2003 Summer Scholar Participants-HHMI Undergraduate Program - Carnegie Mellon University

2003 Summer HHMI-supported Participants

Yuliya Anikanova

Yuliya Anikanova, Carnegie Mellon University (Advisor: Dr. Gordon Rule)

Mutating the H-site of Glutathione S-Transferase Pi

Glutathione transferases (GSTs) are a family of detoxification enzymes that catalyze the nucleophilic attack of glutathione protein (GSH) on the electrophilic centre molecule of xenobiotics. The reaction is as follows: GSH + R-X => GSR + HX, where R-X is the electrophilic substrate and GSH is the glutathione activated by GST for nucleophilic attack. There are at least five classes of GSTs but the work below is concerned with human GST class pi. The mechanism and kinetics of this GST class are being studied in numerous laboratories because of its potential as a marker during carcinogenesis and its alleged role in the antineoplastic drug resistance in cells. Tyr 108 residue in the hydrophobic pocket (the electrophilic substrate binding site), has been shown to have an important role in the kinetics of the GSH and R-X reaction. The mutation of tyrosine 108 to structurally incongruent to alanine can potentially have a significant effect on GST substrate interaction. This mutation has been constructed and the enzyme-bisubstrate interaction is being characterized by studying the kinetics of the mutant enzyme.

Jeffrey Clarke

Jeffrey Clarke, Indiana University (Advisor: Dr. Jonathan Minden)

Protein Capture by a Maleic Anhydride Column for Use in Difference Gel Electrophoresis and Proteomic Analysis

Difference gel electrophoresis (DIGE) for proteomic analysis relies on fluorescently tagging proteins from two different whole cell extracts. These sets of tagged proteomes are then applied to two-dimensional polyacrylamide gels and subsequently imaged under differing wavelengths. Current cell extraction techniques fail to separate proteins from nucleic acids, lipids, and other charged, intracellular entities, resulting in the disruption of isoelectric focusing. This interference is problematic in attaining uniformity between samples. Many common protein precipitation and resuspension protocols demonstrate low recovery efficiencies, which is unfavorable in proteomic analysis. A novel system for proteome separation proposes utilizing a maleic anhydride capable of amide formation with free amino groups under basic conditions. This reaction is completely reversible at low pH levels. By applying whole cell extracts to an agarose column that bears maleic anhydride groups, proteins will covalently bind to the column via their amino-termini and lysine residues at pH > 8. Lowering the pH below 6 will cause the release of the polypeptides. Purifying cell extracts by maleic anhydride will aid isoelectric focusing used in DIGE analysis, while retaining a high protein recovery efficiency.

Jamie Conklin

Jamie Conklin, Carnegie Mellon University (Advisor: Dr. Javier Lopez)

Analysis of Splicing by Characterization of Intron Lariat Populations

Alternative splicing of pre-mRNA can generate functionally distinct protein isoforms from a given gene or control the levels of expressed protein. Previously, I identified three genes in Drosophila whose alternative splicing correlatfor experience-dependent neuronal plasticity and may help to regulate this plasticity. Our bioinformatic analyses suggest that many plasticity-related RNAs may be alternatively spliced during this period. To develop a more complete picture of the role of alternative splicing in neuronal plasticity, I worked on methods for large-scale analysis of splicing based on comparison of intron lariat populations. Lariats are a side product of the 5¹ splice site cleavage reaction, during which the 2¹ hydroxyl group of the branchpoint adenosine 15-40 nt upstream of the 3¹ splice site is joined to the 5¹ end of the intron by a phosphodiester bond. Short fragments flanking the 2¹-5¹ phosphodiester bonds of lariats can be affinity purified, amplified by RT-PCR, and sequenced to identify the ends of the introns and, thus, the exons being joined by each splicing event in a cell. To identify changes in alternative splicing, the frequency of each lariat type in two sequenced populations may be compared, or differentially expressed lariats may first be identified (e.g., by differential display of lariat cDNAs) and then sequenced. In either case, efficient analysis requires that reverse transcription initiated within the lariat not terminate preferentially at the three-way junction, as it does when primed from the 3¹ end of the intron. I tested this with a specially designed RT-PCR assay and found no evidence of preferential termination for cDNA initiated within the lariat loop under my reverse transcription conditions. I then proceeded to test a method for differential display of lariat fragments based ones with a critical period PCR with short arbitrary primers. I used a debranching enzyme mutant yeast strain (dbr-) in which lariats are stabilized. To test the approach, 8 primers of 10-13 nucleotides were designed that could prime on multiple lariats, including those for RPS14A and RPS14B. As predicted, different combinations of these primers amplified the predicted RPS14A and B lariat fragments plus different patterns of unrelated bands. One combination of these primers was then used on cDNA ³doped² with different concentrations of RPS14B lariat to show that differences in relative expression level can be detected. In addition, I implemented methods for synthesis of second strand cDNA from short lariat fragments by template switching, and for subsequent generation of lariat cDNA concatemers for efficient sequencing. Finally, to facilitate lariat analysis in Drosophila, I tested whether a specific lariat accumulates to higher levels in heterozygotes for the lethal deficiency Df(3L)pbl-X1, which deletes the homolog of yeast DBR1. I did not observe enhanced lariat accumulation in the heterozygotes, suggesting that half a dose of dDBR does not limit lariat turnover. Future experiments will make use of transgenes expressing double-stranded RNA to knock down dDBR expression more completely in a stage-and tissue-specific manner.

Rhonita Culver

Rhonita Culver, Alabama State University (Advisor: Dr. David Hackney)

Oligomeric Organization of Kinesin Motor Proteins

Conventional kinesin is a motor protein that is likely to be responsible for the transport and movement of organelles within cellular membranes along microtubules. Its native structure consists of two dimerized globular head groups joined together by a coiled-coil neck region. A protein of interest, BimC, exists within the kinesin superfamily. BimC is a motor protein whose structure is similar to that of kinesin. Like kinesin, it contains a globular head group and a possible coiled-coil region. The head domain of BimC is a monomer that dimerizes when joined with a coiled-coil region. It was our goal to analyze the sedimentation and diffusion coefficients of both kinesin and BimC by the methods of sedimentation and gel filtration, so that a greater knowledge of their oligomeric organization could be obtained. Research on the coiled-coil region of BimC was conducted to see if it was capable of dimerization within heterologous systems. Previously, studies found this region to be dimerized; yet, it was discovered that interaction of the coiled-coil region alone was not effective enough to cause dimerization. In addition to this, an investigation of the stability of dimerized BimC and kinesin in dilutions was performed to view whether or not differing salt concentrations cause the dimers to dissociate. Dimerization was seen at high concentrations, but the unstable dimer dissociated at low concentrations. Because of these findings, confirmation of the oligomeric state of BimC and kinesin constructs was obtained. Due to the fact that oligomeric states establish the manner in which these motor proteins move along microtubules, future studies will involve assessment of ATPase rates.

Kathleen Dean

Kathleen Dean (Advisor: Dr. Alison Barth)

Protein Expression Profiles of FosGFP Transgenic and Wild Type Mice

In order to assess the functionality of the fos-Green Fluorescent Protein (GFP) fusion protein in fosGFP transgenic mice, we will compare the protein expression profiles in the hippocampus of wild type and transgenic littermates under both control and induced conditions. To aid in this assessment, we also aim to analyze the levels of specific proteins and other immediate-early gene products as well as proteins that are induced after c-fos expression in fosGFP transgenic animals. We hypothesize that fosGFP fusion protein levels will be shown to mirror levels of the endogenous protein, indicating 'normal' neuronal function in fosGFP transgenic animals relative to their wild type counterparts.

Anne Dollard

Anne Dollard, Carnegie Mellon University (Advisor: Dr. Elizabeth Jones)

Characterization of Mutations in the Yeast Debaryomyces hansenii

The aim of this project was to learn about the yeast Debaryomyces hansenii, importing the molecular biological techniques usually used to study Saccharomyces cerevisiae. The properties of these two strains were compared. The goals of this project were threefold. First we wanted to analyze the existing tryptophan (trp) mutant strains of D. hansenii (isolated by Tomas Bolumar) to identify the mutated gene. Also, we wanted to determine whether various Trp- strains undergoing starvation secrete anthranilate. Finally, we wanted to isolate fluoroanthranilic acid (FAA) resistant mutants in the strains S. cerevisiae (BJ 8919) and Schwanniomyces castellii (IGC 2827). To analyze the existing trp mutations, several methods were tried. First, we attempted to transform into D. hansenii a S. cerevisiae TRP1 plasmid to complement the mutations. We were unable to obtain transformants by either LiAc or electroporation. The method of spheroplast fusion was then adopted. We hoped to fuse each of the D. hansenii trp mutants with an identified trp mutant of S. cerevisiae. The theory is that if the mutations present in the two strains are in orthologous genes, the fused cells will require tryptophan, but if the mutations present are not in orthologous genes the strains will complement one another and the fused cells will be TRP. Because of inefficiency, this method was also abandoned; instead new vectors were created. Using the TOPO TA cloning kit (Invitrogen), the TRP3 and TRP4 genes from S. cerevisiae were individually cloned, as well as the autonomously replicating sequence (ARS) from D. hansenii. Two vectors were constructed - one containing the ARS and TRP3 and one containing the ARS and TRP4 ­ and were used to complement the D. hansenii mutants by transformation. The TRP4 gene was successfully inserted into a vector. The same procedure is being performed on the TRP3 and D. hansenii ARS genes, and further tests will confirm the outcome. To determine whether various Trp- mutants are secreting anthranilate, the technique of thin layer chromatography was chosen. Trp- mutants were grown in two different media. Initially, cells were cultured in complete­trp broth with a limiting amount of tryptophan. The other way of culturing cells was to allow them to grow overnight in YEPD, then change the medium to complete­trp for several hours to starve the cells. In both cases, 50 uL of the supernatant was loaded onto a piece of chromatography paper and sprayed with a battery of chemicals. A control solution of anthranilic acid was loaded as well. If there is anthranilate in the supernatant, a purple color will appear following the application of the chemicals, because anthranilate contains a diazotizable amine. We confirmed that the Trp4- cells are secreting anthranilate. FAA resistant mutants were isolated by exposing patches of either S. cerevisiae or S. castellii to ultraviolet (UV) light for varying lengths of time. The concentration of FAA was optimized to prevent wild type cells from growing and only allowing FAA resistant cells to grow.

Cesar Guerrero

Cesar Guerrero, Carnegie Mellon University (Advisor: Dr. Peter Berget)

Cloning of Bacterial Phage T4 Gene 53 and Purification of Protein 53

Protein 53 has a significant role in the infectious activity of T4 bacteriophage. Protein 53 is approximately 22 KDaltons and is located on the wedge of the injection apparatus of T4 called the baseplate. The mechanism of action of protein 53 is still unknown, but most likely it polymerizes with other proteins, allowing it to recognize and bind to the surface of bacteria such as E. coli. In the absence of protein 53, T4 is unable to infect bacteria. The goals of this project were to clone gene 53 by PCR amplification, overproduce protein 53, confirm the biological activity of this protein, and purify the protein for structural studies. First, T4 was generated in order to isolate T4 DNA. Custom designed primers were used in conjunction with PCR to amplify gene 53 of the T4 genome. The PCR product was then sequenced, confirming its identity. Cloning of the PCR product into a PET24d+, a bacterial expression vector resistant to kanamycin, allowed introduction of the created plasmid into E. coli BL21 with the hopes of expressing protein 53. An SDS electrophoresis gel of the induction of BL21 cells with PET24d+g53 did not resolve a protein of approximately 22 Kdaltons. A bioassay with T4 53-t- (T4 with an amber mutations in genes 53 and t) and the protein extract from the induced E. coli BL21 cells suggest that protein 53 had not been produced.

David Hill

David Hill, Carnegie Mellon University (Advisor: Dr. Elizabeth Jones)

Systematic Deletions of the Carboxy-Terminal End of Pbn1p in Saccharomyces cerevisiae

PBN1 is an essential gene that encodes an ER protein required for post-translational processing of Prb1p, the precursor of the vacuolar hydrolase protease B (PrB). Thus, pbn1-1 mutants show a defect in the ER processing of Prb1p. The PrB defect alone is not lethal as shown by the viability of prb1 deletion strains. As a result, Pbn1p is suspected to have other critical roles in the cell, presumably in protein folding and ER quality control. In hopes of better understanding the role of Pbn1p, the polymerase chain reaction (PCR) technique was used to generate alleles of PBN1 that had the final, penultimate, and antepenultimate C-terminal amino acids converted to stop codons. The C-terminal end was targeted because it has been shown previously that deletion of the C-terminal cytoplasmic tail-encoding region of PBN1 is lethal, indicating it is essential for the function of Pbn1p. In effect, this process caused the deletion of these amino acids in the protein, which we hope will disrupt function enough to cause a conditional mutation in PBN1. PCR primers were designed to include the mutations. A plasmid encoded wild type PBN1 gene was used as the template. The pbn1* fragment was then gap-repaired back into the original PBN1 plasmid. The PCR primers also incorporated a novel restriction site that aided in screening for proper mutations. Mutant alleles were sequenced to ensure the correct constructs and were transformed into haploid pbn1_ strains carrying plasmid encoded PBN1 for complementation tests. The haploid pbn1_ strain was then selected for the loss of the URA3-marked PBN1 WT plasmid, leaving only the plasmid encoded pbn1* in the deletion strain. These strains were tested for growth temperature sensitivity of pbn1* and for the maturation of PrB.

Zachary Kahler

Zachary Kahler, Carnegie Mellon University (Advisor: Dr. Alison Barth)

Size Doesn't Matter: An Analysis of Interneuron Density in the Barrel Field

In an evolutionary sense, the cortex is the most advanced part of the brain. The cortex is responsible for sensory processing, voluntary motor control, and conscious thought, among other things. In higher forms of life, the cortex expands to cope with similarly increasing processing demands; however, the vertical height of the cortex stays proportionately the same across species. Expansion of the cortex must therefore come laterally. This lateral expansion comes from the addition of cortical columns- discrete units of functionality that are added on as needed. In order to determine the extent of similarity among neuron density across functional groups in the cortex, we performed immunohistochemistry for Somatostatin and Parvalbumin cells versus total cells in small and large barrels in the whisker barrel field of rats and mice. Interneuron density was not found to change across small and large barrels.

Lirona Katzir

Lirona Katzir, Carnegie Mellon University (Advisor: Dr. Adam Linstedt

The Function of COP I Protein Complexes in the Disassembly of the Golgi Complex During Mitosis

One of the most interesting and least understood aspects of the Golgi apparatus, which is responsible for the processing and sorting of proteins throughout the cell, is the mechanism by which it disassembles during mitosis and evenly distributes itself into two daughter cells. Generally, the Golgi apparatus breaks down into hundreds of vesicles at the onset of mitosis that later fuse to evenly reform two separate Golgi complexes in each daughter cell. The mechanism mediating this, however, is still unknown. The most widely accepted model, which can be referred to as the docking-inhibition model, suggests that Golgi disassembly is a result of vesiculation mediated by the COP I coat complex, and as a result, inhibits vesicle docking. However, recent results do not coincide with this model, suggesting that an alternate mechanism may be responsible for Golgi inheritance. Both in vitro and in vivo experiments demonstrate that COP I is not necessary for mitotic Golgi vesiculation. Further, Golgi will breakdown at interphase when COP I is inhibited. Therefore, mitotic Golgi vesiculation might not be a result of COP I complex function, but rather its inhibition. The known mechanism of COP I involves the interaction of the COP I proteins with the peripheral membrane G-protein ARF-GEF. When ARF-GEF interacts with ARF-GDP, GDP is released, allowing GTP to bind, which, in turn, initiates COP I protein recruitment. Therefore, COP I recruitment is dependent on the presence of ARF-GEF. We performed a permeabilized cell COP I recruitment assay, replacing the cytosol, including the COP I proteins, with mitotic(M) cytosol in order to determine if the COP I proteins are functional at mitosis. COP I recruitment can be directly visualized by immuno fluorescent staining for _-COP, one of seven subunits that make up the COP I protein complex. We determined that in the presence of interphase peripheral membrane proteins, mitotic COP I proteins are functional. However, to be sure that COP I recruitment was not due to the presence of the interphase peripheral membrane proteins, high salt washes were performed to get rid of peripheral membrane proteins, specifically the ARF-GEF. The cells were incubated separately with either mitotic cytosol or interphase cytosol and results revealed that the level of recruitment detected in the cells incubated with interphase cytosol were higher than that of mitotic cytosol. As a result, it could be suggested that Golgi disassembly may not be a result of vesiculation mediated by the COP I coat complex, but rather by an alternate mechanism.

 

Justin Ker, Carnegie Mellon University (Advisor: Dr. Chien Ho)

A Biophysical Investigation of Recombinant Hemoglobin rHb (βH117C)

Hemoglobin is a 60kDA tetrameric protein that transports oxygen from the lungs to tissues in humans. Besides serving as an essential component of the circulatory system, hemoglobin also provides a model for studying structure-function relationships in multimeric, allosteric proteins. The objectives of this investigation were to construct recombinant hemoglobins (rHb), each with a single mutation on its exterior surface, and to characterize their functional and structural properties. The Escherichia coli expression system developed in our laboratory was used to substitute b117 His with a Cys residue to create rHb(bH117C). A second mutant rHb(aH20C) was created with the same technique; this mutant had its a20 His replaced by a Cys. Previous studies have shown that mutants with cysteine residues on their surfaces have formed octameric strains of hemoglobin. These octamers are a possible blood substitute, as they do not dissociate readily and are functionally similar to normal adult HbA. Oxygen binding studies and Nuclear Magnetic Resonance (NMR) have revealed that rHb(bH117C) is very similar to HbA in terms of structure and function; it is most likely that rHb(bH117C) did not form an octamer. The oxygen binding studies for rHb(aH20C) contrast markedly with HbA, while its NMR spectra shows some deviation from HbA. rHb(aH20C) is possibly an octamer; this will have to be confirmed by mass spectroscopy.

Meng Lu

Meng Lu, Carnegie Mellon University (Advisor: Dr. Elizabeth Jones)

Finding Functions of the Pbn1 Gene by Tagging

PBN1 is an essential gene that codes for an ER integral membrane protein in Saccharomyces cerevisiae. The only known function of Pbn1p is to facilitate the autocatalytic cleavage of the protease B precursor in the ER. Cells lacking protease B (PRB1) can survive, but deletion of PBN1 is lethal. We are trying to understand the essential role of PBN1 using biochemical, cell biological, and genetic approaches. In this project we have tagged the C-terminus of the genomic copy of PBN1 individually with a hemagglutinin (HA) tag and Green Fluorescent Protein (GFP). These tags were obtained by the virtue of homologous recombination in yeast. The PBN1-HA tag has been confirmed by PCR to be bona-fide for insertion of the tag and by Western blot for expression of the fusion protein. In the case of the PBN1-GFP, PCR amplification showed the right size product, but when we observed the GFP-tagged cells under a fluorescence microscope, there was no GFP signal. Absence of a GFP signal can be attributed to a) the tag being out of frame, b) GFP being misfolded, or c) Pbn1p being of low abundance under physiological conditions. We are doing further tests to clarify the cause of the problem. It has been previously determined that the pbn1-1 allele shows a synthetic lethal interaction with ero1-1. Ero1p is an ER protein involved in oxidation and disulfide bond formations. This knowledge led us to suspect possible genetic interactions between mutant alleles of PBN1 and PDI1. PDI1 is also an essential gene whose protein product is involved in facilitation of protein folding by catalyzing disulfide bond formation. We planned to observe this interaction by checking the viability of a strain carrying both the pbn1-1 allele and the PDI1 allele (pdi1S2S6). The pdi1S2S6 allele contains two serines instead of cysteines in the active site. A selectable marker (HIS3) was successfully gap-repaired upstream of a plasmid-borne copy of pdi1S2S6 to allow selection for incorporation of this allele into the genome replacing the wild type allele. We were able to PCR amplify the HIS3-pdi1S2S6 from the plasmid and are in the process of inserting this fragment into the genome at the wild type PDI1 locus. This new strain will then be crossed with the strain containing pbn1-1 in an attempt to produce a doubly mutant pbn1-1 pdi1S2S6 to check for lethality.

 Shail Mehta

Shail Mehta, Carnegie Mellon University (Advisor: Dr. William Brown)

Ribonuclease S as a Model for Biosensor Development

Biosensors are devices that consist of a sensing biological molecule that is attached to or incorporates a signal transduction element. Upon the interaction of the biological molecule with a particular substrate, the transducer gives a measurable signal. Antibodies, attached to environmentally-sensitive fluorescent dyes, can be employed as biosensors. However, not all antibody-antigen systems are well understood. Ribonuclease S, a very well-characterized system resulting from the digest of Ribonuclease A by subtilisin, may instead serve as a model for biosensor development. RNase S consists of a 20 amino acid-long S-peptide and 104 amino acid-long S-protein, and retains RNase activity. The S-peptide binds to the S-protein with extremely high affinity (Kd = 500 nm), and so RNase S can model antibody-antigen interactions and be used to 'screen' dyes for potential as transducers. Variants of the S-peptide have been studied. Deleting the five carboxy terminal residues of the peptide has little effect on binding, but deleting residues from the amino terminus results in decreased binding and RNase activity. Truncated variants of the S-peptide with cysteine (C) added to either the carboxy or amino terminus (1-15C, C2-15, C4-15) were synthesized and conjugated to fluorescein-5-maleimide. These dye-peptide conjugates were titrated with S-protein, and the dyes monitored for an increase in fluorescence polarization. For all conjugates, the polarization of fluorescein increased significantly upon addition of S-protein. No change was observed upon titration with RNase A. The maximum polarization change appeared to depend on the location of fluorescein, and a higher change (.15) was observed for 1-15C-fluorescein than for the other conjugates (.075). In addition, the concentration of S-protein needed to give the maximum polarization change was lower for those peptide conjugates with a higher affinity for S-protein. Thus, fluorescein has potential as a transducer in this system. In the future a fluorescein-peptide conjugate will be covalently attached to S-protein in such a way that it can be reversibly displaced from the binding pocket by wild-type S-peptide; this would create a model of a Œreagentless¹ biosensor.

Michael Palmer,

Michael Palmer, Carnegie Mellon University (Advisor: Dr. Jonathan Minden)

The Study of cAMP In Heat-Shocked Wild-Type and HSP82 Mutant Saccharomyces cerevisiae

Previous studies of the HSP82 mutants in yeast using difference gel electrophoresis (DIGE) have shown that certain other heat shock proteins are up-regulated as compared to wild-type when both are grown in heat shock conditions (37ýC). It is known that the transcription of the genes encoding for these proteins is governed by Msn2/4 transcription factors. Msn2/4 activity is negatively regulated by the cAMP signaling pathway. The same DIGE studies showed that in these HSP82 mutants, the expression of several metabolic enzymes decreased; the transcription of these is regulated by Rap1 and Abf1 transcription factors. Rap1 and Abf1 activity are positively regulated by cAMP signaling. The goal of this project is to quantify levels of cAMP in wild-type and mutant S. cerevisiae in order to explain certain protein expression levels in HSP82 deletion mutants. Our hypothesis is that cAMP levels decrease in HSP82 mutants compared to wild-type when both are grown under heat shock (37ýC), thus explaining the protein expression level differences. Wild-type and mutant strains were grown exponentially under heat shock conditions; cAMP and proteins were extracted. The levels of cAMP were quantified using an indirect fluorescence assay. Protein levels were determined by a colorimetric assay and used to normalize cAMP results. Due to the presence of an undefined inhibitor, the results obtained to date have provided inconclusive evidence to either support or challenge our hypothesis.

Jennifer Richards

Jennifer Richards, Cedar Crest College (Advisor: Dr. Javier Lopez)

Reduction of Lariat Debranching Enzyme Activity in Drosophila

Eukaryotic genes are composed of introns and exons. Splicing of the primary transcript removes the introns and joins the exons to produce the mRNA. Cleavage of the 5¹ splice site during the first step of intron removal produces an RNA lariat in which the 2¹ hydroxyl group of the branchpoint adenosine 15-40 nt upstream of the 3¹ splice site is joined to the 5¹ end of the intron by a phosphodiester bond. Analysis of lariats provides information about mechanisms of alternative splicing; for example, it can distinguish between regulation at the first versus second step of splicing or between exon skipping versus recursive splicing. Lariats are short-lived because debranching enzyme breaks the 2'-5' phosphodiester bond, but they can be detected in normal Drosophila and mammalian cells using an RT-PCR strategy. However, lariat abundance can be elevated more than 1000-fold in a yeast mutant deficient for the debranching enzyme gene (dbr1). The purpose of this study was to facilitate in vivo analysis of pre-mRNA splicing intermediates in metazoans by increasing the half-life of lariat products. Double-stranded RNA interference was used to knockdown expression of dDBR1 in Drosophila. This was done first by introducing in vitro-synthesized double-stranded dDBR1 RNA into Drosophila SL2 cells. Specific reduction of DBR1 mRNA levels was confirmed by RT-PCR. Lariat stabilization was assessed using an RT-PCR assay to detect the lariats for alternative splicing of the second intron from an Ultrabithorax transgene. Germline transformation constructs are being developed for controlled stage- and tissue-specific knockdown of DBR in whole Drosophila. These constructs contain inverted repeats of the dDBR1 sequence separated by an intron and placed under the control of either a heat shock- or GAL4-regulated promoter. The intron stabilizes the inverted repeat in the plasmid, but is removed from the transcript by splicing. In the future, knockdown of DBR1 in mouse and human cells could be done by using short interfering RNAs.

Joana Ricou

Joana Ricou, Carnegie Mellon University (Advisor: Dr. Charles Ettensohn)

Guidance of Primary Mesenchyme Cell Migration in Sea Urchin Gastrulation by b-Catenin Signalingand by Interaction with ECM3-Containing Fibers

Primary mesenchyme cell (PMC) migration in sea urchin development constitutes a model system for cell-extracellular matrix (ECM) interactions. β-catenin is known to be a part of the adherens junction cadherin complex and to function as a transcription activator in the wingless (Wnt) signaling pathway. In the default Wnt pathway, β-catenin is continuously degraded by proteasomes following phosphorylation by glycogen synthase kinase-3 (GSK-3), but the presence of a signal can allow β-catenin to accumulate until it may transported into the nucleus to function as a transcription factor. Namely, it has been shown to be responsible for axial patterning in several deuterostome organisms (including specifying vegetal cell fates in sea urchins) and to influence PMC migration. The nuclear concentration of this molecule is higher in the vegetal pole where the PMCs converge, but decreases towards the animal pole. LiCl, a known vegetalizing agent, can disrupt the natural distribution of β-catenin by inactivating GSK-3 and allowing β-catenin to enter the nucleus independently of signaling. In the presence of LiCl, β-catenin can be found in a higher nuclear concentration throughout the micromeres and vegetal cells, and an increase of endodermal and mesodermal tissues can be observed. The position of the skeletal ring is known to shift towards the animal pole with increasing LiCl concentration. To study the period of time in which β-catenin signaling influences PMC migration, we incubated embryos in LiCl for periods of varying lengths until the establishment of the skeletal ring. The position of the ring was visualized both by light microscopy and by immunostaining with antibodies against PMCs (6a10) and against ECM3 (2.5c4), which allowed for a comparison of their distribution. We determined that the PMC ring had shifted towards the animal pole in those cases where the embryos were incubated in LiCl for a period of at least 4 hours that encompassed the interval of 8-11 hours after fertilization. This evidence suggests that in this period of time, β-catenin signaling is active and relevant for guiding PMC migration.

Stalia Soto

Stalia Soto, Universidad Metropolitana (Advisor: Dr. Robert Murphy)

Location Proteomics Methods for Polarized Cell Monolayers

Determining the subcellular location pattern of proteins is critical to understanding their function, and an automated recognition and grouping by pattern for all protein is needed as part of the characterization process for all proteins in each cell type. Fluorescence microscopy is the method used for the determination of the protein. An automated system that can recognize all major subcellular structure in 2D fluorescence microscopy images has been previously described using numerical features that can describe the location of a protein. This approach has been extended to 3D images, but only for single, unpolarized cells (cells that do not have distinct top and bottom membrane domains). In this project we will extend the 3D method to polarized cells to see if we can distinguish proteins that localize differently to the apical and basolateral surfaces. To do this we will also need to extend our methods to work on cell monolayers (in which cells are touching each other), whereas our previous work has been done on single cells.

Chris Wong

Chris Wong, Carnegie Mellon University (Advisor: Dr. William Brown)

Using Polarity-Sensitive Fluorescent Dyes as Signal Transducers in Biosensors

Biosensors are detectors of binding interactions. We are interested in finding various polarity sensitive fluorescent dyes that could be used as signal transducers in biosensors to detect antibody-antigen interactions. There were two aspects to this project. First, through UV-Vis and fluorescence spectroscopy, we determined the fluorescence properties of various dyes in solvents of different polarities. These solvents mimicked the various hydrophobic and hydrophilic environments that the transducer could be exposed to in actual binding events. Dyes possessing markedly different fluorescence properties in polar and nonpolar environments were then selected as transducer candidates. Generally, the fluorescence intensity of these dyes significantly increased in nonpolar solvents. Next, using fluorescence spectroscopy we observed the changes in fluorescence properties of the dye when bound to a model protein. A model system of RNase S was employed. RNase S is composed of two distinct parts: a 104 amino acid S-protein and a 20 amino acid S-peptide. S-protein and S-peptide tightly bind each other (Kd = 10-10 M) to form RNase S. This interaction is similar to that of an antibody-antigen binding event. When S-protein is added to the dye solutions, fluorescence intensity increased. This suggested that the dye molecules had bound to hydrophobic binding sites of the protein. Currently, we are trying to attach transducers (fluorescent dyes) to the S-protein using a molecular linker. The transducer will remain in the hydrophobic binding pocket of S-protein. The addition of S-peptide should displace the transducer out of the pocket. Although still linked to the S-protein, the transducer should be in the polar aqueous environment; as a result, its fluorescence properties should change. Eventually, we plan to attach transducers to antibodies to monitor their binding interactions with specific antigens that have been known to cause occupational asthma.