Of the different fields of science, biology is probably the most visual. Physics certainly isn't. Physics is mainly concerned with forces, like gravity or magnetism. The eyes are useless in experiencing either of these -- they can only be felt or imagined. Chemistry is the same way. Molecular interactions are way too small to be seen, so they must be monitored in other ways, perhaps by comparing differences in weights, volumes, or temperatures before and after some event. The results usually involve a graph, plenty of equations, and the odd figure, which is usually abstract as to only be understood by chemists. Pick up an issue of Science and you'll agree right away.

This argument -- that biology is the most visual of the sciences -- isn't black and white, of course, because biology requires graphs and math, and both physics and chemistry put to use microscopes, or sometimes telescopes, to answer questions. But the point is that there is very little to see, speaking literally, in the majority of what physicists and chemists do. And while Newton may have pondered gravity by watching an apple fall from a tree he certainly couldn't start carrying on about this undescribed force without explaining it using well defined physical properties, and -- again -- lots of math.

It may sound like a bold claim, but it is a plain and simple fact that biology is extremely visual, and in many cases biologists have gone out of their way to maintain this state of affairs. That is, they have found ways of using the microscope to answer all sorts of questions that aren't particularly visual. For instance, it is common for a molecular biologist studying some sort of intracellular enzyme to attach a protein on to it which will make the enzyme glow where it was invisible before. Ever look through a microscope at a cell? Your lucky if you can identify anything besides the nucleus. Now imagine deciding whether your lab's special protein is inside that cell -- a protein that is many orders of magnitude smaller than the cell and the nucleus, and is as transparent as water on top of that. It would help if you could say that your protein is the only fluorescent protein that the cell knows how to make, no? It would immediately become clear, as you stared through your microscope, whether or not the cell is making your protein.

But sometimes biologists want to know more than just the location of a protein -- sometimes it's helpful to know what the protein looks like. To know which surfaces contact other proteins, or which surfaces are responsible for binding DNA, or know exactly where a molecule of oxygen is tucked away in the protein is nothing but earth-shattering information. Again, literally. Making a protein glow does little good in this instance. In fact, microscopes are fairly inept at helping visualize things as small as proteins at all, if the details are important. Yet in keeping with biology's obsession with the visual, scientists have ways of making 3D models of proteins and other intracellular objects which are for all intents and purposes unable to be seen. It may seem paradoxical that a model of something that is quite literally invisible would be of any help at all, but it turns out to be a remarkably useful way of explaining how a protein works. Even more remarkable is that these models are true-to-form, physical blow-ups of the molecules that they represent -- they are not abstracted in any sort of way. The models are literally exactly what these molecules would look like if we were able to see them.

The process, which is as painstaking as it is rewarding, is called crystallography, and it strays afield from the traditional biology into the physicist's world of X-ray beams and the chemist's realm of solutions and precipitations. In principle the technique starts off as simple as, say, making rock candy in fifth grade science class. Instead of adding tons of sugar to boiling water, like fifth graders hoping to create big, hunky sugar cubes, crystallographers put tons of their protein into different solutions and hope to have the protein arrange into a regular pattern as the liquid evaporates. The odds are against this from happening. If the solution evaporates too quickly, or if the protein is somehow altered by the solution—say, by too much or too little salt -- it will solidify in a patternless pile of junk that is useless for the upcoming visualization process. If a crystal does form, it may not happen for a week, or it may take a month, or it may take many months. It may form out of the first solution into which the protein dissolves, or it may have to be tested in hundreds or even thousands of slightly different solutions to find one in which it will arrange regularly as the liquid evaporates. As if all this weren't enough, it's possible that the protein may never form crystals at all. The most confounding part of all these variables is the randomness of the process -- there are nothing but general outlines that define the best conditions to form a crystal from any given protein. A protein may form once, just by chance, never to be created again after multiple repetitions in the same conditions. The technique is firmly trusted by scientists when it works, but is as unexplainable as witchcraft in half or more of the cases that it does.

In any event, getting a protein to form a crystal is only the beginning. Purified crystals must be analyzed using X-ray diffraction, and so are taken to a machine called a synchrotron. This device is required to produce a focused, super strong beam of X-rays which passes through the crystal. Computers at the synchrotron measure how the X-ray beam is bent by the sample, or, in other words, how much it is diffracted. The X-rays particles, or photons, change their direction as influenced by the atoms and electrons in the sample, and their new positions can be recorded. A particular crystal has a unique X-ray diffraction pattern which depends directly on the placement of the atoms within the substance and the orientation of the crystals in space. A rudimentary parallel can be drawn to tracing someone's profile by using a shadow that has been cast on the wall -- it is a 2D representation of a 3D object, and it changes as the person rotates in space. So rather than depending on just one orientation of the crystal, the X-ray diffraction pattern is logged again and again when the crystal is in a number of different orientations. It is up to a computer program and a skilled crystallographer to compile these data and transform the patterns of scattered X-rays into a 3D model of what a molecule looks like.

Crystals are essential to creating decipherable X-ray diffraction patterns. Since the atoms on a crystal are oriented in a regular, repeating pattern, the X-ray passing through the sample as a whole will be bent in a similar manner by every repetition of a molecule in the crystal. The effect is to resolve the X-ray beam into a reliable and repeatable pattern. If a crystal doesn't form, but instead the protein randomly assembles as it dries out of the solution, like a pile of garbage, the X-ray beam will be bent inconsistently and the information will be useless. But when it works the information creates patterns that faithfully represent the arrangement of atoms in the molecule.

Now, one doesn't just leave the synchrotron with a number of photographs that show a protein in different spatial orientations. The synchrotron produces images of diffraction patterns, but these are about as useful as the silhouettes are to decipher the 3D structure that produced them. It's up to a computer and a lot of guess-and-check to go from fairly vague information to a complex image which shows the atoms in a protein. But the light at the end of the tunnel is a model which can provide countless clues about a protein's function.

The sense of sight is fundamental in the way that we interact and understand our surroundings, and science has developed ways of making what is not directly tangible and visible -- whether it is too far away, too small, or abstract -- into something that can be easily interpreted with two eyes. In a roundabout way, X-ray crystallography creates models of miniscule substances that can be studied using how they look as a gauge for how they work. It produces realistic images of things that cannot be seen, and though it is as mysterious as black magic, it produces something simple out of something unfathomable.

« Previous Next »