Lab Visit Report: Biophysics

One of the many very cool things going on in the Laser Cooling Empire at NIST is a series of experiments using optical tweezers to study various biological systems. I used to share an office with the biochemist in the group, who was there to handle the wet chemistry that physicists are notoriously bad at.

I've toyed with the idea of setting up an optical tweezers apparatus at Union-- the optical set-up is very simple-- so I spent a little while talking about it with Kris Helmerson, the PI on that project. He mentioned some cool things they were doing, one of which is explained very nicely in this arXiv pre-print. It looks as though this article from Applied Physics Letters is the same text after a few rounds of editing, so we'll even call this a peer-reviewed research blogging post (the arXiv version has more context, though, so if you're only going to read one, read that. If you're a hard-core peer-review fetishist, though, the APL is available free online.)

The basic idea of the work in question is that they have developed a technique for creating femtoliter (that's 10^-15 liters) containers for biological molecules, in the form of very small droplets of water-- called "hydrosomes" in the preprint, but backed off to "aqueous droplets" in the APL-- suspended in another fluid. These droplets can be trapped and manipulated with optical tweezers, allowing them to do some controlled single-molecule chemistry.

The key idea of optical tweezers is that an object with a higher index of refraction than the material surrounding it will act like a lens, and bend light that passes through it. Since the light carries momentum, changing the direction of the light is analogous to exerting a force on the light, and there is a reaction force that acts back on the object. It turns out that this force will tend to pull the object toward regions of higher light intensity, and as a result, if you focus a laser beam down to a very small point, you find that you can trap small objects in the focus. (The laser power required is pretty small, so you won't actually obliterate the object in question)

This technique has been used to trap single cells and small glass beads. If you're clever, you can track the position of the object to within a nanometer or so, and thus look in detail at biological processes like the motion of kinesin molecules towing glass beads. You can also easily determine the force exerted by the tweezers, and use this to measure all sorts of forces at the single cell/ single molecule scale.

One thing it would be nice to be able to do is to study reactions between single molecules of various substances. There are a few tricks people can pull to do approximately this-- attaching one of the molecules to some sort of substrate, or confining them very tightly in solid matrices of some kind-- but many of these involve at least the risk of distorting the molecule in some way that might change the reaction. What you really want to do, particularly with biological molecules, is to find a way to bring together single molecules in solution, as they will be in most of the situations where these things actually react. That's a little harder to manage.

In the papers linked above, they offer a really cool way of doing this sort of thing. They create tiny droplets of water, a micron or two in diameter, suspended in some other fluid (the fluid they use is "fluorinert FC-77," a fluorocarbon liquid that they use as a coolant in some of the other experiments in the Laser Cooling Group-- they need something with a lower index of refraction than water, and they had this stuff lying around). They put a small amount of water containing the molecules of interest into a larger amount of fluorinert, and then shake it up in an ultrasonic cleaner for a few minutes. The water and the fluorinert don't mix, but the sonication causes the water to break up into micron-scale droplets, which can then be trapped using optical tweezers. These droplets serve as convenient containers for molecules in solution, and if the concentration of the original solution is low enough, there will be only a few molecules per drop.

In order to convincingly claim that this is a useful tool for single-molecule chemistry, you need to demonstrate a number of things. The first is that you're really trapping single molecules in the droplets, which they do by looking at the behavior of fluorescent dye molecules in the droplets. They trap a droplet in the tweezer beam, then illuminate it with light that causes the dye molecules to emit green light. They collect green photons on a detector for several seconds, and look at the amount of green light emitted as a function of time.

The dye molecules they use are subject to "photobleaching," which means that they only work for so long, and then they stop emitting green light. In looking at the time behavior of the fluorescence (Figure 2 of the APL, Figure 4 of the pre-print), they see clear and discrete steps in the amount of light, corresponding to individual molecules bleaching at different times. They show the fluorescence signals of droplets containing one, two, and three dye molecules, demonstrating that it's possible to contain and detect small numbers of molecules by this technique.

If this is going to be a useful tool in biochemistry, you would also like to know that the process does not damage the molecules, or change their behavior. They show this in two ways, first by simply encapsulating Red Fluorescent Protein in droplets, and showing that it survives the sonication process intact. They also did some studies of dyes bound to DNA molecules and showed that FRET (Fluorescence Resonance Energy Transfer) reactions are still visible (Figure 3 of the APL, Figure 5 of the pre-print). So chemistry still works inside the droplets.

The final thing you need to make this a useful tool is the ability to controllably combine droplets containing different molecules you would like to react. They demonstrate this using a pair of optical tweezer traps, one fixed and one moving, and show that simply bringing two droplets into contact with one another is enough to cause them to merge and mix their contents (Figure 4 of the APL, Figure 2 of the pre-print).

The pre-print version also contains some discussion of an experiment in which two droplets, one containing a dye that will fluoresce only when bound to DNA, and the other containing DNA strands, are brought together and mixed, demonstrating that the fluorescence shows up only after the mixing. This isn't in the final published version, though, and I'm not sure why.

Kris said that more recently, they've been able to make droplets on demand using a micropipette system to inject a single drop of water into the fluorinert, which is even better than the more complicated sonication process. This really looks like a great technique for doing single-molecule chemistry, which has all sorts of fun applications for basic science. Nobody's likely to be building up designer drugs using this technique, but it might be a useful tool for exploring really basic chemcial reactions on an extremely small scale.

Or it might not. I'm not a biophysicist, after all. It sure sounded cool to me, though, so I figured I'd share it.

J. E. Reiner, A. M. Crawford, R. B. Kishore, Lori S. Goldner, K. Helmerson, M. K. Gilson (2006). Optically trapped aqueous droplets for single molecule studies Applied Physics Letters, 89 (1) DOI: 10.1063/1.2219977