2009 Summer Scholar Participants-HHMI Undergraduate Program - Carnegie Mellon University

2009 HHMI Summer Scholar Participants

Image of Katherine Bonnington

Katherine Bonnington, Carnegie Mellon University
Mentor: Dr. Gordon Rule

Expression and Purification of Activation-Induced Cytidine Deaminase through Fusion Proteins in E. Coli

Activation-induced cytidine deaminase (AID) assists in the production of more diverse and effective antibodies by directly mutating the DNA of B-cells. The AID enzyme accomplishes this antibody diversification by acting on the variable region of an immunoglobulin through somatic hypermutation (SHM), a process which improves an antibody’s affinity for an antigen by introducing random mutations. Additionally, AID allows for B-cell class switch recombination, a change of the antibody’s constant region which alters the body’s response upon encounter with the antibody’s specified antigen. However, the mechanisms, regulation, and extent of AID’s activity in the cell are not well known. To gain more insight about AID’s structure and mechanism, the ability to produce large quantities of the enzyme to perform biochemical and biophysical assays in vitro would be extremely valuable. In order to accomplish this, methods of increasing the solubility of AID in vitro must first be explored through its expression in E. coli. By coupling AID with known solubility enhancers, such as the stress-responsive enzyme SlyD or glutathione S-transferase (GST), aggregates of AID are less likely to form. The creation of these fusion proteins allow for SlyD and GST to help refold AID before protease sites release the solubilized native protein from its coupled state. These methods allow for future biophysical and structural studies of the enzyme, shedding light on questions related to the protein’s interactions, structure, and mechanism of deamination.

Image of Molly Evans

Molly Evans, Carnegie Mellon University
Mentor: Dr. Subha R. Das

Chemo-genetic Analysis of the Hepatitis Delta Virus Ribozyme

The hepatitis delta virus (HDV) uses a ribozyme in its replication cycle. The ribozyme cleavage mechanism involves an active site cytosine (C76) that acts as a general acid. The C76 residue is part of a possible network of hydrogen bonding interactions that may alter the intrinsic pKa of C76 and allow it to act as a catalyst. The purpose of my research is to determine the influence of the putative active site network that promotes the catalytic capability of the C76 residue. Here we replace two non-bridging oxygen atoms in an internucleotide phosphate within the network with sulfur atoms.  Replacement of the oxygen atoms results in both the RP and SP phosphorothioate diastereomers in a synthesized RNA, and these are separated by ion-exchange HPLC. These RNAs are each ligated to another transcribed RNA using T4 RNA ligase to provide the full-length single atom mutant ribozymes. The cleavage rate of each single atom mutant ribozyme will give insight into the influence of the pro-RP and pro-SP oxygen atoms in the active-site network on the catalytic cytosine residue.

Image of Cameron Exner

Cameron Exner, Carnegie Mellon University
Mentor: Dr. Veronica Hinman

Elucidation of Evolutionary Changes in Transcriptional Regulation of the SM50 Gene Allowing Larval Skeletogenesis in Sea Urchins

Transcriptional activation of the SM50 gene and resulting skeletogenesis in the larval stage of development is a feature of sea urchins not shared by other echinoderms, though SM50 and its transcription factors are conserved and direct adult skeletogenesis across the phylum. The purpose of this summer’s research was to determine how the Sm50 pathway gets activated in larval sea urchins (represented by Strongylocentrotus purpuratus), but not in the larvae of ancestral echinoderms (represented by sea stars, specifically Asterina miniata). The three possibilities are that: 1) the cis regulatory module (CRM) of SM50 is mutated in sea urchins compared to the ancestral state, 2) at least one SM50 transcription factor is expressed during the larval stage of sea urchin development, but not in the ancestral larval stage, or 3) all factors are expressed in both sea urchins and the ancestor, but at least one has mutated such that it can activate Sm50 in sea urchin larvae whereas the ancestral form cannot. To determine the mechanism of larval SM50 activation, five candidate genes (transcription factors of SM50: Hnf6, Dri, Alx1, Ets1, and Erg) were amplified by PCR, digested with restriction enzymes, and ligated into the expression vector pCS2+ or the vector pGEM. The resulting plasmids were be transformed into E. coli DH5α cells for amplification. The mMESSAGE mMACHINE Kit was used to obtain mRNA from the inserts that were successfully transformed; this was subsequently be injected into sea urchin (positive control) and sea star embryos with another construct, this one containing the CRM of SM50 and a reporter gene, green fluorescent protein (GFP). Injected embryos were observed during larval development to determine whether transcription of GFP or unusual skeletogenesis occurred. Due to inaccurate sequence data and resulting inefficient amplification, only Hnf6 was able to be injected by the end of the summer.  Dri and Erg are currently in pCS2+ but have not been transcribed or injected, and Alx1 is currently in pGEM.  Ets1 was never successfully transformed.  SM50:GFP is expressed as expected in sea urchin positive control embryos, alone and with Hnf6 mRNA; sea star injections are yet to be performed with conclusive results, but observations suggest that SM50:GFP expression does not occur with or without coinjecting Hnf6 mRNA.

Image of Priyamvada Gupta

Priyamvada Gupta, Carnegie Mellon University
Mentor: Dr. Chien Ho

A Biophysical-Biochemical Comparison of Hemoglobin from Mammoth, Asian Elephant and Human

This study is aimed at understanding the molecular basis of environmental adaptation of hemoglobin from tropical climate to low temperature in the Arctic region. For this, we have carried out a biochemical and biophysical characterization of the structural and functional properties of hemoglobins from woolly mammoth and Asian elephant and compared those to human hemoglobin, Hb A. Hb A consists of two alpha (α) subunits and two beta (β) subunits. Asian elephant hemoglobin (Hb E) was found to contain α subunits and β/δ fusion subunits. Authentic woolly mammoth hemoglobin (Hb M) was synthesized by inserting Asian elephant α-like and β/δ- like cDNA into our Hb expression vector, expressing the hemoglobin in E. coli, and introducing the mammoth-specific residue differences into the Asian elephant plasmid. As Hb E and Hb M contain β/δ fusion chains, we also compared them to human hemoglobin A2 (Hb A2) which contains delta (δ) chains instead of the β subunits present in Hb A. Oxygen affinity, Bohr effect, and cooperativity of the oxygenation process was measured at different temperatures over a range of pH for each hemoglobin to compare their functions and 1H-NMR spectra were measured for structural comparisons. The study indicates that Hb E has the highest O2 affinity as compared to that of Hb M, Hb A2, and Hb A. Hb A2 and Hb A have very similar affinities. It was also found that the effect of an allosteric effector, inositol hexaphosphate (IHP), is the most prominent on Hb A2 as compared to Hb A, Hb E, and Hb M. NMR analysis indicates that the α1δ1 and α1δ2 interfaces are perturbed in both Hb E and Hb M, whereas only the α1δ1 interface is perturbed in Hb A2 compared to Hb A. Hb E and Hb M have structures that are very different from that of Hb A2 and HbA, consistent with the altered functional properties.

Image of Andrew Hersan

Andre Hersan, Carnegie Mellon University
Mentor: Dr. Mark Macbeth

Structural Analysis of the Human Adenosine Deaminase that Acts on RNA via X-Ray Crystallography

The Macbeth Lab is interested in the structure and function of adenosine deaminases that act on RNA (ADARs). This family of enzymes catalyzes the deamination of specific adenosine residues to inosine in double stranded RNA. Unfortunately, these editing reactions are not very well understood as the selectivity mechanisms are still unknown. Preliminary data suggests that the selectivity may be related to a combination of sequence and structure of the RNA substrate as well as the structure of the enzyme itself. Functional ADARs have however been proven to be important for normal nervous system development across many species. In particular, RNA editing by ADARs is required for cognitive function in mammals. As a result, the overall objective is to determine the structure of the catalytic domain of the human ADAR1 in an attempt to correlate structure with function. Knowing the three dimensional structure of this protein could prove useful in providing a model to explain how inositol hexakisphosphate (IP6) helps regulate the selectivity of this reaction as well as the biological role of RNA editing in general. Consequently, there are three major goals to this project. The first of these goals was to clone the catalytic domain of the human ADAR1 into a yeast expression system. Once the yeast was successfully expressing the ADAR1 gene, the next objective was to extract and purify the protein of interest. Currently, and lastly, attempts of crystallization are being made to undergo x-ray crystallography.

Image of Mekail Khan

Mekail Khan, Carnegie Mellon University
Mentor: Dr. Gordon Rule

Expression and Purification of Activation-Induced Cytidine Deaminase in E. coli Using Fusion Proteins

The human body has the ability to generate billions of antibodies whose job is to bind a highly specific antigen. This antibody diversification allows the immune system to eliminate harmful microbes and occurs when activation-induced cytidine deaminase (AID) randomly converts cytosines in single-stranded DNA into uracils. AID triggers somatic hypermutation or gene conversion in a naïve B-cell’s variable region and class switch recombination in its switch region by deaminating the B-cell’s DNA during T cell-stimulated immune responses; these events directly cause the antibody variation and increase its antigen binding properties. Both the structure of AID and its mechanism for deaminating DNA are unknown. Since enzyme structure correlates with enzyme function, solving the structure of AID would shed light on its deamination mechanism. In order to map its structure, AID must first be purified in its native form. My goal was to purify AID by overexpressing it in E. coli, where it is naturally insoluble, by separately fusing it with SlyD and glutathione-S-transferase (GST): two proteins that have been shown to increase protein solubility in E. coli. After successful expression, the fusion protein formed an inclusion body, which had to be solubilized before additional purification could take place. After solubilizing the inclusion body, the fusion protein was purified using column chromatography with a cobalt resin that has a high affinity for the fusion protein’s His tag. Dialysis was performed in order to remove the denaturant required to solubilize the inclusion body. My future plans are to cleave AID from the fusion protein by using a TEV protease. Once AID is purified in its native form, structural studies, specifically NMR analysis, and biochemical assays will be performed to provide insight into this unexplored enzyme.

Image of Sefa Kploanyi

Sefa Kploanyi, Carnegie Mellon University
Mentor: Dr. Mark Macbeth

Structure Function Analysis of Adenosine Deaminases that Act on RNA: Crystallization of  ADAR-RD bound to a synthesized RNA Substrate

In eukaryotes, RNA editing after transcription is essential to creating a matured mRNA molecule and generating RNA and protein diversity. Specifically RNA editing involves the  modification of one or so nucleotides in order to change the information content. One of the enzymes that assist in this molecular process is the Adenosine Deaminase that acts on RNA (ADARs).  ADARs catalyze the deamination of specific adenosine residues to inosine, which mimics guanine, in double stranded RNA.  The full-length human ADAR2 protein (hADAR2) contains two double stranded RNA binding motifs (dsRBMs). Truncated versions of the hADAR2 can also catalyze the deamination reaction with near equal efficiency as the full length ADAR2. The hADAR2- R2D protein is a truncated version of the full length consisting of only the second RNA binding motif and the catalytic domain.  I will attempt to crystallize and determine the structure of the R2D bound to a synthesized RNA substrate.  In order to achieve this goal, I will overexpress and purify the R2D protein and synthesize an RNA substrate. The RNA substrate contains an analog state of an edited adenosine site, 8-azenebularine, which will increase the proteins affinity for the RNA substrate. To aid in crystallization, I will fuse a U1 RNA module to the end of my RNA construct, which readily crystallizes in the presence of the U1A binding protein. This interaction has been shown to facilitate the crystallization of several large catalytic RNAs and RNA-protein complexes. The long-term goal of this project is to determine how ADARs target specific adenosines for deamination.

Image of Kellie Kravarik

Kellie Kravarik, Carnegie Mellon University
Mentor: Dr. Brooke McCartney

Adenomatous Polyposis Coli Disruption in the Drosophila Pupal Wing Leads to Wnt Dependent Changes in Cell Proliferation, Tissue Integrity and Cell Death

Adenomatous Polyposis Coli (APC) is a negative regulator of the Wnt signal transduction pathway and a tumor suppressor whose mutation is implicated in >80% of human colon cancers. To better understand the effect of APC inactivation in tissue morphogenesis and colon cancer development, we are examining patches of APC mutant tissue (APC null clones), in the developing Drosophila wing. Previous work observed a higher clone frequency in the larval wing precursor compared to the adult wing. To ask if programmed cell death is responsible for this discrepancy, we assayed the mutant tissue for apoptosis during pupal development by terminal dUTP nick end labeling (TUNEL). Mid-pupal clones exhibit extensive tissue proliferation between the wing blades, with a sub-set of these cells apoptosing. By late pupal development, preliminary data suggests that clones appear smaller in size and that apoptosis has ceased. These observations are consistent with the hypothesis that apoptosis is responsible for eliminating some internalized APC null tissue, and that the process may be temporally regulated.

Because APC also regulates cytoskeletal organization, we also asked if ectopic Wnt signaling is responsible for the APC mutant phenotype. We examined wild-type pupal wings expressing (1) stabilized Armadillo/ ß –catenin (Arm), a key effector of Wnt signaling and (2) APC null tissue co-expressing a Dominant Negative form of TCF (DN-TCF), the Wnt pathway transcription factor. Preliminary analysis suggests that activation of the Wnt pathway is both necessary and sufficient for the APC null phenotype. These analyses characterize the effects of APC disruption and ectopic Wnt signaling in a epithelial systems, and will ultimately lead to a better understanding of how APC mutation can human initiate cancer development.

Image of Jane Lew

Joohyeon (Jane) Lew, Carnegie Mellon University
Mentor: Dr. Mark Macbeth

DNA Deamination Catalyzed by Adenosine Deaminase That Acts on RNA; and Cloning Overexpression and Purification of a tRNA Editing Enzyme

The mutation resulting from the deamination of adenosine into inosine by adenosine deaminase that acts on RNA (ADAR) is crucial for proper neuronal functions. Knowing that ADAR deaminates RNA, we now want to know if it can catalyze deamination of DNA in a manner similarly to that of the ADAR homologs cytidine deaminases. Cytidine deaminases have similar catalytic domain and proposed mechanism as ADARs. Activation-Induced (Cytidine) Deaminase (AID), a cytidine deaminase, is known to cause mutations in DNA that render yeast resistant to an antibiotic, canavanine. For comparison, we will express an ADAR in yeast and screen for canavanine resistant colonies. To confirm that ADARs are the agent responsible for generating mutations, we will also express a catalytic mutant of ADAR which cannot deaminate a nucleic acid. In addition, Drosophila adenosine deaminease that acts on tRNA (dADAT) will be examined by x-ray crystallization after overexpression and purification of the protein. dADAT allows deamination of adenosine to inosine in tRNA, which enables the tRNA to recognize three different arginine codons in mRNA since the inosine is interpreted as being guanosine. The genes that encode for the dADAT are essential for the proliferation of bacteria and yeast which signifies the magnitude of their biological activity.

Image of Charles Miller

Charles Miller, Carnegie Mellon University
Mentor: Dr. Bruce Armitage

Synthesis of Peptide Nucleic Acids with the Goal of Targeting to Specific DNA and RNA Targets

Peptide Nucleic Acids (PNAs) are DNA analogs that consist of nucleobases on a peptide backbone, instead of a sugar/phosphate backbone (as is found in DNA and RNA). Because of this, the PNAs have unique qualities that make them potential candidates for changing gene expression or mRNA transcription. Two of the most likely targets right now are the short DNA sequences hTelo22 and Myc19, promoters of genes shown to be overexpressed in cancer cells. These sequences form structures known as “G-Quadruplexes.” These structures are formed by guanine-rich sections of DNA; the guanines can bond between the two helices two each other through a process known as Hoogsteen base pairing. This creates an exceptionally stable structure in the form of two interlinked helices, or a quadruplex (because they form due to guanines, they are referred to as G-Quadruplexes). PNAs will bind to these structures preferentially, and favorably, to disrupt their structure and prevent them from promoting their downstream oncogenes; to this end, a PNA known as Peg2 (two groups of three guanine bases, separated by two flexible spacers composed of polyethylene glycol) was synthesized with the idea that it would bind tightly to these structures, as the two guanine groups are well-spaced to bind to the guanine-rich sections of the DNA sections without causing the backbone of the DNA or PNA to bend in sterically unfavorable ways. The goal of this project was to better quantify the binding of Peg2 to these sequences, and to develop a method of causing Peg2 to enter cells (with the goal of creating an effective way to downregulate the expression of the genes these promoters control). To this end, the main experimental methods used were melting curve analysis and fluorescence studies (using fluorescently labeled DNA and PNA). These experiments shows that Peg2 does indeed bind tightly to these structures, and that it is more favorable for the DNA and PNA to form a heteroquadruplex than for the DNA to simply quadruplex with itself. Additional modifications to the PNA (such as additional spacers with fluorescent dyes attached, or the addition of Arginines, which has been shown to increase cell uptake of the PNA) continue to be tested.

Image of Amy Wang

Amy Wang, Carnegie Mellon University
Mentor: Dr. Mark Macbeth

Structural Characterization of ADAR substrates: the R/G and Q/R Editing Sites of the Glutamate Receptor

Adenosine Deaminase that acts on RNA (ADAR) is an RNA-editing enzyme that catalyzes the deamination of adenosines to inosines in double stranded RNA substrates. Thus, they are capable of generating point mutations in the protein for which the RNA encodes. Editing of RNA by ADARs is relatively rare and highly specific, only recognizing and deaminating certain adenosines out of many present in the pre-mRNA.  The mechanism by which ADARs target certain adenosines for deamination is unknown; however, the lack of common sequence motifs among the known ADAR substrates suggests that the mechanism is structure dependent rather than sequence dependent. My goal is to determine the structure of two known ADAR substrates, the Q/R and the R/G editing sites of the glutamate receptor pre-mRNA using x-ray crystallography. Similarities between the two structures may suggest a mechanism by which ADARs recognize their substrates. I have constructed the plasmid containing the gene that expresses the Q/R and R/G RNA and purified the U1A protein by nickel column and gel filtration. The U1A protein binds to the U1 RNA loop that is fused to the Q/R and R/G RNA to facilitate crystallization. I am currently running large-scale transcriptions and purifying the Q/R and R/G RNA in order to obtain enough RNA for crystallization.