FAQ

Study Sections

Results of FARE 2005

Results of FARE 2004

Results of FARE 2003

Results of FARE 2002

Results of FARE 2001

Winning Abstracts from FARE2003

Abraham Zangen, Ph.D., NIDA
Rewarding and Psychomotor Stimulant Effects of the Endogenous Peptide Endomorphin-1: Anterior-Posterior Differences Within the Ventral Tegmental Area and Lack of Effect in Nucleus Accumbens
Endomorphin-1 (EM-1) is a recently isolated endogenous peptide having potent analgesic activity and high affinity and selectivity for the µ-opioid receptor. We’ve studied the rewarding and psychomotor stimulant effects of EM-1 in specific brain regions and searched for anatomical specificity within these regions. We found that rats would learn without priming or response shaping to lever-press for microinjections of EM-1 directly into the ventral tegmental area (VTA); responding was most vigorous for high-dose injections into the posterior VTA. Rats did not learn to lever-press for microinjections of EM-1 into the nucleus accumbens (NAS) or regions just dorsal to the VTA. Lever-pressing for EM-1 in the VTA was extinguished when vehicle was substituted for the peptide, and was reinstated when EM-1 reinforcement was re-established. Conditioned place preference was established by EM-1 injections into the posterior, but not the anterior VTA or the NAS core or shell. Injection of EM-1 (0.1-1.0 nmol) into the posterior VTA induced robust increases in the locomotor activity, while injections into the anterior VTA had very weak locomotor stimulating effects. When injected into the NAS (core or shell), EM-1 (0.1-10.0 nmol) did not affect locomotor activity. Using fluorogold microinjections into the posterior and anterior VTA combined with immunocytochemical detection of EM-1 containing cells, we’ve identified EM-1 containing cells around the arcuate nucleus of the hypothalamus that project to the posterior VTA. Finally, injection of a specific antibody for EM-1 into the posterior VTA (but not the anterior VTA or the NAS) significantly decreased responses for sucrose reward. These findings implicate the posterior portion of the VTA as a highly specific and sensitive site for opioid reward. Moreover, our finding suggest a role for EM-1-containing cells originating in the hypothalamus and projecting to the posterior VTA in the rewarding effects of natural incentives and drugs of abuse.

Adrian Wiestner, MD/PhD, NHLBI
Zap70 expression is a distinctive feature of B-CLL cells with unmutated immunoglobulin genes and can be useful for diagnostic application.
The clinical course of B-cell chronic lymphocytic leukemia is varied. While some patients have a chronic lymphocytosis without any need for therapeutic interventions, other patients may die despite aggressive treatment. The classic staging systems provide only limited prognostic information in newly diagnosed patients. Recently, the presence or absence of somatic mutations in the immunglobulin (Ig) variable region genes has been shown to distinguish between two disease subsets conferring important prognostic information. A median survival of 95 months was found in patients with unmutated Ig genes versus 293 months in patients with mutated Ig genes (Hamblin, 1999). Unfortunately, the ability to sequence Ig genes is not available in most clinical laboratories. More recently, it was shown that the Ig mutational status correlates with two distinct CLL populations as characterized by large scale gene expression profiling (Rosenwald, 2001). Surprisingly, zap70, a tyrosine kinase involved in T cell receptor signaling, was among the strongest distinction genes. To assess zap70 protein expression in the two groups we performed immunoprecipitations followed by Western blotting from CD19 selected peripheral blood lymphocytes in 18 patients. In 7 patients with unmutated Ig there was a strong specific signal on Western blots whereas 10 of 11 patients scored negative. Immunohistochemistry (IHC) confirmed strong expression of zap70 in T cells and B-CLL cells with unmutated Ig genes both in peripheral blood and bone marrow samples. In contrast B-CLL cells with mutated Ig genes were negative for zap70 expression by IHC. These results indicate that testing for zap70 expression can serve as clinically useful assay to distinguish the two groups of CLL and yield important prognostic information. Furthermore, as zap70 characterizes B-CLL with a distinct biology it may be possible to select patients for targeted therapeutic strategies based on zap70 expression.

Ahmad R Hariri, Ph.D., NIMH
FUNCTIONAL GENOMICS OF THE SEROTONIN TRANSPORTER: IMPACT ON THE RESPONSE OF THE HUMAN AMYGDALA
BACKGROUND: Two common allelic variants in the promotor region of the serotonin transporter (5-HTT) gene have been differentially linked to anxiety-related behavioral traits and acquisition of conditioned fear responses. Specifically, individuals possessing the short (s) allelic variant, which has been associated with decreased 5-HTT expression and subsequent 5-HT uptake, are more likely to display abnormal levels of anxiety and develop affective illness than those possessing only the long (l) allelic variant, associated with increased 5-HTT expression and 5-HT uptake. We utilized fMRI to explore the relationship between this 5-HTT gene polymorphism and the response of brain regions implicated in anxiety and fear behavior, namely the amygdala, during the processing of emotional facial expressions
METHODS: Fourteen subjects (l/l = 7 & s/s, s/l = 7) completed a blocked paradigm where a control condition was interleaved with an emotion processing condition where subjects matched the affect of angry and fearful faces. Functional images were acquired on a GE 3T scanner using a gradient echo EPI sequence. Data sets were analyzed for task-specific BOLD responses using second-level random effects approaches in SPM99. Genotyping for the VNTR promotor polymorphism was by RFLP-gel methods
RESULTS: In all subjects, there was a significant BOLD response in the amygdala during perceptual processing of angry and fearful facial expressions. Direct group comparisons revealed that individuals possessing the s allele (s/s or s/l) had a significantly greater amygdala response in comparison to l/l individuals
CONCLUSIONS: Our results implicate a genetically determined link between 5-HTT function and the response of brain regions critical for emotion processing. Specifically, they suggest that individuals with the less efficient s allele variant, who presumably have relatively higher levels of synaptic 5-HT, exhibit an increased amygdala response to fearful stimuli, while those homozygous for the l allele exhibit a decreased amygdala response. These results are the first to demonstrate a relationship between 5-HTT polymorphism and the differential response of brain regions important for emotion processing. Furthermore, they provide insight to the underlying neural substrates for the increased anxiety and affective spectrum disturbances associated with the less efficient s allele variant of the 5-HTT.

Alberto Inga, PhD, NIEHS
Characterization of p53 DNA binding affinity, transactivation capacity and effects of post-translational modifications using a yeast-based system
Cellular responses to stress usually involve broad modulation of gene expression. This is generally achieved by rapid adaptive activation of transcription factors that target genes and biological responses. Sequence-specific DNA binding at response elements (REs) of target genes is a necessary step in the complex process through which transcription factors stimulate gene expression. Thus, differences in relative binding affinity for individual REs can lead to differential transactivation and changes in biological responses. The binding differences could result from sequence divergence between REs, availability of cofactors, regulation of nuclear levels of the transcription factor or epigenetic changes that modify DNA binding affinity. All these possibilities have been invoked to explain functional regulation of p53. While gene expression profiles demonstrate p53 target selectivity, the unique features of chromatin at different target genes make it difficult to assess the specific role of DNA binding affinity and its regulation in achieving transactivation selectivity
We developed an in vivo system based in yeast to determine p53 transactivation capacity under isogenic conditions where i) all factors affecting transcription are kept constant with the exception of 20-22 bp p53 REs and ii) the amount of p53 can be varied using a rheostatable promoter. We determined the transactivation capacity of wild type (WT) p53 towards 30 REs placed upstream of either an ADE2 color reporter gene or a quantifiable luciferase reporter gene. Dramatic differences in intrinsic DNA binding affinity between the REs were found
We are also examining the effect on transactivation of amino acid changes at sites that are phosphorylated (P-) by ATM, ATR, CHK1, CHK2, HIPK2, p38 or acetylated by pCAF and p300/CBP. Ser to Ala or Ser to Asp (amino acids15, 20, 33, 37, 46) and Lys to Arg (320, 373, 382) changes were constructed. Transactivation was evaluated also in the presence of the regulatory protein MDM2. All modifications were distinguishable from WT p53 in the functional assay. In particular, changes at S46 had a stronger effect on function, with S46D being the least active. MDM2 was capable of inhibiting WT p53 or the Ala mutants, but not S15D or S20D. Interestingly, no changes in transactivation specificity was found. This suggests that P- and Ac- modifications relieve the critical inhibition of p53 by MDM2 but do not modify p53 intrinsic DNA binding affinity for target REs.

Alessandro Tessitore, MD, NIMH
Dopamine modulates the response of the human amygdala: A study in Parkinson's disease
In addition to classic motor signs and symptoms, Parkinson’s disease (PD) is characterized by neuropsychological and emotional deficits, including a blunted emotional response. In the present study we explored both the neural basis of abnormal emotional behavior in PD and the physiological effects of dopaminergic therapy on the response of the amygdala, a central structure in emotion processing. PD patients and matched normal controls (NC) were studied with BOLD fMRI during a paradigm that involved perceptual processing of fearful stimuli. PD patients were studied twice, once during a relatively hypodopaminergic state (i.e. at least 12 hours after their last dose of dopamimetic treatment), and again during a dopamine-replete state. The imaging data revealed a robust bilateral amygdala response in NC that was absent in PD patients during the hypodopaminergic state. Dopamine repletion partially restored this response in PD patients. Our results demonstrate for the first time an abnormal amygdala response in PD that may underlie the emotional deficits accompanying the disease. Furthermore, consistent with findings in experimental animal paradigms, our results provide in vivo evidence of the role of dopamine in modulating the response of the amygdala in human subjects

Alicja Mason, MD, PhD, NINDS
Writer's cramp as a disorder of sensory processing
Introduction Writer's cramp, a disorder of task dependent muscle co-contractions, is traditionally considered a movement disorder. Multiple electrophysiological and neuroimaging studies agree that there is simultaneous involvement of motor and sensory circuits, but it is still not clear where the pathology actually originates. There is also evidence that writer's cramp, even though a movement disorder per se, might have as an underlying cause a sensory processing defect
Materials and Methods We performed a PET study with O15 H2O to learn more about the physiology of writer's cramp
We studied ten patients with writer's cramp and ten age- and gender-matched controls. Both groups were scanned while at rest, writing and tapping their index finger for 1 min, 2 min, 3 min, and 4 min. Additionally, we obtained EMG recordings from flexor digitorum superficialis (FDS), extensor digitorum proprius (EIP) and a subjective score of severity of writer's cramp for each task. Analysis of scans, which were realigned, normalized and smoothed, was performed using Statistical Parametric Mapping (SPM99). We performed three types of analysis, intragroup, intergroup and correlation analysis. For the last analysis we used EMG recordings and subjective dystonia score as covariates
Results Random effect analysis of comparison between dystonia patients and controls in the writing task showed overactivity of primary sensory cortex (area S1) and a tendency to overactivity in the right cerebellum. Correlation analysis showed overactivity of primary sensory cortex when normalized EMG score of FDS and subjective dystonia score were used as covariates. When normalized EMG score of EIP and dystonia co-contraction index were used as covariates, overactivity was seen mainly in premotor cortex
Conclusion We feel that overactivity of S1 reflects increasing severity of dystonia correlating with activity in FDS and subjective dystonia score. Secondly, we feel that the activity in the premotor cortex reflects a compensatory mechanism of the antagonistic muscle, in this case EIP. We propose that writer's cramp arises as a dysfunction of sensory processing. Distorted sensory information is subsequently processed by the motor-related structures, resulting in defective motor activity producing co-contractions of muscles and overflow phenomena.

Anandwardhan A Hardikar, PhD, NIDDK
FGF-2 induces directed migration of human pancreatic ductal cells to form islet like cell aggregates.
Development and differentiation of islet precursor cells in vivo involves many carefully orchestrated cellular events. At an early stage, directed migration allows precursor cells to form aggregates; an initiating event in the differentiation process. Although many insights have been gained into the molecular events associated with differentiation, little is known about the migratory process
We have investigated whether Panc-1 cells in culture could serve as a model for this aspect of differentiation. Panc-1 cells maintained in the medium with 10% fetal calf serum grow in monolayers and express the pancreatic ductal cell markers; Cytokeratin-7 (Ck-7) and Cytokeratin-19 (Ck-19), but not islet / neural cell markers. We demonstrate that Panc-1 cells differentiate into islet-like cell aggregates (ICAs) when incubated in a defined DME / F12 medium. Immunocytochemistry and RT-PCR analysis of the resulting aggregates revealed expression of ngn-3, Pdx-1, Isl-1, insulin, glucagon, somatostatin and Glut-2 along with the neural markers nestin and beta-tubulin; TuJ1. ICAs responded to glucose with increase in cytoplasmic free Ca2+ concentration (measured by Fluo-2) and insulin secretion (measured by radioimmunoassay). Panc-1 cells exposed to this defined medium showed on an average, a 7-fold increase in rate of migration and traveled ~13-fold greater distance from their original start point over a 12 hour period, than cells that were incubated in conventional medium. To examine different factors that induced cellular migration, we tested several growth and differentiation factors in a Transwell® system. FGF2 (and also FGF1, but not FGF10) maximally induced cellular migration across the transwell membrane. An inhibitor of the FGF-receptor (FGFR) tyrosine kinase inhibited the migration and aggregation of these cells in a dose dependent manner
We demonstrate the minimal conditions for a precursor /ductal cell to differentiate into hormone producing islet-like aggregate. Panc-1 cells could be used as an important model to understand the process of islet cell development and differentiation.

Andre Fisahn, PhD, NICHD
Gamma oscillations, electrographic seizures and the synaptic location of kainate receptors - Zooming from network activity to synaptic physiology.
The healthy brain exhibits oscillatory electrical activity in different frequency bands that occur during specific behaviors. An example is gamma oscillations (30-80Hz) which play a role in learning and memory. Oscillatory activity can also be found in the pathological brain. An example is electrographic seizures that occur in epilepsy. A common in vivo animal model used to study seizures is the injection of the kainate receptor agonist kainate into the brain. Of interest, kainate (100nM) also induces gamma oscillations in in vitro brain slices of the hippocampus under physiological conditions. What is the connection between those two actions of kainate? Mice lacking the GluR5 kainate receptor are more susceptible to seizures following kainate injections than WT mice. Those lacking the GluR6 receptor, however, are less susceptible. Using extracellular field recordings we have shown that kainate induces gamma oscillations in slices of GluR5-/- mice at lower concentrations than in WT slices. Increasing kainate concentration leads to electrographic seizures in GluR5-/- but not WT. In contrast, in slices of GluR6-/- mice kainate fails to induce gamma oscillations or seizures at any concentration. This is consistent with reported expression patterns of GluR5 on presynaptic boutons of inhibitory interneurons (IN) that synapse onto pyramidal cell (PC) somata and of GluR6 on PC dendrites and somata
Patch clamp experiments in hippocampal slices revealed that the increase of inhibitory currents in PC after kainate (100nM) was much smaller in GluR5-/- than in WT. The higher susceptibility of GluR5-/- slices to kainate can therefore be explained by a lowered inhibitory control of IN over PC in the network. Conversely, the lack of the GluR6 receptor prevents the depolarization of PC by kainate and therefore oscillations or seizures. The kainate-induced inward current that depolarizes PC in WT likely is carried by the ion channels inherent to the ionotropic kainate receptor. Interestingly, we could also show that a recently reported metabotropic action of kainate receptor activation - the reduction of the slow AHP current - is modulated by the GluR6 receptor since this action is absent in GluR6-/- mice
Our data explain the differences in network activity seen both in an in vivo seizure model and in an in vitro gamma oscillation model by uncovering the effects of kainate receptors expressed in different synaptic locations and the ion currents modulated by them.

Andrew L Feldman, M.D., NCI-CCR
Relationship Between Histologic Grade and Outcome in Patients with Peritoneal Mesothelioma
BACKGROUND. Peritoneal mesothelioma is a rare primary tumor of the peritoneal cavity. Correlation between histologic type and outcome has been difficult due to lack of effective treatment, median survival of less than 12 months in most series, and lack of consistent reporting, which often includes patients with pleural mesothelioma. We investigated the association between histologic grade of peritoneal mesothelioma and outcome in patients undergoing treatment with continuous hyperthermic peritoneal perfusion (CHPP)
METHODS. Subjects included patients undergoing laparotomy, tumor debulking, and CHPP on IRB-approved Phase I and II clinical trials at the National Cancer Institute. Tumors were evaluated by histologic grade (low grade, tubulopapillary or adenomatoid vs. high grade, solid/epithelioid or sarcomatoid), the degree of invasion into underlying tissues, and the presence or absence of desmoplasia. Immunostaining of tumor cells for MIB-1, p27, and p53 was scored. Histologic, immunohistologic, and clinical parameters were analyzed with respect to progression-free survival rates
RESULTS. Thirty-six patients (M: 24, F: 12; mean age, 50 years) underwent laparotomy with tumor debulking and 90-minute intraoperative hyperthermic (40ºC) CHPP with cis-platinum (250 mg/m2). Median follow-up was 19 months (range: 1-92). Histologic evaluation revealed 1 adenomatoid, 14 tubulopapillary, 18 solid/epithelioid, and 3 sarcomatoid tumors. Progression-free survival was longer for patients with low-grade tumors (median 25 months) than for patients with high-grade tumors (median 10 months; p=0.006, log-rank test). Similar improvements in progression-free survival were noted for tumors without deep invasion and those without desmoplasia (p=0.01 and p=0.02, respectively). Backwards stepwise regression analysis identified tumor grade as a statistically significant independent predictor of progression-free survival
CONCLUSIONS. Histologic grading of tumors at the time of laparotomy can help predict outcome after treatment of patients with peritoneal mesothelioma, with a substantial (15-month) difference in median progression-free survival in the current series. These findings suggest that tumor histology should be a major factor in the evaluation and follow-up of these patients.

Angelo Del Parigi, M.D., NIDDK
Role of Ghrelin as a Possible Determinant of Hyperphagia and Obesity in Prader-Willi Syndrome
Prader-Willi Syndrome (PWS) is a genetic disorder occurring in 1 out of 10,000-16,000 live births. It is characterized by excessive appetite with progressive massive obesity as well as short stature, muscular hypotonia, and mental retardation. Most patients have growth hormone (GH) deficiency, hypogonadotropic hypogonadism, and develop type 2 diabetes mellitus later in life. The cause of hyperphagia is unknown. Ghrelin is a hormone mainly secreted by the stomach, which induces adiposity by increasing food intake. Peripheral administration of ghrelin increases appetite and food intake in humans indicating that ghrelin is an orexigenic signal from the periphery to the brain. We hypothesized that because ghrelin affects appetite and GH secretion abnormal ghrelin secretion may be present in PWS. We measured fasting plasma ghrelin concentration and body composition (DXA) in 7 subjects (M/F = 6/1, 26 +/- 7y, 39 +/- 11% body fat, mean +/- SD) with PWS (diagnosis confirmed by genetic test) and 30 healthy subjects (M/F = 15/15, 32 +/- 7y, 36 +/- 11%). Plasma ghrelin was negatively correlated with adiposity in both subjects with PWS (r = -0.81, P = 0.027) and controls (r = -0.49, P = 0.006). It was higher in subjects with PWS than controls (307 +/- 164 vs. 109 +/- 24 fmol/mL; P 0.001) and this difference remained significant after adjustment for percent body fat (P 0.001). Plasma ghrelin was also higher (P = 0.0004) in subjects with PWS compared to 5 healthy subjects fasted for 36 hours (M/F = 4/1; 33 +/- 7y, 28 +/- 7% body fat, 106 +/- 62 fmol/mL plasma ghrelin). In both subjects with PWS and controls, there was a positive correlation between plasma ghrelin and subjective ratings of hunger (r = 0.71, P = 0.008). Furthermore, in subjects with PWS, the concentration of the hormone did not significantly change after ingestion of 2 mL and a satiating amount of a liquid meal (307 +/- 164 vs. 306 +/- 205 vs. 260 +/- 134 fmol/mL, respectively; P = 0.56). Our data provide the first evidence that ghrelin is elevated in subjects with PWS and suggest that this hormone is responsible, at least in part, for the hyperphagia observed in PWS. Proof of this hypothesis will require intervention studies using agents that block ghrelin secretion and/or action and its effect on food intake.

Anna E Maciag, Ph.D., NCI-CCR
Expression of mutant K-ras in lung epithelial cells increases generation of reactive oxygen species and oxidative DNA damage.
The K-ras gene is frequently mutated in lung adenocarcinoma. However, wild-type K-ras is a tumor suppressor in lung [1]. This raises a new urgent question: how mutant K-ras protein is so aggressively oncogenic? Several studies demonstrate that expression of a constitutively active mutant of ras leads to increase in intracellular level of reactive oxygen species (ROS) [2,3]. Fibroblasts or keratinocytes stably transfected with a constitutively active isoform of p21Ras H-ras produced large amounts of superoxide [3,4,5]
We have pursued the hypothesis that expression of mutant K-Ras protein may result in increase of ROS and DNA oxidative damage in lung epithelial cells. We demonstrate that stable transfection of constitutively active human K-rasV12 in mouse lung epithelial cell line E10 leads to a significantly increased generation of ROS, as indicated by FACS analysis of intracellular oxidation of CM-H2DCF-DA and by reduction of nitroblue tetrazolium. ROS production by K-rasV12 was blocked by a flavoprotein inhibitor, diphenylene iodonium, implicating NAD(P)H oxidase as generating enzyme, whereas the mitochondrial oxidation inhibitor rotenone was without effect. Furthermore, we observed significant increase in oxidative DNA damage, measured as 8-oxo-2'-deoxyguanosine level, which was proportional to amount of K-Ras protein. Comet assay of single-strand DNA breaks showed a marked increase in this damage in cells expressing K-rasV12 compared with the parental line and vector-only controls (P 0.001). This damage was blocked by the pre-treatment with catalase, confirming involvement of hydrogen peroxide. ROS generated by RasV12 could contribute to transformation by DNA damage and/or by alterations in intracellular signaling. This may explain the active oncogenicity of mutant K-ras
1/ Zhang Z. et al. Nature Genet 29(1): 25-33, 2001
2/ Sundaresan M, et al. Biochem. J.318(Pt.2): 379-382, 1996; 3/ Irani K, et al. Science 275: 1649-1652, 1997; 4/ Lee A.C, et al. JBC 274 (12): 7936-7940, 1999; 5/ Yang, J-Q, et al. Mol. Carcinogen. 26: 180-188, 1999

Anna K Jerebko, PhD, CC
Multi network classification scheme for detection of colonic polyps in CT colonography data sets
A multi-network decision classification scheme for colonic polyp detection is presented. Colon cancer remains a danger for approximately 6% of Americans who may face this problem during their lifetime. Computed tomographic colonography C) or ''virtual colonoscopy'' as an alternative colon cancer screening has progressed rapidly over the past 6 years and so increased the need for powerful automatic polyp detection algorithms. Precise discrimination between real polyps and false positive detections (leftover stool or natural haustral folds) is based on dozens of features such as image texture and density, geometric parameters of the lesions, colon surface and wall characteristics, and their means and standard deviations. The large variety of polyp shapes (e.g. pedunculated, hyperplastic, sessile) and polypoid haustral folds complicate the detection process. High dimensionality of the feature space needed for discrimination unacceptably increases the complexity of the model. The novel approach we propose in this paper is to break the set of features into subsets and use the combination of several simple classifiers each processing a small number of input features. The subsets of variables are weighted by their effectiveness calculated on the basis of the training and test sample misclassification rates. The final decision is based on the majority vote across the classifiers. We investigated several types of classifiers and the best performance was reached using Support Vector Machines (SVM) and Back-propagation Neural Networks (BP) - classifiers that can learn from the data. For comparing models and for assessing a final selected model we used bootstrap estimates of prediction error (.632+ rule) which allowed us to obtain smoother estimates with lower variability compared to traditional cross-validation. The advantage of suggested multi-network classification method is that none of the component classifiers are overloaded with the input data and at the same time the classifier takes into consideration enough distinguishing features to provide sufficiently high sensitivity and specificity rates on test data sets (83% and 95% correspondingly for BP-based classifier and 81% and 98% for SVM-based classifier). This method reduces the false negative and false positive rates by factors of 1.6 and 1.5, respectively, compared to single SVM decisions, and by factors of 1.4 and 1.6 compare to single BP decisions.

Anuradha Sundararajan, Ph.D, NCI-CCR
The Flap-endonuclease Rad27 inhibits Ty1 mobility in Saccharomyces cerevisiae
The structure specific yeast nuclease Rad27 plays an important role in DNA replication and repair. Deleting RAD27 causes a variety of defects such as an increase in spontaneous mutations, elevated short sequence recombination and genome instability. We have examined the effects of various rad27 mutations on the process of Ty1 element retrotransposition and cDNA recombination, termed Ty1 mobility. Ty1 elements are mobile genetic elements that are structurally and functionally related to retroviruses. The effect of Rad27 nuclease on Ty1 mobility was assayed using a chromosomally marked Ty1 element fused to the reporter gene his3-AI (His3 gene interrupted by an Artificial Intron, AI). Transcription, splicing, and reverse transcription of the Ty1 fused reporter gene followed by retrotransposition results in a functional HIS3 gene product that can be phenotypically monitored. Using this assay we found that the point mutants rad27-G67S, rad27-G240D and rad27-E158D that have defects in their enzymatic activities display variable increases in Ty1 mobility ranging from 4 to 22-fold. The frameshift mutant rad27-324 with a possible defect in nuclear localization had by far, the maximum increase in Ty1 mobility (200-fold), unincorporated cDNA and insertion at target sites relative to the other rad27 mutants. The null mutant differed from the other rad27 alleles by displaying increased multimeric forms of Ty1 thereby suggesting unusual intermolecular recombination between unintegrated cDNA. The null mutant also displays a higher level of Ty1 cDNA recombination. The mutants do not significantly alter the levels of Ty1 RNA or the TyA1 gag protein suggesting that RAD27 effects Ty1 mobility post-translationally. However, there is an increase in the stability of unincorporated Ty1 cDNA. Our results suggest a novel role for Rad27 nuclease in inhibiting Ty1 mobility by destabilizing/degrading the cDNA intermediate and preventing the formation of Ty1 multimers

Benfang Lei, Ph. D., NIAID
Extracellular Proteins of Group A Streptococcus: Systematic Identification of Potential New Vaccine Candidates
Group A Streptococcus (GAS) is a Gram-positive human pathogen that causes a variety of diseases such as pharyngitis, cellulitis, bacteremia, streptococcal toxic shock syndrome, necrotizing fasciitis, and post-infection sequelae including acute rheumatic fever, rheumatic heart disease and glomerulonephritis. Although studies on GAS have been conducted for more than half a century, the molecular basis of GAS pathogenesis remains unclear and no licensed GAS vaccine is available. To advance the understanding of GAS pathogenesis and identify new therapeutic targets, systematic studies are required for extracellular proteins which play vital roles in the life cycle and pathogenesis of GAS by mediating pathogen adhesion, evasion of host immune responses, tissue destruction, and nutrient uptake. To this end, bioinformatic analysis was used to identify extracellular proteins and open reading frames (ORF) encoding putative extracellular proteins. A total of 121 ORFs of an M1 GAS genome were identified to encode 74 secreted proteins, 28 lipoproteins, and 19 cell surface proteins. Transcription of these ORFs was analyzed by Taqman (real-time PCR) analysis when GAS grew in vitro, in human blood, and in a mouse infection model. Fifty-six of the identified ORFs were cloned and 47 proteins were successfully overexpressed in E. coli. Twenty-four recombinant proteins were purified. The in vivo production of majority of these proteins was confirmed by the presence of specific antibodies in convalescent sera from humans and mice with GAS infections. One secreted protein was found to have immunoglobulin G-endopeptidase activity, block the interaction between antibody and CD16 on surface of human PMNs, and inhibit opsonophagocytosis and bacterial killing by PMNs. One surface protein bound heme and the gene was co-transcribed with an ABC transporter genes whose lipoprotein component also bound heme, suggesting that this locus is involved in the acquisition of essential iron. To identify potential vaccine candidates, the purified proteins were screened for ability to induce bactericidal activity of blood in immunized mice. Four proteins were found to induce significant bactericidal activity. Further characterization of these extracellular proteins will identify novel virulent factors and therapeutic targets.

Benjamin W Purow, M.D., NCI-CCR
Derivation of glioma-targeting peptides through phage display biopanning
Gliomas are one of the most difficult tumors to treat because of their tendency to infiltrate surrounding brain. A means to selectively target toxic molecules to glioma cells would thus be extremely beneficial for treatment. For this reason, we sought to derive peptides selectively binding the surface of glioma cells through bacteriophage random peptide display biopanning. Four rounds of in vitro biopanning were performed on the U87-MG human glioma cell line with pre-clearance on human astrocytes prior to each round. Two homologous phage peptide sequences resulted, labelled Np and Sp. In addition, seven rounds of in vivo biopanning against glioma vasculature were done in nude mice bearing orthotopic U87-MG xenografts. Several phage clones resulted, and one labelled Aq has a degree of homology to Np and Sp. All three peptides have homology to a region of VEGF-A-206 not previously known to have binding activity. Phage ELISA on immobilized VEGF receptor-1 shows binding of the phages to this receptor. Phage ELISA on glioma lines demonstrates high affinity of all three phages for four of four human glioma lines tested. To assess their selectivity for glioma cells, we performed phage immunohistochemistry with Sp, Aq, or unselected library phage on fixed tissue sections of nude mouse brains with xenograft U87-MG tumors. Both Sp and Aq phage show marked staining of tumors with minimal background staining of surrounding normal brain. Only the choroid plexus and a few very rare blood vessels stain with these phages in normal brain. This work has thus identified targeting peptides to be incorporated into conjugates for testing in therapeutic and diagnostic applications in neuro-oncology. It also suggests a previously unknown receptor-binding site in a VEGF isoform.

Bing-Sheng Li, Ph.D., NIAMS
Cdk5 is required for neuregulin dependent activation of PI-3 kinase and Akt activity in Neuronal survival
Neuregulins(NRGs) and their receptors have been shown to couple to PI-3 kinase and Akt activation. In this study, we reported that cyclin-dependent kinase-5 (Cdk5) regulates Akt activity and cell survival through the neuregulin and PI3-Kinase signaling pathways. We found that the brain of Cdk5-/- mice display a lower PI-3 kinase activity and phosphorylation of Akt compared to that in wild type mice. Moreover, we demonstrated that Cdk5 phosphorylated the neuregulin receptor ErbB2 at Ser-1176 and and ErbB3 at Thr-871 and Ser-1204. We also identified the Ser-1120 of sequence RSPRPR in the ErbB3 is a novel cdk5 phosphorylation consensus sequence by MALDI-TOF spectra. Finally, we found that Cdk5 activity is involved in neuregulin-induced Akt activity and neuregulin-mediated neuron protection. These findings suggest that cdk5 may exert a key role in promoting neuronal survival by regulating Akt activity through neuregulin/PI3-K signaling pathway

Bo Chen, Ph.D., NCI-CCR
Drm, a BMP antagonist, transcriptionally activates p21Cip1 and p27Kip1 expression and can inhibit the tumorigenic phenotype via a novel mechanism
Drm/Gremlin, a member of the DAN family of BMP antagonists, functions in early development, is down regulated following fibroblast transformation, and is rarely expressed in tumor derived cell lines1. To test the hypothesis that Drm might act as a tumor suppressor, we generated constructs expressing Drm under Ecdysone-inducible and constitutively promoters and introduced these constructs into various tumor cell lines. When they express Drm either constitutively or following induction, both Daoy (primitive neuroectodermal) and Saos-2 (osteoblastic) cells exhibited significantly reduced colony formation in agar or Matrigel, as well as a reduction in the rate of cell proliferation in culture. These phenotypic changes induced by Drm/Gremlin appear independent of Drm’s known ability to act as a BMP antagonist, since they cannot be reversed by treatment with BMP. Moreover, the ability of Daoy cells to form tumors in nude mice is also significantly inhibited. In comparison to control cells, Drm-expressing Daoy and Saos-2 cells express increased levels p21Cip1 protein, and also exhibit a reduced level of phosphorylated p42/44 MAP kinase. In inducible cells, p21Cip1 and p27Kip1 show a dose dependent response to the level of Drm, while inducer treatment of control cells which lack the inducible Drm construct show no change in p21Cip1 protein levels. Drm acts at the transcriptional level to induce p21 shown by Real-Time RT-PCR and expression of a luciferase reporter controlled by the p21Cip1 promoter is also induced by Drm expression. Inhibitors of either p42/44 or the p38 MAP kinase pathways do not block induction of p21Cip1 in response to Drm, and induction of p21Cip1 occurs in Daoy and Saos-2 cells, both of which lack functional p53. We conclude that Drm can induce p21Cip1 via a novel p53 and MAP kinase independent pathway. Furthermore, we suggest that Drm/Gremlin, besides its known role as a BMP antagonist, possess additional functions that allow it to inhibit both tumor formation and cell proliferation

Bojana B Beleslin-Cokic, MD, NIDDK
Erythropoietin Induction of NO Synthase and NO Levels in Bone Marrow Endothelial Cells
Erythropoietin (Epo) response extends beyond erythropoiesis and is required for normal embryonic vascular, heart and brain development. Long term Epo administration has been associated with hypertension, attributed in part to increased red cell mass and direct binding of the vasodilator nitric oxide (NO) to free hemoglobin. We determined the response of vascular endothelium to Epo and the effect on NO release. Endothelial nitric oxide synthase (eNOS) produces NO which maintains blood pressure homeostasis and blood flow. Immortalized human bone marrow microvascular endothelial cells (TrHBMEC) were cultured under reduced oxygen tensions to simulate the environment of the bone marrow during hypoxic stress. Proliferation increased at reduced oxygen tension with an optimum at 5% O2. Epo (5U/ml) proliferative response was modest under normoxia, increased at low oxygen tension and was maximal at 2% O2, comparable to the effect of vascular endothelial growth factor (VEGF). Increased Epo responsiveness was due in part to induction of erythropoietin receptor (EpoR) by hypoxia that was further increased with Epo (up to 6-fold). Induction of eNOS expression was also observed at 1.5 fold with hypoxia (2% O2) and 3 fold with hypoxia plus Epo. In addition, an immediate Epo response in NO production was observed. Under hypoxia but not normoxia, NO increased within 30 min. after Epo stimulation and persisted for up to 24 hours. Hypoxia induced VEGF and VEGF receptor, hflt-1, in TrHBMEC. However, eNOS induction by Epo could not be attributed to VEGF as Epo decreased the expression of VEGF and hflt-1 at hypoxic conditions. In HUVEC hypoxia plus Epo also induced EpoR (8 fold), eNOS (2 to 3.5 fold) as well as VEGF. In contrast to HUVEC, induction of VEGF expression by hypoxia plus Epo in TrHBMEC was not observed, suggesting a variable response of different endothelia to Epo. These data provide evidence that endothelial cells exhibit multiple responses to Epo including increased proliferation, EpoR, eNOS expression, and NO production. The results further suggest that in vivo, induction of Epo by hypoxia results not only in increased red cell production, but also in increased endothelial response with induction of EpoR and increased sensitivity to Epo giving rise to increased eNOS and NO production to counter the hypertensive effects of NO binding by increased hemoglobin mass. Rather than contribute to, endothelium may exert a protective effect toward Epo-induced hypertension.

Brian M Ward, PhD, NIAID
Poxvirus Wrapping and Movement: Interaction of a 20-Amino Acid Region of A36R with the Cytoplasmic Tail of A33R
All viruses must traverse the viscous cytoplasm in order to escape the cell and subsequently infect other cells. Poxviruses employ a two-stage mechanism for viral egress. Enveloped virions move to the cell periphery via microtubules, fuse with the plasma membrane, and then utilize thick actin tails to propel themselves away from the cell surface. Recently, we demonstrated the separation of these two mechanisms by constructing a recombinant vaccinia virus containing an A36R protein that had the two critical tyrosines required for actin tails mutated to phenylalanine. This mutation abolished actin tail formation while still allowing the intracellular virion movement of a B5R-GFP labeled virions. In contrast, we were unable to detect intracellular virion movement with a recombinant virus that had the entire A36R gene deleted and which formed small plaques. Among several possibilities, we considered that the phenotype was due to a disruption of the A33R-A36R interaction. Using the yeast two-hybrid assay, we confirmed previous reports that the A36R and A33R proteins interact and extended these results to show that the interaction takes place between the cytoplasmic tails of the two proteins. A series of C- and N-terminal truncations defined a region from residues 91-111 on A36R that is sufficient for interaction with the cytoplasmic tail of A33R in the yeast two-hybrid system. Five recombinant vaccinia viruses were constructed each containing B5R-GFP. Four had the full-length A36R gene replaced with various length C-terminal truncations, (1- 80, 1-91, 1-111, and 1-123). The fifth had the normal copy of A33R replaced with a mutated A33R gene in which most of the cytoplasmic tail had been deleted. Analysis of the mutants indicated that the presence of both regions (residues 91-111 of A36R and the cytoplasmic tail of A33R) is required for a large plaque phenotype. Immune-precipitation showed that residues 91-111 of A36R and the cytoplasmic tail of A33R are required for interaction during viral infection. Using time-lapse microscopy, viruses containing the interaction region displayed normal intracellular movement while we were unable to detect movement of the three recombinant viruses in which A33R and A36R were unable to interact. Virion quantification determined that the interaction is required for efficient envelopment of virions. It is this disruption of envelopment that is likely responsible for the absence of movement and the small plaque phenotype.

Carl G Feng, PhD, NIAID
Transgenic mice expressing human IL-10 in the antigen presenting cell compartment show increased susceptibility to infection with Mycobacterium avium associated with decreased macrophage effector function and apoptosis
Interleukin-10 (IL-10) is thought to play an important role in the regulation of microbial immunity. While T cell-derived IL-10 has been shown to suppress cell-mediated immunity, there has been debate as to whether antigen presenting cell (APC)-derived cytokine can perform the same function in vivo. To assess the influence of APC-produced IL-10 on host resistance to mycobacterial infection, transgenic mice expressing human IL-10 under the control of the MHC class II promotor (hu10Tg) were infected with Mycobacterium avium and bacterial burdens and immune responses compared with those observed in wild type (wt) animals. Hu10Tg mice harbored substantially higher numbers of M. avium and abruptly succumbed 16-18 wk post-infection. The granulomas in the infected hu10Tg mice showed marked increases in both acid-fast bacilli and host foamy macrophages. In addition, these animals displayed a dramatic increase in hepatic fibrosis. The increased susceptibility of the hu10Tg mice to M. avium infection is independent on T cell-produced endogenous murine IL-10 as bacterial burdens in mice derived by crossing the hu10Tg with murine IL-10-deficient mice were not significantly different to those in hu10Tg mice. Importantly, IFN-g responses were not decreased in the infected transgenic vs wt animals suggesting the normal development of Th1 effector cells. In contrast, mycobacterium-induced macrophage apoptosis as well as production of TNF, nitric oxide and IL-12p40 were strongly inhibited in hu10Tg mice. Together, these data indicate that APC-derived IL-10 can exert a major inhibitory effect on control of mycobacterial infection by a mechanism that appears to involve the suppression of macrophage effector function.

Chad A Ellis, PhD, NCI-CCR
Identification and Characterization of Novel Ras-Related GTPases Exhibiting Tumor Suppressor Properties
Ras superfamily proteins are structurally similar yet functionally diverse, having roles in regulation of cell growth, differentiation and survival. Ras oncoproteins serve as molecular switches, shuttling between active GTP-bound and inactive GDP-bound states. To function as oncogenes, Ras proteins must be processed at their carboxyl-terminus by a farnesyl isoprenoid lipid. This rare, post-translational modification has been successfully targeted for clinical drug development. As the precise cellular target of these compounds, Farnesyl Transferase Inhibitors, remains unidentified, an attempt was undertaken to discover novel farnesylated GTPases and characterize their role in tumorigenesis. As a result of a bioinformatic search, two novel Ras-related proteins, Rig1 and Rig2, were identified. Amino acid sequence analysis revealed that the Rig proteins (84% homology) are highly homologous (~60%) to the Arhi1 tumor suppressor as well as Rap (54%) and Ras (47%) proteins. Biochemical examination of recombinant proteins demonstrated that Arhi1, Rig1 and Rig2 undergo post-translational modification by farnesyl and geranylgeranyl (GG) isoprenoid lipids in vitro. Plasma membrane localization of Arhi1, Rig1 and Rig2 is isoprenylation-dependent, as determined by mutational analysis. In NIH3T3 fibroblasts, ectopic expression Arhi1, Rig1 or Rig2 antagonized Ras-mediated signaling and transformation. In addition, Arhi1 and Rig1 inhibit cell growth and initiate apoptosis in isoprenylation-dependent, caspase-independent manner. Overexpression of a GG-specific Rig1 mutant antagonizes Ras without inhibiting NIH3T3 cell growth indicating that de-farnesylation alters Rig1 function. Also, Rig interacts with Raf-1 and RPIP8 in vivo, effector molecules for Ras and Rap2A, respectively. Rig1 expression is limited to normal cardiac and neural tissue whereas Rig1 protein expression is down regulated in primary human astrocytomas suggesting that, like Arhi1, Rig functions as a tumor suppressor. Reintroduction of Rig1, via induced expression, inhibited astrocytoma cell growth. Therefore, Rig1 is a novel Ras superfamily member that inhibits normal and tumor cell growth as well as oncogenic transformation by interacting with known and possibly novel effector molecules. We are further characterizing the tumor suppressor potential of Rig1, the role of isoprenylation and potential effectors of Rig1-mediated growth inhibition.

Charles A Scanga, Ph.D., NIAID
MyD88 is required for resistance to Toxoplasma gondii infection and regulates parasite-induced IL-12 production by DC
Host resistance to the intracellular protozoan Toxoplasma gondii is highly dependent on early IL-12 production by antigen presenting cells (APC). Toll-like receptors (TLR) are a family of evolutionarily conserved transmembrane receptors that stimulate production of IL-12 by APC in response to defined structures present on bacterial, fungal and viral pathogens. We examined host resistance to T. gondii in mice lacking the adaptor molecule MyD88, an important signaling element used by TLR family members. Infection of MyD88-/- mice with T. gondii resulted in uncontrolled parasite replication and fatal infection by 16 days. Plasma IL-12 levels were greatly reduced but not abolished. Interferon-gamma levels were also reduced both in plasma from T. gondii-infected MyD88-/- mice and in splenocyte cultures restimulated with T. gondii antigens (STAg). Defective in vitro IL-12 responses to STAg were observed in MyD88-/- macrophages, neutrophils and splenic dendritic cells (DC). Importantly, DC from TLR2- or TLR4-deficient animals developed normal IL-12 responses to STAg. Signaling through the chemokine receptor CCR5 by a G protein-coupled mechanism also plays a major role in T.gondii-induced IL-12 production by DC. In vivo treatment of MyD88-/- mice with pertussis toxin abolished the residual IL-12 response displayed by STAg-stimulated DC. Taken together, these findings implicate TLR as critical in innate immunity to T.gondii and in the case of DC argue that full induction of IL-12 by this pathogen involves both G protein-coupled and MyD88-dependent signaling pathways.

Chiju Wei, Ph.D., NIDDK
Human Pancreatic Ductal Cells Rapidly Convert to Islet-like Cell Aggregates (ICAs) In Vitro
In order to better understand pancreatic islet cell differentiation, a number of laboratories have studied conversion of putative islet cell precursors into insulin-producing cells. As there is evidence that islet precursors may be found in pancreatic ducts, conversion of cells derived from ducts into hormone-producing cells in vitro has been used as a model of normal differentiation. In this study, we report the establishment of a cell culture system using cells derived from human cadaveric pancreases using a protocol reported to obtain ductal cells primarily. In medium supplemented with 10% fetal calf serum (''growth medium''), these cells proliferate in culture as monolayers of flattened, polygonal cells that express cytokeratins 7 and 19. Interestingly, when trypsinized and incubated in a serum-free medium, most of the cells rapidly form aggregates. Over the next several days in culture, these aggregates attain a phenotype approaching that of islet-like clusters. More importantly, insulin, glucagon and somatostatin were up-regulated as determined by RT-PCR. Insulin expression at the protein level was confirmed by confocal microscopy and quantified by ELISA to be approximately 78 ng/mg of cell protein. We also found that islet-like clusters could be formed on the surfaces of poly-L-Lysine treated plates in a two-step process: first, cell migration or chemokinesis, which is dependent on protein kinase A signaling; second, formation of clusters, which is inhibited by Br-cAMP, IBMX, and forskolin. Lastly, in addition to stimulation of insulin secretion, GLP-1 can also facilitate ICA formation. In summary, our studies demonstrate that human pancreatic ductal cells can rapidly differentiate into islet cells in a serum-free medium. These studies define a simple culture system in which to study islet cell differentiation in vitro.

Christopher C Stebbins, PhD, NIAID
Vav1 is the key substrate of the tyrosine phosphatase SHP-1 targeted during KIR mediated inhibition in NK cells
Natural killer (NK) cells function as a first line of defense against tumors and virally infected cells. To prevent lysis of normal cells, NK cells are blocked from killing through a process of inhibition. In humans the killer Ig-like receptors (KIR) are phosphorylated on immunoreceptor tyrosine-based inhibition motifs (ITIM) after engagement by MHC class I molecules expressed on normal cells. These ITIMs, located in the KIR cytoplasmic tail, recruit the phosphatase SHP-1 resulting in a blockade of NK cell cytotoxicity. The mechanism of SHP-1 mediated inhibition remains ill defined and previous attempts to identify SHP-1 substrates have lacked the specificity necessary to directly identify proteins targeted by SHP-1 during KIR mediated inhibition. It was therefore of great interest to identify these SHP-1 substrates not only to understand the next step in the inhibitory pathway but also to gain insight into NK cell activation. To identify only SHP-1 substrates targeted after KIR engagement we created a KIR/SHP-1 chimera and we show that this chimera reproduces the normal inhibitory function of KIR when expressed in NK cells. Since binding of SHP-1 substrates is a transient association, we further mutated the catalytic site of SHP-1 in the KIR/SHP-1 chimera to create a ''trapping'' mutant which prolonged substrate binding. Using this approach we identified a single 95 kDa phosphoprotein bound to KIR/SHP-1 after engagement by MHC class I on target cells. This protein was identified as Vav1, a Rac1 exchange factor essential for regulation of cytoskeletal reorganization. To test whether cytoskeletal reorganization was critical for NK cell activation we asked if cytochalasin D, an inhibitor of actin polymerization, would block signaling trough the activating receptor 2B4. Treatment of cells with cytochalasin D did block 2B4 phosphorylation but did not block the trapping of Vav1 suggesting that the SHP-1/Vav1 interaction occurs upstream of cytoskeletal reorganization. Based on these results we suggest a new model for KIR mediated inhibition in which SHP-1 dephosphorylates Vav1 which blocks actin polymerization. Failure to reorganize the actin cytoskeleton at the NK/target cell interface inhibits the amplification of activating signals and ultimately blocks target cell lysis. These results suggest that Vav1 may represent the key SHP-1 substrate targeted during KIR mediated inhibition.

Christopher D Reiter, Ph.D., NIDDK
CELL-FREE HEMOGLOBIN LIMITS NITRIC OXIDE BIOAVAILABILITY IN SICKLE CELL DISEASE.
While the deleterious effects of heme deposition and heme catalyzed membrane oxidation have been appreciated in sickle cell disease, the systemic effects of chronic hemolysis on nitric oxide (NO) bioavailability have not been considered or quantified. Central to this investigation is the understanding that NO reacts 1000 times more rapidly with free hemoglobin (Hb) solutions than with erythrocytes and that decompartmentalization of Hb into plasma would divert NO from homeostatic vascular function. We demonstrate herein that plasma heme concentrations are elevated in patients with sickle cell disease (4.2 +/- 1.1 uM) compared to normal volunteers (0.2 +/- 0.1 uM; p=0.004). Consistent with a scavenging effect of NO by cell-free Hb, we found that plasma from patients with sickle cell disease consumes 6.8 +/- 1.2 uM NO compared to 2.2 +/- 0.1 uM NO by plasma from normal volunteers (p 0.01), as measured with an electrode-based NO consumption assay. Additionally, the amount of NO consumed by plasma from patients with sickle cell disease correlated with plasma heme levels (r=0.92; NO:heme=0.86). Immunoprecipitation with anti-Hb antibodies normalizes sickle cell plasma NO consumption. Consistent with our hypothesis that cell-free Hb will limit NO-mediated vasodilation, we found that forearm blood flow responses to infusions of nitroprusside, a direct NO donor, were blunted in sickle cell patients with high concentrations of plasma Hb (59 +/- 17% increase in flow from baseline) compared with the patients with low plasma Hb levels (162 +/- 18%; p 0.01). Finally, we hypothesized that inhaled NO would selectively oxidize cell-free Hb in the pulmonary circulation and inhibit the ability of plasma Hb to consume NO in the peripheral circulation. Indeed, plasma collected from the brachial artery of patients with sickle cell anemia following the inhalation of 80 ppm NO gas for 1.5 hr consumed less NO (4.4 +/- 0.4 uM) than plasma collected before NO inhalation (6.8 +/- 1.2 uM; p=0.015). The reduction in NO consumption by plasma of sickle cell patients after NO inhalation paralleled an increase in oxidized plasma Hb, detected by electron paramagnetic resonance. In conclusion, the decompartmentalization of Hb and subsequent scavenging of NO may explain the vascular complications, such as pulmonary hypertension and cutaneous ulceration, shared by chronic hemolytic disorders. Therapies that reduce hemolysis as well as inactivate ferrous plasma Hb may restore NO bioavailability.

Clemens Scheinecker, M.D., NIAID
Direct visualization of dendritic cell presentation of a tissue-specific autoantigen in normal and autoimmune mice.
The MHC-dependent presentation of processed tissue-specific self-antigens is believed to be critical to the development of peripheral (extra-thymic) tolerance. Although hematopoietic antigen presenting cells have been shown to present transgenic analogs of self-antigens, the isolation and detailed characterization of the specific cells responsible such presentation has not yet been accomplished. Towards this end, we have analyzed the presentation of an endogenous, gastric-specific H+/K+-ATPase peptide in the context of the MHC class II molecule I-Ad. Visual evidence for the uptake of H+/K+-ATPase by dendritic cells (DC)in the stomach and their migration to the gastric lymph node (LN) was obtained by immunofluorescence staining and confocal microscopy analysis. Anti- H+/K+-ATPase staining was exclusively detected in vesicular compartments of CD11c+ DC in gastric but not in peripheral or mesenteric LN, or spleen. Likewise, only gastric, but not mesenteric or peripheral DC stimulated an H+/K+-ATPase specific T cell clone TXA-23 to produce interferon gamma without addition of H+/K+-ATPase protein or specific peptide. This presentation occurred in the presence of a lysosomotropic agent, indicating that the antigen not only was present in the gastric DC, but also already processed and bound to MHC class II molecules in vivo before gastric DC isolation. Such constitutive self-antigen presentation was seen with both CD8ahi and CD8alo gastric DC subsets. Upon the induction of experimental autoimmune gastritis (AIG) by the adoptive transfer of CD25-depleted CD4+ T cells into BALB/cnu/nu mice, DC were found to be among the first cells to infiltrate the stomach mucosa. Moreover, proportions as well as absolute numbers of DC were substantially increased in the gastric, but not in the peripheral LN. In line with this, the frequency of H+/K+-ATPase staining in CD11c+ cells selectively increased in the gastric LN, as did overall gastric DC presentation of H+/K+-ATPase. These results provide the first clear identification of DC as the cells involved in the uptake and presentation of a tissue-specific antigen in normal animals and the augmentation of such presentation during the development of overt autoimmune disease. Such direct evidence for constitutive presentation of processed self-antigens in normal mice raises questions about current models of peripheral tolerance proposed to be induced by immature DC presentation of self-antigens

D.C. Ghislaine Mayer, Ph.D., NIAID
Polymorphisms in Plasmodium falciparum Erythrocyte-binding Ligand Change its Receptor Specificity
The binding of merozoites from Plasmodium species, the causative agent of malaria, to erythrocytes is dependent on two families of parasite ligands, the reticulocyte binding-like (RBL) and the Duffy binding-like (DBL). The DBL family in P. falciparum includes erythrocyte binding antigen-175 (EBA-175) and its paralogue, BA erythrocyte binding ligand (BAEBL). Previously, we have characterized the binding specificity of BAEBL. We demonstrated that BAEBL from P. falciparum clone, Dd2Nm, recognizes glycophorin C/D. Characterization of BAEBL by another group using the P. falciparum clone E12, isolated from Papua New Guinea, revealed that its red cell receptor is likely to be a glycolipid. We sequenced the erythrocyte-binding domain of BAEBL also called Region II from 24 clones isolated from parts of the world where P. falciparum malaria is highly prevalent. We observed 5 different sequence variants with the polymorphisms always occurring at the same position among all clones. We have found that these polymorphisms are not restricted to any geographical areas and are the results of identical nonsynoymous base pairs changes at specific positions in Region II of BAEBL.
We investigated the functional significance of these nonsynonomous substitutions in the erythrocyte-binding domain of BAEBL by expressing Region II of each of the five polymorphic groups transiently on the surface of COS cells. Each polymorphism led to a different erythrocyte-binding specificity as demonstrated by different binding patterns to enzyme treated and Gerbich negative erythrocytes. Region II from some P. falciparum clones was sensitive to neuraminidase and trypsin and was the only sequence to bind poorly to Gerbich-negative erythrocytes; the other three sequences bound equally well to Gerbichnegative erythrocytes as to normal erythrocytes. At least three other patterns of binding specificity were observed: clones that bound trypsin-treated but not neuraminidase-treated erythrocytes; clones that bound both neuraminidase- and trypsin-treated erythrocytes and those that bound neuraminidase- but not trypsin-treated erythrocytes. We showed that a single base change leading to a single amino acid substitution in Region II of BAEBL led to recognition of different molecules on human erythrocytes. These findings describe an additional mechanism for recognition of different erythrocyte receptors by various forms of a single P. falciparum erythrocyte-binding ligand.

Daniel J Tomso, Ph.D., NIEHS
The Influence of Local Sequence Context on Human Single Nucleotide Polymorphisms
Recent progress in genomic sequencing has generated a flood of new data on human genetic variation. Although it is assumed that polymorphisms, particularly single nucleotide polymorphisms (SNPs), account for the majority of variation between individuals, little is known about most reported SNPs. Predicting the biological impact of variation is of singular importance in the study of human genetics and will require, in conjunction with detailed biological investigation, a rigorous analysis of SNP DNA sequences. Furthermore, an analysis of the distribution and frequency of SNPs will provide insight into the origin and dynamic status of genetic variation, which in turn will help illuminate essential processes such as mutagenesis and genome maintenance.
DNA sequence context influences many biological events, including protein binding, mutagenesis, and repair. I predicted that local sequence structure would be of great value in classifying individual SNPs, as well as in understanding the distribution and maintenance of human variation. To assess the impact of sequence context on SNPs, I developed a set of PERL programs to collect local sequence information from the dbSNP database. I then performed frequency analysis for over 1.9 million human SNPs, using local sequences ranging from 1 to 8 base pairs in size. By comparing the frequency of short sequences at polymorphic versus non-polymorphic positions, I have measured the relative over- or underrepresentation of short sequences at polymorphic positions.
The dinucleotide CpG is dramatically overrepresented at polymorphic positions (6.6 fold). Although this sequence is known to be highly mutagenic, its pronounced association with human polymorphisms has not been previously documented. Additional effects are observed in longer sequences. For example, TACG is the most hyperpolymorphic 4-base sequence in the genome, and is several fold more overrepresented than other CpG-containing sequences, suggesting a role for flanking bases in modulating CpG mutagenesis. Other overrepresented sequences contain adjacent microsatellite repeats, also presumably hybermutable. Still other sequences are overrepresented without obvious correlation to known mutagenic sequences, implying the existence of undescribed mutagenesis or maintenance pressures. Collectively, these data yield insight into the forces that govern genetic variation, and provide a tool for predicting important biological phenomena.

Dario C Ramirez, PhD - MSc, NIEHS
IMMUNO-SPIN TRAPPING
Protein-derived free radicals, especially those derived from hemoproteins, play important roles in pathophysiological conditions, such as aging cancer, inflammation, atherosclerosis, neurodegenerative diseases and exposure to environmental hazards. Certainly, the earlier detection and characterization of hemoprotein-derived free radical could improve the human life quality and expectancy. In order to probe the involvement of a hemoglobin radical in the human oxyhemoglobin (oxyHb) or methemoglobin (metHb)/ H2O2 system, we have developed and utilized a new approach, ''Immuno-Spin Trapping, (IST)'' which combines the specificity and sensitivity of both spin trapping and antigen:antibody interactions. Previously, a novel rabbit polyclonal anti-DMPO nitrone adduct antiserum, which specifically recognizes protein radical-derived nitrone adducts, was developed and validated in our laboratory. In the present study, the formation of nitrone adducts on hemoglobin depended on the oxidation state of the iron heme, the concentrations of H2O2 and DMPO, and time as determined by an Enzyme-Linked Immuno-Sorbent Assay (ELISA) and Western Blotting. The presence of reduced glutathione or L-ascorbate significantly decreased the level of nitrone adducts on metHb in a dose-dependent manner. Western blotting analysis showed that only the complete system (oxy- or metHb/DMPO/H2O2) generates epitopes recognized by the antiserum. The specific modification of tyrosine residues on metHb by iodination produced a significant decrease in antibody binding, while the thiylation of cysteine residues did not affect it at all. In comparison with either direct ESR or ESR spin-trapping detection of the tyrosine-centered protein-derived free radical in biological systems, IST showed the following advantages: i) exhibits higher sensitivity, ii) requires smaller samples, iii) gives the approximate molecular weight of the protein radical-derived nitrone adducts, iv) permits the assay of many samples at the same time, and v) does not require expensive ESR instruments to detect the radical adducts. In conclusion, IST showed the formation of tyrosyl radical(s) in human oxyHb and metHb exposed to very low concentrations of H2O2 and we reported here for the first time the detection of intracellular nitrone adducts in red blood cells, as a first application.

David Berrigan, PhD MPH, NCI-CCR
Patterns of health behavior in US adults
Background. Associations between health-related behaviors are important for two reasons. First, disease prevention and health promotion depend on understanding both prevalence of health-related behaviors and associations between such behaviors. Second, behaviors may have synergistic effects on disease risk.
Methods. We document patterns of adherence to recommendations concerning five health-related behaviors (physical activity, tobacco use, alcohol consumption, fruit and vegetable consumption, and dietary fat intake) in US adults (n = 15,166) using data from the Third National Health and Nutrition Examination Survey (NHANES III). Division of individuals into categories associated with adherence or non-adherence to lifestyle recommendations results in 32 patterns of adherence/non-adherence.
Results. Proportions of US adults with 21 of 32 behavior patterns characterized here deviated from proportions expected if health behaviors are independent of each other. The two extreme patterns, all adherence (5.86%), and all non-adherence (4.86%), were found in about double the proportion expected. Age, gender, race/ethnicity, education and income were associated with a number of patterns, including the two extremes.
Conclusions. This analysis of behavior patterns highlights population sub-groups of public health importance, provides a benchmark for studies of multivariate associations between health behaviors, and supports a multidimensional model of health behavior.

Dawn R Clifton, Ph.D., NIAID
Identification of tyrosine phosphorylated proteins in response to in vitro infection by Chlamydia trachomatis
Chlamydia trachomatis is the etiological agent of several significant human diseases. This obligate intracellular bacterium has a biphasic developmental cycle that involves cell types adapted for extracellular survival (elementary bodies, or EBs) and intracellular multiplication (reticulate bodies, or RBs). Chlamydiae undergo their developmental cycle entirely within a parasitophorous vacuole termed an inclusion. Although studies describing tyrosine phosphorylation in response to chlamydia infection have been published, these phosphorylated proteins have not been identified. Thus the functional importance of this signaling event, in the entry or developmental cycle of the organism or in the response of the host cell to infection is unknown. The antibody 4G10 targeted against phosphotyrosine-containing proteins was used to screen uninfected or chlamydia-infected cells by Western blotting and immunofluorescence (IF). Western blot analysis revealed a unique high molecular weight protein (~180kDa) in infected cell lysates that was phosphorylated within 5 minutes post-infection (p.i.) and remained phosphorylated through 18hr p.i. Inhibition of bacterial transcription and translation did not affect phosphorylation, suggesting this protein is either derived from the host or is a pre-existing bacterial protein. IF studies revealed colocalization of the antibody with the EBs/inclusion throughout the cycle of infection. The staining appeared asymmetric with reactivity localized to one side of the EB/nascent inclusion. By immunoelectron microscropy, the reactivity was localized to the cytoplasmic face of the plasma membrane at the site of attachment of surface-associated EBs. This transduction of phosphotyrosine signaling across the plasma membrane was associated with actin-rich pedestal-like structures at the site of EB entry, suggesting that the phosphorylation event is associated with actin-reorganization during entry of the organism. Interestingly, infection with a different C. trachomatis serovar also lead to the tyrosine phosphorylation of a slightly higher molecular weight protein, suggesting it may be a homologous protein of chlamydial origin. Two-dimensional electrophoresis followed by MALDI-TOF mass spectroscopy tentatively identified this protein as a member of a family of chlamydial membrane proteins of unknown function. Studies are underway to confirm the identity of the protein and investigate its function during chlamydial infection.

Paolo A Muraro, MD, PhD, NINDS
Molecular Tracking of Myelin Basic Protein - Specific T Cell Expansion in Multiple Sclerosis Exacerbations
T cell responses to myelin antigens such as myelin basic protein (MBP) may be important in the pathogenesis of multiple sclerosis (MS). The extent of variation over time of MBP-specific T cell responses and the relationship with the course of disease, however, are poorly understood. Our goal was to define the kinetics of frequency of MBP-specific T cells in patients with MS during clinical exacerbations. We developed a highly specific and sensitive method for measuring the frequency of individual T cell clones by real-time PCR quantification of clonotypic T cell receptor transcripts. We applied our technique for tracking in a patient with relapsing-remitting MS a CD4+ T helper 1 clone, P2-10, which recognized the immunodominant myelin epitope MBP(83-99) with extremely high functional antigenic avidity. We monitored the frequency of P2-10 in the patient’s peripheral blood before, during and after clinical exacerbations. As controls, we tracked the frequency of antigen-specific T cell clones in peripheral blood from a healthy donor and from patients with other neurological disorders (OND: HTLV-I associated myelopathy and chronic Lyme neuroborreliosis). The frequency of P2-10 in the patient’s peripheral blood increased 5-fold compared to baseline at one week before the clinical onset of a severe disease exacerbation, reaching an absolute frequency of 1 cell in 1,800 PBMC. The clone’s frequency already decreased at baseline levels on the day of onset of clinical symptoms of the relapse, when magnetic resonance imaging of the brain showed a sharp increase in the number of contrast-enhancing lesions. In contrast, tracking of control T cell clones in peripheral blood from a healthy subject and the patients with OND showed stable clonal frequencies over time. The kinetics of frequency of an autoreactive Th1 clone during an exacerbation of MS closely resembled those of encephalitogenic T cells in experimental animal models, which reach their highest frequency before the onset of clinical symptoms and rapidly decrease in the peripheral compartments following massive migration into the central nervous system. An important implication of our results is that significantly increased frequencies of autoreactive effector T cells might only be detectable during a short period of time preceding the clinical onset of relapses.

Dganit Danino, PhD, NIDDK
Conformational Changes of Dynamin-Lipid Tubes upon GTP Addition: A Time-Resolved Study using Digital-Imaging Cryo-TEM
Dynamin is a large GTPase that plays a key role in receptor-mediated endocytosis. It assembles into helical structures at the necks of clathrin-coated pits, and mediate fission of endocytic vesicles from the plasma membrane upon GTP binding and hydrolysis. Although studied extensively, the exact mechanism of action of dynamin in the pinching event remains unclear. One model proposes that dynamin acts as a mechaochemical enzyme, while another suggests it possibly acts as a signaling switch, activating down-stream effectors.
It was shown by negative-stain electron microscopy that wild type dynamin and a mutant missing its proline rich domain transform anionic liposomes to long helical tubes that rapidly constrict and fragment into small vesicles upon addition of GTP. Light-scattering studies revealed a sharp and rapid decrease in the intensity of the scattered light following GTP addition, supporting the microscopy findings. This study showed that dynamin can use GTP hydrolysis to generate a force on the membrane, and cause vesiculation. However, the mechanism of the vesicle formation remained unknown.
The stimulated GTPase activity of dynamin is Kcat ~ 3 per sec, therefore a rapid ''fixing'' mechanism is required to reveal the formation of intermediate structures following GTP addition. We applied a novel approach, and studied the dynamic process by cryo-transmission electron microscopy (cryo-TEM), which is based on thermal (physical) fixation that preserves the assemblies at their native state. In a time-resolved study of the dynamin/lipid tubes, conformational changes were followed at sequential time points after addition of GTP. The microscopy experiments show that the assembled dynamin functions in a cooperative mechanism to constrict the tubes through their full length. However, free vesicles were not observed. The vesiculation seen previously may therefore be an artifact caused during preparation of negative stain samples. We show further that the rapid decrease in the scattered light reflects conformational changes of the tubes during the GTP reaction, not vesiculation, as the level of the scattered light remains much higher than what is observed with small vesicles.
Our study clearly shows that dynamin is a force-generating enzyme capable of constricting an underlying membrane. However, there is no evidence that dynamin alone can cause destabilization of the membrane and directly mediate release of endocytic vesicles from the plasma membrane.

DIANE M KLOTZ, Ph.D., NIEHS
In Vivo Evidence for IGF-1/Estrogen Receptor Cross-talk: Estrogen Receptor-alpha is a Required Intermediate in IGF-1-Stimulated Uterine Proliferative Responses
The endogenous steroid hormone estradiol (E2) exerts its mitogenic effects on the uterus by binding to its nuclear estrogen receptor (ER) and activating its transcriptional activity, resulting in the production of mitogens such as insulin-like growth factor-1 (IGF-1). However, in vitro studies have demonstrated another level of interaction between IGF-1 signaling and ER transcriptional activity. In those cell culture studies IGF-1 alone, in the absence of E2, stimulated the transcriptional activity of ER, identifying a cross-talk mechanism between cell membrane-generated growth factor signaling and nuclear hormone receptor signaling pathways. To investigate IGF-1/ER cross-talk and the role of ER in uterine responses to IGF-1 in vivo, both wild-type (WT) and transgenic mice lacking ER alpha (aERKO) were ovariectomized (to remove E2) and administered IGF-1, and several uterine responses were compared. Although IGF-1 treatment resulted in activation of the uterine IGF-1R signaling pathway in both WT and aERKO mice, DNA synthesis and proliferation, as measured by BrdU incorporation and PCNA expression, were stimulated only in the WT uterus. These data demonstrate that ERa is necessary for the proliferative effects of IGF-1 in the uterus and suggest that nuclear ER activation is a target of IGF-1 signaling. To confirm that ER is activated in vivo by IGF-1, transgenic mice with a luciferase reporter gene that is transcribed in response to activated ER (ERE-luciferase mice) were utilized. Treatment of ovariectomized ERE-luciferase mice with IGF-1 resulted in an increase in reporter gene expression in the uterus that was attenuated by the ER antagonist ICI 182, 780. These are the first data to demonstrate that 1) ERa expression is necessary for IGF-1-induced uterine proliferative responses, and 2) stimulation of E2-independent ER transcriptional activity by IGF-1 indeed occurs in vivo. Additional studies currently underway utilizing gene arrays indicate that the expression patterns of IGF-1-modulated genes differ in WT and aERKO uteri, supporting the hypothesis that ER is a target for IGF-1 signaling and suggesting that this interaction may be necessary to modulate a subset of genes required for mitogenesis in the uterus. Finally, elevated levels of IGF-1 are implicated in hormonally-responsive cancers, and activation of ER by IGF-1 may represent a mechanism for the stimulation of E2-responsive tumor growth even in the absence of endogenous E2.

Dimiter G Demirov, Ph.D., NIAID
THE ROLE OF TSG101 IN HIV-1 BUDDING
HIV-1 budding is promoted by a Pro-Thr-Ala-Pro (PTAP) sequence near the N-terminus of the p6 Gag protein. Analogous motifs, collectively termed ''late'' or ''L'' domains to reflect their role late in virus budding, have been identified in a number of other retroviral Gag proteins.
Retroviral L domains appear to interact with the host ubiquitination machinery. The cellular protein TSG101, which contains a domain related to E2 ubiquitin-conjugating enzymes, binds HIV-1 Gag via the p6 L domain. TSG101 assembles into a complex called ESCRT-1 that functions in endosomal sorting. Recently, we found that overexpressing the E2-like domain of TSG101, denoted TSG-5’, suppresses HIV-1 budding by blocking p6 late domain function. This budding inhibition is specific for HIV-1, as TSG-5’ does not disrupt the release of other retroviruses (e.g., murine leukemia virus and Mason-Pfizer monkey virus)
To elucidate the role of TSG101 in HIV-1 budding, we have addressed the following questions: 1) Does inhibition of virus release by TSG-5’ depend upon its interaction with Gag? 2) Does overexpression of TSG101 disrupt virus budding, and if so, does the effect require Gag binding and is it specific for HIV-1? 3) Does overexpression of TSG-5’ or full-length TSG101 impair the host endosomal sorting pathway? Finally, 4) does HIV-1 Gag recruit TSG101 to the site of virus budding at the plasma membrane?
We have observed that a mutation that disrupts TSG101/Gag binding abolishes the ability of TSG-5’ to inhibit virus release. Biochemical and EM analyses indicate that overexpression of full-length TSG101 also impairs virus budding, but in a manner that does not require interaction with Gag and is not specific for HIV-1. Characterization of the endosomal sorting pathway in TSG-5’- and TSG101-overexpressing cells by confocal microscopy reveals that overexpression of full-length TSG101 but not TSG-5’ disrupts endosomal sorting. Membrane flotation studies indicate that HIV-1 Gag recruits TSG101 to plasma membrane lipid rafts in a late domain-dependent manner. These results establish a critical role for TSG101, and the cellular endosomal sorting pathway, in HIV-1 budding. Furthermore, to our knowledge, TSG-5’ is the first example of a polypeptide that specifically inhibits HIV-1 budding. Its high specificity towards HIV-1 and low cytotoxicity upon overexpression make it a potential candidate for the development of antiviral drugs.

Dirk Daelemans, Ph.D., NCI-CCR
A new inhibitor of the CRM1-mediated Rev export as a tool for studying nucleocytoplasmic transport
The HIV-1 Rev protein is an essential regulator of the HIV-1 mRNA expression that promotes the export of unspliced and partially spliced mRNAs from the nucleus to the cytoplasm of the cell. These mRNAs are responsible for the production of the viral structural proteins. Nuclear export of Rev is mediated by its leucine-rich NES, which binds to the cellular CRM1 export receptor. We identified a low molecular weight compound, PKF050-638, as a specific inhibitor of HIV-1 Rev function. This drug inhibits Rev-dependent mRNA expression in a dose-dependent fashion. Using GFP fusion proteins we showed that PKF050-638 is an inhibitor of the Rev nuclear export. PKF050-638 induced nuclear accumulation of Rev and of cytoplasmic Rev mutants. Furthermore, it inhibited colocalisation of CRM1 with Rev in the nucleolus, suggesting that PKF050-638 blocks the interaction of Rev with its export receptor CRM1. Using a quantitative in vitro CRM1-NES cargo-binding assay, we showed that PKF050-638 disrupts CRM1-NES complex formation. The compound displayed strict structural requirements for its activity, since its enantiomer was inactive in all assays tested. These results show that we identified a new drug interfering with the CRM1-mediated nuclear export through inhibition of the CRM1-NES complex formation. The inhibitor has a distinct site of interaction with CRM1, which extends beyond the site of interaction of the other known CRM1 inhibitor leptomycin B (LMB). This interaction is not covalent, which affords the reversal of drug effects upon removal of the drug. The reversible effects of PKF050-638 make it an attractive alternative for studies of nucleocytoplasmic trafficking of macromolecules through the CRM1 pathway. This pathway is essential for cell function and responsible for the export of many important factors including cancer-related genes.

Dolena R Ledee, Ph.D., NEI
Muskelin, a specific interacting protein of the CDK5 activating protein p39
Cyclin dependent kinase 5, CDK5, is a member of the small serine/threonine cyclin-dependent kinase (CDK) family. Its activity, which is highest in neuronal cells, requires association with either of its two regulatory proteins, p35 or p39. Functional roles of CDK5 in neuronal cells include cytoskeletal regulation, axon guidance and membrane transport. Although CDK5 activity was initially considered neuron specific, active CDK5 has now been detected in several non-neuronal tissues, including the lens. Our lab has reported on CDK5 involvement in cell-cell and cell-matrix adhesion in lens epithelial cells. Specifically, over-expression of CDK5 in these cells reduced cadherin dependent cell-cell aggregation and promoted binding between the cells and fibronectin, an extracellullar matrix (ECM) component. To discover other roles of CDK5 and it activating proteins we utilized the yeast-two hybrid system to identify interacting proteins. CDK5, p35 and p39 were cloned into a bait vector and a rat lens library, made from 18-day old embryonic rat lenses, was cloned into the target vector. The library was screened with each bait vector. The activating protein, p39, was found to specifically interact with a clone we identified as muskelin. Muskelin is a novel intracellular mediator of cell adhesion and cytoskeletal responses to thrombospondin-1 (TSP-1), another ECM component. Binding specificity of muskelin to p39 and not CDK5 or p35 was confirmed using GST-pulldowns, co-immunoprecitations, and the yeast-two hybrid system. Transient dual transfections of fluorescent-tagged proteins, CFP-muskelin and YFP-p39, confirmed co-localization. Both proteins had a diffuse cytoplasmic pattern with concentrated overlap at the marginal ruffles. The presence of a potential CDK5 phosphorylation site in muskelin suggests a role for p39 as a bridge between CDK5 and muskelin, and co-immunoprecipitation assays have confirmed the presence of CDK5, muskelin and p39 in a complex. Muskelin represents the first protein reported to interact specifically with p39 and not p35. This binding specificity raises the possibility that p35 and p39 may serve distinct functional roles by targeting CDK5 to different cellular substrates.

Dong-Wan Seo, Ph.D., NCI-CCR
Tissue inhibitor of metalloproteinase-2 (TIMP-2) inhibits FGF-2-induced human microvascular endothelial cell growth
The tissue inhibitors of metalloproteinases (TIMPs) are multifunctional regulators of cell proliferation, migration and invasion associated with matrix metalloproteinase (MMP)-mediated turnover of the extracellular matrix. We have previously reported a direct role for TIMP-2 in regulating tyrosine kinase-type growth factor receptor activation. In this study, we have investigated TIMP-2 signaling pathways and cellular responses in FGF-2-stimulated human microvascular endothelial cell growth in vitro and in vivo. TIMP-2 inhibits the phosphorylation of FGFR as well as the binding of FRS-2 and PLC gamma to FGFR. This results in the suppression of activation of p42/44MAPK and Akt. TIMP-2 inhibition of these signaling pathways subsequently blocked the cell cycle progression via the inactivation of NF-kB, stabilization of cyclin-dependent kinase inhibitor p27kip1, and dephosphorylation of pRb. Both TIMP-2 and the MMP inhibitor null mutant Ala+TIMP-2 can completely abrogate FGF-2-induced angiogenesis in vivo. The inhibitory effect of TIMP-2 on the mitogenic response to FGF-2 was reversed by sodium orthovanadate, demonstrating that TIMP-2 effect be mediated through a protein tyrosine phosphatase. TIMP-2 treatment of human microvascular endothelial cells results in preservation of SHP-1 (a protein tyrosine phosphatase) association with FGFR-1. In addition, TIMP-2 treatment increased cAMP levels and TIMP-2-dependent inhibition of the FGF-2-stimulated cell proliferation is prevented by the treatment with the adenylate cyclase inhibitor SQ22536. However, the blockade of p42/44MAPK and Akt activation by TIMP-2 was not dependent on the cAMP production since treatment with dibutyryl cAMP failed to reduce p42/44MAPK and Akt activity. These results suggest the possibility that TIMP-2 might modulate the cell growth through a novel mechanism, involving both direct SHP-1 association with FGFR, that is independent of adenylate cyclase activation, as well as cAMP-dependent events. We also find that TIMP-2 induces the association of FGFR with integrin alpha5beta1, and, using function-blocking alpha5beta1 antibodies, demonstrate that TIMP-2 binding to the cell surfaces is markedly reduced. Collectively, our findings suggest that the TIMP-2 induces crosstalk of FGFR with integrin alpha5beta1, and that this mechanism may play a pivotal role in the control of angiogenesis.

Eric C Greene, PhD, NIDDK
Insights into site-specific DNA recombination through direct observation of single Mu transposition reactions
DNA recombination is a fundamental process required by virtually all organisms to maintain the integrity of their genomes. DNA replication and repair, chromosome segregation, and development of the eukaryotic immune system are all processes that rely on DNA recombination. Deficiencies in any part of the machinery involved in these recombination reactions can lead to DNA lesions, which have the potential for developing into cancerous phenotypes. In addition, retroviruses, such as HIV and Rous sarcoma virus, exploit site-specific DNA recombination to integrate a copy of their genome into the host cell chromosomes. These examples underscore the importance of elucidating the molecular mechanisms involved in DNA recombination
We have designed and assembled a fluorescence-based system for the direct observation of a single site-specific recombination event catalyzed by the bacteriophage Mu. To observe single reactions fluorescently labeled Mu proteins were injected into a flowcell containing molecules of DNA tethered to the flowcell surface. Interactions between the proteins and the DNA can be directly observed, in real-time, using a total internal reflection fluorescence (TIRF) microscope equipped with a highly sensitive intensified CCD camera. This system can detect emission from single fluorescent molecules and it has allowed us to examine Mu-directed site-specific recombination at a level of detail that had not previously been possible. We have shown that the MuB protein helps select a site on the DNA molecules for insertion of its genome by assembling into an oligomeric complex on A/T-rich sequences of DNA. We have also demonstrated that Mu proteins bound to different DNA sites separated by up to several kilobases can interact with one another through a DNA looping mechanism. We are currently developing a novel TIRF microscope with two different laser excitation sources and a pair of intensified CCD cameras. This microscope will allow detection of the fluorescence signal from two different color single fluorophores, thus allowing us to simultaneously monitor the interactions between any two different reaction components.
The approaches we are developing in our study of Mu can be readily applied to a variety of DNA recombination reactions as well as other cellular processes involving the assembly and regulation of nucleoprotein complexes.

Erik L Snapp, Ph. D., NICHD
Ribosomes Organize and Maintain Fully Assembled Translocons at the Mammalian Endoplasmic Reticulum
We have used fluorescence resonance energy transfer (FRET) to probe the organizational state in cells of the translocon, a protein complex of over 20 membrane proteins. Eukaryotic secretory proteins are cotranslationally inserted into the endoplasmic reticulum through the translocon. The standard model of protein translocation begins with assembly of mRNA with the large and small subunits of the ribosome in the cytoplasm, translation of a signal sequence which interacts with the signal recognition particle and its receptor (SRP receptor) to bring the ribosome and signal sequence into contact and alignment with the translocon where protein synthesis resumes and the nascent peptide chain is inserted into the ER. At the end of protein synthesis in vitro, the ribosome releases from the translocon and repeats the cycle. In one model, the translocon assembles in response to ribosome binding and disassembles at the end of protein synthesis when the ribosome dissociates. An alternative model proposes that the translocon is constituitively assembled and is gated open by the ribosome and the lumenal chaperone BiP. Translocon assembly and dynamics have been difficult to study by traditional methods. Using FRET microscopy, we have determined the organizational state of the translocon during the translocation cycle in vivo. FRET is highly sensitive and measures protein proximities below 70 angstroms. Surprisingly, the entire translocon remains assembled even after the completion of protein translocation, with small conformational changes distinguishing the active and quiescent states. We observe both biochemically and by antibody accessibility assays that ribosomes stay translocon-bound in vivo even in the absence of ongoing translocation. Biochemically, we find that ribosome binding facilitates the co-association of TRAM and SRP receptor with a stably assembled core translocon composed of the Sec61, translocon-associated protein and oligosaccharyl transferase sub-complexes. These results point to an organizing role for ribosomes in the maintenance of translocon structure and suggest that the perpetual juxtaposition of the protein synthesis, targeting, and translocation machineries in vivo may facilitate efficient initiation of protein translocation despite a highly crowded cellular environment. Our methods can be applied to other large soluble and membrane protein complexes in vivo.

Estela M Munoz, Ph.D., NIMH
CHARACTERIZATION OF A CIRCADIAN E-BOX
The E-box (CACGTG) is a widely used DNA control element. In spite of its brevity and broad distribution, the E-box is a remarkably versatile regulatory sequence. The circadian clock is one of many pathways that use E-Box elements to modulate transcription of their target genes. Since a single 6-bp consensus sequence is unlikely to encode enough transcriptional information, we searched for additional requirements in an attempt to arrive at a functional definition of a circadian E-Box. We compared the performance of a perfect E-Box in different contexts: the minimal promoters of two rhythmically expressed genes, vasopressin (AVP) and cyclin B1 (CYC), only the first of which is controlled by the circadian clock in vivo. Expectedly, the AVP promoter was strongly stimulated by the circadian transcription factor BMAL1/CLOCK (B/C) in transfected cells whereas the CYC promoter was completely refractory. Starting from these prototypic regulatory regions we generated an extensive series of chimeric and mutant promoters through the exchange of discrete domains and site-directed mutagenesis. The systematic analysis of these reporters allowed us to identify several qualifying factors that can contribute to the differential activation of a perfect E-Box by B/C in this system. Specifically, we discovered three levels of control affecting the circadian competence of a perfect E-Box. First, we identified novel as well as previously suspected cis-acting elements, upstream and downstream of the AVP perfect E-Box, that are absolutely required for strong activation by B/C. Second, we found via electrophoretic mobility shift assays that the high affinity binding of the CYC E-Box toward constitutive factors results in nearly constant occupancy and high constitutive levels of transcription. We hypothesize that this situation could prevent B/C activation of the CYC E-Box. Third, we determined the effect of key bases flanking the perfect AVP E-Box, such as the guanine at position –6, as playing a key role in activation of the promoter by B/C. Based on these results we propose that the response of a perfect E-Box to the clock input can take many values, from total refractoriness to high sensitivity, and that this is a function of which and how many of these modular elements are combined around a perfect E-Box. This study is important because it sharpens the focus on the subtle mechanisms that can either enable or disqualify a perfect E-Box for control by the biological clock.

Fanis Missirlis, PhD, NICHD
Mutations that lead to oxidative stress activate Iron Regulatory Proteins of Drosophila.
Iron overload and oxidative stress parallel neurodegeneration in human neurological disorders including Alzheimer’s disease, Parkinson’s disease and ALS. Cells react to intracellular iron changes through Iron Regulatory Proteins (IRPs). IRP-1 has two distinct and mutually exclusive biological functions. IRP-1 holoprotein contains a cubane iron-sulfur cluster and exhibits cytosolic aconitase activity. Conversely, IRP-1 lacking the iron-sulfur cluster specifically binds to mRNA stem loops known as Iron Response Elements (IREs). This latter interaction post-transcriptionally coordinates expression of proteins involved in iron metabolism. Assembly and disassembly of the iron-sulfur cluster depends on the bio-available iron levels, thus allowing IRP-1 to function as a cellular iron sensor. In addition, both hydrogen peroxide treatment and quinone-induced oxidative stress of cultured cells promote a transient increase in the fraction of IRP-1 which binds to IREs. Expression of iron metabolism proteins is also altered accordingly. An intracellular signal other than hydrogen peroxide is presumed to mediate this response, because the activation cannot be reconstituted in vitro. We show that two functional homologues of IRP-1 are conserved in Drosophila melanogaster. Fly extracts possess cytosolic aconitase and IRE-binding activities, which are inversely regulated in response to iron levels. Therefore IRP-1 function in iron metabolism seems to be evolutionarily conserved between invertebrates and mammals. We tested aconitase and IRE-binding activities of the IRP-1s in flies lacking catalase or copper-zinc superoxide dismutase and compared them with those of wild-type control flies. In homogenate extracts from both mutant animals we observed a profound decrease in enzymatic aconitase activity, accompanied by enhanced IRE-binding activity. Thus, the transient activation of IRP-1 IRE-binding activity which follows prooxidant treatment of mammalian cells can be sustained in vivo under conditions of chronic oxidative stress. We are currently investigating the consequences that this response has on iron regulated proteins in order to clarify the molecular components that orchestrate oxidative stress and iron metabolism interactions. Furthermore, we are screening for mutants that will suppress IRP-1 activation, in anticipation of candidate genes that may mediate IRP-1’s response to prooxidant environment. Such genes can be used us drug targets for human disease.

Farideh Beigi, Ph.D, NIA
Immobilization of G Protein-Coupled Receptors on Liquid Chromatographic Stationary Phase for On-line Studies of Drug-Receptor Affinities
The G-Protein Coupled Receptor (GPCR) superfamily is characterized by the presence of seven transmembrane helices. For the first time, G-protein coupled receptors have been immobilized on silica-based chromatographic stationary phases to produce liquid chromatographic columns. The resulting systems can be used as part of drug discovery and drug development programs for the on-line determination of compounds that bind to the immobilized GPCR from simple or complex chemical and biological mixtures. The system can be used for the isolation, identification and characterization of both antagonists and agonists to the immobilized GPCR.
In our laboratory we have approached a novel method in studying the G-protein coupled beta-adrenergic and the opioid receptors and their ligands. Individual receptor subtypes has been solubilized from cultured cell membrane with the anionic sodium cholate detergent. The partially purified receptor complex has been immobilized in the phospholipid monolayer of an immobilized artificial membrane (IAM) chromatographic stationary phase. For immobilization procedure cell membrane pellet produced from 100 million cells was taken and homogenized, for removal of large cell particles and organelles the suspension was centrifuged stepwise. The resultant membrane pellet was suspended in solubilization buffer containing sodium cholate detergent. In order to immobilize the receptor into the gel particles, 0.2 mg of IAM beads was added to the solubilized receptor suspension. This mixture was stirred for several hours at room temperature and dialyzed. Upon removal of detergent the lipids and the proteins reorganize and form a layer with the hydrocarbon chains of the IAM beads. The IAM support containing fragments of the immobilized membrane and the receptor were collected and packed into a HR 5/2 chromatographic glass columns to obtain 0.4 ml of gel bed. The column was connected to an HPLC pump and the ligands were injected separately and detected on-line by a radio-flow detector in both zonal and frontal chromatographic mode. The immobilization efficacy and stability of the receptor into the gel beads was enhanced upon addition of extra lipids.
By frontal affinity chromatography 147 pmole binding sites were calculated for U69593 agonist for the K receptor with Kd = 84 nM and also 0.02 pmole of binding sites were calculated for the beta2-adrenergic receptor with Kd = 0.11 nM for the CGP 12177 antagonist.

FENG YANG, PhD and MD, NICHD
Ca2+ Influx Independent Synaptic Potentiation Mediated by Mitochondrial Na+-Ca2+ Exchanger and Protein Kinase C
Activity-dependent modulation of synaptic transmission is a fundamental mechanism for the development and functions of the brain. Information processing often leads to changes in the efficacy of transmission that last for a short or long period of time, a general phenomenon known as synaptic plasticity. Experimentally, this could be mimicked by repetitive stimulation of presynaptic neurons. Long-term forms of synaptic plasticity that last for at least one hour, particularly the long-term potentiation (LTP) observed at the CA1 synapses in the hippocampus, are thought to be mediated mostly by postsynaptic mechanisms. LTP has been hypothesized to be a cellular correlate for learning and memory. In contrast, short-term forms of synaptic plasticity last from seconds to minutes results from changes in presynaptic transmitter secretion. At the neuromuscular junction (NMJ), a brief, high frequency stimulation (tetanus) results in an enhancement of transmitter release that lasts many minutes called post-tetanic potentiation (PTP). While it is generally believed that the presynaptic forms of plasticity are due to a prolonged elevation of intracellular concentrations of free Ca2+ after the termination of the tetanus, the precise molecular mechanisms for the enhancement of transmitters secretion remain unclear. Here we report an unusual form of synaptic modulation that depends on mitochondrial Na+-Ca2+ exchange, but not on Ca2+ influx. A brief tetanic stimulation of Xenopus motoneurons induced a rapid and dramatic potentiation of transmitter release at neuromuscular synapses in Ca2+-free medium. Blockade of either Na+ influx or the rise of intracellular Ca2+ concentrations ([Ca2+]i) dramatically reduced the effect of tetanus. Inhibition of Ca2+ release from mitochondrial Na+-Ca2+ exchanger significantly attenuated the synaptic potentiation. Tetanic stimulation also elicited an increase in [Ca2+]i which is prevented by inhibition of Na+ influx or mitochondrial Ca2+ release. Inhibition of protein kinase C (PKC) blocked the tetanus-induced synaptic potentiation as well as mitochondrial Ca2+ release. These results reveal a novel form of synaptic plasticity, and suggest a role of PKC in mitochondrial Ca2+ release during synaptic transmission. Our studies may also help understand the contribution of mitochondria and PKC in transmitter release, and provide a useful model to investigate molecular mechanisms for transmitter release without the interference of Ca2+ influx.

FENG HONG, MD, NIAAA
Elevated interleukin-6 during ethanol consumption acts as a potential endogenous protective cytokine against ethanol-induced apoptosis in the liver: involvement of induction of multiple anti-apoptotic proteins
Elevation of serum interleukin-6 (IL-6) levels is always associated with alcoholic liver disease (ALD), but the significance of such elevation is not clear. Here we show that chronic ethanol consumption induces significant apoptosis in the liver of IL-6 (-/-) mice but not IL-6 (+/+) mice. IL-6 (-/-) hepatocytes are more susceptible to ethanol- and tumor necrosis factor Ą- (TNFĄ-) induced apoptotic killing, which can be corrected by IL-6. Expression of both anti-apoptotic (such as Bcl-2 and Bcl-xL) and proapoptotic (such as Bax) proteins is markedly elevated in the liver of human ALD and chronically ethanol-fed IL-6 (+/+) mice. On the contrary, induction of Bcl-2 and Bcl-xL is not observed in the liver of chronically ethanol-fed IL-6 (-/-) mice, whereas expression of Bax protein remains elevated. Injection of IL-6 markedly induces expression of Bcl-2 and Bcl-xL but not Bax in the liver. Finally, high concentrations of ethanol inhibit IL-6-activated anti-apoptotic signal, but increasing the concentrations of IL-6 is able to overcome such inhibitory effect. These findings suggest that elevated serum IL-6 levels in ALD may overcome the inhibitory effect of ethanol on IL-6-mediated anti-apoptotic signals and prevent alcohol-induced hepatic apoptosis by induction of Bcl-2 and Bcl-xL.

Fernando L Pagan, MD, NINDS
Multivoxel Multimetabolite 1H-Magnetic Resonance Spectroscopy in Essential Temor
Abstract Background The pathological substrate of essential tremor (ET) remains unknown although several studies have suggested that the cerebellum and olivocerebellorubral loop are involved. Magnetic resonance spectroscopy is a technique that can access brain pathology noninvasively
Methods We studied 10 patients with ET and 10 age-matched volunteers using a multislice magnetic resonance spectroscopy imaging sequence developed at NIH on a GE 1.5 Telsa scanner to obtain spectra for N-acetylaspartate (NAA), creatine (Cr), choline, and lactate (lac). Routine MRIs were also performed to localize cerebellar regions of interests (ROIs) by coregistering anatomical slices to spectra slices. Phase encoding procedures produced a 32 x 32 array of spectra from voxels having a nominal dimension of 7.5 mm x 7.5 mm x 15 mm, TE=280ms. Cerebella were assessed with a qualitative atrophy scale. ET assessment and quality of life measurements were obtained for each patient
Results Mean cerebellar atrophy scores were similar for ET patients and normal subjects. Left and right cerebellar hemisphere NAA/CR and NAA/Cho ratios were significantly smaller in the ET patients than normal subjects. Mean NAA/CR and NAA/Cho ratios of the vermis and dentate nuclei were not statistically different between the two groups. Comparison of NAA/CR and NAA/Cho ratios to age, gender, tremor assessment, and quality of life measurements were not significant
Conclusions NAA/Cr and NAA/Cho levels are lower in the lateral cerebellar hemispheres in ET patients. This is the first documented anatomical abnormality in these patients. Previous neuropathological data show no neuronal degeneration within the cerebellum of ET patients, and no atrophy difference was found between our two groups. The data suggest that the decreased ratios within the cerebellum may represent an abnormality in synapses or neuronal size, rather than neuronal degeneration

Frederic CATEZ, PhD, NCI-CCR
HMGN chromosomal proteins modulate the binding of Histone H1 to chromatin in live cells.
The structure of the chromatin fiber and the activity of the underlying DNA are affected by numerous factors that need to bind chromatin to exert their function. The linker histone H1 promotes chromatin condensation and is considered as a general repressor of transcription while the High Mobility Group N (HMGN) proteins unfold the chromatin fiber and enhance transcription and replication from chromatin templates. Footprinting analysis revealed that HMGN and H1 binding sites in chromatin partially overlap, and in vitro studies with SV40 minichromosomes indicated that HMGN proteins alleviate histone H1-mediated transcriptional repression and chromatin condensation. These findings suggest that an interplay between histone H1 and HMGN proteins exist and affect the structure of the chromatin fiber and DNA related activities in the context of chromatin
In this study we combined microinjection and Fluorescence Recovery After Photo-bleaching (FRAP) to study competition between proteins for common binding sites in live cells. FRAP allows to monitor the intranuclear mobility of proteins. The intranuclear mobility of chromosomal proteins is related to their chromatin binding properties. We found that HMGN and H1 are dynamically bound to chromatin and that the chromatin residence time of H1 is significantly longer than that of HMGN. Competition experiments between unlabelled proteins and GFP-fusion constructs demonstrated that HMGN proteins compete with H1 in vivo. Various HMGN proteins increased the mobility of heterochromatin bound H1 and euchromatin bound H1 to a different extent. The effect of HMGN on the mobility of H1 involves competition for chromatin binding sites since point mutants of HMGN, which did not bind to chromatin properly, do not affect the intranuclear mobility of H1. HMGN affected the dynamically bound fraction of H1 but did not displace the statically bound fraction of H1
In conclusion, our results show that HMGN proteins can modulate the interaction of histone H1 with chromatin in living cells. The effect of HMGN on H1 binding involves competition for shared binding sites. A local concentration of HMGN could weaken the interaction of H1 with chromatin, thereby ''unfolding'' chromatin and providing an opportunity for the binding of regulatory factors to their chromatin target. Thus, such a dynamic interplay between nucleosomal binding proteins may lead to changes in chromatin structure and plays a role in regulating gene expression.

Gabriela Lavezzari, PhD, NINDS
Differential endocytic sorting of NMDA receptor NR2 subunits
Glutamate is the major excitatory neurotransmitter in the central nervous system. The ionotropic glutamate receptors, which include AMPA NMDA, and kainate receptor subtypes, mediate fast excitatory neurotransmission and are of particular interest because they play a central role in several models of synaptic plasticity such as long-term potentiation. AMPA receptors have been shown to rapidly cycle in and out of the postsynaptic membrane, and these dynamic properties are thought to mediate changes in synaptic strength at excitatory synapses. In contras, NMDA receptors were thought to be a stable component of the postsynaptic density. There is now evidence that NMDA receptor surface expression is also tightly regulated, such that NMDA receptors undergo robust internalization early in development, but become more stable as neurons mature. The cytosolic C-terminus of NR2B, an NMDA receptor subunit which is expressed early in development, contains endocytic motifs contributing to the dynamic properties of NMDA receptors. However NR2A, unlike NR2B, is only expressed later in development. To elucidate subunit-specific properties regulating NMDA receptor surface expression, I have used a chimera which contains the interleukin-2 receptor alpha chain (Tac) and the distal portion of the NR2A tail (TacNR2A) or the NR2B tail (TacNR2B). I have analyzed the different endocytic sorting signals within the NR2A and NR2B C-termini in both HeLa cells and in neurons, using immunofluorescence and ELISA techniques. I found that like NR2B, the NR2A C-terminus was sufficient to induce the internalization of the stable surface protein Tac. Interestingly, although both NR2A and NR2B possess internalization motifs, they regulate trafficking through distinct intracellular pathways. While both TacNR2A and TacNR2B colocalize with early endosomal markers at early time points (5 min), TacNR2A moves into Rab9- and Rab7-positive late endosomal compartments at later time points (15, 30 min). In contrast, TacNR2B continues along the recycling endosomal pathway at later time points. There is also evidence that the degradation rates differ between TacNR2A and TacNR2B. These data demonstrate the presence of different targeting motifs encoded within the C-termini of the NR2A and NR2B receptor subunits and suggest that subunit-specific trafficking contributes to NMDA receptor surface expression throughout development

Gerard W Dougherty, Ph.D., NIDCD
CP3 Links Prestin to the Actin Cytoskeleton of Auditory Hair Cells: Implications for Electromotility
Auditory outer hair cells (OHCs) undergo voltage-dependent elongation and contraction in response to sound, a process known as ''electromotility.'' The phenomenon of electromotility is unique to these cells only, and represents the only known biological motor that converts voltage to membrane-based movement, in contrast to cytoskeleton-based motors which are ATP-dependent (myosin, actin, kinesin). This cellular movement is believed to act as the cochlear amplifier to heighten sensitivity of hearing. A recently cloned integral membrane protein, Prestin, is exclusively expressed in OHCs and confers voltage-dependent capacitance and motility to heterologous cells, and is thus the likely ''motor protein'' that drives electromotility. To investigate the molecular components of electromotility and to identify potential binding proteins to Prestin, we screened a mouse organ of Corti yeast two-hybrid library. One clone, designated CP3, bound to Prestin in yeast under high-stringency conditions. CP3 encodes a novel 27kD protein that contains a single amino-terminal calponin homology (CH) domain. CP3 and Prestin were colocalized to the lateral wall of outer hair cells by immunofluorescence and immunogold electron microscopy. Interestingly, the CH domain of CP3 predicts binding to F-actin, and we have colocalized endogenous CP3 with actin in cultured cell lines. Freeze-fracture of OHCs reveal an intricate network of spectrin and actin intimately associated with the lateral wall, where Prestin and CP3 are present. These results suggest that CP3 may link Prestin to the actin cytoskeletal network of OHCs and participate in a unique mechanism that relays membrane-based, voltage-dependent motor activity to the actin cytoskeleton to engage cellular movement.

Gheorghe Iordanescu, Eng, CC
Automated Centerline Detection Algorithm for CT Colonography
Introduction: CT colonography (CTC, also known as virtual colonoscopy) is a relatively new diagnostic test for the colon. CTC consists of processing CT slices of the patient rather than the traditional invasive colonoscopic inspection. CTC could be very useful for colon cancer screening since it increases the patient’s comfort and reduces the time necessary for the patient to be investigated. Our purpose is to obtain the centerline of the reconstructed colon surface by splitting the colon surface into rings or small cylinders of equal length whose axes are parallel to the colon’s centerline
Method: Our method involves six steps applied sequentially
1. Decimation: The 3D colon surface is preprocessed to reduce the number of triangles and vertices composing it
2. Vertex classification: Every vertex is assigned a class number, based on its Z coordinate 3. Thinning: The thinned colon is computed by iteratively averaging the distances between colon vertices
4. Modeling: The thinned colon is modeled by an ordered group of 3D points by taking approximately equally-spaced distanced vertices from the thinned colon
5. Remapping: We use the model to compute rings perpendicular to the thinned colon. Using the one-to-one mapping of the vertices on the thinned and original colon, the indices of the vertices from the same ring are used to get a slice (or a ring) in the original colon
6. Centerline computation: The ring-like areas are processed to compute the local centerline point
Results: CT colonography was performed on 10 patients. The centerline algorithm was tested by visual inspection of the resulting colonic surfaces. We obtained a centerline that stayed inside the colon except in extreme cases such as nearly collapsed segments having a wall convex to the lumen. Total average execution time is 6 minutes
Discussion: We have developed a new method to segment topologically cylindrical shapes. Our algorithm uses only the reconstructed 3D surface. We do not need the original CT images. Our method is also simple, fully automated (for well distended colons) and relatively quick (average processing time is 6 minutes). Our algorithm provides virtual fly-path elements (projection of the rings onto a plane perpendicular to their axis). Potential applications of our method include navigation path definition and mapping the surface to the centerline, which could improve polyp detection.

Giancarlo Pascali, Chemistry, CC
Use of 1,8-bis(dimethylamino)-naphtalene / H18F complex as new radiofluorinating agent for the production of radiopharmaceuticals for PET
INTRODUCTION Positron Emission Tomography (PET) is an imaging methodology of high impact on both clinical and biomedical research. It relies on radiotracers labeled with positron emitting short-lived radionuclides, such as 18F (T1/2 109 min) and 11C (T1/2 20 min). The investigation of new methods for the introduction of 18F atoms into biologically interesting molecules may widen PET applications. We have explored the use of the complex between 1,8-bis(dimethylamino)-naphtalene (Proton Sponge, PS) and H18F in nucleophilic substitution reactions
RESULTS We studied nucleophilic aromatic substitution on a variety of halogenated aromatic model compounds and a few aliphatic tosylates (OTs) and mesylates (OMs), using both non radioactive HF and H18F. The following reaction parameters were considered: dehydration of aqueous fluoride, stoichiometry of reagents and substrates, molarity of substrate, reaction times, solvents (including ionic liquids), heating conditions (thermal vs microwave). Yields based on fluoride of non-radioactive fluorinated products were generally very low, failing in case of aliphatic OTs and OMs. This was not always the case with the corresponding reactions with radioactive 18F which afforded the labeled products in modest to excellent yields. We optimized reaction conditions for 2 substrates, 2-chloro-3-nitropyridine (2CNP) and pentamethylbenzylic ester of 4-fluoro-3-nitrobenzoic acid (FNE). Optimized yields (based on 18F total radioactivity and corrected for decay) were 61% for [18F]-2-fluoro-3-nitropyridine and 91% for pentamethylbenzilic ester of [18F]-4-fluoro-3-nitrobenzoic acid. The optimized conditions required addition of a small quantities of an ionic liquid (a new class of high boiling solvents) as a co-solvent to CH3CN, thermal heating at 100C degrees for 30 min. The concentration of substrate was optimized at 0.1M for FNE and 0.2M for 2CNP. The utilization of microwave gave lower labeling yields but in considerably shorter times
CONCLUSION A new radiofluorination method using of PS/H18F complex with an ionic liquid as cosolvent was developed. The method provides the efficient 18F-fluoride incorporation into 2CNP and FNE. This method provides an alternative approach for 18F-labeling that may be useful for the preparation of new PET-radiopharmaceuticals.

Giovanni Di Pasquale, Ph.D, NIDCR
Identification of PDGFR as a receptor for AAV5 transduction by comparative gene expression analysis using cDNA microarrays
Gene transfer vectors based on adeno associated virus (AAV) have been gaining popularity due to a combination of attractive features. AAVs are naturally defective for replication and considered to be nonpathogenic. They can transduce both mitotic and post-mitotic cells, efficiently transfer genes to a number of cell types, and mediate long-term gene expression. Most pre-clinical AAV studies and current phase I clinical trials use vectors derived from AAV serotype 2 (AAV2). However, other serotypes have gained attention as potentially useful vectors for gene transfer applications because of improved transduction efficiencies in vivo compared to AAV2. Vectors based on AAV serotype 5 (AAV5) more effectively transduce cells in the lung, CNS, muscle, and eye compared with AAV2
In order to understand the natural tropism of this virus and identify proteins necessary for AAV5 transduction we have compared the gene expression profiles in 42 cell lines that were defined as permissive or non-permissive for AAV5 transduction using cDNA microarrays
A statistically significant correlation was observed between expression of the platelet derived growth factor receptors (PDGFRs) and AAV5 transduction. The importance of this protein in AAV5 transduction was confirmed by the specific inhibition of AAV5 transduction following treatment with SB203580, an inhibitor of PDGFR expression. The change in transduction observed following treatment with SB203580 also correlated with a change in virus binding to the cell surface. Transfection of HeLa, PAE or A549 cells, either weakly or non-permissive for AAV5, with either PDGFR-Alpha or PDGFR-Beta expression plasmids increased AAV5 transduction but had no effect on AAV2 transduction. Similarly, transduction of the myloid progenitor cell 32D-AlphaR or 32D-BetaR, which express the PDGFR-Alpha or PDGFR-Beta respectively, dramatically increased compared to the parental cells or 32D cells expressing the EGFR tyrosine kinase. Additional experiments demonstrated that incubation with PDGF-AA, the ligand specific for PDGFR-Alpha, could inhibit transduction of the 32D-AlphaR cells. Addition of ligand also blocked binding of AAV5 to the cell surface but did not inhibit virus binding or transduction of 32D-BetaR cells
The identification of PDGFR as a receptor for AAV5 should provide new insights not only into AAV5 biology and natural tropism, but also in the optimal use of AAV5 vectors for human gene therapy.

Greg A Snyder, HopefullyPhD, NIAID
Oligomer DC-SIGNR binds avidly to HIV-1 gp120 and can be blocked by specific DC-SIGN antibodies
The dendritic cell specific ICAM-3 grabbing nonintegrin (DC-SIGN) has been found to bind high mannose oligosaccharides on HIV-1 viral particle envelope glycoprotein gp120 greatly enhancing virus infection efficiency by a mechanism that is not understood. DC-SIGN and closely related DC-SIGNR are C-type lectins consisting of four major domains, a short intracellular signaling domain, a transmembrane domain, a repeat domain, which contains 7.5 copies of a 23 amino acid repeat and a functional carbohydrate recognition domain (CRD). Using size exclusion chromatography, we show that by regulating the number of repeat domains included with the CRD we are able to titrate receptor oligomerization from monomer, to dimer and higher oligomers. We hypothesize that oligomers of DC-SIGNR will bind more avidly and may provide specificity through recognizing not only high mannose carbohydrates but also their number and spatial distribution. We show using surface plasmon resonance that oligomerized DC-SIGNR bind more avidly to HIV-1 envelope glycoprotein gp120 than monomer or dimer species as seen by an increase in dissociation constant (KD) from 10-20nM for oligomer to ~10uM for monomer species. We show binding is carbohydrate dependent using a nonglycosylated variant of gp120 that binds oligomer DC-CISGNR with a 1000 fold decrease in KD. Since DC-SIGNR is calcium dependent and selectivity binds high mannose oligosaccharides we show binding can be inhibited by incubating DC-SIGNR (monomer, dimer or oligomer) with mannose or by chelating calcium ions using EDTA. In addition to nonspecific methods to inactivate DC-SIGNR activity we identify two DC-SIGNR antibodies that specifically block the DC-SIGNR and gp120 interaction. We also tested the gp120 carbohydrate specific antibody 2G12. Interestingly, binding was only partially blocked, possibly indicating that oligomerized DC-SIGNR bind to more than one high mannose carbohydrate found on gp120. We believe that DC-SIGNR oligomerization which increases binding avidity may provide specificity by recognizing not only high mannose oligosaccharides but also their number and specific spatial distribution pattern on envelope glycoprotei