from RIIIS/J genomic DNA revealed no differences compared to C57BL6/J. Analysis of Galgt2 flanking sequences has also failed to identify any large segments of potential regulatory sequence shared with the VWF gene. Current efforts are focused on direct comparison of the RIIIS/J and C57BL6/J sequences throughout this candidate genomic region. A block of highly divergent sequence (2-3% variation between RIIIS/J and C57BL/6/J) has been identified spanning a segment beginning 25kb upstream of exon 1 and extending over 10kb into intron 1. Differences include single base pair substitutions, short deletions and insertions, and a 120bp imperfect duplication present in RIIIS/J and absent in C57BL6/J. This highly divergent RIIIS/J Galgt2 allele and its endothelial expression pattern are shared by at least 3 other inbred mouse strains, suggesting a common origin in a single, ancient mouse founder allele. Candidate mutations identified by our analysis will be shuttled into the C57BL/6J Galgt2 BAC described above and tested for a shift in expression pattern when introduced as transgenes into the appropriate mouse genetic background. Continued investigation of this unique regulatory mutation should provide important new insights into the requirements for endothelial-specific gene expression, as well as the more general mechanisms underlying tissue-specific gene regulation.

MCP-1 alters astrocyte survival and induces TNF-alpha in microglia: Implications for HIV dementia

Yogeshwar Kalkonde, Randolph Shelton, William Pate, Lisa Adams, Jason Schaefer, Bysani Chandrasekar, Seema Ahuja, and Sunil Ahuja

Veterans Administration Research Center for AIDS and HIV-1 infection and University of Texas health Sciences Center, San Antonio, TX 78229.

Monocyte chemoattractant protein-1 (MCP-1), a potent chemoattractor and activator of mononuclear phagocytes (MP) has been implicated in the pathogenesis of several inflammatory brain diseases, including HIV-1 associated dementia (HAD). We recently demonstrated a role for MCP-1 in HAD. The MCP-1 allele (-2578G) was linked to elevated serum MCP-1 levels and enhanced MP recruitment to inflammatory sites and was associated with 5- fold higher risk of developing HAD. We hypothesized that MCP-1 influences the activation/apoptotic state of microglia and astrocytes, important cellular mediators of inflammatory brain damage. In a murine system, CC chemokine receptor 2, the cognate receptor for MCP-1 was expressed on astrocytes and microglia. MCP-1 at a concentration of 50ng/ml decreased the survival of murine astrocytes whereas at higher concentrations (200ng/ml) it moderately increased astrocyte survival and was associated with enhanced nuclear translocation of NF-kappaB. MCP-1 did not significantly alter astrocyte proliferation. TNF-alpha, a neurotoxic cytokine, was induced by MCP-1 in microglia cells. Collectively, these results indicate that MCP-1 may play a detrimental role in the nervous system during inflammation. Studies are underway to evaluate mechanism(s) by which MCP-1 alters astrocyte survival and the effect of chemokines on neuronal survival in a co-culture system of astrocytes, microglia and neurons.

Therapeutic targeting of prohibitin in adipose tissue vasculature

Mikhail G. Kolonin¹, Pradip K. Saha², Lawrence Chan², Renata Pasqualini¹, and Wadih Arap¹

¹The University of Texas M.D. Anderson Cancer Center, Houston, Texas. ²Baylor College of Medicine, Houston, Texas.

Obesity is an increasingly prevalent disease in modern society. Despite major progress in the understanding of the molecular mechanisms leading to obesity, no safe and effective treatment has yet been found. Here, we report a novel strategy for anti-obesity treatment. We used phage display to isolate a peptide motif (sequence CKGGRAKDC) that homes to the white fat vasculature in vivo. We show that in white fat the CKGGRAKDC peptide associates with prohibitin, a multifunctional membrane protein, and establish a previously unknown function of prohibitin as an adipose vascular marker. Targeting a pro-apoptotic peptide to prohibitin in the adipose vasculature caused a marked apoptosis-mediated resorption of white fat. The targeted ablation of white adipose tissue, associated with increased lipid metabolism, resulted in rapid weight loss without detectable adverse effects. This study shows that accumulation of adipose tissue can be controlled by targeted destruction of blood vessels and opens a new approach to development of drugs for specific treatment of obesity.

Enhanced ethanol sensitivity in adenylyl cyclase knockout mice

James W. Maas Jr.¹, Yuki Izumi², Charles Zorumski², and Louis J. Muglia¹

Departments of ¹Pediatrics and Molecular Biology and Pharmacology & ²Psychiatry, Washington University School of Medicine, St. Louis, MO.

The cyclic AMP (cAMP) and calcium signaling pathways play important roles in the function of central nervous system (CNS) neurons. Consequently, the calcium stimulated isoforms of adenylate cyclase -- adenylate cyclase 1 (AC1) and adenylate cyclase 8 (AC8) -- which convert ATP to cAMP in response to elevations in intracellular calcium concentration, are positioned to play a key role in CNS function and response to drugs of abuse. Ethanol, the most common drug of abuse, has two known effects on neurons -- inhibition of the excitatory NMDA-type glutamate receptor (NMDAR) and potentiation of the inhibitory GABA_A receptor. We hypothesize that mice with a genetic deletion of AC1 and/or AC8 will show an increased sensitivity to the behavioral effects of ethanol due to the role of AC1 and AC8 in signaling downstream of the NMDA receptor. To test this hypothesis, we performed loss-of-righting-reflex assays after ethanol administration in wildtype (WT) mice or mice deficient for AC1 and/or AC8. Mice missing both AC1 and AC8 (DKO) display a markedly increased sensitivity to the sedative effects of ethanol, exhibiting a nearly two-fold increase in sleep time following ethanol administration, as compared to WT mice. Mice missing AC1 alone (AC1 KO) show a 1.5 fold increase in sleep time compared to WT mice, while mice missing AC8 alone (AC8 KO) do not differ from WT mice in ethanol-induced sleep time. In another measure of ethanol sensitivity, the two-bottle ethanol preference test, DKO and AC8 KO mice, but not AC1 KO mice, consumed significantly less ethanol than WT mice. The blood alcohol levels after administration of ethanol do not differ from WT in any of the genotypes, indicating that loss of AC1 and/or AC8 does not alter the metabolism of ethanol. There was no difference in the sleep time of DKO and WT mice following administration of pentobarbital, a GABA-mimetic agent, or ketamine, a NMDAR antagonist, indicating that the enhanced sensitivity of the DKO mice to ethanol is not due to enhanced sensitivity to GABAergic potentiation or NMDAR antagonism by ethanol. Electrophysiologic recording from hippocampal slices demonstrated no differences between WT and DKO slices in the ethanol-mediated inhibition of the NMDAR component of the field (EPSP) potential. Our results indicate that AC1 and AC8 are critical determinants of ethanol sensitivity and that the mechanism is distinct from both the GABA potentiating effects and the NMDAR antagonism of ethanol. Furthermore, our results indicate that AC1 and AC8 are not redundant in their function, but in fact mediate distinct components of the response to ethanol.

Bradykinin and Bradykinin 1-5 Inhibit Thrombin-Induced Platelet Aggregation in Humans

Hector A. Malave,¹ Italo Biaggioni,² Douglas E. Vaughan,¹ Laine J. Murphey,² and Nancy J. Brown²

Divisions of Cardiovascular Medicine¹ and Clinical Pharmacology,² Vanderbilt University Medical Center, Nashville, TN.

Background: Bradykinin 1-5, the major stable metabolite of bradykinin, is formed by the proteolytic action of angiotensin converting enzyme. Previous studies in vitro and in animals suggest that bradykinin 1-5 possesses biological activity. This study tests the hypothesis that bradykinin 1-5 affects endothelial vasodilation, fibrinolysis or platelet aggregation in human subjects. Methods and Results: Graded doses of bradykinin (47-377pmol/min) and bradykinin 1-5 (47-18,850pmol/min) were infused in the brachial artery in random order in 25 healthy subjects. Forearm blood flow (FBF) was measured using strain-gauge plethysmography and simultaneously obtained venous and arterial plasma samples were analyzed for tissue plasminogen activator (t-PA) antigen. In 7 patients, alpha-thrombin-induced platelet aggregation was measured in platelet rich plasma obtained from antecubital venous blood at baseline, and during bradykinin and bradykinin 1-5 infusions. Bradykinin caused significant dose-dependent increases in FBF and net t-PA release (from 3.6+0.4 to 21.0+2.2 mL/min/100 mL and from -1.9+1.1 to 71.9+16.9 ng/min/100 mL at the 377 pmol/min dose, respectively, P<0.001). Bradykinin 1-5 did not affect either FBF (P=0.13) or net t-PA release (P=0.46) at concentrations more than 1500 times higher than baseline values. In contrast, both bradykinin and bradykinin 1-5 inhibited alpha-thrombin-induced platelet aggregation (P<0.01 versus baseline). Bradykinin and bradykinin 1-5 did not affect thrombin receptor activating peptide-induced platelet aggregation. Conclusions: This study is the first to demonstrate biological activity of bradykinin 1-5 in humans. The data are consistent with the hypothesis that bradykinin and bradykinin 1-5 inhibit alpha-thrombin-induced platelet aggregation by preventing the cleavage of the thrombin receptor and liberation of thrombin receptor activating peptide.

In vivo genetic manipulation of SLP-76 in T lymphocytes

Jonathan S. Maltzman, Peggy S. Myung, and Gary A. Koretzky

Abramson Family Cancer Research Institute, Departments of Medicine and Pathology, University of Pennsylvania School of Medicine, Philadelphia, PA.

The adaptor protein SLP-76 is required for normal T lymphocyte development, mast cell function and lymphatic versus blood vessel specification. SLP-76 is expressed in T cells, mast cells, NK cells, platelets and macrophages but not in B lymphocytes. Traditional homologous recombination has been used to generate SLP-76 deficient mice resulting in an absolute blockade in thymic development and a lack of peripheral T lymphocytes. Transgenic reconstitution of the T cell compartment with mutant forms of SLP-76 rescues this defect to various extents depending on the exact nature of the mutation employed. Transgenic animals have peripheral T lymphocytes that show differential requirements for specific domains of SLP-76. However, conclusions about the role of SLP-76 in peripheral T cells from these mice is complicated by the selective pressures on thymic development in the absence of normal SLP-76 expression. SLP-76 has been shown to be critical for signaling in a cell line model of mature T cells and previous work has demonstrated that SLP-76 protein levels are regulated in peripheral T cells in an activation dependent fashion. Developing a system to modulate expression of wild type or mutant variants of SLP-76 in mature primary T cells after normal thymic development has occurred will be invaluable in the understanding of SLP-76 function. We are currently developing two complementary bitransgenic approaches to achieve this goal. The first utilizes a tetracycline inducible system which will allow for the titration of SLP-76 levels in T lymphocytes. In the second system, SLP-76 has been flanked by loxP sites and can be excised by use of a drug-inducible Cre recombinase. By utilizing these genetically modified mice, we will be able to address the role of SLP-76 in primary T cell survival and function.

Anti-Ro Autoimmunity in Human SLE Begins By Targeting an Epitope Which Cross-Reacts with the Epstein-Barr Virus Nuclear Antigen-1

Micah T. McClain, John B. Harley, and Judith A. James

Oklahoma Medical Research Foundation, University of Oklahoma Health Sciences Center, and US Department of Veterans Affairs Medical Center Oklahoma City, OK.

In order to examine potential immunologic links between humoral autoimmunity and infectious agents, this study seeks to analyze the earliest autoimmune specificities available, including those which have been shown to commonly be present for years before the diagnosis of SLE, such as anti-60kD Ro. Serum samples from 24 lupus patients, taken both before and after the onset of anti-Ro antibodies and as early as six years prior to the onset of SLE, were tested for reactivity with the 531 maximally overlapping octapeptides of 60kD Ro by a modified solid phase ELISA. Nine of these patients have an early serum sample in which only one epitope of Ro is bound. Initial serum binding to octapeptides of Ro in eight of the nine patients consistently targets the amino terminal sequence TKYKQRNGWSHK (aa 169-180) before spreading to include additional regions of the protein at later serum dates. Affinity-purified antibodies to the Ro 169-180 peptide comprise greater than 95% of the anti-Ro reactivity in these early serum samples. Interestingly, these affinity-purified anti-peptide antibodies show strong reactivity to the Epstein-Barr virus nuclear antigen 1 (EBNA-1) by immunoblot, and this anti-viral reactivity is inhibited by pre-incubation with purified Ro 169-180 peptide. Conversely, we have demonstrated that the binding of anti-Ro positive SLE patient serum to the Ro 169-180 peptide is inhibited by recombinant EBNA-1. Furthermore, we have mapped the cross-reactive epitope on EBNA-1 utilizing recombinant EBNA-1 fragments and octapeptide assays and show that this EBNA-1 specificity does not become antigenic until the same date that anti-Ro 169-180 antibodies appear. Thus, the earliest major humoral epitope of 60kD Ro in systemic lupus shows strong cross-reactivity with a human pathogen that has been associated with SLE, providing further reason to suspect a pathologic and/or etiologic role for EBV in SLE.

The Impact of Medicare Coverage on Basic Clinical Services for Previously Uninsured Adults

J. Michael McWilliams, Alan M. Zaslavsky, Ellen Meara, John Z. Ayanian

From the Division of General Medicine and Primary Care, Department of Medicine, Brigham and Women's Hospital and Harvard Medical School (JMM, JZA); Department of Health Care Policy, Harvard Medical School (JMM, AMZ, EM, JZA), Boston, MA; and National Bureau of Economic Research (EM), Cambridge, MA.

Context: Uninsured adults receive less appropriate care and suffer more adverse health consequences than insured adults. Objective: To assess the differential impact of gaining Medicare coverage on the use of basic clinical services by previously insured and uninsured adults. Design and Setting: Household survey data from the nationally representative Health and Retirement Study were used to analyze differences in the receipt of basic clinical services by adults in 1996 and 2000, before and after becoming eligible for Medicare at age 65. Participants: 2,203 adults aged 60 to 64 in 1996 classified as continuously uninsured, intermittently uninsured, or continuously insured in 1994 and 1996. Main Outcome Measures: Individuals� reports of receiving cholesterol testing, mammography, prostate exams, and treatment of arthritis and hypertension in the prior two years. Results: The absolute difference in cholesterol testing between continuously uninsured and insured adults was significantly reduced after Medicare eligibility (35.4% vs. 17.7%, p=0.003), and the reduction was substantially greater among those with hypertension or diabetes than among other adults (29.8% vs. 7.7%, p=0.048). Differences in use were similarly reduced after Medicare eligibility for mammography in women (30.3% vs. 15.0%, p=0.04) and prostate exams in men (45.2% vs. 20.0%, p=0.01). Continuously uninsured adults with arthritis reported significantly greater increases in arthritis-related medical visits and limitations of activity than continuously insured adults after Medicare eligibility, but not greater increases in arthritis treatments. Among adults with hypertension, differences in the use of anti-hypertensive medications between continuously uninsured and insured adults were essentially unchanged after Medicare coverage. Conclusions: Medicare coverage of previously uninsured adults substantially increased their use of basic clinical services but not medications, thereby demonstrating that gaining coverage reduces unmet needs of uninsured adults for covered services. An affordable option for near-elderly uninsured adults to purchase Medicare coverage may have similar effects.

ADAMTS13 Mutations Identified in Familial TTP Patients Result in Loss of VWF-Cleaving Protease Activity

David Motto, Gallia Levy, Beth McGee, Han-Mou Tsai, and David Ginsburg

University of Michigan, Ann Arbor, Michigan

Thrombotic thrombocytopenic purpura (TTP) is a life threatening systemic illness often associated with a decreased capacity of a proteolytic activity in plasma to cleave the blood clotting protein von Willebrand Factor (VWF). Recently the gene responsible for the familial form of TTP was identified as ADAMTS13, a novel member of the ADAMTS metalloproteinase family; and to date, 25 TTP-causing ADAMTS13 mutations have been found. Here we report 10 additional mutations, bringing the total number of TTP-causing ADAMTS13 mutations to 35. In order to explore the structure and function of ADAMTS13 we prepared full-length ADAMTS13 cDNA expression vectors. Additionally, site directed mutagenesis was used to introduce into the expression vectors each of nine missense mutations we identified previously in patients with familial TTP (H96D, R102C, T196I, R398H, R528G, R692C, C951G, C1213Y, and R1219W). These wild-type and mutant constructs were transiently transfected into CHO and COS-1 cells, and cell lysates and conditioned media were analyzed by SDS-PAGE. Western blotting was performed using antisera raised against a peptide derived from the predicted ADAMTS13 sequence. VWF-cleaving protease activity was measured using purified VWF as a substrate as described previously. Recombinant wild-type ADAMTS13 has an apparent molecular mass of 160-kD to 200-kD with similar electrophoretic mobilities observed for all nine mutants studied. Recombinant wild-type ADAMTS13 exhibits VWF-cleaving protease activity similar to that observed in normal human plasma. In contrast, all nine missense mutations associated with familial TTP exhibit markedly reduced activity ranging from 5-25% of wild-type. In conclusion these findings demonstrate the functional activity of recombinant wild-type ADAMTS13 and confirm its identity with the VWF-cleaving protease observed in normal plasma. These data also confirm the classification of all nine missense substitutions studied here as authentic familial TTP disease-causing mutations. Finally, some or all of these ADAMTS13 mutants may retain low or trace levels of VWF-cleaving activity, consistent with the possibility that complete ADAMTS13 deficiency may be lethal. Ongoing studies of mice genetically deficient for the ADAMTS13 gene will directly test this hypothesis and may provide further insight into the pathogenesis of TTP.

NEURONAL ATP7A, PEPTIDYLGLYCINE alpha-AMIDATING MONOOXYGENASE (PAM) AND AMIDATED NEUROPEPTIDES IN THE MURINE MODEL OF MENKES DISEASE

Mark J. Niciu, Richard E. Mains and Betty A. Eipper

The University of Connecticut Health Center, Department of Neuroscience, Farmington, CT.

Menkes disease (MD) is an X-linked copper-deficiency disorder resulting from mutations at the ATP7A locus, which encodes a P-type ATPase responsible for supplying reduced copper [Cu(I)] to the lumen of the secretory pathway. The mottled brindled (Mo^Br) mouse, housing a two amino acid deletion in Atp7A, serves as the closest murine model of MD. The neurological manifestations of MD have been attributed to the diminished function of cuproenzymes like cytochrome c-oxidase, dopamine-beta-monooxygenase and peptidylglycine alpha-amidating monooxygenase (PAM). PAM contains two enzymatic domains that perform the sequential reactions of C-terminal alpha-amidation, converting glycine-extended neuropeptides into bioactive amidated products. Many amidated neuropeptides display critical regulatory roles in neuronal proliferation, survival and differentiation, e.g. pituitary adenylate-cyclase activating polypeptide (PACAP) and neuropeptide Y (NPY). We therefore explored the possibility that attenuated delivery of Cu(I) to the lumen of the secretory pathway in hemizygous affected male (Mo^Br/y) mice impairs production of amidated neuropeptides and may contribute to the neurological degeneration observed in MD.

In the central nervous system, Atp7A is preferentially expressed in the cerebellum at P4. Atp7A expression is also prominent in the pituitary gland, adrenal gland and kidney in postnatal and adult mice. Atp7A immunostaining was observed in the trans-Golgi network (TGN) area of neurons, resembling the TGN marker, TGN38, and a less robust signal in GFAP-positive astrocytes. The level of amidated cholecystokinin (CCK) is depressed in P4 and P12 Mo^Br/y neocortex and P4 Mo^Br/y hypothalamus relative to age matched wild-type mice. A severe depression of amidated PACAP38 radioimmunoreactivity is also detected in P12 Mo^Br/y neocortex and cerebellum. Research supported by DK-32949 and MSTP GM08607.

Genetic anthropology of the colorectal cancer susceptibility allele APC I1307K

Bethany L. Niell^{1, 2}, Jeffrey C. Long³, Gad Rennert⁴, Stephen B. Gruber^{1, 2, 3}

¹Department of Internal Medicine, Division of Molecular Medicine and Genetics, University of Michigan Medical School, Ann Arbor, MI. ²Department of Epidemiology, University of Michigan School of Public Health and ³Department of Human Genetics, University of Michigan Medical School, ⁴CHS National Cancer Control Center, Carmel Medical Center, Haifa, Israel

The APC I1307K allele is found in 6% of the Ashkenazi Jewish population and in 1-2% of Sephardi Jews, conferring a relative risk of 1.5-2.0 for colorectal cancer (CRC). The ethnic distribution of the I1307K polymorphism can be better understood by examining its genetic anthropology and allele age. From the Molecular Epidemiology of Colorectal Cancer study, a population-based case-control study in Israel, we selected 83 matched pairs in which one or both of the pair carries I1307K. Genotyping was completed on all 166 individuals at three microsatellites and two SNPs. Haplotypes were statistically constructed using PHASE and by genotyping homozygous carriers. Single marker age estimates for I1307K were calculated using the approach described by Risch et al. A common progenitor haplotype was identified in 76/83 (91.6%) of the I1307K carriers, and was observed in individuals of Ashkenazi, Sephardi, and Arab descent. This haplotype spans across APC I1307K from the centromeric marker D5S135 to the telomeric marker D5S346, constituting a physical distance of approximately 500,000 base pairs. The most recent common ancestor of present day I1307K alleles existed between 87.9 and 118 generations ago, or approximately 2200 to 2950 years ago. These age estimates indicate that I1307K existed close in time to the beginning of the Jewish Diaspora, therefore explaining the presence of I1307K in non-Ashkenazi populations. The high prevalence of I1307K within the Ashkenazim is likely due to a subsequent founder effect. This research suggests that future studies of APC I1307K should extend beyond the Ashkenazim.

Protection from nitric oxide enhances microbial virulence

Jesse C. Nussbaum^1,2, Marisol de Jes�s-Berr�os², Limin Liu^3,6, Gary M. Cox^2,3, Jonathan S. Stamler^3,4,6, Joseph Heitman^2,3,5,6

School of Medicine¹ (Medical Student Research Fellow), Departments of Molecular Genetics and Microbiology², Medicine³, Biochemistry⁴, and Pharmacology and Cancer Biology⁵, Howard Hughes Medical Institute⁶ Duke University Medical Center, Durham, NC

Oxidative and nitrosative challenge are key components of the innate immune defense against intracellular pathogens. Microbial pathogens are known to employ a number of enzymes such as catalase and superoxide dismutase to detoxify reactive oxygen intermediates and counteract the toxic host environment during infection. It is also known that nitric oxide (NO) produced by macrophages serves not only as an important signaling molecule but also as an antimicrobial toxin; however, the existence and physiological importance of microbial enzymes that counteract nitrosative challenge has not been determined. Two enzymes, flavohemoglobin (FHB1) and glutathione-NO reductase (GNO1), were recently identified in Saccharomyces cerevisiae and shown to detoxify nitrosative radicals. These enzymes have orthologs in Cryptococcus neoformans, an opportunistic fungal pathogen and model organism for studying intracellular microbial infection. The FHB1 and GNO1 genes in C. neoformans were individually disrupted by site-directed mutagenesis; when compared to the wild-type strain, the fhb1 mutant strain was hypersensitive to NO in vitro, had attenuated virulence in a murine infection model, and grew poorly in an alveolar macrophage cell line. These effects were directly attributable to the absence of the NO-scavenger flavohemoglobin and were confirmed by an independent insertional mutagenesis screen for genes that confer NO resistance; two mutant strains have been isolated using this functional screen, each with a unique FHB1 mutation. Our studies establish for the first time the relevance of agents that protect against nitrosative stress in pathogenesis. Moreover, because the flavohemoglobin gene is conserved in many pathogens, but not in humans, it represents an ideal target for antimicrobial therapy.

Belly spot and tail: Identification and Characterization of a Mouse Minute

Edward R. Oliver, Thomas L. Saunders, Susan A. Tarl�, and Tom Glaser

University of Michigan, Ann Arbor, MI

Diamond-Blackfan anemia is caused by mutations in the human riboprotein gene RPS19, one of 80 genes encoding proteins that compose the mammalian ribosome. Little is known about riboprotein mutations in vertebrates. These mutations in Drosophila melanogaster, termed Minutes, are abundant and have been extensively studied. Heterozygous Minute flies display shorter bristles and a profound delay in development. Flies homozygous for the Minute mutations die during the larval stage. This interesting class of mutants was instrumental in elaborating the concepts of cell autonomy, compartmental development, and cell competition.

Belly spot and tail (Bst) is a semidominant, homozygous lethal mutation that spontaneously arose at the Jackson Laboratory. Bst/+ heterozygotes display a white ventral midline spot, skeletal and vertebral anomalies (kinked tail), a reduction in retinal ganglion cell (RGC) number, overall retinal disorganization, and retinal neovascularization. We mapped Bst to a 0.5cM genetic interval on mouse chromosome 16 and identified a mutation in Rpl24, a gene encoding a ribosomal protein component of the large ribosomal subunit.

The Bst mutation deletes 4 bp from the first intron splice branchpoint of Rpl24. It impairs splicing of the first and second exons, resulting in premature truncation of the L24 peptide. Approximately 25% of Bst transcripts are correctly spliced, pointing to a partial loss of function mechanism. The defect in Rpl24 leads to a relative block in rRNA processing. Thus, L24 deficiency impairs ribosome biogenesis.

Bst causes a growth disadvantage at both the cellular and organismal levels. Bst/+ MEFs (mouse embryonic fibroblasts) have a decrease in rates of protein synthesis and cell proliferation compared to wildtype cultures. The slower growth rate is due to a prolonged G₁ phase of the cell cycle. The differential growth rate is evident in chimeras derived from ROSA26 (lacZ+) ES cells and Bst/+ or +/+ blastocysts. The ROSA26 contribution is significantly greater in Bst/+ ROSA26 adult chimeras compared to +/+ ROSA26 adult chimeras. This cell competition is more pronounced in certain tissues, suggesting a mechanism for tissue-specific findings in the Bst/+ mouse and Diamond-Blackfan anemia patients.

Cardiac-specific Gain-of-function Mutation in the Natriuretic Peptide A Receptor is Associated With Improved Diastolic Function in Presence and Absence of Pressure Overload Hypertrophy

Allison M. Pritchett¹, Maria L. Valencik², John A. McDonald², Gerald E. Harders¹, and Margaret M. Redfield¹

¹ Mayo Clinic Foundation, Rochester, MN. ² Salt Lake City Veterans Administration Medical Center, Salt Lake City, UT.

Background: Few therapies are known to enhance diastolic function. Systemic administration of natriuretic peptides (NP) improves diastolic function. However, this may be due to effects on load, rather than a direct myocardial action. We generated mice with a cardiac-specific (alpha-MHC promoter) gain-of-function mutation (GoF, HCAT E974A) in the natriuretic peptide A receptor (NPRA) resulting in myocyte-specific constitutive NPRA activation. We hypothesized that GoF mice would exhibit enhanced diastolic function and develop less diastolic dysfunction with pressure-overload. Methods: GoF mice (n=12) and their wildtype (WT -/-, n=18) littermates underwent abdominal aortic banding to generate pressure-overload hypertrophy. LV structure and function (echo, catheterization, autopsy) were assessed three weeks after banding. Non-banded GoF (n=15) and WT -/- (n=13) groups were controls. Results: See table (mean + SD): With aortic banding, both GoF and WT mice developed LV hypertrophy. While load and systolic function were similar between WT and GoF mice without and with banding, the time constant of LV relaxation (Tau) was decreased in GoF mice as compared to WT indicating improved diastolic function. Conclusion: Chronic cardiac-specific NPRA activation via this GoF mutation enhanced diastolic function in mice and markedly attenuated diastolic dysfunction associated with pressure-overload hypertrophy. These data suggest that NP improve diastolic function via a direct myocardial effect.

Variable HCAT E Control HCAT E Banded

GoF (+/-) Littermate (-/-) P Value GoF (+/-) Littermate (-/-) P Value

LV End-diastolic dimension, mm 3.5+0.2 3.4+0.4 0.45 3.7+0.3 3.5+0.2 0.03

Fractional shortening, % 36+5 34+4 0.33 33+4 34+3 0.89

LV Mass, g 0.09+0.02 0.09+0.02 0.55 0.12+0.03 0.11+0.01 0.49

Heart Rate (anesthesia), bpm 450+60 398+45 0.04 409+72 399+83 0.76

Max LV Pressure, mmHg 94+14 84+14 0.25 96+15 86+12 0.07

End-diastolic LV Pressure, mmHg 6+6 5+6 0.58 7+8 8+5 0.60

Tau, ms 13+2 16+3 0.04 16+4 21+5 0.04

Heart weight / Body weight, mg/g 4.2+0.3 4.1+0.3 0.36 4.8+1.0 4.5+0.5 0.30

The Crumbs3/PALS1 Complex Participates in Mammalian Epithelial Polarization

Michael Roh¹, Shuling Fan^2,3, Albert Liu², and Ben Margolis^1,2,3

¹Departments of Biological Chemistry and ²Internal Medicine, ³Howard Hughes Medical Institute, University of Michigan, Ann Arbor, MI.

In Drosophila, the Crumbs/Stardust/Discs Lost complex is required during the establishment of polarized epithelia. Embryos that lack a component of this complex or overexpress Crumbs exhibit defects in epithelial morphogenesis. We recently cloned a novel mammalian epithelial Crumbs isoform, Crumbs3 (CRB3). CRB3 exists in a complex at tight junctions (TJs) with Pals1 and PATJ, the mammalian homologues of Stardust and Discs Lost, respectively. Here, we observe that overexpression of CRB3 leads to delayed TJ formation in MDCK epithelial cell monolayers and disruption of polarity in MDCK cysts cultured in collagen. Both phenomena require the last four residues of CRB3. Next, we expressed, in MDCK cells, a dominant negative Myc-Lin-2/Pals1 chimeric protein, where the PDZ domain of Lin-2 was replaced with that of Pals1. TJ and apical polarity defects were also observed in these cells. Collectively, this suggests that the CRB/Pals1 interaction is important for formation of TJs and polarized epithelia. These results provide insight into the function of the mammalian Crumbs complex during TJ formation and epithelial polarization.

Interferon-gamma receptor 2 (IFNgammaR2): Molecular localization and functional characterization of regions regulating accumulation, distribution and IFNgamma response

Sergio D. Rosenzweig and Steven M. Holland

Laboratory of host Defenses, National Institute of Allergy and Infectious Diseases, NIH, Bethesda, MD.

IFNgamma responsiveness regulates immune function ranging from mycobacterial susceptibility to autoimmunity to immune control of tumors. However, the intrinsic regulation of the function of the IFNgamma receptor (IFNgammaR) complex is still largely unknown. The IFNgammaR is composed of IFNgammaR1 and IFNgammaR2; IFNgammaR1 is constitutively expressed on all nucleated cells, while IFNgammaR2 membrane display is tightly regulated. We identified a patient with a mutation in the transmembrane domain of IFNgammaR2 (791delG), leading to autosomal dominant repression of IFNgamma signaling. This mutant IFNgammaR2 had a novel subcellular distribution without membrane expression. The finding of a membrane-excluded receptor with a dominant effect focused our attention on the transmembrane and membrane-proximal domains of IFNgammaR2. To study this domain and the factors leading to control of the IFNgammaR response, we created a series of labeled IFNgammaR1 and IFNgammaR2 molecules, into which we introduced specific mutations. The LI residues at 255-256 control IFNgammaR2 membrane accumulation and IFNgamma response: their deletion (255delta2) or alanine substitution (LI255-256AA) leads to plasma membrane overaccumulation and a dominant gain of function for IFNgamma signaling. In contrast, introduction of a stop immediately before this region (254X) leads to overaccumulation and a robust dominant negative effect. Therefore, by investigation of these natural (791delG) and created (LI255-256) mutants, we have shown that IFNgammaR2 receptor display is indeed the bottleneck in IFNgammaR responsiveness. The creation of dominant gain of function IFNgammaR2 mutants indicates that these pathways can be exploited to enhance host immune response, which may be of value in the treatment of intracellular infections and cancer. However, receptor display is neither necessary nor sufficient for dominant inhibition of IFNgamma signaling. These novel mutations indicate that further critical cytoplasmic components in the IFNgamma signaling pathway remain to be identified.

Bone Marrow Stromal Cells Express Two Distinct Splice Variants of ER-alpha that Are Regulated by Estrogen

A. Sanyal, B. L. Riggs, T. C. Spelsberg, and S. Khosla

Mayo Clinic, Rochester, MN, USA

Estrogen plays a critical role in bone metabolism in both sexes. While the major action of estrogen is to inhibit bone resorption, it is now clear that early osteoblastic (or stromal) cells are a target for estrogen action, mediating the effects of estrogen on bone formation as well as resorption. However, little is known about the expression or regulation of the estrogen receptor (ER)-alpha in these cells. The expression of ER-alpha is regulated by a complex set of promoters and ER-alpha splice variants are presents in different tissues. Thus, we sought to define the ER-alpha splice variants and their regulation by estrogen in the mouse bone marrow stromal cell line, ST-2, which can be induced to differentiate into mature osteoblasts. ST-2 cells expressed the mRNAs and proteins for both the 66 and 46 kD forms of ER-alpha; the latter lacks the AF-1 domain and can transduce estrogen signaling in some tissues, while serving as a dominant negative receptor in others. Using primers specific for each of the five 5�-untranslated exons of ER-alpha, we found that ST-2 cells utilize only the promoters upstream of exons F and C (in contrast to most reproductive tissues, which utilize promoters upstream of virtually all the five exons). Moreover, 17beta-estradiol (10^-8 M) treatment of ST-2 cells markedly diminished levels of the 66 kD (by 78%) as well as the 46 kD (by 86%) ER-alpha proteins, largely through suppression of the transcript arising from the F1 promoter, which contains an ERE half-site and a consensus AP-1 sequence. These data thus indicate that: (1) bone marrow stromal cells express at least two variants of ER-alpha; (2) promoter usage in these cells is more restricted than in classical reproductive tissues; and (3) estrogen down regulates the ER-alpha mRNA and protein in these cells. Our findings also suggest that bone marrow stromal cells may regulate sensitivity to estrogen through one or more of these mechanisms.

Winner, Poster Session Award

Multipotent Adult Progenitor Cells from Bone Marrow Differentiate into Functional Hepatocyte-like Cells

Robert E. Schwartz¹, Morayma Reyes¹, Lisa Koodie¹, Yuehua Jiang¹, Mark Blackstad¹, Troy Lund¹, Todd Lenvik¹, Sandra Johnson¹, Wei-Shou Hu², and Catherine M. Verfaillie^1,3

Stem Cell Institute¹, Department of Chemical Engineering², Department of Medicine³, University of Minnesota, Minneapolis, MN.

We have identified a population of primitive cells in normal human, mouse and rat post natal bone marrow that we have termed Multipotent Adult Progenitor Cell or MAPC that have multipotential differentiation and extensive proliferation potential. MAPC have been shown to differentiate in vitro into most mesodermal and neuroectodermal lineages and in vivo into all embryological lineages. We have recently demonstrated that hepatocyte-like cells can be derived from MAPC in vitro. Rodent MAPCs differentiated into epithelioid cells that expressed hepatocyte nuclear factor-3beta (HNF-3beta), GATA4, cytokeratin 19, transthyretin, alpha-fetoprotein, CK18, HNF-4, and HNF-1alpha in a time dependent manner consistent with hepatocyte development. These cells also acquired functional characteristics of hepatocytes: they secreted urea and albumin, had phenobarbital-inducible cytochrome p450, could take up LDL, exhibited polarization and stored glycogen. We are in the process of examining whether MAPC, or MAPC derived hepatocyte-like cell stem cells or MAPC derived hepatocyte-like cells can rescue an animal model of acute liver failure.

beta-Blockers Do Not Inhibit The Progression Of Systolic Dysfunction In A Transgenic Model Of Heart Failure

Anisha N. Shah, Jennifer Dawson, Erdal Gursoy, Barry London

UPMC Pittsburgh, PA

beta-blockers decrease mortality and sudden death in patients with congestive heart failure. We sought to test whether and by what mechanisms long-term beta-blocker therapy would affect heart failure progression and arrhythmogenesis in a mouse model of heart failure. Subcutaneous sustained release propanolol (5mg over 60 days) pellets were implanted in eight week old female wild-type (WT) FVB and TNF-alpha mice. Echocardiograms were performed after four weeks. We observed that compared to placebo, therapy with propanolol did not reverse the increase in left ventricular end-diastolic dimension (LVEDD) in TNF-alpha mice (LVEDD: TNF/placebo 2.7+0.4, TNF/drug*2.8+0.3, WT/placebo 2.4+0.2, WT/drug*2.3+0.4) Propanolol also caused a decrease in fractional shortening (FS) of similar magnitude in TNF-alpha and wild-type mice (FS: TNF/placebo 34+11, TNF/drug*23+8, WT/placebo 47+5, WT/drug*38�0.1). {n=5, *p<0.05 TNF vs. WT}Thus, in this study, long term beta-blocker therapy was found to not prevent the progression of systolic dysfunction. The precise mechanism(s) by which beta-blockers act to improve mortality and sudden death in humans has not been established. Optical mapping and holter studies to assess the effect of long-term beta-blocker therapy on arrhythmogenesis in this mouse model are ongoing at this time.

Levels of nitrotyrosine, an inflammatory marker generated by nitric oxide derived oxidants, are associated with risk of cardiovascular disease and modulated by statin therapy

Mehdi H. Shishehbor, Noyan Gokce, John F. Keaney, Jr., Marc S. Penn, Dennis L. Sprecher, Joseph A. Vita, and Stanley L. Hazen

Cleveland Clinic Foundation

Context: Formation of nitric oxide-derived oxidants may serve as a mechanism linking inflammation to development of atherosclerosis. Nitrotyrosine, a specific marker for protein modification by nitric oxide-derived oxidants, is enriched in human atherosclerotic lesions and low density lipoprotein (LDL) recovered from human atheroma. Objectives: To determine whether systemic levels of nitrotyrosine correlate with the prevalence of coronary artery disease (CAD) and are modulated by hydroxymethylglutaryl coenzyme A reductase inhibitor (statin) therapy. Design and Setting: A case-control study and an interventional study at two tertiary-care referral centers. Patients: (Case control study) - 208 patients comprised of 100 with established CAD and 108 without clinically evident CAD. (Interventional study) � 35 patients with hypercholesterolemia and no known CAD monitored at baseline and following 12 weeks of atorvastatin (10 mg/d) therapy. Main Outcome Measures: (Case control study) Systemic levels of protein-bound nitrotyrosine, along with other traditional CAD risk factors. (Interventional study) Nitrotyrosine levels, along with lipoprotein and C-reactive protein (CRP) levels. Results: Nitrotyrosine levels were significantly higher among patients with CAD (median 9.1 micromol/mol tyrosine vs. 5.2 micromol/mol; P < 0.001). Patients in the upper quartile of nitrotyrosine levels had higher risk of CAD relative to the lowest quartile (unadjusted odds ratio, 6.1; 95% confidence interval, 2.6 to 14; P < 0.001). In multivariate models adjusting for Framingham Global Risk Score and CRP, upper quartiles of nitrotyrosine remained predictive for CAD risk (odds ratio, 5.4; 95% confidence interval, 1.7 to 11.6; P < 0.001). Statin therapy caused a significant reduction (25%; P<0.02) in nitrotyrosine content of plasma proteins that was similar in magnitude to reductions in total cholesterol and LDL particle number (25% and 29%, respectively; P<0.001 each), yet independent of alterations in lipoproteins and inflammatory markers like CRP. Conclusion: Nitrotyrosine serves as a significant and independent marker of CAD, and is modulated by statin therapy. These results suggest a potential role for nitric oxide-derived oxidants as inflammatory mediators in CAD and may have important implications for atherosclerosis risk assessment, diagnosis and the monitoring of anti-inflammatory actions of statins.

S100A8 and S100A9, two ubiquitous proteins, act as leukocyte repellents through classical CC chemokine receptors

H. Sroussi, J. Berline, J. Palefsky

University of California in San Francisco

The recruitment of immune competent cells from the circulation is a crucial step in developing a localized efficient immune response. It is important to mount such a response when needed and to repress it when necessary to limit deleterious immune activity. Most of the research on the factors controlling leukocytes infiltration has focused on the activity of chemo-attracting cytokines; recent evidence has however pointed to the contribution of chemo-repulsion in the delicate balance controlling localized inflammation. Two members of the S100 calcium binding proteins, S100A8 and S100A9 were isolated from a subtractive library of hairy leukoplakia (HL) against normal tissue. One of the characteristic features of HL is a lack of leukocyte infiltrate and depletion of Langerhans cells. Based on prior reports implicating other S100 proteins such as the murine homologue of the human S100A8 in the modulation of leukocyte migration, we sought to investigate the possible contribution by these two S100 proteins to down regulation of the localized immune response. We conducted under-agarose and Transwell migration assays with freshly isolated peripheral leukocytes. S100 proteins were expressed as GST-tagged proteins in E coli, purified in a glutathione column and thrombin-cleaved. The data demonstrated that S100A8 and S100A9 caused a repulsion of peripheral monocytes and neutrophils down a concentration gradient. Further analysis demonstrated the ability of RANTES and not IL8 or MCP1 to inhibit the effect of the S100 proteins on peripheral monocytes. The functional relationship between RANTES and the two S100 proteins was further investigated with a murine lymphoma cell line transfected with two of RANTES' receptors: the classical chemokine receptors CCR1 and CCR3. Our data showed that despite the lack of a signature cysteine structure, the two S100 proteins act via CCR1 and to a lesser extent CCR3. Additional analysis of the regulation of S100A8's functions demonstrated that like its murine homologue, the effects of S100A8 are regulated by oxidation and are inhibited by the formation of a disulfide bridge between conserved cysteine residues. Taken together, the data support a model in which mucosal epithelial expression of S100A8 and S100A9 create a chemical barrier against inflammatory infiltrates. In this model, the oxidative burst associated with acute inflammation, contributes to the inflammatory process through functional inactivation of the S100 proteins and their fugetactic effects. Signaling of S100A8 and S100A9 through classical chemokines and the high level of expression of these proteins in many inflammatory conditions suggest that they play an important role in regulation of inflammation.

Estrogen enhances efficacy of an oncolytic herpes simplex viral mutant in estrogen receptor positive breast cancer cells

Brendon M. Stiles, Amit Bhargava, Stephen F. Stanziale, Teresa H. Kim, Prasad Adusumilli, and Yuman Fong

Department of Surgery, Hepatobiliary Division, Memorial Sloan-Kettering Cancer Center, New York, New York

Introduction: Estrogen is well known to increase cellular proliferation and decrease apoptotic fractions in estrogen receptor positive (ER+) human breast cancer cells. These effects have been demonstrated to decrease the efficacy of numerous chemotherapeutic agents and of radiotherapy in vitro, mediated in part by upregulation of the anti-apoptotic protein, Bcl-2. Oncolytic Herpes Simplex Virus-1 (HSV-1) mutants selectively replicate in and lyse tumor cells. Because enhanced host cell proliferation and protection from apoptosis increase viral replication and survival, we hypothesized that oncolysis of ER+ breast cancer cells with an HSV-1 mutant, NV1066, would be improved in the presence of estrogen. Methods: NV1066 is a multi-mutated, replication competent HSV-1 that carries a transgene for enhanced green fluorescent protein (eGFP). MCF-7 cells (ER+) were grown either in the presence of 10nM B-estradiol or with vehicle alone and infected with NV1066 at multiplicities of infection (MOI; number of viral particles per tumor cell) between 0.1 and 10.0. Viral replication was determined by plaque assay, cell-to-cell proliferation by FACS analysis of eGFP expression, and cell kill by a lactate dehydrogenase assay. TUNEL and Hoescht staining were used to determine apoptotic fractions, while a Bcl-2 ELISA was used to determine Bcl-2 protein expression. Results: In the presence of estrogen, MCF-7 cells proliferated more rapidly, had lower apoptotic fractions, and higher Bcl-2 levels. Treatment with NV1066 resulted in two mechanisms of cell death, oncolysis and apoptosis. Cells grown with estrogen had significantly lower apoptotic fractions after infection with NV1066 (p < 0.03, t-test). As measured by FACS analysis of eGFP expression, cell-to-cell spread of virus was faster in cell populations grown with estrogen, particularly at an MOI of 0.1. Differences as high as 20-25% were seen in the first few days after infection (p < 0.01, t-test). Cells grown with estrogen also had significantly higher rates of Bcl-2 expression both prior to and after viral treatment. Infection with NV1066 decreased Bcl-2 expression in both groups. By plaque assay, peak viral replication titers were 3.7 and 2.3 fold higher respectively in infected cell populations grown with estrogen at MOIs of 0.1 and 0.5, compared to titers of cells infected in the absence of estrogen (p < 0.05, t-test). Correspondingly, there is a significantly greater cytopathic effect in cells treated with NV1066 in the presence of estrogen. Cell kill with estrogen was 95% and 97% at MOIs of 0.1 and 0.5, compared to 53% and 87% respectively without estrogen (p < 0.001, t-test). A similar phenomenon was also seen in T47D cells (ER+), but not in a hormone receptor negative breast cancer cell line. Both cell lines were sensitive to NV1066 at similar MOIs. Conclusions: In contrast to other in vitro therapies, estrogen enhances efficacy of an oncolytic HSV-1 mutant, NV1066, in the treatment of ER+ breast cancer cells. Therapy with replication-competent HSV-1 mutants may be an attractive option for patients whose tumors remain responsive to the effects of endogenous estrogen. By decreasing Bcl-2 levels in infected cells, NV1066 may sensitize resistant tumors to treatment with chemotherapy or radiation. These findings should influence the timing of viral therapy with respect to selective estrogen receptor modulators clinically. Hormonal milieu is an important consideration in the treatment of hormone-responsive cancers with viral-based therapeutics.

Engineering small diameter arterial prostheses: production of collagen and elastin-mimetic fibers and fabrics

Ben S. Thomas¹, Karthik Nagapudi¹, William T. Brinkman¹, Vince P. Conticello², and Elliot L. Chaikof^1,3

Departments of Surgery¹ and Chemistry², Emory University. Department of Chemical Engineering, Georgia Institute of Technology³.

Finding suitable replacements for the current generation of vascular grafts (ePTFE, Dacron^R, etc.) is critical in the fields of cardiac, plastic, and vascular surgery. Construction of small diameter arterial prosthesis can be accomplished by the mimicry of morphological and physiological characteristics of some or all of the components of the arterial wall. Hence, for our initial assessment, we have targeted collagen and elastin, which are two of the most important components of the arterial wall.

In connective tissues, elastin is present as a network of fibrils that are crosslinked through available lysine residues. Naturally occurring elastin is comprised predominantly of valine (Val), proline (Pro), glycine (Gly), and alanine (Ala). Moreover, these amino acids exist as characteristic repeat sequences of penta-, tetra-, and nanopeptides. Thus, the generation of protein-based polymers containing any or all of the aforementioned peptide sequences provide a useful starting point for the design and development of elastomeric biomaterials.

An elastin-mimetic protein polymer, poly((Val-Pro-Gly-Val-Gly)₄(Val-Pro-Val-Lys-Gly))₃₉, was synthesized by recombinant techniques and modified through the lysine residue to incorporate an acrylate moiety. Fibers (diameter range: 300nm � 1.5micro-m) and fabric samples were prepared by electrospinning and were crosslinked by visible light mediated photoirradiation. The effect of inverse transition temperature, flow rate, and solution concentration on the ensuing fiber morphology were studied. ¹³C solid-state NMR confirmed complete crosslinking with a concomitant increase in modulus and tensile strength. A typical hydrated non-woven fabric exhibited a young modulus of 0.45 MPa and an elongation to break of 115%. These values are commensurate with that of naturally occurring elastin.

It is believed that these elastin networks, when used in conjunction with collagen fiber networks, can lead to the creation of an arterial substitute that would possess good resilience, tensile strength, and durability. In addition, this would provide an ideal biomechanical environment for normal endothelial and smooth muscle cell function. The generation of such engineered crosslinked elastin-mimetic/collagen fabrics will provide a rationale for the design and development of a variety of tissue like scaffolds with well defined mechanical and physiochemical properties, and will significantly enhance the utility of these materials in biomedical applications.

Erythroid Cell Cycle Regulation by GATA-1

John J. Welch¹, Marcin B. Rylski², Gerd A. Blobel¹, J. Alan Diehl³, Lewis A. Chodosh³, and Mitchell J. Weiss¹

¹Hematology, Children�s Hospital of Philadelphia and the University of Pennsylvania, Philadelphia, PA; ²Genetics, University of Warmia and Mazury, Olshtyn, Poland; and ³Abramson Family Cancer Research Institute, University of Pennsylvania, Philadelphia, PA.

GATA-1 is an essential transcription factor for erythroid and megakaryocytic maturation. Mutations in GATA-1 are associated with hyperproliferation of hematopoietic precursors, suggesting a role in cell cycle control. While numerous GATA-1 target genes specifying mature hematopoietic phenotypes have been identified, how GATA-1 regulates proliferation remains unknown. We used a complementation assay based on synchronous inducible rescue of GATA-1� erythroblasts to show that GATA-1 promotes both erythroid maturation and G₁ cell cycle arrest. Molecular studies combined with microarray transcriptome analysis revealed a distinct GATA-1-regulated program of cell cycle control in which numerous growth inhibitors were upregulated and mitogenic genes repressed. GATA-1 inhibited expression of cyclin dependent kinase (Cdk) 6 and cyclin D2, and induced the Cdk inhibitors p18^INK4C and p27^Kip1 with associated inactivation of all G1 Cdks. These effects were dependent on the repression of the c-myc (Myc) protooncogene which occurred within three hours after GATA-1 restoration. Enforced expression of Myc prevented GATA-1-induced cell cycle arrest but had minimal effects on erythroid maturation demonstrating that these two genetic programs can be uncoupled. Chromatin immunoprecipitation (ChIP) studies revealed that GATA-1 binds to the Myc proximal promoter in vivo. Together, our studies suggest a regulatory hierarchy in which GATA-1 directly inhibits Myc transcription to block cell cycle progression. More generally, these results illustrate how GATA-1, a lineage-determining transcription factor, coordinates proliferation arrest with cellular maturation through distinct, but interrelated genetic programs.

Impact of HOXB4 expression levels on cellular growth

Elke Will¹; Zheng Wang²; Daniel Speidel³; Bernhard Schiedlmeier²; Hannes Klump²; Melanie Galla¹; Wolfram Ostertag² and Christopher Baum^1,2

¹Department of Experimental Hematology, Children�s Hospital Medical Center, Cincinnati, OH; ²Department of Hematology and Oncology, Hannover Medical School, Hannover, Germany; ³Department of Tumor Virology, Heinrich-Pette-Institute for Experimental Virology and Immunology, Hamburg, Germany

The ectopic expression of HOXB4 in murine and human hematopoietic cells leads to expansion of the repopulating stem cells and progenitors, but also impairs myeloid differentiation when expressed at high levels (Schiedlmeier et al., 2003). In order to define the HOXB4-dosage optimal for stem cell expansion, a variety of retroviral vectors expressing different levels of HOXB4 was generated. Based on the observation that ectopic HOXB4 expression improves colony-formation in soft agar (Krosl et al., 1998), we used Rat1 fibroblasts to measure HOXB4 activity. While stimulating anchorage independent growth, enforced HOXB4 expression impaired the growth of Rat1 cells in a monolayer in a striktly dosage dependent manner. A growth disadvantage conferred by high amounts of HOXB4 protein was also observed in other murine and human cell lines, implicating a common mechanism of HOXB4 activity which may also impair differentiation and survival of primary hematopoietic cells.

Combined Bisulfite Restriction Analysis (COBRA) of Alu Elements to Detect Demethylation in Leukemia Patients Treated with 5-aza-2�-deoxycytidine (Decitabine)

Allen S. Yang, Guillermo Garcia-Manero, Hagop M. Kantarjian, and Jean-Pierre Issa

UT M.D. Anderson Cancer Center

DNA methylation is believed to play an important role in tumorigenesis, and many loci have been shown to be aberrantly methylated in tumors. Current methods of detecting global DNA methylation use high pressure liquid chromatography or capillary electrophoresis to examine the 5-methylcytosine content of total genomic DNA. This method, however, is labor intensive and requires large amounts of DNA. PCR based methods using bisulfite treated DNA (COBRA, MSP) use less DNA, but can only examine one locus at a time. We have developed a global genome methylation assay that uses COBRA analysis of Alu repetitive elements. Total genomic DNA is bisulfite treated and then a non-specific PCR of Alu repetitive elements is performed. The PCR product, which is a pooled mixture of Alu elements, is then digested with a restriction enzyme whose cutting is dependent upon the methylation status of the Alu element. This method can simultaneously assess the methylation of several thousand Alu elements, and thus assess the DNA methylation status of several thousand loci simultaneously.

This method was used to study samples from patients treated with a demethylating agent, 5-aza-2�-deoxycytidine (Decitabine). Peripheral blood samples from patients with leukemia have been collected before, during and after treatment with Decitabine. Initial studi

Variable	HCAT E Control			HCAT E Banded
Variable	GoF (+/-)	Littermate (-/-)	P Value	GoF (+/-)	Littermate (-/-)	P Value
LV End-diastolic dimension, mm	3.5+0.2	3.4+0.4	0.45	3.7+0.3	3.5+0.2	0.03
Fractional shortening, %	36+5	34+4	0.33	33+4	34+3	0.89
LV Mass, g	0.09+0.02	0.09+0.02	0.55	0.12+0.03	0.11+0.01	0.49
Heart Rate (anesthesia), bpm	450+60	398+45	0.04	409+72	399+83	0.76
Max LV Pressure, mmHg	94+14	84+14	0.25	96+15	86+12	0.07
End-diastolic LV Pressure, mmHg	6+6	5+6	0.58	7+8	8+5	0.60
Tau, ms	13+2	16+3	0.04	16+4	21+5	0.04
Heart weight / Body weight, mg/g	4.2+0.3	4.1+0.3	0.36	4.8+1.0	4.5+0.5	0.30