2002 ASCI/AAP Joint Meeting

Signal Transducer and Activator of Transcription 3 (STAT3) has been shown to function as an oncogene and is constitutively activated in a variety of cancers including squamous cell carcinoma of the head and neck (SCCHN). We hypothesized that a transcription factor decoy approach could be used to block activated STAT3 thereby inhibiting SCCHN proliferation. The STAT3 decoy contained the double-stranded DNA binding site of activated STAT3, the high-affinity serum inducible element (hSIE). Mutant controls were designed by mutating selected base pairs in the STAT3 decoy. Electrophoretic Mobility Shift Assay (EMSA) of the binding of the hSIE to STAT3 showed that the decoy bound to activated STAT3, whereas control mutant decoys did not bind to STAT3. Live cell imaging demonstrated nuclear incorporation of the decoy by six to nine hours. Treatment of SCCHN cells in vitro with STAT3 decoy showed a decrease in the level of activated STAT3 when compared to the mutant control. After treatment with 25 microM STAT3 decoy for six days there was growth inhibition of a SCCHN cell line compared to those treated with control mutant decoy. Treatment of a SCCHN cell line with STAT3 decoy showed a decrease in the cellular RNA level of the apoptotic inhibitor bcl-xL RNA compared to control mutant as shown by RNAse protection assay. These preliminary results suggest that a transcription factor decoy approach may be used to block STAT3 activation, and thus serve as a potential novel therapy for SCCHN.

Winner, Poster Session Award

Epidermal Growth Factor Receptor and Ink4a/Arf: Convergent mechanisms governing terminal differentiation and transformation along the neural stem cell to astrocyte axis

Trainee: Robert Bachoo

Robert M. Bachoo^1,2,4, Elizabeth A. Maher^1,2, Keith L. Ligon^3,5, Norman E. Sharpless², Suzanne S. Chan², Mingjian James You², Yi Tang⁷, Jessica DeFrances², Elizabeth Stover², Ralph Weissleder⁷, David Rowitch³, David N. Louis⁶, and Ronald A. DePinho^2,8

¹Center for Neuro-Onocology, Departments of ²Adult Oncology and ³Pediatric Oncology, Dana-Farber Center Institute; Departments of ⁴Neurology and ⁵Pathology, Division of Neuropathology, Brigham and Women's Hospital; ⁶Department of Pathology and Neurosurgocal Service, and ⁷Center for Molecular Imaging Research, Department of Radiology, Massachusetts General Hospital; ⁸Departments of Genetics and Medicine, Harvard Medical School, Boston, MA

Ink4a/Arf inactivation and epidermal growth factor receptor (EGFR) activation are signature lesions in high-grade gliomas. How these mutations mediate the biological features of these tumors is poorly understood. We demonstrate that the combined loss of p16^INK4a and p19^ARF, but not of p53, p16^INK4a or p19^ARF, enables astrocyte dedifferentiation in response to EGFR activation. Moreover, transduction of Ink4a/Arf-/- NSCs or astrocytes with constitutively active EGFR induces a common high-grade glioma phenotype. These findings identify NSCs and astrocytes as equally permissive compartments for gliomagenesis and provide evidence that p16^INK4a and p19^ARF synergize to maintain terminal astrocyte differentiation. These data support the view that dysregulation of specific genetic pathways, rather than the cell-of-origin, dictates the emergence and phenotype of high-grade gliomas.

The enzymatic activity of the GDP-mannose 4,6-dehydratase protein is dependent on cellular fucosylation

Trainee: Daniel J. Becker

Daniel J. Becker¹, Peter L. Smith^2,3, Bronia Petryniak², Robert J. Kelly², Jay T. Myers², Bingyuan Wu⁴, James A. Alford, III², Peng G. Wang⁴, and John B. Lowe^1,2,3

Graduate Program in Cellular and Molecular Biology¹, Howard Hughes Medical Institute² and Department of Pathology³, University of Michigan Medical School, Ann Arbor, MI, and Department of Chemistry⁴, Wayne State University, Detroit, MI

Fucose is a six-carbon sugar that is an important component of many complex carbohydrates decorating proteins and lipids at the cell surface. Specific fucosylated glycans are essential for proper trafficking of leukocytes to inflammatory sites and lymphoid organs, and there is evidence for the involvement of fucose in a number of developmental processes including modulation of Notch-mediated signaling events by the Fringe family of glycosyltransferases. Fucosylated glycans are constructed by fucosyltransferases, which transfer fucose to nascent carbohydrate chains from the substrate GDP-fucose. There are two main pathways for the synthesis of GDP-fucose in mammalian cells. In the de novo pathway, GDP-mannose is first converted by the GDP-mannose 4,6-dehydratase protein (GMD) to an intermediate product, which is then transformed to GDP-fucose by a bifunctional enzyme known as the FX protein. An alternative, quantitatively less important salvage pathway utilizes free fucose from the diet or from catabolism of fucosylated molecules to synthesize GDP-fucose. Cell lines with mutations affecting GDP-fucose biosynthesis have been produced from several mammalian species. Biochemical examination of these cell lines implies that the process by which GDP-fucose is synthesized may be more complex than the pathways outlined above. To explore this apparent complexity, we have generated and characterized four independent Chinese hamster ovary cell (CHO) mutant clones with defects in the de novo GDP-fucose biosynthetic pathway. Cytosolic extracts from these clones exhibited deficient GMD activity, despite normal GMD transcript and protein expression and the absence of coding region mutations in GMD. Surprisingly, the cell surface fucosylation defect of the mutant clones was restored by transfection with a cDNA encoding the human FX protein, but was not corrected by transfection with a GMD cDNA. Accordingly, mutations in the coding region of FX were found in all four clones, none of which contained detectable levels of FX protein. This unexpected discordance between the biochemical and genetic data suggested interdependence between the GDP-fucose biosynthetic enzymes. Biochemical investigation of the CHO clones and embryonic fibroblasts from mice with a targeted mutation at the FX locus did not reveal a specific requirement by GMD for the presence of the FX protein. However, culture of FX mutant cells in exogenous fucose not only restored cell surface fucosylation via the salvage pathway, but also restored GMD activity, indicating that cellular fucosylation is necessary for normal GMD function. This correction of GMD activity was not mediated by an increase in GMD protein expression, and cytosol mixing experiments did not provide evidence for control of GMD activity by a fucose-dependent soluble activator. The finding that fucose availability is a determinant of GMD activity demonstrates a role for cellular glycans in the regulation of GDP-fucose biosynthesis.

RNA editing of the Drosophila Dalpha6 nAChR transcript

Trainee: Tarun Bhalla

Tarun Bhalla and Robert A. Reenan

Department of Genetics and Developmental Biology, University of Connecticut Health Center, Farmington, CT

Adenosine-to-inosine (A-to-I) editing hydrolytically deaminates A to I in target molecules through the action of a denosine deaminases that act on RNA (ADARs). ADARs are capable of site-specific conversion of A-to-I in precursor messenger RNAs (pre-mRNAs). The best characterized targets of A-to-I editing are nervous system specific: the mammalian glutamate receptor genes (GluRs) and the serotonin receptor gene (5HT_2C-R). Because inosine has base-pairing properties like those of guanosine (G), the translation machinery interprets I as G. Thus, A-to-I conversion in pre-mRNA has the potential to recode genomic information resulting in single amino acid changes that can have dramatic effects on protein function.

We describe here seven editing sites in the Dalpha6 transcript, which encodes an alpha-subunit of a nicotinic acetylcholine receptor (nAChR) in Drosophila, where RNA editing occurs. Editing at these sites is evolutionarily conserved across orders between Diptera (D. melanogaster) and Lepidoptera (H. virescens). However, the level of editing at each site appears to be species-specific. In the case of site 7, a highly conserved region was found in the intron upstream of the editing site; this region was shown to be complementary to the region of the exonic editing site. Thus, editing at site 7 would appear to involve a mechanism whereby the edited exon forms a base-paired secondary structure with the distant conserved noncoding sequences located in adjacent introns, similar to the mechanism shown for A-to-I RNA editing of mammalian glutamate receptor subunits. Transgenic Drosophila which express constructs that will elucidate the intronic requirements for editing of sites 1-5 are currently under investigation. In addition, transgenic Drosophila which over-express the completely edited and unedited forms of Dalpha6 will be used to investigate the functional significance of RNA editing of nAChRs in the nervous system.

TGF-beta1 signaling in glomerulosclerosis

Trainee: Markus Bitzer (ASCI new member: Erwin Paul B�ttinger)

Albert Einstein College of Medicine, Bronx, NY

Transforming growth factor-beta (TGF-beta) is a central mediator of fibrosis, but not much is understood about the role of particular signaling cascades activated by TGF-beta. TGF-beta activates TAK1/p38, PI3-kinase, and the Smad-cascade in which Smad2/Smad3 are phosphorylated by ligand-activated TGF-beta receptor type I, followed by nuclear translocation and transcriptional regulation of target genes. In order to link distinct features of renal fibrosis to individual signaling cascades, we studied mice carrying a TGF-beta1 transgene (TG), and develop progressive glomerulosclerosis and tubulo-interstitial fibrosis, beginning at 2 weeks of age. Total kidney RNA from 0.5 to 6 week-old TG and wildtype littermates (WT) mice was purified, and hybridized to cDNA microarray chips. The other kidney was used to examine apoptosis, extracellular-matrix and collagen type I deposition, tubulo-interstitial inflammation, and macrophage infiltration. Applying high-stringency statistical analysis we identified transcripts with highly significant differential expression profiles (TG vs. WT) which correlated with histological phenotypes (including Mmp2, Tnc, Col6a1, Col3a1, Gabbr1, and Hey1) or preceding development of renal disease. Later genes are candidate genes for initiation of progressive glomerulosclerosis and tubulo-interstitial fibrosis.

To study the role of Smad3-dependent signaling in TGF-beta1-induced glomerulosclerosis, we generated Smad3-deficient TGF-beta1-transgenic mice by crossing TG with Smad3 knockout mice (Smad3KOxTG). Smad3KO mice die early secondary to defective immune system, Smad3 heterozygous mice (Smad3HT) are normal. We found that Smad3HTxTG exhibit prolonged survival compared to TG. In Smad3KOxTG we found less apoptosis, infiltration of macrophages, and matrix deposition, compared to Smad3WTxTG littermates. In cultured podocytes some genes over- or underrepresented in TG are regulated by TGF-beta dependent on Smad3.

By correlating gene expression with distinct features of renal fibrosis and establishing Smad2/3-deficient renal cell lines or using specific inhibitors for the TAK1/p38 and PI3-kinase pathways, we propose to link TGF-beta-inducible signaling pathways with specific features of renal fibrosis.

Cyclooxygenase -1 and -2 knockout mice demonstrate increased cardiac ischemia/reperfusion injury, but are protected by acute preconditioning

Trainee: Michael G. W. Camitta (ASCI new member: Darryl C. Zeldin)

Michael G. W. Camitta^1,2, Scott A. Gabel¹, Patricia Chulada¹, J. Alyce Bradbury¹, Robert Langenbach¹, Elizabeth Murphy¹, Darryl C. Zeldin¹

¹NIEHS, NIH, Research Triangle Park, North Carolina, and ²Division of Pediatric Cardiology, Duke University Medical Center, Durham, NC

Previous studies have shown both detrimental and beneficial effects of cyclooxygenase (COX) inhibition on recovery of cardiac function following ischemia. The purpose of this study was to examine cardiac recovery after ischemia in genetically and chemically prostaglandin deficient mice. Methods: Langendorff perfused hearts from wild type (WT), COX-1^-/- and COX-2^-/- mice were subjected to 20min global ischemia followed by 40min reperfusion and measurements of left ventricular developed pressure (LVDP) were made. In some experiments, mice were pretreated with indomethacin (indo) (4 mg/kg/day po x 2 days) or indo (10 microM) was added to the perfusate. Results: Hearts from COX-1^-/- and COX-2^-/- mice recovered less post-ischemic function than WT (LVDP as % baseline: 14.1�2.4, 13.3�2.3, and 29.9�3.3, respectively, p=0.001). LVDP recovery was also significantly lower in hearts of WT mice pretreated with indo (6.2�2.4, p<0.001); however, hearts acutely perfused with indo were not significantly different than control (29.0�3.9). Addition of PGE₂, PGD₂, PGF_2alpha, or carbaprostacyclin (5 microM each) to the perfusate of COX-1^-/- and COX-2^-/- hearts did not significantly alter functional recovery. Conclusions: Genetic disruption of COX-1 or COX-2, or 2 day chemical inhibition of COX, negatively affect post-ischemic recovery of cardiac function; whereas acute inhibition of ongoing prostaglandin biosynthesis does not. We postulate that the beneficial effects of COX products occur indirectly and require time-dependent changes in gene expression.

Immune selection against stress inducible Hsp70 expression in tumors

Trainee: Margaret Callahan

Margaret Callahan¹, Ursula Munoz-Najar¹, Hong Zheng¹, Paul Clark², Antoine M�noret¹

¹Center for Immunotherapy of Cancer and Infectious Diseases, University of Connecticut School of Medicine, Farmington, CT; ²Department of Laboratory Medicine, Yale University, New Haven, CT

The relationship between the host immune system and growing tumor is an escalating competition. Point: the host immune system recognizes and eliminates the most immunogenic tumor cells. Counterpoint: tumors are selected for modifications which make them less immunogenic, commonly by down regulation of molecules involved in antigen presentation or immune stimulation. Often, it seems, the tumor wins when it becomes invisible to the immune system. In this context, we have identified the stress inducible Hsp70 as a molecule which is down-regulated in clones of a parental tumor line, diminishing their immunogenicity in a T cell dependent fashion. Despite differences in immunogenicity all clones retained the ability to cross immunize. Furthermore, the differences in immunogenicity were independent of MHC-I (H2-L^d, H2-K^d and H2-D^d) surface expression or killing by CTLs specific for a known tumor antigen. Most surprisingly, expression of 7 other heat shock proteins were normal and were unable to compensate for the loss immunogenicity associated with loss of Hsp70 expression. All together these results confirm and extend the unique role of stress inducible Hsp70 in the T cell mediated rejection of tumors in immuno-competent hosts.

Adenovirus E1A inhibition of inducible nitric oxide synthase expression

Trainee: Wangsen Cao (ASCI new member: Charles J. Lowenstein)

Wangsen Cao, Clare Bao, Charles J. Lowenstein

Division of Cardiology, Department of Medicine, The Johns Hopkins University School of Medicine, Baltimore, MD

Nitric oxide (NO) is an anti-viral effector of the innate immune system. Viruses that can interfere with NO synthesis may be able to replicate more rapidly than viruses that cannot limit NO synthesis. We show that the adenovirus E1A protein can inhibit NO production by decreasing expression of the inducible NO synthase (NOS2). The amino terminal portion of E1A decreases transactivation of the NOS2 5' flanking region that contains a functional NF-kappaB binding site. With an stable E1A-expressing cell line we also show that inducible expression of E1A limitis the DNA binding activity of nuclear factor-kappa B (NF-kappaB), and inhibitis NOS2 expression. E1A is thus able to deactivate a critical component of the host defense against viral infection. Viral inhibition of NO production is a novel mechanism that may enable certain viruses to evade the host innate immune system.

Investigating the role of gp96 in the processing of MHC Class I epitope precursors

Trainee: Marissa Caudill

Marissa Caudill and Zihai Li

Center for Immunotherapy of Cancer and Infectious Disease, University of Connecticut Health Center, Farmington, CT

We are investigating the potential contribution of an endoplasmic reticulum (ER) chaperone, gp96, in the processing of N-terminally extended precursors of known MHC class I epitopes. This investigation is based on the hypothesis that gp96 is an aminopeptidase that contributes to the generation of the diverse array of peptide epitopes presented by MHC class I molecules. There are several pieces of evidence supporting a potential involvement of gp96 in this process: gp96 has been shown to possess aminopeptidase activity in vitro, it is an ATPase, it binds peptides delivered into the ER by TAP, it is upregulated in response to stress and IFN-gamma exposure, and we have shown that gp96 co-immunoprecipitates with MHC class I molecules. These data, together with several studies which implicate an ER aminopeptidase in the generation of MHC class I epitopes from N-terminally extended precursors, have led us to begin an investigation of the in vivo aminopeptidase activity of gp96. Utilizing a cell line that lacks functional gp96 protein, we are examining the contribution of this ER protein to the processing of N-terminally extended precursor peptides. To this end, we have characterized the class I surface expression patterns of cells containing and lacking gp96 by flow cytometry and analyzed the overall surface peptide profile of the cells by HPLC. We are using cellular assays to measure the antigen presentation efficiency of model peptides in cells that have or lack gp96, as well as microsomal systems to monitor the processing of peptides in vitro. These experiments will allow us to definitively state whether or not gp96 possesses a detectable aminopeptidase activity in its natural environment, the ER, and whether or not that aminopeptidase activity contributes to the MHC class I peptide repertoire.

Cross-talk between two species: The how and why of an autocrine viral-human feedback circuit

Trainee: Malini Chatterjee (ASCI new member: Patrick S. Moore)

Malini Chatterjee, Julie Osborne, Giovanna Bestetti, Yuan Chang, Patrick S. Moore

Columbia University College of Physicians and Surgeons, New York, NY and the University of Pittsburgh Cancer Institute, Pittsburgh PA

Primary effusion lymphoma (PEL) cells infected with Kaposi's sarcoma-associated herpesvirus (KSHV) are autocrine-dependent on the virus-encoded vIL-6 but not on hIL-6. These two cytokines activate the same downstream signaling pathways but differ in their membrane receptor usage: hIL-6 binds to the gp80 IL-6 receptor (IL-Ralpha) to activate the gp130 transmembrane signal transducer, whereas vIL-6 directly binds to gp130 to initiate signaling independent of gp80. Since both IL-6 receptor components are expressed on PEL cell lines, the reasons for this difference are unclear.

Interferons (IFNs) are induced as antiviral cytokines after virus infection. KSHV vIL-6 but not hIL-6, protects the PEL cell line, BCP-1, from IFN-alpha induced growth arrest and inhibits IFN-alpha induced upregulation of the cyclin-dependent kinase inhibitor, p21^WAF1. Despite the presence of both gp80 and gp130 on resting PEL cells, electromobility gel shift assays of PEL cell lines demonstrate that bioequivalent amounts of vIL-6 are far more active in initiating IL-6 signaling than hIL-6. Thus, vIL-6 acts as the effector arm of the autocrine loop by inhibiting the action of IFN induced during virus infection.

vIL-6 mRNA transcription itself is activated by IFN-alpha in absence of de novo protein synthesis. Microarray analysis using 87 KSHV gene products demonstrate that after 12 hours IFN treatment, vIL-6 mRNA is induced up to 30 fold and it is the only cycloheximide-independent viral transcript induced by IFN. The vIL-6 promoter has two interferon-responsive elements, ISRE1 and ISRE2, demonstrable through promoter-reporter deletion studies. When the full-length promoter-gene cassette is stably reconstituted in a heterologous system using the IFN-responsive Daudi cell line C11, wild-type vIL-6 promoter protects C11 cells from IFN-induced growth arrest. This protection is markedly diminished when single and double mutations are introduced into the promoter at ISRE1 and ISRE2, recapitulating the interferon-vIL-6 autocrine loop found in PEL cell lines.

If vIL-6 and hIL-6 are indistinguishable in their downstream signaling pathways, then why doesn't the human cyokine prevent IFN-induced arrest of PEL cells? Using western blots and flow cytometry, we find that IFN-alpha immediately downregulates gp80 membrane expression making PEL cells unresponsive to hIL-6 but not to vIL-6. Thus, KSHV has evolved a mechanism to sense and inhibit IFN-alpha responses by circumventing IL-6 receptor regulation. This KSHV-human autocrine feedback loop shows that viruses monitor and actively modify their environment, a fundamental property of living biological systems.

Obesity resistance and enhanced glucose metabolism in mice transplanted with DGAT1-deficient white adipose tissue

Trainee: Hubert C. Chen

Hubert C. Chen^1,2,3, Heather M. Myers¹, and Robert V. Farese, Jr.^1,2,3

¹Gladstone Institute of Cardiovascular Disease, ²Cardiovascular Research Institute and ³Department of Medicine, University of California, San Francisco, CA

Recent studies have identified the white adipose tissue (WAT) as an important endocrine organ that regulates energy and glucose metabolism through secreted factors. Mice lacking acyl CoA:diacylglycerol acyltransferase 1 (DGAT1), a key enzyme in mammalian triglyceride synthesis, are protected against diet-induced obesity and glucose intolerance due to increased energy expenditure and enhanced insulin sensitivity. Because DGAT1 is highly expressed in WAT, we hypothesized that DGAT1 deficiency affects the expression of an adipocyte-derived factor that regulates energy and glucose metabolism. Transplantation of DGAT1-deficient WAT into wild-type mice decreased adiposity and enhanced glucose disposal in two models of obesity and insulin resistance: high-fat fed C57BL/6J mice and Agouti yellow mice. Analysis of DGAT1-deficient WAT revealed a twofold increase in the expression of adiponectin, a molecule that enhances fatty acid oxidation and insulin sensitivity, and this increase may account in part for the transplantation-induced metabolic changes. Our results highlight the importance of the endocrine function of WAT and suggest that an alteration in this function contributes to the increased energy expenditure and insulin sensitivity in DGAT1-deficient mice.

Multiple dose dependent effects of Lis1 on cerebral cortical development

Trainee: Dawn L. Darling (ASCI new member: Anthony Wynshaw-Boris)

Michael J. Gambello¹, Dawn L. Darling¹, Jessica Yingling¹, Teruyuki Tanaka², Joseph G. Gleeson² and Anthony Wynshaw-Boris¹

¹Departments of Pediatrics and Medicine; ²Department of Neurosciences, University of California, San Diego, La Jolla, CA

Humans with heterozygous null mutations or deletions of the LIS1 gene suffer from Type I lissencephaly, a severe form of cortical dysplasia hypothesized to result from abnormal neuronal migration. Previously we reported the construction of an allelic series of the Lis1 gene in mice to analyze the effects of graded reduction of LIS1 protein on the pathogenesis of lissencephaly, demonstrating a defect in neuronal migration (Hirotsune et al., 1998). Here we present a systematic examination of the consequences of dosage reduction of LIS1 on neocortical development using wild-type, null heterozygous (70% LIS1 protein) and compound null/hypomorphic (40% LIS1 protein) mice. The development of the preplate and Cajal-Retzius cells appeared to be unaffected by LIS1 levels. However, dose-dependent morphologic changes in the radial glial scaffold were noted, and the subplate was diffusely organized. LIS1 dose-dependent defects in neuronal migration were found in vivo and in vitro. The position and number of mitotic cells in the ventricular zone was more aberrant as LIS1 levels decreased, suggesting defects in interkinetic nuclear migration and neuroblast proliferation. LIS1 dose-dependent progressive thinning of the cortex and ventricular zone was observed as development progressed, consistent with neuronal loss. TUNEL analysis demonstrated an increase in cell death as LIS1 reduced. Thus, in addition to its requirement for cell autonomous neuronal migration, LIS1 influences radial glial scaffold morphology, and generation and survival of cortical ventricular zone neuroblasts. These studies reveal the importance of LIS1 levels in cerebral cortical morphogenesis and suggest new insights into the pathogenesis of type I lissencephaly.

The role of SLP-76 and ADAP (SLAP-130/Fyb) in cytoskeletal rearrangements and integrin activation in T cells

Trainee: Geo S. Derimanov

University of Pennsylvania, Philadelphia, PA

Background: The hematopoietic-specific adaptor proteins SLP-76 and ADAP (also known as SLAP-130/Fyb) mediate signal transduction events in T cells after T cell receptor (TCR) stimulation. SLP-76 is required for the TCR-induced PLCgamma phosphorylation, calcium flux, ERK activation, and the resulting changes in gene transcription as well as for thymocyte development. It has also been implicated in cytoskeletal rearrangements following TCR engagement. However, the structure-function relationship of the different domains of SLP-76 has not been completely elucidated. The role of ADAP in signaling through the TCR is less well understood. Preliminary data from our and other laboratories had suggested that ADAP may be important for cell migration and adhesion.

Results: To address the role of the SLP-76 domains in cytoskeletal rearrangements, we studied TCR clustering, a process dependent on an intact cytoskeleton, in double positive (DP) thymocytes from transgenic mice expressing mutant forms of SLP-76 on SLP-76 null background. We found that a mutation of the amino terminal portion of SLP-76, preventing its phosphorylation on tyrosine residues, leads to a significant decrease in TCR clustering. This defect is similar to that described for VAV-deficient thymocytes.

The SLP-76 mutant thymocytes also demonstrate hypophosphorylation of a protein with molecular weight similar to VAV.

To investigate the function of ADAP in cyoteskeletal and integrin activation, we assessed actin polymerization and integrin clustering at the site of TCR engagement in peripheral T cells from ADAP null mice generated in our laboratory. While the ADAP-/- T cells exhibited normal actin capping, they were deficient in their ability to recruit LFA-1 to the site of TCR activation.

Conclusions: Our results suggest that the effect of SLP-76 on cytoskeletal mobilization is mediated through its phosphorylated tyrosine residues and may involve a cooperation with VAV. On the other hand, ADAP seems dispensable for actin polymerization. However, it appears to be important for coupling of the cytoskeletal rearrangements to integrin activation upon TCR stimulation.

Deletion of macrophage ACAT1 decreases cholesterol ester storage, free cholesterol efflux, and the TNF-alpha response to inflammation

Trainee: Dwayne E. Dove

Dwayne E. Dove, Amy S. Major, MacRae F. Linton, and Sergio Fazio

Vanderbilt University Medical Center, Nashville, TN

Atherosclerosis and its complications are the major cause of morbidity and mortality in the United States. Inflammation and accumulation of cholesterol by macrophages within the artery wall are hallmarks of atherosclerotic lesions. Acyl-coenzyme A: cholesterol acyltransferase (ACAT) is an enzyme that is important in cellular cholesterol homeostasis because it converts free cholesterol into cholesterol esters for storage in lipid droplets. Recent studies in our laboratory have shown that the deletion of the ACAT1 gene which encodes the macrophage cholesterol storage enzyme, ACAT1, results in an increase in atherosclerotic lesion area in the aortas of mice. These results are surprising given that ACAT1 mediated cholesterol ester accumulation is viewed as a crucial step in foam cell formation. Because the effects of ACAT1 deletion on cholesterol balance and inflammatory functions of macrophages had not been reported, studies of cholesterol ester storage, free cholesterol efflux, cell morphology, and Tumor Necrosis Factor (TNF)-alpha responses to lipopolysaccharide (LPS) were performed with macrophages treated with low density lipoproteins (LDL) that were chemically modified. Compared to ACAT1(+/+) macrophages, ACAT1(-/-) macrophages cholesterol-loaded with acetyl-LDL showed an 88% decrease in cholesterol ester storage (p<0.001) while the mass of free cholesterol was not changed (p=0.679). Cholesterol-loaded ACAT1(-/-) macrophages also had a 38% decrease in free cholesterol efflux (p<0.01). Morphologically, cholesterol-loaded ACAT1(-/-) macrophages were characterized by a drastic reduction in lipid droplets but had increased surface activity with more endosomal vesicles. Compared to identically treated ACAT1(+/+) macrophages, ACAT1(-/-) macrophages treated with oxidized-LDL had a 32% decrease in the TNF-alpha response to LPS stimulation (p<0.01). Together, these results suggest that ACAT1(-/-) macrophages have a diminished ability to respond to cholesterol loading and inflammatory stimuli after exposure to modified lipoproteins. These diminished responses may be due to changes in macrophage viability or changes in the membrane distribution of free cholesterol in association with the loss of cholesterol ester storage. Because cholesterol efflux is the first step of the reverse cholesterol transport system and is critical in preventing atherosclerosis, the mechanisms of decreased cholesterol efflux from ACAT1(-/-) macrophages should be investigated further as a potential cause of atherogenesis.

Selective activation of PPARgamma increases PTEN expression and inhibits pancreatic cancer invasion

Trainee: Buckminster Farrow (ASCI new member: B. Mark Evers)

The Univ of Texas Med Branch, Galveston, TX

Peroxisome proliferator activated receptor-gamma (PPARgamma) is a ligand-activated transcription factor known to have anti-inflammatory, insulin-sensitizing, and anti-tumor effects. The purpose of this study was to define the inhibitory effect of PPARgamma ligands on pancreatic cancer invasion and determine potential mechanisms for this effect. Methods. The human pancreatic cancer cell lines, AsPC-1 and SUIT-2, were treated for 48h with non-toxic doses of PPARgamma ligands (15d-PGJ₂, troglitazone or rosiglitazone) then placed into Matrigel Invasion chambers for 24h; the number of treated cells invading the membrane was compared to vehicle treatment. To further investigate the anti-invasive properties of the PPARgamma ligands, AsPC-1 cells were treated with rosiglitazone and Western blotting performed for PTEN, a tumor suppressor protein known to have inhibitory effects on invasion; GW9662, an inhibitor of PPARgamma activation, was used to determine the specificity of rosiglitazone-induced PPARgamma activation. To identify other anti-invasive targets of PPARgamma, a focused array for genes that contribute to invasion (Clontech Human Cell Interaction Array) was used following 48h rosiglitazone treatment (10 microM); analysis of the radiolabeled hybridized filters was performed using ImageQuant software. Results. Treatment with 15d-PGJ₂ (AsPC-1 and SUIT-2) and rosiglitazone (AsPC-1) decreased cell invasion compared to control. PTEN expression was increased by rosiglitazone at both low (1 microM) and high (10 microM) concentrations; GW9662 (10 microM) reversed the effect of rosiglitazone on PTEN expression. The expression of integrins (alpha4, beta7, and beta5) and matrix metalloproteinases (MMP-3 and 7) was decreased more than 3-fold by rosiglitazone treatment compared with control. Conclusions. PPARgamma ligands inhibit pancreatic cancer invasion, suggesting that these agents may represent novel strategies to treat pancreatic cancer through increased expression of PTEN and subsequent downregulation of integrins and MMPs.

Retroviral-mediated gene transfer (RMGT) reduces cutaneous inflammation and granuloma formation in a murine model of X-linked chronic granulomatous disease (X-CGD)

Trainee: W. Scott Goebel

W. Scott Goebel¹, Jeffrey B. Travers², Nancy Pech¹, and Mary C. Dinauer¹

Depts. of ¹Pediatrics (Hematology/Oncology) and ²Dermatology, and Herman B Wells Center for Pediatric Research, Indiana University, Indianapolis, IN

Our laboratory has previously described the correction of neutrophil NADPH oxidase function by RMGT in murine X-CGD. We have also previously shown that X-CGD mice have exaggerated inflammatory responses and form granulomas following challenge with sterile Aspergillus fumigatus hyphae (AF). In this study, we examined the cutaneous response to injected sterile AF in wild-type and gene-corrected X-CGD mice chimeric for differing numbers of NADPH oxidase-positive neutrophils, which were created by transplanting varying amounts of wild-type or RMGT-treated X-CGD marrow cells into X-CGD hosts. The gene-corrected and wild-type chimeric mice were divided into two groups based on the percentage of NADPH oxidase-positive neutrophils as assessed by nitroblue tetrazolium (NBT) testing: those with <20% (gene-corrected range 4-15%, mean 10.7 � 3.7%; wild-type range 6-16%, mean 11.1 � 3.9%) and those with >25% (gene-corrected range 26-78%, mean 51.1 � 16.2%; wild-type range 20-34, mean 26.1 � 4.6%). 2.5 micro g of AF or PBS vehicle was injected subcutaneously into the ears of wild-type (n=10), female X-CGD carrier (n=6), X-CGD (n=6), chimeric wild-type (n=8), or chimeric gene-corrected X-CGD (n=22) mice. Inflammation was assessed by measuring ear thickness at 3 and 30 days post-injection, by weighing 5 mm punch biopsies from the injection sites at 30 days post-injection, and by histologic examination of biopsy specimens. Intradermal AF induced marked inflammation, as measured by edema and ear thickness, at both 3 and 30 days in the X-CGD mice, but not in the carriers or the wild-type mice. The inflammatory responses of chimeric mice with <20% NBT positive neutrophils were significantly greater than wild-type mice, but were less than that of X-CGD mice. Wild-type chimeric mice with <20% NBT positive neutrophils, which exhibit approx. 3-fold more NADPH oxidase activity than neutrophils from gene-corrected mice, still showed an increased inflammatory response to AF challenge that was not statistically different from that of the gene-corrected chimeras. Chimeric mice (both wild-type and gene-corrected) with >25% NBT positive neutrophils displayed inflammatory responses similar to those of wild-type mice. No wild-type or carrier animals displayed histologic evidence of granuloma formation 30 days after AF injection; in contrast, X-CGD mice showed marked chronic inflammation and granuloma formation, often throughout the biopsy specimen. All but two (n=16) chimeric animals with <20% NBT positive neutrophils showed some histologic inflammation and granuloma formation, whereas chimeras with >25% NBT positive neutrophils had normal histology. The gene-corrected chimeras with <20% NBT positive neutrophils displayed more inflammation and granuloma formation than the wild-type chimeras with similarly low levels of normal neutrophils, perhaps reflective of the lesser amounts of NADPH oxidase in the gene-corrected neutrophils. This work describes a new experimental model system in which to study cutaneous granuloma formation in CGD, and supports our prior work which suggested that partial correction of NADPH oxidase activity by RMGT may be beneficial for patients with CGD if sufficient numbers of gene-corrected neutrophils are obtained.

Zebrafish Cyclooxygenases 1 and 2

Trainee: Tilo Grosser

Tilo Grosser¹, Shamila Yusuff², Ellina Cheskis¹, Michael A. Pack², Garret A. FitzGerald¹

¹Center for Experimental Therapeutics and ²Department of Medicine, GI Division, University of Pennsylvania Medical School

Study of the cardiovascular biology of cyclooxygenases (COXs) has been limited by the critical role of COX-2 in murine reproduction and renal organogenesis. We sought to characterize COX expression and function in zebrafish (z), a model organism for vertebrate genetics and development. Full-length cDNAs of zCOX-1 and zCOX-2 were cloned and assigned to conserved regions of chromosomes 5 and 2 respectively. The deduced proteins are 67% homologous with their human orthologs. Prostaglandin (PG) E2 is the predominant zCOX product detected by mass spectrometry. Pharmacological inhibitors demonstrate selectivity when directed against heterologously expressed zCOX isoforms. Zebrafish thrombocyte aggregation ex vivo and hemostasis in vivo are sensitive to inhibition of zCOX-1, but not zCOX-2. Both zCOXs were widely expressed during development and morpholino based knock down of zCOX-1 causes growth arrest during early embryogenesis. zCOX-1 is widely evident in the embryonic vasculature, while zCOX-2 exhibits a more restricted pattern of expression. Both zCOX isoforms are genetically and functionally homologous to their mammalian orthologs. The zebrafish affords a tractable model system for the study of COX pharmacology and the role of COX isoforms in cardiovascular biology and development.

Cross-talk between cPLA₂ and group sPLA₂s in hydrogen peroxide-induced arachidonic acid release in murine mesangial cells

Trainee: Won K. Han

Won K. Han, Adam Sapirstein, Alessandro Alessandrini, and Joseph V. Bonventre

Massachusetts General Hospital East, Harvard Medical School, Charlestown, MA

Oxidant stress and phospholipase A₂ (PLA₂) activation have been implicated in numerous pro-inflammatory responses of the mesangial cell. Mammalian cells contain more than one form of PLA₂. Two types of murine mesangial cells (MC) were used to investigate the cross-talk between group IVa cytosolic PLA₂ (cPLA₂) and secretory PLA₂s (sPLA₂s) during H₂O₂-induced arachidonic acid (AA) release: i) MC^+/+, which lack group IIa and V PLA₂s and ii) MC^-/-, which lack group IIa, V and cPLA₂s. H₂O₂-induced AA release was greater in MC^+/+ compared with MC^-/-. Group IIa, V or cPLA₂ was expressed in MC^-/- and MC^+/+using recombinant adenovirus vectors. Expression of cPLA₂ in H₂O₂-treated MC^-/- increased AA release to a level approaching that of H₂O₂-treated MC^+/+. Infection with adenovirus expressing either group IIa PLA₂ or V PLA₂ enhanced AA release in MC^+/+, but had no effect on MC^-/- AA release. Inhibition of the ERK signaling pathway with MEK-1 inhibitor, U0126, significantly reduced H₂O₂-induced AA release in MC^+/+ infected with adenoviruses expressing group IIa or V PLA₂. By contrast, H₂O₂-induced AA release was not enhanced when ERK 1/2 was activated by infection of MC ^+/+ with constitutively active MEK-1DD. We conclude that the effect of group IIa and V PLA₂s on H₂O₂-induced AA release is dependent upon the presence of cPLA₂ and the activation of ERK1/2. The activation of ERK1/2 is necessary but not sufficient to augment H₂O₂-induced AA release.

Winner, Poster Session Award

The Ly-49H activation receptor of murine natural killer cells recognizes a murine cytomegalovirus-encoded ligand

Trainee: Jonathan W. Heusel

Jonathan W. Heusel^1,2, H.R.C. Smith¹, I.K. Mehta¹, S. Kim¹, B.G. Dorner¹, O.V. Naidenko², K. Iizuka¹, H. Furukawa¹, D.L. Beckman¹, J.T. Pingel¹, A.A. Scalzo³, D.H. Fremont² and W.M. Yokoyama¹

¹Howard Hughes Medical Institute, Division of Rheumatology, Department of Internal Medicine, and ²Department of Pathology and Immunology, Washington University School of Medicine and Barnes-Jewish Hospital, St. Louis, MO; ³Department of Microbiology, University of Western Australia, Queen Elizabeth II Medical Centre, Nedlands 6907, Australia

Natural Killer (NK) cells represent a vital component of the innate immune system, and are especially important for control of certain viral infections. Ly-49H is a DAP12-associated activation receptor expressed on murine NK cells. Previously, we have shown that resistance of inbred mice to murine cytomegalovirus (MCMV) infection is associated with functional Ly-49H expression. Here, we extend these studies in vitro and show that MCMV-infected cells directly activate Ly-49H-expressing BWZ.36 reporter cells, which contain an inducible lacZ reporter gene cassette. MCMV-induced activation is shown to be specific for Ly-49H and dependent upon DAP12, since BWZ.36 reporters expressing the related activation receptor, Ly-49D, or co-expressing Ly-49H in association with a mutant form of DAP12, fail to become activated. Importantly, among related herpesviruses that infect mice, including herpes simplex (HSV) and gamma-herpesvirus 68 (HV68), the Ly-49H receptor is specific for MCMV, and Ly-49D recognizes none of these herpesviruses. The Ly-49H ligand expressed in MCMV-infected targets is detected as early as 8 hours after infection (early-phase kinetics), but is not expressed by cells infected with UV-inactivated virus, or by cells treated with type I interferon, suggesting that the Ly-49H ligand is encoded within the MCMV genome. Interestingly, the presence of MHC class I on MCMV-infected cells is not required, but may modulate or facilitate optimal activation mediated by Ly-49H. Heterologous expression of PCR-amplified MCMV-encoded genes revealed that Ly-49H specifically recognizes the MCMV gene product, m157. Ly-49H⁺ (but not Ly-49H^-) NK cells are directly activated in vivo by MCMV infection (or in vitro by m157-expressing cells) to secrete IFN and other cytokines, and both fresh C57BL/6 NK cells and IL-2-activated NK cells specifically kill m157-expressing targets in vitro. Finally, m157 is shown to be one of eleven MCMV-encoded structures that are predicted to fold with MHC class I-like topology, raising the potential for additional MCMV ligand-NK cell receptor interactions. These results are an important advance in understanding NK cell activation and mechanisms of cellular innate immunity to pathogens.

Pu.1 is expressed by early granulocytic progenitors in the zebrafish

Trainee: Karl Hsu

Karl Hsu,^1,2 Jennifer Rhodes,¹ Ting Xi Liu,¹ John Kanki,¹ Barry Paw,³ Leonard I. Zon,³ and A. Thomas Look¹

¹Department of Pediatric Oncology, Dana Farber Cancer Institute, Boston, MA; ²Division of Hematology/Oncology, Beth Israel Deaconess Medical Center, Boston, MA; ³Howard Hughes Medical Institute, Children's Hospital, Boston, MA

Pu.1 is a member of the ets family of transcription factors that in the mouse, plays an essential role in the development of both myeloid and lymphoid systems. We analyzed Pu.1 expression in the zebrafish. Pu.1 expression was observed by 16 hours post fertilization (hpf) in the anterior yolk region of the developing embryo, as well as in the posterior ICM. The posterior expression was lost between 22 and 24 hpf whereas anterior expression persisted and was observed in cells spreading anteriorly over the yolk. By 28 to 30 hpf, Pu.1 expression was no longer observed. Co expression of Pu.1 and the zebrafish homologue of the myeloperoxidase gene (mpo) were observed in some cells of the anterior yolk sac at 22 to 24 hpf. In order to use Pu.1 regulatory sequences to program gene expression in early myeloid cells, we linked genomic sequences upstream of the pu.1 initiator codon to green fluorescent protein and microinjected them into single-cell zebrafish embryos. Fluorescent cells were evident in microinjected embryos 20 hpf and were absent by 48 hpf, in a pattern similar to Pu.1 mRNA expression detected by RNA in situ hybridization. DNA constructs containing 15 kilobases (kb) or 9 kb of the putative zebrafish promoter conferred similar expression patterns. These embryos were raised to sexual maturity and offspring screened to identify founder fish with germ-line transmission of GFP-positive myeloid cells. We have generated seven lines with the 9 kb promoter element. Expression patterns in five of these lines are similar to that seen in transient injections. Thus, Pu.1 is specifically expressed in early myeloid cells in the zebrafish similar to the mouse, except that Pu.1 becomes undetectable after 30 hpf in granulocytes in the zebrafish.

Proteolytic control of Cdc20p expression in S. cerevisiae

Trainee: Jim Huang (ASCI new member: David Pellman)

Huang, J.¹, Pellman, D.², Plon, S.¹

¹Texas Children's Cancer Center, Department of Pediatrics, Baylor College of Medicine, Houston, TX; ²Departments of Pediatric Oncology, The Dana-Farber Cancer Institute and Pediatric Hematology, The Children's Hospital, Harvard Medical School, Boston, MA

The eukaryotic cell cycle is driven by waves of cyclin dependent kinase activity coupled with regulated proteolysis. The Anaphase Promoting Complex/Cyclosome (APC/C) is a highly conserved multi-component ubiquitin ligase that controls the proteolysis of a number of key cell cycle proteins. The APC/C is thought to have two functionally distinct forms - APC^Cdc20 and APC^Cdh1 - based on its association with either Cdc20p or Cdh1p, two tryptophan-aspartic acid (WD) repeat containing proteins. In yeast, APC^Cdc20 targets the anaphase inhibitor Pds1p for proteolysis at the metaphase to anaphase transition, and APC^Cdc20mediated degradation of S phase cyclins is required for the activation of APC^Cdh1 in mitosis. APC^Cdc20 activity itself is inhibited by the spindle damage checkpoint pathway, and Cdc20p expression is tightly regulated by both transcriptional and proteolytic mechanisms. We found that Cdc20p degradation in late G1 is dependent upon the presence of APC^Cdh1 and ensures that Cdc20p level remains low, and thus APC^Cdc20remains off, until after S phase cyclins have accumulated. Surprisingly, unlike other APC^Cdh1 substrates, Cdc20p continues to be rapidly degraded in S phase, when APC^Cdh1 is inactive. This raises the possibility for an alternative pathway regulating Cdc20p expression. We show that Cdc20p degradation in S phase is not dependent upon APC^Cdh1. We have defined a region of Cdc20p which contains the cis-acting signal for APC^Cdh1 mediated degradation. Additionally, we have defined a region in Cdc20p which provides a novel cis-acting signal necessary for mediating the degradation of Cdc20p during S phase. Identification of this novel degradation signal provides a means to identify trans-acting factors needed for Cdc20p degradation in S phase, and to determine the functional significance of that regulation.

Ferredoxin reductase affects p53-dependent, 5-fluorouracil-induced apoptosis in colorectal cancer cells

Trainee: Paul M. Hwang (ASCI new member: Toren Finkel)

Paul M. Hwang^1,2,3, Fred Bunz², Jian Yu², Carlo Rago¹, Tim A. Chan², Kenneth W. Kinzler² & Bert Vogelstein^1,2

¹The Howard Hughes Medical Institute, ²The Johns Hopkins Oncology Center, The Johns Hopkins University School of Medicine, Baltimore, MD; ³The Cardiovascular Branch, NHLBI-NIH, Bethesda, MD

Loss of p53 gene function, which is observed in the majority of colon cancer cells, has been shown to abolish the apoptotic response to 5-fluorouracil (5-FU). To identify genes downstream of p53 that might mediate these effects, we assessed global patterns of gene expression following 5-FU treatment of isogenic cells differing only in their p53 status. Mitochondrial ferredoxin reductase (FR) was one of the few genes significantly induced by p53 after 5-FU treatment. The FR protein was localized to mitochondria and suppressed the growth of colon cancer cells when over-expressed. Targeted disruption of the FR gene in human colon cancer cells showed that it was essential for viability, and partial disruption of the gene resulted in decreased sensitivity to 5-FU induced apoptosis. These data, coupled with the effects of pharmacologic inhibitors of reactive oxygen species, suggest that FR contributes to p53-mediated apoptosis through the generation of oxidative stress in mitochondria.

Trainee: D.A. Ingram

D.A. Ingram and D. W. Clapp

Cancer Research Institute, Indiana University, Indianapolis, IN

Mutations in the NF1 tumor suppressor gene cause neurofibromatosis type I, a disease characterized by the formation of cutaneous neurofibromas infiltrated with a high density of degranulating mast cells. Neurofibromas also have elevated mRNA transcripts for stem cell factor (SCF), a chemoattractant and potent mitogen for mast cells. Recruitment and local proliferation of mast cells is a potentially critical event in tumor formation as recently shown in genetically engineered Nf1 deficient mice and other murine models of skin cancer (Coussens, Genes and Development 1999). Neurofibromin, the protein encoded by NF1, negatively regulates p21^ras activity by accelerating the conversion of Ras-GTP to Ras-GDP. Specifically, Nf1 +/- mast cells have increased survival and proliferation in response to SCF compared to wildtype cells (Ingram, JEM 2000, JEM 2001). Using transwells and videomicroscopy, Nf1 +/- mast cells also show a two-fold increase in migration on fibronectin compared to wildtype cells in response to SCF. However, identification of alterations in specific p21^ras effector pathways that control proliferation and migration of NF1 deficient cells is incomplete and critical for understanding disease pathogenesis. Recent studies have suggested that the proliferative and cytoskeletal mediated effects of p21^ras may depend on signaling outputs from the small Rho GTPases, Rac and Rho, but the physiologic importance of these interactions in an animal disease model has not been established. Utilizing a genetic intercross between Nf1 +/- and Rac2 -/- mice, we now provide genetic evidence to support a biochemical model where hyperactivation of the extracellular regulated kinase (ERK) via the hematopoietic specific Rho GTPase, Rac2, directly contributes to the hyperproliferation of Nf1 deficient mast cells in vitro and in vivo. We also demonstrate that Rac2 functions as mediator of crosstalk between phosphoinositide 3-kinase (PI-3K) and the classical p21^ras-Raf-Mek-ERK pathway to confer a distinct proliferative advantage to Nf1 +/- mast cells. Further, previous studies show that activation of PI-3 kinase and Rac2 are necessary for wildtype mast cell migration to SCF. Nf1 +/- mast cells show a two-fold increase in PI-3 kinase and Rac activity compared to WT cells with SCF stimulation. To genetically test whether hyperactivation of these pathways is also responsible for the increased migration of Nf1+/- mast cell, we intercrossed Rac2 -/- or p85alpha+/- mice (a regulatory subunit of Class 1_A PI-3 kinase) with Nf1 +/- mice. The increase in migration of Nf1 +/- mast cells was completely attenuated in Nf1 +/-; Rac2 -/- and Nf1 +/-; p85alpha-/- mast cells. Thus, these studies identify Rac2 and Class I_A PI3-K as important downstream effectors of p21^ras in Nf1 +/- mast cells which function to alter the cellular phenotype of a cell lineage involved in the pathologic complications of a neurofibromatosis type 1.

Management of chronic kidney disease in an academic primary care clinic

Trainee: A. Israni

Israni, A.¹, Korzelius C.², Mesler D.², Feldman, H.¹

¹Department of Renal, Electrolyte and Hypertension, Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania School of Medicine, Philadelphia, PA; ²Department of Nephrology, Boston University of School of Medicine, Boston, MA

Three million people in the United States are estimated to have chronic kidney disease (CKD) with a serum creatinine over 1.7 mg/dl. Little is known about how these CKD patients are being managed in the outpatient, primary care setting in an academic center. We undertook this study to evaluate the clinical management of CKD using published literature and available guidelines. We analyzed CKD care in a primary care site of an academic hospital with 23,000 annual visits and 8,300 patients. Patients seen in this clinic between 10/1/97 to 3/25/99 with an elevated SCr > 1.7 mg/dl on two separate measurements at least 6 months apart had charts reviewed for pre-defined indicators of CKD management. We found good compliance with hemoglobin A1c assessment and management. However, other indicators of CKD management were suboptimal including: control of blood pressure, use angiotensin converting enzyme inhibitors or angiotensin receptor blockers, assessment of proteinuria (qualitative and quantitative) and rate of renal consultation. CKD management was similar for diabetic and nondiabetic patients. CKD management also did not differ substantially between patients with an estimated creatinine clearance in ml/min (CrCl) > 50 and those with CrCl between 50 to 30 and those with CrCl < 30. Opportunities exist to improve management of CKD in an academic primary care setting.

The Kruppel factor KLF15 regulates the insulin-sensitive glucose transporter GLUT4

Trainee: Mukesh K. Jain (ASCI new member: Daniel I. Simon)

Susan Gray, Mark W. Feinberg, Sarah Hull, Chay T. Kuo, Masafumi Watanabe, Sucharita Sen(Banerjee), Mukesh K. Jain

Cardiovascular Division, Brigham and Women's Hospital, Boston, MA

Resistance to the stimulatory effects of insulin on glucose utilization is a key feature of type 2 diabetes, obesity, and the metabolic syndrome. Recent studies suggest that a primary cause of insulin resistance is a defect in glucose transport. GLUT4 is the main insulin-responsive glucose transporter and is expressed predominantly in muscle and adipose tissues. Despite the importance of GLUT4 in glucose homeostasis, the mechanisms regulating GLUT4 gene expression remain poorly understood.

We cloned KLF15, a novel member of the Kruppel-like family of transcription factors (KLF). KLF15 is highly expressed in adipocytes and myocytes in vivo. In vitro, KLF15 expression is induced when 3T3-L1 preadipocytes are differentiated into adipocytes. Overexpression of KLF15 in adipose and muscle cell lines potently induces GLUT4 expression. As a consequence both basal and insulin-stimulated glucose uptake are increased approx. 3.0 fold in adipocytes. This ability of KLF15 is specific as overexpression of another family member -- LKLF -- potently inhibits adipogenesis and GLUT4 expression.

To determine the mechanism of KLF15's ability to induce GLUT4 we performed gene reporter assays using the GLUT4 promoter. Previous studies highlight the importance of a MEF2 binding site within the proximal regulatory region as essential for GLUT4 expression in vivo. We found that KLF15 induces the GLUT4 promoter approx. 9-fold. Promoter deletion and mutational analyses demonstrate that this induction is mediated by a KLF binding site just upstream of the MEF2A element. Co-transfection and co-immunoprecipitation studies suggest that KLF15 and MEF2A interact and synergistically transactivate the GLUT4 promoter. Finally to understand the role of KLF15 in vivo we generated KLF15 null mice by homologous recombination. Initial studies reveal that KLF15 deficiency results in a reduction in GLUT4 expression. Taken together, these studies support a role for KLF15 as a regulator of GLUT4 expression. Furthermore, these studies introduce Kruppel-like factors as regulators of adipocyte biology.

Mechanisms of therapeutic failure with dendritic cell vaccine in colon cancer-bearing mice

Trainee: John Y. Kao

John Y. Kao, Chuan-Min Chen, Yusong Gong, Qiong-Duan Zheng, Jian-Jun Chen

Division of Gastroenterology, University of Michigan Medical Center, Ann Arbor, MI

Purpose: colon cancer is the second leading cause of cancer-related mortality in the United States. Most patients with colon cancer present with advanced disease for which no effective therapy is available. Dendritic cell (DC) vaccine is a promising antitumor immunotherapy that has the potential to eradicate metastatic cancer. Although the majority of na�ve mice immunized with DC vaccine can reject lethal tumor challenges, the vaccine is much less effective in tumor-bearing mice, a situation that resembles patients with cancer. Therefore, it is critical to understand why DC vaccine fails to suppress tumor growth in tumor-bearing mice. We will address the following questions: 1) can DC vaccine induce tumor-specific CTL activity in tumor bearing mice? 2) Is a loss of tumor antigens responsible for failures to reject tumors? And 3) if CTL activity is found in tumor bearing mice, is it able prevent growth of metastatic lesions? Methodology: mouse colon adenocarcinoma cell line, CT26, grown in BALB/c, served as a mouse model of colon cancer. DC vaccines were generated by pulsing DCs with tumor lysate. Mice (3 groups) were inoculated s.c. with CT26 on day 0. CT26 lysate-pulsed DCs (5x10⁵ cells) were injected i.p. on day 5 and 10. In the first group, CTLs were isolated from spleen on day 21 and killing activity against CT26 was assessed. In the second group, tumors were harvested on day 21 and used as target cells for CT26-specific CTL killing assay to assess tumor antigen loss. In the third group, mice were challenged with tumor at a second site on Day 13. This group was designed to investigate whether DC vaccine can prevent distant metastasis. Results: 1) DC vaccine was able to induce tumor-specific CTL activity in animals with pre-established tumors, but failed to eradicate pre-established colon cancer. 2) Persistent tumors after DC vaccination were susceptible to CT26-specific CTL killing in vitro, suggesting no loss of tumor-specific antigens. 3) DC induced CTL activity in tumor-bearing mice did not protect animals against tumor challenge at a second site. Conclusions with implications: DC vaccine is able to induce tumor-specific CTL activity in tumor-bearing mice. These CTLs were not able to inhibit primary tumors in vivo, but has strong killing activity in vitro. Tumor growth despite existing tumor-specific CTLs suggests the presence of inhibitory factors suppressing in vivo CTL activity. Neutralization of immunosuppressive factors such as TGF-beta may lead to the development of a highly effective DC vaccine for the treatment of advanced colon cancer.

Calmodulin kinase II Activity is required for beta-adrenergic effects on sino-atrial impulse formation and atrioventricular conduction

Trainee: Michelle S.C. Khoo (ASCI new member: Mark E Anderson)

Michelle S.C.Khoo, Rong Zhang, Wei Zhang, Roger Colbran, James B. Atkinson, Mark E. Anderson

Vanderbilt University, Nashville, TN

L-type Ca²⁺ current (I_Ca) regulates sinoatrial node (SAN) and atrioventricular conduction (AVC); L-type Ca²⁺ channel activity is increased by beta-adrenergic signaling (betaAS) and the multifunctional Ca²⁺/calmodulin-dependent protein kinase II (CaMK II). betaAS hastens SAN impulse formation and AVC, but the role of CaMK II in modulating the SAN and AVC is poorly understood. We engineered mice with cardiac targeted CaMK II inhibition by expression of a specific CaMK II inhibitory peptide, AC3-I. Control mice had cardiac expression of an inactive peptide, AC3-C. Both peptides were linked to green fluorescent protein (GFP) to reveal widespread expression throughout the atria and the AV node. Cardiac homogenates from AC3-I mice had significantly lower CaMK II activity than AC3-C controls. 30 min ECG recordings in unanaesthetized and unrestrained mice with implanted telemeters showed no episodes of AV block in AC3-I (0/9) or AC3-C (0/5) at baseline. In contrast, 2^nd or 3^rd degree AVC block were observed in all AC3-I mice (9/9). But no AC3-C mice (0/5, P<0.001) after isoproterenol (Iso, 100 micro g i.p.). Signal-averaged ECG recordings showed baseline PP and PR intervals were not different between AC3-I and AC3-C mice, but AC3-I mice had a significant paradoxical increase in PP and PR intervals after Iso. Iso effects were not a non-specific consequence of transgenic protein expression because of the experimental design, and because the P wave and QRS intervals were not different in AC3-I and AC3-C mice. These studies show that CaMK II activity is important for normal manifestations of betaAS on SAN and AVC.

Enhancing DNA vaccine potency by co-administration of DNA encoding anti-apoptotic proteins

Trainee: Tae Woo Kim (ASCI new member: T.-C. Wu)

Tae Woo Kim, Chien-Fu Hung, Morris Ling, Liangmei He, J. Marie Hardwick, and T.-C. Wu

The Johns Hopkins University School of Medicine, Baltimore, MD

Intradermal vaccination via gene gun efficiently delivers DNA vaccines into dendritic cells (DCs) of the skin, resulting in their activation and priming of antigen-specific T cells in vivo. However, DCs have a limited lifespan, hindering their ability to prime antigen-specific T cells. We reason that a strategy that prolongs the survival of DNA-transfected DCs will enhance priming of antigen-specific T cells and DNA vaccine potency. Here we show that co-delivery of DNA encoding inhibitors of apoptosis with DNA encoding model antigens is able to prolong the survival of transfected DCs. More importantly, vaccinated mice exhibited significant enhancement in antigen-specific CD8+ T cell immune responses, resulting in a potent antitumor effect against antigen-expressing tumors. Among these anti-apoptotic factors, BCL-xL DNA demonstrated the greatest enhancement in antigen-specific immune responses and antitumor effects. Thus, co-administration of DNA vaccines with DNA encoding anti-apoptotic proteins represents an innovative approach to enhance DNA vaccine potency.

Mice with blood cells deficient in GPI-anchored proteins: a genetically precise murine model of Paroxysmal Nocturnal Hemoglobinuria (PNH)

Trainee: Shashikant Kulkarni (ASCI new member: Monica Bessler)

Washington University School of Medicine, St. Louis, MO

PNH is an acquired blood disorder characterized by intravascular hemolysis, cytopenia, and an increased propensity to thrombosis. PNH is caused by the clonal expansion of a hematopoietic progenitor cell that has acquired mutation in the X-linked PIGA gene. PIGA encodes a subunit of an enzyme essential in the biosynthesis of glycosyl phosphatidylinositol (GPI) anchors. Blood cells derived from the PNH clone are therefore deficient in all GPI-linked proteins. To study the biology of blood cells deficient in GPI-linked proteins, and to investigate the growth advantage of the PNH clone we generated mice that genetically mimic the human disease. By the use of the Cre-loxP system we inactivated the Piga gene in the early embryo, in the erythroid/megakaryocytic lineage and in hematopoietic stem cells. In mice mosaic for PIGA- cells the proportion of blood cells deficient in GPI-linked proteins was stable over an observation period of 18 months, indicating that a Piga gene mutation does not confer an intrinsic growth advantage to hematopoietic cells. PNH is found in 40-50% of patients with aplastic anemia. We therefore hypothesize that blood cells deficient in GPI-linked proteins have a growth or survival advantage when normal hematopoietic stem cells die. To test whether a Piga gene mutation alters the susceptibility to apoptosis we subjected thymocytes, granulocytes and bone marrow cells to specific apoptotic stimuli, including gamma-irradiation, anti-FAS antibody, dexamethasone, etoposide and serum starvation. For all stimuli tested the rate of apoptosis in PIGA- cells was similar when compared to PIGA+ cells from the same mouse or when compared to cells from a wild type control mouse. These findings indicate that a Piga gene mutation does not alter the susceptibility to cell death and reinforce the notion that additional factors combined with a Piga gene mutation are necessary to promote the clonal outgrowth of PIGA- cells.

Cbfb, a well-known regulator of hematopoiesis, cooperates with Cbfa1 and plays a critical role in bone development

Trainee: Mondira Kundu

Mondira Kundu¹, Amjad Javed², Yingze Yang³, Glen Nuckolls⁴, Maximillian Muenke⁵, Jane Lian², Paul Liu¹

¹Genetics and Molecular Biology Branch, ³Genetic Disease Research Branch, and ⁵Medical Genetics Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD; ²Department of Cell Biology, University of Massachusetts, Worcester, MA;⁴Cartilage Biology and Orthopaedics Branch, National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health, Bethesda, MD

Cbfa1 is essential for bone development in mouse models, and mutations in CBFA1 are found in 65-80% of patients with cleido-cranial dysplasia (CCD), a developmental syndrome characterized by a variety of skeletal and dental defects. Although all Cbfalpha family members (encoded by Cbfa genes) can interact with Cbfbeta (encoded by Cbfb) and form heterodimeric transcription complexes, a role for Cbfbeta in bone development has not been demonstrated. Cbfb^-/- embryos die during mid-gestation at 12.5 dpc due to severe hemorrhage and absence of definitive hematopoiesis, precluding the evaluation of Cbfb function at later time points. We fused the gene encoding green fluorescent protein (GFP) with the Cbfb gene by homologous recombination in murine embryonic stem (ES) cells. Heterozygous Cbfb-GFP mice had normal life spans and appeared normal. However, homozygous pups died within the first day after birth. While no major defects in hematopoiesis were detected and no bleeding diathesis was apparent, homozygous embryos exhibited a delay in endochondral as well as membranous ossification that was especially evident in the cranial and pelvic bones. The phenotype of these mice was similar to, but more severe than that of Cbfa1^+/- mice. In situ hybridization and GFP detection showed Cbfb expression in the developing bones. In addition, we demonstrated that Cbfbeta interacts with Cbfalpha1 and enhances its transactivation ability. Biochemical studies suggest that Cbfb-GFP represents a hypomorphic allele that maintains sufficient function in hematopoietic cells to bypass the early embryonic lethality, but uncovers a novel role for Cbfb in bone development. Thus, this study provides the first conclusive evidence that Cbfb plays a critical role in bone development and raises the possibility that mutations in CBFB may be responsible for some of the cases of CCD that are not linked to CBFA1 mutations.

Winner, Poster Session Award

RANKL:RANK signaling in the osteoblast induces bone formation

Trainee: Jonathan J. Lam

Jonathan J. Lam, F. Patrick Ross, and Steven L. Teitelbaum

Department of Pathology and Immunology, Washington University School of Medicine, Barnes-Jewish Hospital, Saint Louis, MO

Bone mass in adults is maintained by the coordinated action of osteoblasts, which deposit bone matrix, and osteoclasts, which resorb it. The cytokine RANKL (ligand for RANK, the receptor activator of NFB) governs bone homeostasis by regulating osteoclast development, activity, and survival. Surprisingly, we find that systemic treatment of mice with an oligomeric, internalization-resistant form of the osteoclastogenic cytokine RANKL profoundly stimulates new bone formation and enhances bone mineral density. RANKL:RANK signaling in pluripotent mesenchymal stem cells upregulates the master osteoblast transcription factor Cbf1 (core binding factor 1), thereby prompting osteoblast differentiation. In the mature osteoblast, RANKL:RANK signaling enhances proliferation and anabolic function. Blockade of RANKL:RANK-mediated ERK (extracellular signal-regulated kinase) activity abrogates the anabolic effect. Mice deficient in RANK exhibit significant impairment in the basal rate of bone anabolism; therefore, RANKL:RANK signaling plays a physiological role in osteogenesis. RANKL may prove an effective anabolic agent in the treatment of osteopenic disorders. We have solved the 3-dimensional structure of RANKL at a resolution of 2.6 � by x-ray crystallography. The crystal structure of RANKL reveals determinants of receptor-ligand specificity, enabling the rational design of pharmacologic modulators of the RANKL:RANK pathway to act as novel anabolic and anti-resorptive agents for bone.

Protein kinase CK2 promotes oncoprotein stability and inhibits the PTEN tumor suppressor

Trainee: David Y. Lou (ASCI new member: David C. Seldin)

David Y. Lou,, Diane H. Song, Padma L.Channavajhala, David C. Seldin

Department of Medicine, Boston University Medical Center, Boston MA

Protein kinase CK2 is a highly conserved, ubiquitous serine-threonine kinase that has over 170 cellular substrates. CK2 is overexpressed in many human tumors, and transgenic overexpression of CK2 in lymphocytes or the mammary gland produces tumors. To investigate the mechanism of transformation dependent upon CK2 overexpression, we have examined the effect of CK2 phosphorylation on several cellular oncoproteins. C-Myc is a substrate of CK2 and synergistic with it in lymphocyte transformation. Inhibition of CK2 activity reduces c-myc half-life in cells, and CK2 overexpression increases c-myc levels. Thus, the C-terminal phosphorylation of c-myc by CK2 appears to stabilize the protein. Similarly, CK2 inhibition reduces beta-catenin levels in cells. Threonine 393 of beta-catenin appears to be the major site of CK2 phosphorylation, and alanine mutagenesis of this residue reduces beta-catenin half life. CK2 potentiates beta-catenin/TCF-dependent transcription and Wnt signaling. PTEN is a tumor suppressor gene that encodes a phosphatase that antagonizes PI3K-dependent growth signaling, reducing cellular levels of phospho-AKT. PTEN is phosphorylated by CK2, and pharmacologic inhibition of CK2 reduces phospho-AKT in PTEN+ cells. Thus, these three critical transformation pathways are regulated by CK2 activity and targeted inhibition of CK2 may be useful in cancer therapy.

Antioxidant properties of PPAR-gamma ligands protect myocardium against reperfusion injury

Trainee: B. Molavi

B. Molavi, H.J. Chen, D.Y. Li, and J.L. Mehta

Division of Cardiovascular Medicine, University of Arkansas for Medical Sciences, Little Rock, AR

Reperfusion injury is being increasingly recognized as a cause of left ventricular dysfunction following prompt revascularization by pharmacologic or catheter-based interventions during acute myocardial infarction. This phenomenon can be defined as further deterioration of cardiac function despite restoration of blood flow. Reperfusion injury is believed to be caused by expression of pro-inflammatory and reactive oxygen species (ROS). Peroxisome proliferator activated receptor-gamma (PPAR-gamma) ligands exert potent anti-inflammatory and anti-oxidant effects in addition to their insulin-sensitizing effect. We postulated that these agents, accordingly, might be useful in limiting reperfusion injury even in non-diabetic settings. We tested this hypothesis in Sprague-Dawley rats fed either regular rat chow or a diet rich in PPAR-gamma ligand rosiglitazone (3mg/kg/day for 7-10 days). Rats were subjected to left coronary occlusion for 1 hr followed by reperfusion for 1 hr with continuous monitoring of left ventricular dp/dt, arterial pressure and heart rate. At the end of ischemia-reperfusion, infarct size (as % of area at risk) was measured by Evans blue perfusion/TTC staining, NADPH oxidase (P67) expression by immuno-histochemistry and Western analysis, and myocardial lipid peroxidation, as MDA. Ischemia alone caused similar decline in cardiac hemodynamics in both groups of rats, with further decline in control-diet group during reperfusion. The decline in cardiac hemodynamics during reperfusion was 75% less in rats fed rosiglitazone-rich diet (P<0.01 vs. control rats). Infarct size was also less in the rosiglitazone-fed rats (10�4% vs. 55�8% in the control group, P<0.01). Importantly, myocardial MDA levels and NADPH oxidase (P67) expression were dramatically reduced in the rosiglitazone-fed rats (P<0.01 vs. control group). These observations indicate that PPAR-gamma ligands have potential to limit myocardial infarct size and functional deterioration following ischemia by reducing NADPH oxidase activity and lipid peroxidation.

Mutations in ADAMTS-13 cause thrombotic thrombocytopenic purpura

Trainee: David Motto

Gallia Levy, David Motto, William Nichols, Han-Mou Tsai, and David Ginsburg

University of Michigan, Ann Arbor, MI

Thrombotic thrombocytopenic purpura (TTP) is a life threatening systemic illness of abrupt onset characterized by the pentad of fever, hemolytic anemia, thrombocytopenia, neurological symptoms, and renal dysfunction. Without treatment such as fresh frozen plasma infusion or plasma exchange, mortality may exceed 90%. Despite advances in therapy, the pathogenesis of TTP remains elusive, and the specific factors in plasma that mediate recovery from acute episodes have yet to be identified. Proteolysis of the blood clotting protein von Willebrand Factor (vWF) is observed in normal plasma, and this proteolyic activity is decreased in patients with TTP. The identity of this protease and its potential role in the pathophysiology of TTP remain unknown. We performed genome wide linkage analysis in four pedigrees of individuals with inherited TTP and mapped the responsible genetic locus to a 2.3 centimorgan interval on chromosome 9q34. A predicted gene in this interval was found to encode a previously unidentified member of the ADAMTS zinc metalloproteinase family, which we have designated ADAMTS-13. Sequence analysis of patient genomic DNA identified 12 distinct mutations in the ADAMTS-13 gene, accounting for 14 of the 15 disease alleles studied. These data identify deficiency of ADAMTS-13 as the underlying molecular mechanism responsible for TTP, and suggest that the physiologic proteolysis of vWF and/or other ADAMTS-13 substrates is required for normal vascular homeostasis.

NOD2 genotype-phenotype correlations in Canadian Crohn's Disease families

Trainee: Bill Newman

Bill Newman, Mark Silverberg, Kathy Siminovitch

Departments of Medicine, Immunology and Medical Genetics and Microbiology, Mount Sinai Hospital, Toronto, Canada

To investigate the phenotypic consequences of NOD2 mutations, 507 Crohn's disease (CD) patients were genotyped for the CD-associated R702W, G908R and 1007fs NOD2 mutations. This group included 238 CD patients with a first-degree relative affected by inflammatory bowel disease (IBD), and 269 CD patients with no family history of IBD and 100 controls. Among the CD patients, the average age of diagnosis was 20.7 years (range 3-66), 26.2% were Ashkenazi Jewish and 70.6% were non-Jewish Caucasian. The analysis revealed at least one of the three mutations to be present in 31.5% of the CD patients (of these 25.8% had one mutation and 5.7% two mutations) compared to 20% of controls. There was no difference between the mutant allele frequencies in familial and sporadic cases. Interestingly, there was no difference in mutant allele frequency in CD patients with a first-degree relative with CD (17.1%) or a first-degree relative with ulcerative colitis (UC) (18.8%, p=0.75). However, the allele frequency in UC-affected siblings was lower (10%) with significant lack of concordance between CD and UC affected siblings (p=0.01). The relative risk of CD associated with R702W, G908R and 1007fs were 1.7, 2.2 and 1.9 respectively. No significant association was found between presence of mutation and age of diagnosis, gender, smoking status, ethnicity (Caucasian versus Jewish) or disease severity. However, a significant positive association was detected between NOD2 status and the presence of distal ileal involvement (p=0.001) and significant negative associations were detected between NOD2 status and involvement of the large bowel: transverse colon (p=0.003), descending colon (p=0.0008) and sigmoid colon (p=0.009). Our results suggest that NOD2 mutations account for a smaller proportion of the genetic susceptibility for CD compared to other CD patient populations and confirm the significant influence of NOD2 mutations on the site of bowel involvement in CD.

Genetic variation in NOD2 confers susceptibility to Crohn's disease

Trainee: C. Piraka

Ogura Y, Bonen DK, Inohara N, Nicolae DL, Chen FF, Piraka C, Cho JH, and Nunez G.

Department of Pathology and Comprehensive Cancer Center, The University of Michigan Medical School, Ann Arbor, MI; The Martin Boyer Laboratories, Gastroenterology Section, Department of Medicine, The University of Chicago Hospitals, Chicago, IL

Crohn's disease is a chronic inflammatory disorder of the gastrointestinal tract, which is thought to result from the effect of environmental factors in a genetically predisposed host. A gene location in the pericentromeric region of chromosome 16, IBD1, that contributes to susceptibility to Crohn's disease has been established through multiple linkage studies, but the specific gene(s) has not been identified. NOD2, a gene that encodes a protein with homology to plant disease resistance gene products, is located in the peak region of linkage on chromosome 16. Using the transmission disequilibrium test and case-control analysis, two point mutations and a frameshift mutation caused by a cytosine insertion, 3020insC, which is expected to encode a truncated NOD2 protein, are associated with Crohn's disease. Wild-type NOD2 activates nuclear factor NF-kappaB, making it responsive to bacterial lipopolysaccharides; however, this induction was deficient in mutant NOD2. These results implicate NOD2 in susceptibility to Crohn's disease, and suggest a link between an innate immune response to bacterial components and development of disease.

beta-adrenergic receptor blockers restore cardiac calcium release channel (ryanodine receptor) structure and function in heart failure

Trainee: Alfonso Prieto

Columbia University, New York, NY

Introduction: beta-adrenergic receptor blockade has emerged as one of the most effective therapies for heart failure (HF). Calcium (Ca²⁺) release channels on the sarcoplasmic reticulum (SR) in cardiomyocytes are required for excitation-contraction (EC) coupling. The cardiac isoform of the ryanodine receptor (RyR2) is a tetramer comprised of four 565,000 dalton subunits. In animal as well as human HF, RyR2 is PKA-hyperphosphorylated, and as a consequence, FKBP12.6 is dissociated from the complex, resulting in abnormal channel function. Methods: Isolation of cardiac SR, preparation of homogenates, immuno-precipitation, back-phosphorylation, single channel recordings, and western blots were performed as previously described. Canine Heart Failure Model and beta-Blocker Therapy: After baseline measurements were recorded, rapid LV pacing was initiated at 210 bpm for 3 weeks, followed by an additional week of pacing at 240 bpm. Metoprolol was strated at 2 weeks after the initiation of pacing and continued for 2 weeks. Results: 1) Scatchard analysis of [³H]DHA binding data demonstrated that treatment of paced animals with metoprolol increased both the affinity and concentration of beta-adrenergic receptors. 2) beta-Blocker treatment protects RyR2 from hyperphosphorylation by PKA. 3) beta-blocker treatment normalizes amount of the phosphatases that co-immunoprecipitate with the channel. 4) Single channel recordings of RyR2 channels demonstrated that beta-adrenergic receptor blockade normalizes RyR2 function in failing hearts. Conclusions: This data demonstrates that systemic beta-blocker therapy can reverse the defective regulation of RyR2 associated with HF, providing a potentially important molecular target for this disease. Treatment of HF canines with metropolol reduced the PKA phosphorylation of RyR2 to that found in non-failing hearts. PKA hyperphosphorylated channels exhibited an increased open probability and/or increased frequency at low cytosolic Ca²⁺. The finding that this improved beta-adrenergic signaling leads to a reversal of the PKA hyperphosphorylation of channels from failing hearts is suggestive that local regulation of kinases and phosphatases control the phosphorylation of RyR2.

Genetic correlates of protective immunity against TB

Trainee: Gurvaneet Randhawa (ASCI new member: William R. Bishai)

Randhawa, GS¹; Guo, QB²; Bishai, WS³

Preventive Medicine¹, Oncology², and Medicine³, Johns Hopkins University, Baltimore, MD

Incomplete understanding of the host response to Tuberculosis (TB) is a major barrier to vaccine development. The completion of the human, mouse and M. TB genome projects has enabled the use of DNA microarrays to ascertain transcription expression profiles that can be used for TB vaccine research.

We are using mouse genome-based DNA microarrays to ascertain the temporal changes in host gene expression profile that occur both in vitro (alveolar macrophages) and in vivo (mouse) models of infection with either BCG or virulent M. tuberculosis.

Our initial studies of mouse alveolar macrophages reveal that over 200 genes are activated and over 300 genes are repressed after 25 hours of BCG infection. Nearly half of these genes do not have a known function. The activated genes with known function include TNF-alpha, iNOS2, IRF7, IL6, C/EBP beta, matrix metalloproteinase 9 etc. The repressed genes include elastase inhibitor, gelsolin, CXCR4 etc. The expression profile changes with time, for example a greater number of metabolic genes are activated at 6 hours as compared to 25 hours after infection. A semi-quantitative RT-PCR assay confirmed the data obtained by microarray.

Data from in vivo BCG infection in a mouse indicates that gene expression profile in the spleen is quite different from that of alveolar macrophages. A large number of genes involved in G-protein signaling, and a different group of cytokines, are activated.

We plan to challenge immunized mice with virulent M. tuberculosis to evaluate efficacy of different candidate vaccines and ascertain those genes that correlate with protective immunity. The valid correlates may be used to evaluate new TB vaccines as well as in novel diagnostic tests that evaluate the host immunity to TB.

Purification, characterization and differentiation of multipotent adult progenitor cells from post-natal human bone marrow

Trainee: Morayma Reyes

Reyes M., Verfaillie C.

Stem Cell Institute, Department of Medicine, and Cancer Center, University of Minnesota Medical School, Minneapolis, MN

We report here the isolation, characterization and differentiation of a human bone marrow stem cell that can at the single cell level give rise to mesodermal, neuroectodermal and endodermal lineages. We named these cells Mutltipotent Adult Progenitor Cells or MAPC. MAPC were selected by depleting bone marrow mononuclear cells (BMMNC) from >30 normal human donors of CD45/Glycophorin-A (GlyA) cells. Cells were cultured on fibronectin with epidermal growth factor (EGF) and platelet derived growth factor (PDGF)-BB and =/< 2% fetal calf serum (FCS). 1 / 5x10³ CD45^-GlyA^- cells, or 1 / 10⁶ BMMNC, gave rise to clusters of small adherent cells. Cell doubling time was 48-60 hours, and cells have been culture expanded for >100 cell doublings and are karyotypically normal. MAPC are CD34^-, CD44^low, CD45^-, HLA-class-I^- and HLA-DR^-. MAPC differentiated into cells of limb-bud mesoderm (osteoblasts, chondrocytes, adipocytes, stroma cells and skeletal myoblasts) as well as visceral mesoderm (endothelial cells). MAPC derived endothelial cells can make vascular tubes and participate in neoangenesis when transplanted in NOD/SCID mice. MAPC can also, depending on the cytokines used, differentiate into oligodendrocytes, astrocytes, and functional, dopaminergic, serotoninergic, or gamma-amino-butyric-acid (GABA)-ergic neurons. Following coculture with the glioblastoma cell line U-87, MAPC-derived neurons matured further and generated tetrodotoxin (TTX)-sensitive Na-gated voltage action potentials. Finally, we determined that MAPC can differentiate into endodermal epithelial cells with morphological and phenotypic features of hepatocytes. These hepatocytes like cells can produce urea, albumin, glycogen and are polarized. Differentiation methods were selected from an extensive trial and error approach based on embryogenesis and ES cell in-vitro differentiation literature. Retroviral marking was used to definitively prove that a single MAPC and its progeny can differentiate into endothelial cells (mesoderm), neurons (ectoderm) and hepatocytes (endoderm). Thus, MAPC that proliferate without obvious senescence under clinically applicable conditions and differentiate at the single cell level not only into mesodermal but also neuroectodermal and endodermal cells, may be an ideal source for treatment of many genetic and acquired diseases.

In vivo selection as a gene therapy method to correct a red cell disorder

Trainee: Robert E. Richard (ASCI new member: C. Anthony Blau)

Robert E. Richard, Jessica Ieremia, C. Anthony Blau

Division of Hematology, University of Washington School of Medicine, Seattle, WA

Gene therapy for red blood cell disorders has proven difficult. Low rates of gene transfer into stem cells translate into a low percentage of transduced erythrocytes. Given the present technology, patients will need to undergo full myeloablation with expected toxicities and low possibility of therapeutic success. We have developed a method to selectively expand the red cell population that is expressing a transgene. We transduce bone marrow cells with a retrovirus that encodes GFP (a fluorescent marker) and a thrombopoietin receptor (mpl) fusion protein that is activated using a chemical inducer of dimerization (CID). The fusion protein contains the signaling portion of the mpl protein and is dependent on the CID for its activity. In this way cell expansion can be regulated in vivo (Jin, et al. Nat. Genet., 2000). We have combined this technique with low dose radiation to selectively expand red cells in mice with pyruvate kinase deficiency. Bone marrow cells were isolated from the wild type mouse (CBA/N), transduced with a retroviral vector that encodes GFP and the mpl fusion protein. Mutant pyruvate kinase deficient mice were conditioned with low dose radiation and infused with transduced cells. Mice were then followed for improvements in their rbc number and reticulocyte count. All mice that received 400 cGy and up to 10 x 10⁶ transduced bone marrow cells have maintained the donor graft as evidenced by expression of GFP in the erythrocytes and platelets. This cohort of mice has shown partial correction of the hemolytic anemia as evidenced by red blood cell number and reticulocyte count. By treating the mice with CID (AP20187) we can selectively expand the transduced cell population. These experiments demonstrate the potential of manipulating chimerism in the red cell lineage using in vivo selection.

Role of signal transducer and activator of transcription 3 (STAT-3) in immune tolerance mediated by antigen-presenting cells (APCs)

Trainee: Eduardo Sotomayor (ASCI new member: Hyam I. Levitsky)

F. Cheng ¹, A.G. Cuenca¹, H. Wang¹, M. Huang¹, H. Yu¹, R. Jove¹, H. Levitsky² and E.M. Sotomayor¹

¹H. Lee Moffitt Cancer Center & Research Institute, University of South Florida, Tampa, FL, and the ²Sidney Kimmel Cancer Center at Johns Hopkins University, Baltimore, MD

Bone marrow derived antigen-presenting cells (APCs) play a critical role in priming T-cell responses against tumor antigens. Our recent demonstration that APCs are also required for the induction of T-cell tolerance to tumor antigens places these cells at the crossroads of highly divergent T-cell outcomes. Although emerging evidence points to the state of activation/differentiation of the APC as the determinant of T-cell priming versus tolerance, the signaling mechanism(s) -- at the APC level -- involved in this critical decision remains to be elucidated.

Recent studies have shown that Stat3 may act as a negative regulator of inflammatory responses. We therefore explored whether inhibition of JAK-Stat3 by the tyrosine kinase inhibitor, tyrphostin AG490, may modulate APC's function. In vitro treatment of peritoneal elicited macrophages (PEM) with AG490+LPS resulted in an enhanced presentation of HA-peptide to na�ve CD4⁺ T cells specific for a MHC class II restricted epitope of influenza hemagglutinin (HA). These clonotypic T-cells display an increased antigen-specific proliferation and cytokine production in response to cognate antigen presented by AG490-treated APCs as compared to those T cells encountering HA-peptide on untreated or LPS-treated PEM. This enhancement in APCs' function correlates with inhibition of STAT3 DNA-binding activity. More importantly, tolerant CD4⁺ T-cells (isolated from tumor bearing mice) exposed to AG490-treated APCs, recover their ability to proliferate and produce IL-2 and IFN-gamma in response to cognate antigen stimulation. In sharp contrast, tolerant T cells encountering antigen presented by untreated APCs remained fully unresponsive. To confirm these results, we evaluated the antigen-presenting capabilities of PEM from mice with a cell type-specific disruption of the Stat3 gene (LysMcre/Stat3^flox/- mutant mice). In response to stimulation with LPS, Stat3^-/- PEM breaks anergy of T-cells isolated from tumor bearing mice. Phenotypic analysis of Stat3^-/- PEM, revealed an increased expression of MHC class II as well as the co-stimulatory molecules B7.1, B7.2 and B7-DC. Furthermore, utilizing multi-template RNA probes we found that LPS-treated Stat3^-/- PEMs display higher mRNA levels of the pro-inflammatory mediators IL-12, MIP-1alpha, MIP1-beta, IP-10, RANTES, but not IL-10. Neutralization experiments with anti-IL-12 and anti-IP10 antibodies points to these two pro-inflammatory mediators as playing a critical role in the ability of Stat3^-/- PEM to break T-cell anergy.

These findings demonstrate that STAT3 signaling pathway negatively regulates the antigen-presenting capabilities of APCs. Furthermore, these studies provide a molecular target to overcome immune tolerance, a significant barrier limiting the efficacy of cancer immunotherapy.

Insulin stimulates skeletal muscle toward a phenotype with greater oxidative capacity

Trainee: Craig S. Stump

Craig S. Stump, Kevin R. Short, Maureen L. Bigelow, Jill M. Schimke, and K. Sreekumaran Nair

Department of Internal Medicine, Division of Endocrinology and Metabolism, Mayo Clinic, Rochester, MN

Insulin plays a key role in post-prandial fuel metabolism by skeletal muscle, where fuel oxidation and ATP production occur primarily in the mitochondria. However, the effect of insulin on skeletal muscle mitochondrial function has not been defined. We hypothesized that insulin stimulates the capacity of mitochondria to produce ATP by enhancing mitochondrial protein synthesis. Healthy subjects (N=6; ages 18-43 y) were studied during eight hours of euglycemic insulin infusion (1.5 mU/kg FFM/ min) to high physiological levels (57�6 mU/L), somatostatin (SRIH) suppression of endogenous insulin, and glucagon and growth hormone replacement; and on a separate day during saline only infusion. A similar group of subjects (N=6) was studied during a lower replacement rate of insulin infusion (0.15 mU/kg FFM/ min) with SRIH. We obtained vastus lateralis muscle biopsies (approx. 350 mg) at baseline and after four and eight hours. Increases in mRNA transcripts from baseline were significantly greater (p<0.05) for the high dose insulin compared to the saline condition at eight hours for NADH dehydrogenase (NADH) IV and cyotchrome oxidase (COX) IV (164-180%). Insulin also preferentially increased myosin heavy chain (MHC) I mRNA expression (68%). Furthermore, mitochondrial protein synthesis rates using L-[1, 2 ¹³C]-leucine as a tracer were significantly increased (20-24%) between four and eight hours of high dose insulin compared to the saline or low dose insulin replacement conditions. These changes were associated with comparable increases in citrate synthase (28%) and cytochrome oxidase (29%) enzyme activities and mitochondrial ATP production rates (32-42%). The data indicate that insulin enhances the expression of mRNA transcripts encoding mitochondrial proteins from both the nuclear (COX IV) and mitochondrial (NADH IV) genomes, as well MHC-I which is the predominant myosin isoform found in slow-oxidative muscle fibers. Insulin also increased mitochondrial protein synthesis, oxidative enzyme capacity, and mitochondrial ATP production potential. These findings suggest that insulin stimulates skeletal muscle toward a phenotype with greater oxidative capacity. This work was funded by NIH RO1 DK41973 and M01 RR00585.

Molecular portraits of diabetic nephropathy

Trainee: Katalin Susztak (ASCI new member: Erwin P. B�ttinger)

Katalin Susztak², Kumar Sharma¹, Akiva Novetsky², Dan Liang², Dona Wu², Stephen Dunn¹, Yanquing Zhu¹, Wenjun Ju², and Erwin P. B�ttinger²

¹Department of Medicine / Nephrology, Thomas Jefferson University Philadelphia, PA, and ²Department of Medicine / Nephrology, Albert Einstein College of Medicine, Bronx, NY

Diabetic nephropathy is the leading cause of end stage renal disease. Our insight into the molecular mechanisms that underlie the origin and progression of diabetic nephropathy remains limited largely because conventional research tools restricted investigators to focus on single genes or isolated pathways. We used the cDNA microarray technology, which allows rapid screening of almost the entire transcriptome to identify new genes with expression profiles, that are characteristic to diabetic nephropathy.

Animal models have been of great benefit in testing various pathways that may contribute to diabetic nephropathy. The db/db mouse model of type 2 diabetes and the streptozotocin-induced type 1 diabetic mouse models were examined. Total RNA from whole kidneys was used for microarray experiments. All diabetic animals manifested hyperglycemia and albuminuria, when sacrificed at 22 weeks of age. Glomerular mesangial matrix expansion was also evaluated in all animals.

We implemented specialized T-statistics using Statistical Analysis for Microarray Program and identified 484 transcripts with statistically-significant and at least two-fold difference in mean expression levels in multiple group comparisons. We have found known e.g. angiotensinogen, cyclin dependent kinase (p21), protein C, and novel e.g. CD36 genes, that are differentially expressed in diabetic animals. By using specialized statistical learning methods we identified novel genes those expression could discriminate animals with hyperglycemia, albuminuria, and mesangial matrix expansion. Based on expression level of CD36, kidney androgen regulated protein and phosphoadenosine phosphosulfotransferase we were able to discriminate animals with diabetes from control mice. The expression of 3-beta-hydroxysteroid dehydrogenase-delta-4,5 isomerase were predicative for glomerular matrix expansion. Expression patterns of these top predictor genes was confirmed by real time quantitative reverse transcription PCR analysis. These genes have also been used to predict phenotypes (hyperglycemia, glomerular disease) of unknown samples. Further analysis of individual genes will determine their functional role in mediating specific disease parameters.

Calcium-modulating cyclophilin ligand regulates receptor trafficking

Trainee: David D. Tran (ASCI new member: Richard J. Bram)

David D. Tran¹, Helen Russell², Gotz von Bulow³, Shari L. Sutor^1,4, Richard J. Bram^1,4

¹Department of Immunology, ⁴Department of Pediatrics and Adolescent Medicine, Mayo Medical and Graduate Schools, Mayo Clinic, Rochester, MN; ²Department of Genetics, St. Jude Children's Research Hospital, Memphis, TN; ³Department of Microbiology and Immunology, Indiana University, Indianapolis, IN

Calcium-Modulating-cyclophilin Ligand (CAML) is a ubiquitously expressed intracellular protein implicated in signal transduction events that regulate early lymphocyte activation. Its exact mechanism of action and physiologic significance are unknown, however. CAML has been implicated in the function of TACI (Transmembrane Activator and CAML Interactor) by identification of a physical association between the two proteins in B-lymphocytes. TACI is of interest because it is required for a normal antibody response to T-independent type II antigens (TI-2 Ag), such as bacterial capsular polysaccharides, thus playing a role in host defense against certain types of bacterial infections. Conversely, enhanced activity of TACI or related receptors is thought to contribute to some autoimmune diseases, such as Systemic Lupus Erythematosis. Understanding the in vivo functions of TACI will in part depend upon elucidation of the mechanism of action of CAML, its primary binding partner.

Here, we report that CAML gene targeting results in embryonic lethality at day 6-6.5. CAML-null embryos share many similarities with those of Epidermal Growth Factor Receptor (EGFR)-deficient embryos, suggesting a potential connection between the 2 proteins. Using normal human cell lines expressing a dominant negative form of CAML or cells derived from CAML-null embryos, we show that CAML is critically important for EGFR-induced mitogenesis. Our data suggests that CAML regulate EGFR intracellular trafficking and EGF-dependent proliferation through its binding both to the receptor and F-actin. We conclude that CAML is an important mediator of receptor function through its ability to link certain cell surface proteins to the actin cytoskeleton. Development of a tissue-specific CAML KO mouse will help further define the specific role of CAML in immunity.

Inhibition of PDGF receptor alpha tyrosine kinases by imatinib mesylate (Gleevec) in human soft-tissue sarcoma cell lines

Trainee: Jonathan C. Trent

Jonathan C. Trent, Jonathan A. Hannay, Elizabeth Buchdunger, PhD., Lan Li, Shreyaskumar Patel, Robert S. Benjamin, Raphael E. Pollock, and Dihua Yu

Department of Sarcoma Medical Oncology and Division of Surgery, The University of Texas M.D. Anderson Cancer Center, Houston, TX

In the U.S. there are 8,100 new cases of tumors arising from connective tissues leading to approximately 4,600 deaths per year. The majority of these patients in the younger years of their lives; thus, the disease is a significant public health problem despite its low incidence. Doxorubicin and ifosfamide are the two most active agents in the treatment of soft-tissue sarcomas (STSs). After failure of these, other agents have response rates of 10%-15%. Thus we investigated the potential therapeutic benefits of the tyrosine kinase inhibitor STI-571 in soft-tissue sarcomas. The objectives of our laboratory investigation are (1) to characterize the anti-tumor activity of STI-571 alone and with doxorubicin, in vitro; (2) to determine whether the inhibition of signal transduction molecules by STI-571 causes cell cycle arrest, apoptosis, inhibition of the invasive phenotype and/or abrogation of angiogenesis; (3) to determine the anti-tumor activity of STI-571 alone and with doxorubicin in tumor xenografts. Proliferation assays were performed with exposure to incremental drug concentrations and viability assessed by MTS reagent. We have investigated the efficacy of STI-571 (0.4-20 microM) alone and combined with doxorubicin (0.1-10 microM) in two rhabdomyosarcoma cell lines, A204 and RD. A204 cells treated with doxorubicin, STI-571, or the combination, exhibit a 40%, 20%, and 60% absolute reduction in percent viability, respectively. STI-571 increased the relative percentage of A204 cells in G1 (by 57%) and decreased S-phase (by 63%) when compared to untreated cells, by flow cytometry. TUNEL assay of STI-571-treated A204 cells revealed an increase in apoptosis from 4% to 15%. Soft agar colony formation assay revealed a 48% reduction in the anchorage independent growth of STI-571-treated A204 cells. Conditioned media from STI-571-treated A204 cells provided less support of endothelial cell growth. Zymogen assay revealed no effect of STI-571 on the in vitro activity of matrix metalloproteinases MMP-2 and MMP-9. In order to determine the molecular mechanism by which STI-571 inhibits tumorigenic properties of A204 but not RD, we initially screened these cell lines for expression of STI-571 targets. We found that both cell lines expressed PDGFR-beta, Abl, and Arg. Neither cell line expressed detectable levels of Kit. Interestingly, A204 but not RD expressed PDGFR-alpha. PDGF-mediated phosphorylation of PDGFR, ERK-2, and AKT-1 was completely abrogated by pretreatment of A204 cells with STI-571. In vivo tumorigencity assays to evaluate the anti-tumor activity of STI-571 against the A-204 rhabdomyosarcoma cell line are ongoing. In vitro results suggest an additive, non-overlapping effect of doxorubicin+STI-571, indicating a potentially useful combination. Together, these efforts should result in new mechanism-driven therapies that are efficacious in soft-tissue sarcoma.

Cox-2 and diacyglycerol kinase - delta may cooperate to shed epidermal growth factor ligands

Trainee: Scott C. Ulmer

Scott C. Ulmer, Matthew K. Topham, and Stephen M. Prescott

The Huntsman Cancer Institute at the University of Utah, Salt Lake City, UT

Prostaglandins, among other activities, regulate cell growth and differentiation. Cyclooxygenase-2 (Cox-2) is an enzyme that catalyzes the conversion of arachidonic acid to prostaglandins when induced by inflammatory cytokines, growth factors, or proto-oncogenes. Cox-2 has been shown to be overexpressed in many epithelial malignancies specifically colorectal cancer where the enzyme is the rate limiting step for neoplastic transformation. The epidermal growth factor receptor (EGFR) is also required for cell growth and differentiation and is overexpressed in most human malignancies. Recently (Pai, R. et al. Nature Medicine 2002, 8(3):289.), prostaglandin-E2 was shown to transactivate the EGFR through an undefined mechanism requiring transforming growth factor - alpha (TGF-alpha). We have found that diacylglycerol kinase - delta (DGK-delta), an isoform of mammalian diacyglycerol kinase that catalyzes the phosphorylation of diacylglycerol to produce phosphatidic acid, is necessary for TGF-alpha release and may be involved in TGF-alpha shedding mediated by Cox-2. Overexpression of DGK-delta may be important in Cox-2 mediated malignant transformation. DGK-delta would then be a novel, specific target for pharmacological inhibition with potential therapeutic benefit in a wide range of human malignancies.

Homeostatic proliferation antagonizes tolerance induction

Trainee: Tracy L. Vanasek (ASCI new member: Daniel L. Mueller)

Department of Medicine and Center for Immunology, University of Minnesota Medical School, Minneapolis, MN

We postulate that Ag-driven cell cycle progression in BALB/c nu/nu recipients is enhanced by the pressure for homeostatic proliferation and that this prevents or reverses clonal anergy induction. CFSE-labeled DO11.10 T cells transferred into either wild-type BALB/c (WT) or T cell deficient BALB/c nu/nu (NU) recipients were stimulated with OVAp. Both Ag-driven cell cycle progression (as determined by CFSE dilution) and recall IL-2 production were higher in the NU recipients as compared to WT recipients. Nevertheless, the addition of the immunosuppressent Rapamycin, a potent anti-proliferative agent, resulted in decreased proliferation in response to initial stimulation and redu