Skip header content and main navigation Binghamton University, State University of New York - Undergraduate Research Old
 
Cohort One

Cohort One Projects

Departments of Faculty mentors

Chart

Majors of 27 undergraduate participants

1/3   Biology
1/3   Interdisciplinary majors
Biochemistry, Neuroscience
Bioengineering, Environmental Studies
1/3   Mathematics, Engineering
Computer Science, Physical Sciences

 Project Titles

  1. Computer Recognition: Apoptotic vs. Normal cells | Read more
  2. Image Processing and Estimation Theory of Multiple Bacterial Species in Biofilm | Read more
  3. Image Stitching-Devonian Forest Fossil Site | Read more
  4. Meta Analysis of DNA Microarray Studies on p53 | Read more
  5. Improvements to a Computer Program for Peptide Identification |  Read more
  6. Statistical Analysis of Genes Affecting Female Remating in Fruit Flies | Read more
  7. Application of Statistical Machine Learning Paradigms to the Prediction of Cancer and
    Efficacy of Treatments | Read more
  8. Microorganisms Communities in Fluid Inclusions in Halite and Gypsum | Read more
  9. Minimization of Nutrients to the Chesapeake Bay: A Wetlands and Watershed Study | Read more
  10. Photophysical Characterization of the Fluorescent Conjugated Polymer Chemosensor tmeda-PPETE in Solution and Solid State | Read more
  11. Nociception and Learning Modulation in an Adolescent Animal Model of Alcohol Abuse: Effects of NMDA Antagonist MK801 and EAAT2 Inhibitor (2R,3R)-2-amino-3-benzamido-4-phenylbutanoic acid | Read More
  12. Vibrational Communication in Insects | Read more
  13. Single Stranded DNA Binding to Single-Walled Carbon Nanotubes | Read more
  14. p53 Binding to DNA in Nanofluidic Devices: Analyzing "The Genome Protector" | Read more

Project Descriptions:

1. Computer Recognition: Apoptotic vs. Normal Cells:

There comes a time in a cell's life when it may have to die. Cells often die in a highly-regulated process known as apoptosis or programmed cell death. Many processes can trigger cells to undergo apoptosis such as the cell death associated with tissue remodeling in development (e.g. the removal of the "webbing" between digits on hands or paws), normal tissue replacement (such as the death of cells in the upper layers of the skin), conditions which induce a stress on the cells (growth factor removal), the direct killing associated with the immune response (the killing of virus infected or cancer cells), or even the killing of cancer cells by chemotherapeutic agents. Scientists need to be able to tell which are normal cells and which are cells going through apoptosis. They do this by looking at the way cell's DNA is clustered. First the cells are stained with a fluorescent DNA-binding dye. The normal cells have nuclei that are evenly stained, while the nuclei of the apoptotic cells show up as smaller and brighter blobs.   The problem is that scientists usually have to figure this out by staring into a microscope and counting the cells and blebs, a task both tedious and prone to error and bias. So researchers from the Biology department and the Computer Science department are collaborating to develop a computer program that will examine the stained images, distinguish normal from apoptotic cells, and count the cells in each group.   This will save time and effort, and it eliminates potential experimenter bias from the equation.

Faculty Mentors: Dr. Lijun Yin (Computer Science) and Dr. Dennis McGee (Biology)
Graduate Mentor: Michael Reale (Computer Science)

Back to top

2. Image Processing and Estimation Theory of Multiple Bacterial Species in Biofilms

Chronic Human diseases are mostly due to biofilm infections and are a direct result of their persistence/recalcitrance. This project has two main objectives: (1) the determination of the chance of the bacterial cells to remain adhered to a surface once they attach to it and (2) the determination of the bacterial cell distribution of dual-species biofilms within a cell cluster. To achieve these objectives single and dual species bacterial biofilms are cultured in both flow cell and chamber slide reactors. When studying the initial bacterial attachment flow cell reactors are inoculated with the bacteria under static conditions and allowed to adhere for a period of 1 hour, after which the flow is started and medium is continuously perfused. Immediately after the flow initiation, the attachment of the bacteria to the glass coverslip of the flow cell reactor is observed using an epifluorescence microscope and images are taken with a digital camera at 1.5 seconds intervals for a period of 1 hour. Image data is then analyzed to locate individual cells and track their position and movement over time. To determine the bacterial cell distribution in cell clusters, biofilms are cultured in chamber slides for a period of 6 days and subsequently stained. The chamber slide is then dismantled and a coverslip is placed on top of the slide containing the biofilms. Bacterial cell distribution is then observed by taking z-stack images of the cell clusters using confocal scanning electron microscopy. This z-stack data is then loaded into a computer program capable of estimating various spatial properties of the biofilm, including biofilm volume and density, diameter and height, and differences between relative distributions of bacterial species, as well as how these properties change over time. The statistics are computed using ANSI standard C, and the resulting code will be packaged and distributed as computer software that biologists can use to scan through thousands of microscope images in a short period of time, computing these statistics in batch; an improvement over the traditional (manual) method of attempting to determine significance of the images and what is seen within them. In addition, these studies will enable the better understanding of the bacterial distribution within biofilms thus, being a means to the improvement of the treatment of diseases.

Faculty Mentors: Dr. Claudia Marques and Dr. Scott Craver
Graduate Mentors: Enping Li (Electrical Engineering) and Diana Amari (Biology)

Back to top

3. Image Stitching-Devonian Forest Fossil Site

In Gilboa, NY, there exists a famous forest fossil site, which has been proved imperative to further study the evolution of earth's plant life. Around the time this site was preserved, trees were in their infancy. These primitive trees could tell biologists many things about how earth's environment has changed. In order to process the site to find these trees, it is important that a picture is created to represent this fossil forest, with close attention to detail. The only way to achieve such detail in a image, is to stitch multiple pictures together. Our goal for this program is to create a high resolution image that respects all aspects of the original forest using an image stitching based algorithm.

In technical terms, image stitching is the process of combining multiple images, each of which contain some form of overlapping region, in order to create one large image that preserves the detail of all of the images. Our current work is based on pixel representations of images. We regard a square that consists of pixels as a unit of similarity. While stitching two pictures, we must check for the existence of a square in our second image that is most similar to a square in our first image. If the most similar square exists, we can conclude that these two pictures can be stitched together; from the offsets we have ascertained that we are able to decide how the stitching process can be implemented. Since pictures are not all taken at the same distance or angle from the viewing focus, rotation and scaling may be applied while processing the stitching between images in order to stay true to the image.

Our future work includes the attempt of using feature-based methods to handle the concept of similarity. Similarity is no longer based on raw pixels of images, but based on the detection of features of objects in the image. This will be used in order to more efficiently find the overlapping area of both images being stitched, as well as the rotation, scaling and translation that need to be applied. This will require data mining and machine learning knowledge to be put into the algorithm design. Once we have completed the general case, many images must be stitched in order to make the complete panoramic effect. We would also like to provide a user-friendly software-like interface for biological uses. Most importantly, our work can be applied to more general purposes about image stitching in computer vision.

Faculty Mentors: Dr. William Stein (Biology) and Dr. Lijun Yin (Computer Science)
Graduate Mentor: Xing Xhang (Computer Science)

Back to top

4. Meta Analysis of DNA Microarray Studies on p53

The gene coding for the tumor suppressor protein p53 is one of the most commonly found mutated genes in cancer. These mutations often result in deformations in the binding site of p53, causing this tumor suppressor to malfunction during its binding process. In our studies, we aim to analyze the mechanism this protein uses to both find these specific sites along DNA strands as well as to test the protein's affinity (i.e. ability to bind). We do this in the hopes that a better understanding of its binding habits can eventually be used in medical applications, especially regarding the case when the binding site of the protein is not in optimal conformation.

The technique we will use here will utilize nanofluidic devices (nano in depth not necessarily in width) to trap and stretch long DNA concatemers (DNA molecules with several copies of the same binding site inserted throughout the entire contour length of the molecule). These sites are placed at specific areas in the DNA sequence thanks to Restriction Enzymes used for the cutting and insertion process used here. These devices will employ an electric field by applying a positive and equal potential to two parallel electrodes. The field lines of these like charged potentials will be the source of what draws the negatively charged molecule down to the center of the device and what deforms the molecule to a more linear state. The p53 protein and the ends of the DNA molecules will be fluorescently tagged so that fluorescence microscopy can be used to analyze the proteins as it moves across the polymer in search of its target sites. Videos of this movement will be taken so that the trajectories of the proteins can be mapped over time. The mean squared displacement of these trajectories over frames can then be used to solve for the diffusion coefficient of the proteins and varying the salt concentration will allow further investigation regarding the kinetics of this protein. For instance, one thought is to plot the diffusion coefficient against the change in salt concentration. This should enlighten us on whether the movement technique used by this protein is primarily one dimensional sliding or a hopping mechanism. Our results will be compared with others such as the study done by Anahita Tafvizi, Fang Huang, Jason Leith, Alan Fersht, and Leonid Mirny in their 2008 paper "Tumor Suppressor p53 slides on DNA with Low Friction and High Stability". By comparing our results with others such as those found in Tafvizi's paper we will be able to both help solidify these findings as well as provide a more stable ground to then go further with our studies into how this protein works on a molecular level.

Faculty Mentors: Dr. Susannah Gal (Biology), Dr. Craig Laramee (Bioengineering) and Dr. David Schaffer (Bioengineering)

Back to top

5. Improvements to a Computer Program for Peptide Identification

Proteomics is a rapidly growing field concerning large-scale study of proteins – their structure, function, and identification. It has become increasingly important to obtain a means of accurate and efficient protein identification. Our goal is to develop a computer program that can accurately identify peptide sequences without reliance on a pre-existing database of known proteins. One typical technique to sequence proteins is through tandem mass spectrometry. Proteins are first split into smaller peptides using an enzyme, such as trypsin, that breaks peptide bonds after specific amino acids. These fragments are run through a mass spectrometer, which calculates the time of flight to separate peptides based on their mass to charge ratio (m/z). These peptides are then bombarded with particles that break them up further into ions. The instrument records the number of particles encountered at each m/z value and creates a spectrum of numerous peaks.

Spectra can be analyzed in several different ways. Database searching correlates data to theoretical decomposition of previously sequenced proteins. We are using a method known as de novo sequencing. The m/z values from the fragmented peptide are equated to masses of amino acids and possible sequences can be proposed. Once several peptides are sequenced, a protein can be identified. Potential sequences are scored on how well the m/z of peaks match to the expected ion sizes. The candidate solutions with the highest scores represent the peptides most likely present.

Experimental spectra, however, are far from ideal. Background noise makes it harder to detect the important peaks. Peaks may be missing, and the m/z differences between adjacent peaks may match one or more amino acid residues. Peaks representing b and y ions do not have distinguishing features. Hundreds of possible amino acid modifications and neutral losses of H2O and NH3, each with a different mass, complicate the problem.

With de novo sequencing, we can study proteins that have not yet been sequenced in the database. Easier protein identification advances research in numerous fields. For example, some cancer cells release specific proteins; identifying and screening for these proteins could aid in earlier cancer detection.

Faculty Mentors: Dr. Patrick Madden (Computer Science) and Dr. Anna Tan-Wilson (Biology)
Graduate Mentors: Jason Gallia (Computer Science) and Yuanyuan Wang (Biology)

Back to top

6. Statistical Analysis of Genes Affecting Female Remating in Fruit Flies

Studies have shown that after insemination by certain males, female Drosophila are less likely to mate with subsequent males. The goal of this project is to identify the genes that are differently expressed between females that are mated to males which are proficient in preventing females from re-mating and males that are not.

The general method that we applied was the DNA microarray experiment. Males that were proficient in preventing females from re-mating were labeled as Class1 while males that were not were labeled as Class2. We then performed eight mating trials with high and low lines of each class of male with genetically identical females. With the help of fluorescent dyes, we got intensities of each feature on the Microarrays after the hybridization of the labeled extracts.

Once the intensities are obtained, we need to go through the data processing procedure to apply statistical methods to our experimental results and get significant genes that are differently expressed in the two classes of samples. The general idea is to conduct Significance Analysis of Microarray on our data under the environment R, which is a language for statistical computing and graphics. In our analysis, we need two packages—affy and siggenes. The former is used to transform the original data to an object which is in a format that can be analyzed with the method of Significance Analysis of Microarray. The latter is used to actually identify the significant genes in our two-class, unpaired sample. In the affy package, we first transform the original data to an AffyBatch Object with the help of library files that map different sets of features to genes. Then we conduct background correction, normalization, perfect match correction and summarization on the AffyBatch Object and further transform it to an ExpressionSet Object, which is in the format that can be analyzed by the siggenes package. In the siggenes package, we first do permutations on the class labels and then calculate characteristic statistics of original sample and mean characteristic statistics of all the permutations. Based on the difference between these two statistics, we can get the identified genes.

Faculty Mentors: Dr. Anthony Fiumera (Biology) and Dr. Xingye Qiao (Mathematics)
Graduate Mentor: Nan Bi (Mathematics)

Back to top

7. Application of Statistical Machine Learning Paradigms to the Prediction of Cancer and Efficacy of Treatments

Lung cancer is the leading cause of cancer related deaths worldwide and contributes to the loss of millions of lives each year. Despite the fact that early-stage cancer patients have a longer overall survival than late stage patients, 35% to 50% of these early stage patients will recur cancer and die within five years of surgery. There are currently no reliable methods for determining if a patient is at high risk of recurring cancer. The successful prediction of high-risk patients would allow for early therapeutic intervention and improved survival among those suffering from lung cancer. Our project focuses on using a newly developed method for predicting survival time for patients with non-small cell lung cancer (NSCLC). We use statistical machine learning paradigms to find the non-linear trends in our data in hopes of finding characteristics that differ between long and short term survival patients. The data we use for our analysis consists of 22,283 probes (also referred to as features) from a U133A Affymetrix micro-array gene chip. Each patient in our data set has gene information represented by each of the micro-array probes. We perform a coarse feature selection consisting of a hard cut off t-test (0.05 significance) and variance pruning methods to determine a subset of probes that show potential significance in relation to cancer progression. This process reduces our probe count from 22,283 to roughly a few hundred. The next step is to use a fine feature selection technique and regression analysis we developed called kernalized partial least squares (K-PLS). This process selects the top 30 features that contribute to differences in survivability among patients by mapping the highly dimensional feature space to a latent space. This allows us to develop a regression equation that captures the non-linearity in our data and to predict survival time for new patients that have just been diagnosed. We evaluate our program's predictive power using receiver operating curves (ROCs) as well as Cox hazard ratios and Kaplan-Meier analysis.

Faculty Mentors: Dr. Walker Land (Bioengineering), Dr. Xingye Qiao (Mathematics) and Dr. David Schaffer (Bioengineering)
Graduate Mentor: William Ford (Bioengineering)

Back to top

8. Microorganisms Communities in Fluid Inclusions in Halite and Gypsum

The goals of this project are to extract, sequence, and identify archaebacteria, algae, and possibly fungal communities within fluid inclusions from salt crystals extracted from basins that have been formed from the deposition of seawater over time. We plan on being able to characterize the communities of organisms that were thriving at the time the crystals formed. If we have luck with these, we would like to move into older samples that span different eras in Earth's history. We also want to try a few samples, which are very close in age, to see the short-term evolution of these communities.

In studying samples from different eras in Earth's history for archaebacteria, we will also determine the chemistry of seawaters during certain time periods. Determining the chemistry will allow us to better document the fluctuation in the composition of seawaters over time, which will allow us to better understand the evolution of calcifying organisms.

Faculty Mentors: Dr. Tim Lowenstein (Geology) and Dr. J. Koji Lum (Anthropology/Biology)
Graduate Mentor: Krithivas Sankaranarayanan (Geology)

Back to top

9. Minimization of Nutrients to the Chesapeake Bay: A Wetlands and Watershed Study

Wetlands and watersheds are greatly affected by anthropogenic activities. Pollution from industry, agriculture, and infrastructure can have harmful effects on the watershed, and wetlands are one of the tools that help keep watersheds clean. By studying pH, conductivity, and nitrogen levels in wetlands and watersheds, and by studying the flow and pollution levels of streams, we can help determine the causes of high nutrient and sediment loads, and therefore help abate them. We are studying different aspects of watersheds (wetlands and streams) and different treatments on each of them (grazing and storm water retention, respectively) to gain a greater understanding of the systems as a whole.

Faculty Mentors: Dr. John Titus (Biology) and Dr. Joseph Graney (Geology)
Graduate Mentors: Rebecca Heintzman (Biology) and Jason Johnson (Geology)

Back to top

10. Photophysical Characterization of the Fluorescent Conjugated Polymer Chemosensor tmeda-PPETE in Solution and Solid State

There is a rising environmental need for the detection of water pollutants released by industrial practices. The pollutants we are specifically trying to identify are metal cations – small, positively charged particles in solution. In order to identify these metals, we are working with polymers that contain a receptor which responds to the metal ions. These polymer sensors are large chemical compounds which contain small repeating units. We are primarily working with tmeda-PPETE, a polymer the Jones Group has synthesized. When metal cations are introduced to the polymer, the polymer will either begin fluorescing (turn on) or stop fluorescing (turn off).   Depending on the concentration of the metal cations, these sensors will fluoresce at different measurable intensities. Furthermore, the fluorescence will change depending on which metal cations are present. In this way, we can create sensors that are both sensitive and selective to specific metal ions. Our ultimate goal is to create solid, portable sensors to be used in the field. Much of our research is comparing the solution phase with the solid phase, and measuring the response of both with different metal ions. We are working to characterize these polymers, and hopefully find an efficient sensor to be tested in the field.

Faculty Mentor: Dr. Wayne Jones
Graduate Mentor: Megan Fegley

Back to top

11. Nociception and Learning Modulation in an Adolescent Animal Model of Alcohol Abuse: Effects of NMDA Antagonist MK801 and EAAT2 Inhibitor (2R,3R)-2-amino-3-benzamido-4-phenylbutanoic acid

Signal transmission by the neurotransmitter glutamate, as well as the neuronal systems that control its activity, are important for normal function in the brain, and for learning and behavior. Modulation of the neuronal processes that regulate glutamate neurotransmission––including altered expression or activity of pre-synaptic glutamate transporter proteins necessary for glutamate re-uptake and post-synaptic binding receptors­­­­­­­––can result in changes of learning and memory ability. In a psychiatric disease model such as alcohol dependence, increases in the expression of the Excitatory Amino Acid Transporter 2 (EAAT2 or GLT1-located pre-synaptically), the N-methyl D-aspartate receptor (NMDA receptor-located post-synaptically), Kainate receptor (KA receptor-located post synaptically), and the alpha amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPA receptor-located post-synaptically), are evident (Gass and Olive 2008). When the expression of the EAAT2 is increased, it is assumed that glutamate is taken up from the synapse and transported back into the pre-synaptic neuron at a faster rate. Chronic alcohol consumption has been shown to trigger a hyperglutamatergic state of brain function in both humans and experimental animal models through activation of neuroadaptive processes that attempt to counter the acute neurochemical actions of alcohol.   Thus, chronic consumption can trigger new expression levels of glutamate transporter and/or glutamate receptor proteins, which can play a key role in altered learning, memory function and/or addiction development. In 2005, research was published on the neuro-protective potential of beta-lactam antibiotics (cephalosporins and penicillins)(Rothstein, Patel et al. 2005). Hypothetically, administration of the beta-lactam antibiotics may alleviate problems in learning and memory abilities in alcohol dependent subjects because these drugs increase the expression of the EAAT2, potentially restoring the normal glutamate homeostasis at the synapse. In addition, exposure to the known EAAT2 inhibitor (2R,3R)-2-amino-3-benzyl-4-oxo-4-(phenylamino)butanoic acid may improve the memory and learning abilities in alcohol-naive subjects by making more glutamate available in the normally functioning synapse. These hypotheses will be tested using a chronic alcohol exposed Wistar rat model in comparison to alcohol-naïve control animals. Different doses of a generic beta-lactam antibiotic, the EAAT2 inhibitor (which will be synthesized), and a saline control will be injected subcutaneously for a pre-determined period of time. During drug administration, the rats will be subjected to nociception tests using a hot plate, locomotor tests using a photobeam chamber, and object recognition tests using an open field setup. Following the behavioral testing the rats will be sacrificed to obtain brains. These brains will be sectioned and Westerns blots will be used to access the relative concentration change of each type of transporter and receptor.

Faculty Mentors: Dr. Christof Grewer (Chemistry) and Dr. Jilla Sabeti (Neuroscience)

Back to top

12. Vibrational Communication in Insects

The project we are working on this summer is called Vibrational Communication in Insects. It is a biology and mechanical engineering interdisciplinary project. The goal is to understand the neurophysiological and mechanical systems that allow small insects to determine the direction of vibrations with which they communicate. The insect of study,Umbonia Crassicornis, commonly called the thorn bug, utilizes a mating ritual involving substrate vibrations along plant stems. A male will sit on a tree branch and vibrate his abdomen, sending a bending wave through the branches of the tree. If a female is interested in the signal, she will 'call' back. The duet will continue until the moving male locates the stationary female. The wavelength produced by these insects is significantly longer than the insect itself and the insect is too small to rely on intensity or reception time differences. The intrigue here is understanding how an insect localizes a signal; meaning, how wave propagation direction is identified. A bending wave simulator, or 'bugstick', is used to mimic a directional bending wave produced byU. Crassicornis. This involves a rigid wooden dowel equipped with two actuators on each end. The actuators are controlled by a computer program that causes them to simulate a bending wave when measured from the center of the dowel. Currently, only one type of wave signal can be sent to the bugstick so the reaction of the bug depends greatly on which direction it is facing. Our summer plans are to modify the bugstick such that different signals can be selected in real time based on the dynamic response of the insect. This can be done by lengthening the bugstick and writing a Matlab program that allows the controller to choose the correct signal based on the insects location along the stick and orientation. The methods by which we are modifying the experimental setup are extensive Comsol work that allows finite element modeling of the system, Matlab applications to generate and select different signals in a graphical user interface, and materials study which will allow the use of a different, less rigid material. This material would propagate the bending wave naturally versus the 'simulation' of a bending wave using controlled displacement algorithms. Additionally, a goal of this research is to help characterize the neural response of the insect to directional bending waves. This will be done using extracellular electrodes to record from the thorn bug's nervous system as it is presented with vibratory stimuli.

Faculty Mentors: Dr. Carol Miles (Biology) and Dr. Quang Su (Mechanical Engineering)br /> Graduate Mentor: Aaron Payne (Biology)

Back to top

13. Single Stranded DNA Binding to Single-Walled Carbon Nanotubes

Carbon nanotubes are made of a network of carbon atoms forming a hollow cylindrical structure with unique intrinsic properties. They can range up to several centimeters in length with a diameter of less than one nanometer. And as one of the most abundant elements on earth, they serve as a promising low-cost material for a wide range of biomedical and engineering applications. What makes nanotubes interesting is the flexibility to modify their surface properties (i.e. adding metal ions and functional groups) to accommodate a specific reaction or environment. Though nanotubes are naturally found as soot and have been in the research spotlight for quite some time, it has been a challenge to increase their compatibility with nature. They are not dissolvable in water, let alone in the human body. It has been shown recently that these materials are capable of passing through the cell membrane, but while toxicity is still a major question, it may also be a problem to destroy or circulate the material out of the system. For our project, we hope to understand the potential medical applications of carbon nanotubes for gene and drug delivery in cells. Since it is possible for them to enter cells, they can perhaps serve as carriers. We are working towards a goal to determine whether pH affects the binding between the nanotubes and DNA. Publications from the past have reported hydrophobic and Van der Waals interactions that drive DNA to wrap around the carbon structure. We hypothesized that pH variation will influence the binding affinity of single-stranded DNA to single-walled nanotubes.

During these past several weeks, we have been researching on the differences between poly-adenine and poly-thymine bases to study the influence of their structure/length on their binding activity in water. In the upcoming month, buffers of various pH will be prepared. We first purified our DNA stock using High Performance Liquid Chromatography to eliminate miscellaneous compounds that could skew results. The concentration was then quantified by Ultraviolet Spectroscopy for the number of moles of DNA in solution. For the nanotubes, a surfactant (NADDBS) was dispersed onto the surface of the nanotubes so that they dissolved in water. This is one of best methods to keep nanotubes suspended in aqueous solutions for long periods of time. It works like a micelle where the NADDBS ligands have both hydrophobic and hydrophilic tails. With increasing amounts, the dissolved nanotubes were combined with the quantified DNA. Circular Dichroism was used to determine the shape of the complex and to see if the addition of CNT's will change the pattern of the absorption peaks; meaning there was a binding between the CNT's and the DNA. Unfortunately, there was no significant change in the documented graphs other than a series of peak amplification, so binding was not confirmed by this technique. Further tests will have to be constructed to confirm that the bands on the CNT's were in fact single-stranded DNA. The next major problem is finding a buffer that works well with the nanotubes and DNA. As of now, the nanotubes do not stay in solution very well in buffers containing certain ions. Once this issue solved we will be able to make significant headway on our project.

Faculty Mentors: Dr. Changhong Ke (Mechanical Engineering) and Dr. Eriks Rozners (Chemistry)
Graduate Mentors: Meng Zhang (Mechanical Engineering) and Thomas Zengeya (Chemistry)

Back to top

14. p53 Binding to DNA in Nanofluidic Devices: Analyzing "The Genome Protector"

The gene coding for the tumor suppressor protein p53 is one of the most commonly found mutated genes in cancer. These mutations often result in deformations in the binding site of p53, causing this tumor suppressor to malfunction during its binding process. In our studies, we aim to analyze the mechanism this protein uses to both find these specific sites along DNA strands as well as to test the protein's affinity (i.e. ability to bind). We do this in the hopes that a better understanding of its binding habits can eventually be used in medical applications, especially regarding the case when the binding site of the protein is not in optimal conformation.

The technique we will use here will utilize nanofluidic devices (nano in depth not necessarily in width) to trap and stretch long DNA concatemers (DNA molecules with several copies of the same binding site inserted throughout the entire contour length of the molecule). These sites are placed at specific areas in the DNA sequence thanks to Restriction Enzymes used for the cutting and insertion process used here. These devices will employ an electric field by applying a positive and equal potential to two parallel electrodes. The field lines of these like charged potentials will be the source of what draws the negatively charged molecule down to the center of the device and what deforms the molecule to a more linear state. The p53 protein and the ends of the DNA molecules will be fluorescently tagged so that fluorescence microscopy can be used to analyze the proteins as it moves across the polymer in search of its target sites. Videos of this movement will be taken so that the trajectories of the proteins can be mapped over time. The mean squared displacement of these trajectories over frames can then be used to solve for the diffusion coefficient of the proteins and varying the salt concentration will allow further investigation regarding the kinetics of this protein. For instance, one thought is to plot the diffusion coefficient against the change in salt concentration. This should enlighten us on whether the movement technique used by this protein is primarily one dimensional sliding or a hopping mechanism. Our results will be compared with others such as the study done by Anahita Tafvizi, Fang Huang, Jason Leith, Alan Fersht, and Leonid Mirny in their 2008 paper "Tumor Suppressor p53 slides on DNA with Low Friction and High Stability". By comparing our results with others such as those found in Tafvizi's paper we will be able to both help solidify these findings as well as provide a more stable ground to then go further with our studies into how this protein works on a molecular level.

Faculty Mentors: Dr. Stephen Levy (Physics) and Dr. Susannah Gal (Biology)

Back to top
Connect with Binghamton:
Twitter icon links to Binghamton University's Twitter page YouTube icon links to Binghamton University's YouTube page Facebook icon links to Binghamton University's Facebook page Instagram

Last Updated: 10/7/13