De Novo Design: Ribose Binding

Adenosine triphosphate, ATP, is essential for life as it is the energy source of a cell, however not much is known about how it binds to protein. ATP is comprised of three chemically distinct parts: adenine, phosphate, and ribose. By engineering three individual proteins to bind to each part of ATP, the fundamental biophysical interactions driving ATP binding can be found. In previous research, a four-helix bundle protein was designed to bind ribose through computationally guided de novo design. Ribose was docked into the protein and was given a binding score of -5.68 kcal/mol. For the summer I intended to express the protein and verify the computational results. The gene was ordered and then expressed for further protein characterization experiments. To understand binding, two main experiments were done, circular dichroism, CD, and glucose-quantification assays. The CD experiments have shown that the mutated protein might have a beta-sheet without ribose present, a sign of misfolding. Furthermore, CD, experiments of the control protein show the predicted alpha helix with and without ribose present. More CD experiments are being performed to show if there is a possible secondary structure conformational change when the mutant protein is in the presence of ribose. Secondly, the current glucose assay was shown to detect ribose by using a UV-spectrum. Further experiments will use this assay to quantify the amount of ribose the mutated protein can bind. Simultaneously in lab work, more computational work was done using the GMU hopper and the program Rosetta. Using Rosetta, and the hopper’s computational power, thousands of proteins with different mutations could be made and scored. Currently, over 25,000 proteins have been made however, none have scored higher than the previous best mutation. Continuing refinement of the Rosetta program will design an equivalent and possibly better protein.

Hi. My name is Annie McAllister and this summer I have been doing de novo design ribose binding. What that specifically means is I have been trying to engineer a protein to bind ribose. A little bit of background. Well what are proteins? Proteins are one of the macromolecules found within your cell. They do a lot of different things and they are very complicated. However, what you really need to know is that structure equals function. The base unit of proteins are called amino acids. Those amino acids are then linked together to form a chain. This is called a primary structure that primary structure starts to interact with itself and folds within itself That’s called the secondary structure. There is two different ways it folds within itself. The alpha helix which is shown on the right on the left and the beta-sheet. Which is shown on the right Now, what is ribose? Well ribose is a subsection of ATP or adenine triphosphate. Ribose is specifically what we would call a sugar Now overall what is ATP? ATP is the energy source in the cell We are trying to build three different proteins to- One to bind phosphate, one to bind ribose, and one to bind adenine in hopes of finding the fundamental binding forces for each of them and then coming together and making one giant protein structure to bind ATP to harness its energy How is this done? Well there is a lot of different methods of protein engineering. The one we decided to go with is the de novo design method. This method is very powerful because the protein we are engineering is 100% computational. Meaning we are doing it on the computer. This protein is not found in the naturally occurring world. Then once you have the protein designed in the computer, you order it you buy it and you now characterize it in the real world or in the wet lab to see how it actually functions Before this summer I have been working on this project on the computational side. The photo on the right is Chimera and that is the protein structure 1M3W that I was hand mutated and so I would go into Chimera, mutate it and then put it into autodock tools and that’s where I got this binding energy and I would repeat opening it up in Chimera editing it, mutating it, and putting it in autodock tools until I got the energy that I wanted This brings me to what I have been doing this summer which is the majority has been the wet lab portion. The first step is to get a large quantity of my protein. This is a long and tedious process however the short and condensed version is I had a little bit of my protein, I fed it to my bacteria bacteria then multiply very rapidly while also multiplying and producing my protein. I then take it back from the bacteria and now I have a large quantity of my protein Now that I have my protein, its time to figure out the actual structure of it. As I said earlier structure equals function so I need to make sure that the actual structure that I saw one the computer is the same structure as was happening in real life to do this I used an experiment called CD, or circular dichroism. Essentially, this is used for a lot of different purposes but what I’m showing you now is the secondary structure. In both instances of my protein, the unmutated and the mutated version of the protein, this is at 5 degrees Celsius if you see the blue line which is our control, the nonmutated you will see these two different humps which this indicated an alpha-helical structure which is what we want. However, the orange structure only has a single hump, which means, that it is in fact a beta-sheet and not an alpha helix Now the next portion of our characterization is to see how well it actually binds To do this we are using assays. Unfortunately, there is no- there was no ribose binding assay kit available so instead we bought a glucose assay kit. Now the main difference between ribose and glucose is that it’s a sugar and its size. So we thought they are similar enough that the ribose- that there is a possibility that the ribose could work the way this assay works- it has the ability to quantify how much glucose is in the solution based off the color change the darker the color the higher the concentration. Which is what you see here. Now to understand if ribose would work for it. I set up the assay and to run over the weekend and I measured it every single hour. Now the equipment that I used the UV-vis isn’t really reliable over 2.5 absorbance so anything really on this right side isn’t really useable or reliable data however, this does show that I can use ribose for this assay I would just have to wait and incubate it for roughly 20 hours for it to workParallel to my wet lab work, I have also been doing computational work. The main focus being efficiency. it takes me roughly 1 hour to get a mutated protein along with its binding score using Chimera and autodock tools. However, this summer I switched programs using Rosetta. Rosetta can make me 1,000 mutant protein in under 5 minutes and then I can score them using Vina all 1,000 of them also under 5 minutes. So instead or taking one hour to get one protein, I can now spend roughly 10-15 minutes to get 1,000 mutant proteins and 1,000 scores which as greatly increased my efficiency Now that I have summarized all of my work from this summer, what’s next? Well for the computational side I am still going to continue using Rosetta and Vina in hopes of making an even better binding protein. For the wet lab stuff, I am going to continue running my CD experiments to see if it’s actually a beta-sheet and how that beta-sheet is affecting binding and in the same vein as binding, now that I know my assay that I ran does work for ribose, I am going to couple it with another assay, to quantify how much ribose my protein actually binds. Here are my references, and finally I would like to thank Dr. Solomon and Dr. Foreman the two PIs on this project, also Rob who is the phd student who is tasked with training me on all the wet lab experiments. He’s been great and a fantastic teacher. Also, Daphine who is an ASSIP intern and was my extra set of hands this summer. She has been fantastic training and collecting all that data for me. Especially a lot of the CD data. Also, Jacklin another phd student who helped me a lot and answered a lot of my questions. She was really great and really sweet. And finally Ralston another undergraduate student who is actually in the URSP program as well. He did very similar projects also under the same mentor Dr. Solomon, and he was great to talk to and sort of bounce ideas off That was my presentation. Thank you for watching