Research Plan.docx

Cao 1 Ryan Cao Science Fair Paperwork Intel ISEF July 23, 2015 Research Plan Analytical ultracentrifugation and various analyses were used in this project to accurately determine whether the salt concentration of a solution affects the partial specific volume and/or frictional ratio of the protein solute. Large, complex biological macromolecules have many amino acid side chains and residuals on the outer, interacting portion of the molecule. These side chains often are locally charged, and thus attract polar water molecules in solution, creating an envelope of water that changes the apparent shape, volume, and mass (and therefore density) of the protein as it is recorded sedimenting. Salt ions added into solution will be more strongly attracted to these local charges and cancel them, causing the envelope of water to disperse and revealing the true form of the protein in solution. Analyzing this type of species using an analytical ultracentrifuge, highly accurate results for sedimentation and diffusion coefficients can be determined and thus highly accurate frictional ratio and partial specific volumes can be deduced using known relationships between the various measurable properties. In terms of applicability, this experiment actually serves a triple purpose. Firstly, it provides accurate experimental data for validation against current computer-simulated bead modeling values. Secondly, it tests and proves the technique of using analytical ultracentrifugation to determine important hydrodynamic properties that cannot be directly measured such as frictional coefficient and partial specific volume, and lastly, it proves

Cao 2 that for multiple well studied proteins, the addition of salt into solution does make a significant, consistent impact on the partial specific volumes and frictional coefficients of all the proteins. The tested hypothesis for this project was that if proteins hemoglobin, cytochrome c, carbonic anhydrase, lysozyme, chymotrypsinogen a, and ribonuclease and placed into TRIS and sodium phosphate solutions with various salt concentrations ranging from 5mM to 155mM, then the partial specific volume will decrease with increasing salt concentration and the frictional ratio will increase, because as the water envelope, initially forming a smooth, low-density cover around the protein (artificially increasing partial specific volume by making the protein overall much larger and less dense and smoother, artificially decreasing frictional ratio) is removed, the true form of the protein, which will be less smooth and more dense than its water-coated fourme, will give back readings showing a higher density (lower partial specific volume) and more resistance to sedimentation (higher frictional ratio). (Mrs. Dickerson-research questions are implied in the first two paragraphs… I feel it is redundant to re-state that we’re trying to find the f/f0 and v-bar of proteins. Also, I don’t believe that the engineering goal(s) portion is applicable for my research. If it is, please tell me and I’ll try and fix this. Thanks! Additionally, I believe that the expected outcomes are addressed sufficiently in the hypothesis, along with the reasoning. If the ISEF people really really like redundancy though, I guess I can just say it again…) Procedures45 mg of dry, powdered carbonic anhydrase was measured out and mixed with 0.9mL of water to make 900µL of a 50mg/mL stock solution.

Cao 3 200 mg of dry, powdered hemoglobin was measured out and mixed with 4mL of water to make 4mL of a 50mg/mL stock solution. 200 mg of dry, powdered cytochrome c was measured out and mixed with 4mL of water to make 4mL of a 50mg/mL stock solution. 200 mg of dry, powdered lysozyme was measured out and mixed with 4mL of water to make 4mL of a 50mg/mL stock solution. 200 mg of dry, powdered chymotrypsinogen A was measured out and mixed with 4mL of water to make 4mL of a 50mg/mL stock solution. 60 mg of dry, powdered ribonuclease A was measured out and mixed with 1.2mL of water to make 1.2mL of a 50mg/mL stock solution. 12.5 mg of dry, powdered carboxypeptidase was measured out and mixed with 250µL of water to make 250µL of a 50mg/mL stock solution. Twenty approx. 30cm long sets of porous dialysis tubing were cut, rinsed with ddH2O, boiled in a solution of 50% ethanol for 10 minutes, rinsed again thoroughly with ddH2O to prepare for dialysis, boiled again in ddH2O for another 10 minutes, rinsed one last time with ddH2O, and stored in a 50% ethanol solution in the fridge (insert temp. here) The 1M stock TRIS buffer was made by mixing 60.55g of TRIS with 400mL ddH2O and 100mL of 0.5M EDTA solution, then adding HCl to lower the pH to 8.0. The 1M sodium phosphate buffer was made by mixing 305mL of 1M di-basic solution with 195 mL of 1M mono-basic solution, then adding NaOH to raise the pH to 7.0. Afterwards, the stock buffer solutions were autoclaved.

Cao 4 5L of 20mM TRIS buffer solution (i.e. the one that the proteins would actually be in) was made by mixing 100mL of the 1M TRIS solution with 4.9L of ddH2O, while 5L of 5mM NaPO4 solution was made by mixing 25mL of the 1M NaPO4 solution with 4.975L of ddH2O. The total volume of each 50mg/mL protein solution was halved, and that amount was placed into a dialysis tube. The dialysis tubes were dropped into 4- and 6-liter Erlenmeyer flasks containing 4 liters of phosphate and TRIS buffer, respectively, and left to dialyze for 16 hours. The buffer was then changed (i.e. another 4 liters of 20mM TRIS and 5mM EDTA solution was made, then replaced the old buffer) and the proteins dialyzed for another 22 hours. The proteins were then extracted from the dialysis tubing and placed into 15mL Falcon test tubes. 300µL of 1mg/mL protein solution was made from each dialyzed 50mg/1mL solution. These samples were run through the spectrophotometer and repeatedly diluted with ddH2O in order to determine spectra curves for a multitude of concentrations. These spectra curves were then transferred into the Ultrascan II global extinction fit program and fit with 25 Gaussians to determine a molar extinction curve, and fitted to the given extinction coefficient produced for that specific protein sequence by Ultrascan. After using the molar extinction curve to determine the factor of dilution needed to make a 0.9OD solution, 800µL of 0.9OD samples in all salt concentrations were made, and subsequent 0.3OD dilutions were made as well.

Cao 5 The AUC cells were cleaned, built, and loaded once to twice a day. (PPE was worn and sharps were disposed of in the designated container) After each run, samples had to be extracted and cell windows and centerpieces had to be cleaned by the sonicator. After each run, the data was transferred to a local database (run by UTHSCSA) and copied over to analytical computers with Ultrascan installed. Solution objects were created in Ultrascan by mixing analytes (proteins as defined by their sequence from the Protein Data Bank) and buffers (TRIS and sodium phosphate). Each run (datum) was associated with a solution, rotor, centerpiece, and instrument. The data (runs) were then edited, specifying a meniscus position, data range, and plateau range, removing spikes, and deleting baseline/light-scattering scans as well as end scans all showing complete sedimentation. The data was then analyzed. The analysis included the followingA first 2DSA (2 Dimensional Spectrum Analysis) fit was performed to remove time-invariant noise from the experimental data. A meniscus fit was then run to determine the true meniscus position, as well as remove time- and radially-invariant noise. An iterative 2DSA was performed to optimize fitting for the experimental data curve. A van Holde-Weischet analysis was performed to estimate the sedimentation and diffusion coefficient ranges in preparation for the Monte-Carlo analysis.

Cao 6 A 100 iteration Monte-Carlo analysis was performed using a reduced 2DSA grid with 100 different random noise plots being generated on top of the iterative fit, taking the fits from the Monte-Carlo and averaging them to produce a Monte-Carlo fit for the initial experimental data. Finally, if the Monte-Carlo fits were close (root mean square deviation of 5x10^-2 or less) to the experimental data, a PCSA (parametrically constrained spectrum analysis) was performed using the sedimentation and diffusion coefficients measured from the Monte Carlo analysis as well as a molecular weight derived from mass spectrophotometry. The resulting PCSA values for frictional coefficient and partial specific volume were then recorded and graphed, revealing an extremely consistent trend among three of the proteins-cytochrome c, carbonic anhydrase, and lysozyme. Potential risks involved the centrifuge rotor being unbalanced and thus exploding, as well as the very real danger of loading needles causing unwanted harm. The plots of the v-bar and f/f0s show a consistent decrease in the v-bar and a consistent increase in the frictional ratio, the exact opposite of what had been hypothesized. Conclusions based on biophysics cannot be drawn yet about why the salt may have affected the properties of the proteins in such a way, although the trend is extremely consistent among the measurable single-species solutes.

Cao 7 Bibliography: Brooks, Emre, and Borries Demeler. "Parallel Computational Techniques for the Analysis of Sedimentation Velocity Experiments in UltraScan." Colloid Polym Sci (2008): n. pag. Http://www.demeler.uthscsa.edu/ultrascan-publications. Borries Demeler, 11 July 2007. Web. 15 July 2015. Brooks, Emre, and Borries Demeler. "Parsimonious Regularization Using Genetic Algorithms Applied to the Analysis of Analytical Ultracentrifugation Experiments." Http://www.demeler.uthscsa.edu/ultrascan-publications. Borries Demeler, 7 July 2007. Web. 8 July 2015. Brooks, Emre, Borries Demeler, Suresh Marru, Marlon Pierce, Mattia Rocco, and Raminderjeet Singh. "UltraScan Solution Modeler: Integrated Hydrodynamic Parameter and Small Angle Scattering Computation and Fitting Tools." Http://www.demeler.uthscsa.edu/ultrascan-publications. Borries Demeler, 18 June 2012. Web. 1 July 2015. Brooks, Emre, Weiming Cao, and Borries Demeler. "A Two-dimensional Spectrum Analysis for Sedimentation Velocity Experiments of Mixtures with Heterogeneity in Molecular Weight and Shape." Http://www.demeler.uthscsa.edu/ultrascan-publications. Borries Demeler, 29 Jan. 2009. Web. 19 June 2015. Demeler, Borries, and Kensal E. Van Holde. "Sedimentation Velocity Analysis of Highly Heterogeneous Systems." Http://www.demeler.uthscsa.edu/ultrascan-publications. Borries Demeler, 28 Oct. 2004. Web. 24 June 2015.