Wednesday, January 11, 2012

My Ph. D. Thesis for the non scientist

I spent 7 and a half years (from June 2000 - January 2008) at Boston University studying to get my Ph. D. in physics. However, most people I know (including my fellow classmates) have no idea what I did there and why it took me so long. So in this post, I will explain my time there including what I did and why. Hopefully I can do it in a way that anyone with a elementary school education can understand. If you are up for it, read the complete version of my thesis here.

Where to begin...

First, I should tell you about my motivation to be a scientist. I remember very clearly Christmas of my 6th grade year, my parents gave me a copy of the "1989 World Book Science Year." In this book, there were two articles the profoundly effected my career choices. First was an article about Stephen Hawking. His ability to make discoveries despite his considerable handicap inspired me. Second was an article about ozone depletion. This article made me decide that I wanted to pursue a career helping to solve environmental problems.

I spent the years after focusing on my science studies and I was fortunate enough to have good biology, chemistry, and physics teachers in high school. In particular, my chemistry teacher gave me a book called "Surely You're Joking, Mr. Feynman" which let to my eventual decision to major in physics in college.

I did some research projects as an undergrad related to the environemnt. One was using some techniques to determine the amount of lead in paint samples while the other was related to determining the amount of radioactivity in water filters used to remove radon. These projects were OK, but I wanted to do something bigger. So when applying to graduate school, I tried to get into programs involved in nuclear fusion research. Sadly, I was not accepted in to these programs.

I ultimately settled on Boston University for my graduate degree. I bounced around among a few research groups, not really satisfied with the projects as they really weren't that related to environmental problems. In my second year, I found the Molecular Biophysics lab, which although was not ideal, was intriguing. They were doing research on a class of photoactive proteins that seemed like they might have some relationship to environmental problems and renewable energy in particular.  Our simple task was to figure out how these proteins work.

So the story begins with the question: What is a protein? You can look at the Wikipedia definition of a protein, but I prefer a simpler explanation. A protein is a collection of molecules that are arranged such that they perform a particular function. The molecules in proteins are not just any molecules, but they are amino acids. Living things use 20 different amino acids in a plethora of different combinations and patterns to make proteins that have a huge variety of functions. Proteins allow us to see, hear, touch, smell, and taste. They allow us to move and they allow us to think. Pretty amazing, huh? So how do we figure out how they work? There are a number of techniques, and I'll get to that eventually, but first let me introduce the protein that I studied.

I studied green-absorbing proteorhodopsin, or GPR, for short. This GPR is a member of a large class of proteins called microbial rhodopsins. You may have heard the word "rhodopsin" before. It is the protein found in the cells of our eye, that allows us to see! So you might think that these microbial rhodopsins are eyes for single celled microorganisms. In a sense, that is correct. There are actually two types of microbial rhodopsins, one of which serves as a light sensor (similar to our eyes, but not nearly as complex - these are microbial organisms, after all), and the other serves as an ion pump. I focused on the ion pumps rather than the light sensors, although the mechanism is surprisingly similar.

What's an ion pump? Good question. As you know, molecules are composed of different kinds of atoms. Most of those involved in living things contain carbon, oxygen, nitrogen, hydrogen, and small amounts of several others. Furthermore, atoms are composed of electrons, protons and neutrons. Electrons have a negative charge while protons have a positive charge and neutrons have no charge. In atoms, the protons and neutrons are concentrated in the nucleus of an atom at it's center while the electrons surround the nucleus. Usually, an atom or molecule has equal number of electrons and protons. However in certain reactions between various atoms and/or molecules, they can gain or loose electrons becoming "ions" - molecules or atoms with a few extra electrons or a few missing electrons.

So you know what an ion is; what is a ion pump? As you know living organisms are composed of cells. Cells are little sacs of goo surrounded by a membrane composed of lipids (special fat molecules). This membrane keeps what is inside the cell separate from what is outside the cell. This is great unless you want something inside the cell to be moved outside or vice versa. In this case, nature has developed proteins called transmembrane proteins that sit inside the membrane and can be used to move stuff across the membrane. Therefore, ion pumps move ions through the cell membrane! One particularly interesting ion pump is found in nerve cells.  It is called the sodium potassium pump and is involved in transmitting signals between nerve cells. This is one of the proteins involved in allowing you to move, think, feel, taste, touch, and pretty much everything else. GPR is a much simpler pump. It simply moves a positively charged hydrogen atom - just a proton - from inside the cell to outside the cell.  Here is a cartoon depicting the GPR protein inside the cell membrane.

Cartoon depicting GPR inside the cell membrane. The cell membrane is the lipid bilayer while the cytoplasm indicates the inside of the cell and the extracellular medium is of course outside the cell. 
How do ion pumps pump ions? That is the main question. This kind of process requires some energy. In humans and other animals, we eat food, which we then digest and convert to an energetic molecule called ATP through the citric acid cycle. The proteins can then use the energy stored in ATP to help ion pumps. In fact, pretty much any energy requiring process in living things makes use of ATP, so it is a pretty important moelcule. Making ATP via the citric acid cycle is a complicated process that requires oxygen. Many microorganisms can do this process, but they aren't so good at farming, hunting and going grocery shopping, so they need an alternative method to make ATP if they cant find food or oxygen (in fact some microorganisms don't like oxygen). They have come up with an ingenious solution! They use the energy from light to make ATP (this is actually just one solution among many that exist).  This is a two step process. First, the ion pump absorbs light, and using the energy it absorbs from the light, it moves a proton from inside the cell to outside the cell. After a while, there exists a lot more protons outside the cell than inside. Essentially, the energy from the light is now stored in the concentration difference between the number of protons inside and outside the cell. Another protein called ATP-synthetase allows the protons back into the cell using the energy released to make ATP. This is called Mitchell's Theory of Chemiosmotic Copuling. In this way, the proteins are converting light energy to chemical energy.

So our task during my thesis studies was to figure out how the GPR uses light to move a proton across the cell membrane. Seems an impossible task, right? Well, it was pretty hard, to be sure, but before I give you the answer, I should tell you why it is an interesting question, or, maybe it isn't interesting. That's up to you to decide.

In the early 1970's research groups discovered a purple protein in the microorganism halobacterium salinarium. This is a special type of microorganism belonging to a class of organisms called archaea. These kinds of organisms are very primitive and like to live in extreme environments. Halobacterium salinarium likes to live in extremely salty areas such as the salt flats in San Francisco Bay. See the map below. The orangish color is due to the salt loving bacteria. They discovered that this pruple protein, which they dubbed bacteriorhodopsin was an ion pump involved in producing ATP. In addition, they also discovered 3 other similar proteins, one of which was a chlorine ion pump and the other two acted as light sensors.

View Larger Map

Many research groups studied these proteins in great detail using a huge variety of techniques, and they came up with a model of how the proteins function. In addition, they also came up with a bunch of hairbrained ideas of technological applications using bacteriorhodopsin in things from fuel cells, to solar cells (this is what got my interest), to memory devices, holograms, and secruity devices, none of which really got very far in the market.

Then beginning in the year 2000, a number of groups, including one led by Craig Venter, father of the human genome project, discovered through sequencing tons of sea water, similar proteins to bacteriorhodopsin found in almost 13% of all the microorganisms in the ocean's photic zone (photic zone means where sunlight is still  present, deeper it's just dark).  If you read Venter's biography, you will learn he is a huge sailing fan, so he combined his gene sequencing with sailing for these projects. Given the density of these organisms throuought the ocean, these organisms convert sunlight to ATP on a huge scale - on the order of the amount of energy used by humans. So if we could figure out how they do it, maybe we could use it in renewable energy applications. That was my motivation for taking on this project. I like the big picture, the potential for society changing results.

In addition, as I mentioned, these rhodopsin proteins are similar in many ways to the rhodopsin in your eyes. Furthermore, the rhodopsin protein in your eye is a member of a a class of proteins called G-Protein coudpled receptors which are involved in vision, smell, behavior and mood regulation, inflammation and immune response, nervous system response, and homeostasis. Therefore, if we can understand how the simpler protein from the microorganism works, we can gain insights into how the more complex ones work.

Some researchers, though, just like puzzles. I'm not one of them.

So, as I mentioned above, groups have been using a variety of techniques such as protein NMR and x-ray crystallography to figure out how these microbial rhodopsins work (and many other proteins, of course), but we used spectroscopy. What is spectroscopy? Simply, it is the study of how light and matter interact. Since these proteins are activated by light, spectroscopy seems like the most natural tool to study them.

Let's talk a bit more about spectroscopy. Light is an electromagnetic wave caused by the oscillation of charges. Depending on the amplitude of the oscillation, the light can be intense or dim, and depending on the frequency or wavelength (the wavelength is essentially the inverse of the frequency, so think of them as the same thing) of oscillation, the light can be visible or invisible to our eyes. In fact, the spectrum of light visible to our eyes is only a tiny fraction of the entire spectrum of light. You can associate light of different wavelengts with different things. Visible light has wavelengths ranging from about 400-700 nanometers. A nanometer is 1/1000000000 of a meter. Microwaves, which are used to cook your food in a microwave, have wavelengths around a centimeter, and radio waves are just light of wavelengths around a meter. Ultraviolet light which gives you a sunburn is just a little bit shorter than visible light, while x-rays are around 1 nanometer. Even shorter than that are gamma rays from nuclear radiation.

Spectroscopy is the study of how light interacts with things. For example, studying how UV light gives you a sunburn is spectroscopy. When you tune your radio, you are doing spectroscopy as you are changing how the radio waves interact with the metal circuits in your radio. For the experiments I did, we were concerned with visible and infrared light and how they interacted with the protein. From this interaction we hoped to figure out how the protein works.

Just as an aside, as one reads this, they might start to think: "Hey, this sounds like biology, not physics!" Well, this is just not true. Indeed, the problem is a biological one, but spectroscopy is about as physicsey as physics can get. Describing how light interacts with matter involves both electromagnetism and quantum mechanics and I can inundate you with fancy looking mathematical formulas describing spectroscopy which would make most biologists and chemists faint, so don't go accusing me of being a biologist instead of a physicist!

Just for fun, here are a few of those fancy formulas from a tutorial lecture on spectroscopy  I give occasionally.

OK, back to how light interacts with the protein. To explain this, lets do a little thought experiment. Think of a collection of a bunch of different atoms. To make a molecule, we connect the atoms together with a bunch of springs. Then surrounding the molecule is a cloud of electrons. They are moving so fast you can't see them. Think of them like a fuzzy cloud. In addition, the atoms are vibrating about in between their springs because the molecule isn't frozen at absolute zero temperature. If we shine visible light on this molecule, most likely, nothing will happen.  However, if we pick just the right frequency (and the frequency depends on the number and arrangement of the atoms in the molecule), the electrons will absorb some of the light. When this happens, simply think of the fuzzy cloud expanding slightly, and maybe changing shape a little. As the electron cloud changes, the amplitude and frequency of the atoms vibrating on their springs might also change. We can think of the molecule in an excited state when it is like this. When we turn off the light, after some time, the molecule will go back to its initial state.

What happens if we use infrared light instead of visible? Well, it so happens that the molecules are vibrating because they are warm. They are vibrating at frequencies in the infrared. In fact, infrared light and heat are the same thing! If we shine light of a certain infrared frequency at the molecule, we might increase the amplitude of vibration around one or more of the springs between the atoms in the molecule.  If we look closer, we can see that the springs connecting the atoms in the molecule are not all the same, some are long, some are short, some are very flexible and some are very stiff.  In addition, not all of the atoms have the same weight.  All of these factors can change the frequency of the vibrating atoms. Therefore, in theory, if we can make some kind of table relating the frequency of vibration to the type of molecule and type of bond (think type of spring here), we can find out a lot about the molecule and how things are connected and change upon external stimulus without actually seeing a picture of the molecule. Ingenious, huh? That is the basis for infrared spectroscopy.

So think of our protein as a complex arrangement of balls connected by strings with an electron cloud around it. Our job is to figure out how the arrangement of balls and springs change when we shine visible light on the protein. So basically, we shine infrared light on the protein and figured out where it absorbed infrared light - we take an infrared absorption spectrum. Then we excited the protein with visible light. Upon excitation the electron cloud changed a bit and also caused some extra vibrations. In some cases these extra vibrations can lead major changes (like breaking springs and making new ones). This is the case with GPR. After shining visible light on the protein wend took another infrared absorption spectrum. By examining the differences in the infrared spectrum before and after shining visible light on the protein, we could determine what in the molecule was changing as a result of the visible light and eventually how it uses the energy from the light to move a proton from inside the cell to outside the cell.  Here is a simple schematic of what I just described.

Schematic of the infrared spectroscopy experiment to determine what happens to the protein. Our job is to assign all of the squiggly lines in the bottom graph to changes occurring in the protein.
It turns out that quite a lot of things are changing in the protein after exposing it to visible light and to make a complete picture, you have do take spectra as a function of time. In fact, it appears that things are changing on a time scale as short as 100 femtoseconds (1 femtosecond = 1/1000000000000000 s) and as long as 1 second. That's a lot of orders of magnitude to cover, so for my thesis I just focused on the earliest changes from femtoseconds to picoseconds (1 picosecond = 1000 femtoseconds).

How can you measure something that lasts only 1/1000000000000000 of a second? The best DSLR cameras have a minimum shutter speed of around 1/8000 s. I still need 12 orders of magnitude faster. The solution is to use ultrafast lasers, the bane of my existence and why my thesis study took so long. These incredibly expensive and complex instrumets can produce light pulses that last only 100 femtoseconds (but only when they feel like working - which isn't very often). How does a femtosecond laser work? Maybe I'll explain later...  So, I built an apparatus that would produce 100 femtosecond infrared laser pulses and 100 femtosecond infrared pulses that would allow me to take infrared spectra before and after excitation with the short visible pulse. Building the apparatus was an incredible headache - one of the laser components was replaced 5 times by the company over the 6 years I worked on the project. In addition, there was no expertise in the group on these measurement techniques, so I had to learn everything on my own by trial and error. I also spent an inordinate amount of time trying to make my data look as good as that of another group doing a similar experiment. Come to find out, they applied smoothing to their data, so it wasn't actually as good as it looked. Cheating!!!  Mine however, looked as good as their smoothed data as I ran the same experiment thousands of times over the course of a week and averaged everything.  It was a rare day when everything (laser, protein, weather, planetary alignment) worked at the same time.  Over the 6 years, I really only had a few weeks were I was able to take good enough data to publish.
Picture of the femtosecond laser aparatus - this isn't the one I used, but the new one made possible by the NSF grant (see below)

So after lots of hard work (and many hours of frustration relieving punching bag sessions), I managed to get some data and found out some details of the molecular changes occurring immediately after absorption of light in GPR that no one else had discovered before. Was it worth the effort? That's to be determined. I suppose you can never really tell the value of scientific research - The paper has been cited 14 times, so I suppose it was useful to at least 14 people anyway. In addition, our work let to the award of a million dollar grant from NSF to take this kind of measurement to the next level. However, I certainly don't feel as I contributed much, if at all to the search for a renewable energy solution, or to society in general. I have to wonder if there are a number of other non puzzle loving scientists who feel similarly about their doctoral research and if there is anything we can do to change the way doctoral research is done to increase its value to society. However, I did learn a lot about the scientific method and how to design apparatus. I also learned how to deal with frustration and failed experiments.

Thanks for reading. I hope you learned something.
Post a Comment