A great effort has been spent for the past several years develop analytical chemistry, particularly of biochemical molecules, on smaller and smaller amounts and on faster and faster times. The reasons for this are clear. In medical settings it is becoming important to be able to perform noninvasive in situ monitoring (for example glucose sensing in a diabetic) in real time and then to respond (for example by dispensing insulin). It has also become more and more important to be able to test for many things at once and to do so in an automated manner. In principle, one would like to be able to implant a small device in the body that would constantly monitor a large number of important aspects of your health and give you, or your doctor, a readout as a function of time. How might one do this sort of thing?
The general idea of an automated micro lab on a chip.
If we are going to have the equivalent of a small clinical laboratory on a chip, we are going to have to accomplish a number of interesting tasks on this scale.
We need to be able to take samples (e.g., blood) and transfer them to different places where reactions or separations are done.
We need to be able to separate components.
We need to be able to detect and quantitate specific compounds in the mixture.
Microfluidics. The first problem is to move liquids around on this scale. A number of cleaver approaches have been developed. Essentially all of them are based on the microelectronics industry's great accomplishments in making smaller and smaller features on surfaces that can be controlled in some way electronically. There are two general ways fluid is moved. One involves the development of moving mechanical parts on very small scales. This is called MEMS (micro electromechanical systems). For more information, see the links on the following page:
Various types of pumps, valves, pipes, mixing chambers, etc. have been manufactured on the micron scale, allowing one to bring in a sample and divide it up for a series of different analytical procedures. Another way of moving liquid around is electroosmosis. Basically, this is a technique which utilizes electric fields to cause fluid flow. The fields can be changed as necessary to move small volumes of liquid along specific channels. This is nice because it is completely solid-state (no mechanical parts). Professor Mark Hayes in our department is an expert in this field, if you are interested in learning more.
Separation. Most of the small scale (microliter or less) separation methods that I am familiar with are based on molecules moving differently in electric fields. Small volume capillary electrophoresis became very popular a little over a decade ago, when it was found the one could separate analytes effectively in very small volumes inside of capillaries by putting a charge across them. The same ideas can be applied to microchannels generated on chips. Also, one can vary the type of molecules which run faster or slower by changing in the surface characteristics of the channel. In principle, all of the chromatography techniques used at large scale could be used at very small scale by driving the flow with either a MEMS pump or electroosmotically.
Detection. Detection is often performed optically, with fluorescence giving the highest sensitivity. Alternatively, it could be used to directly generate an electric signal. In either case one usually needs to have something that the analyte of interest specifically interacts with, resulting in a change in optical density, electronic or fluorescence propoerties. It is now possible to put light sources (light emitting diodes and laser diodes) directly on the chip as well as light detectors (photodiodes and avalanche photodiodes). Thus, one can in principle build simple versions of absorbance spectrophotometers and fluorimeters directly onto chips.
Biochemistry one molecule at a time.
In the past decade there has also been a tremendous interest in investigating the properties of individual molecules or molecular complexes, rather than large numbers of molecules together. This is enabling biochemists to ask questions about how much molecular properties vary within a population and potentially what order complex reactions or assembly sequences occur in. It also has allowed direct measurements of interaction energies between molecules and molecular dynamics.
Optical methods. There are two kinds of methodology that come into play here. One is optical microscopy, in various forms, and the other scanning probe microscopy. In the optical microscopy realm, there are two approaches. One is just to look at a highly magnified field under the microscope with a CCD (charge coupled device) camera. They make some sensitive low noise cameras these days that work pretty well for clusters of fluorophores. Scientists in Japan have taken pictures of an ATPase complex turning round and round during proton pumping and ATP hydrolysis this way. See:
for additional information on this and
for a neat site that another professor put together on molecular motors in general which includes a link to a movie made of the ATPase working. Other types of molecular motors have been investigated in this way as well including RNA polymerase, DNA polymerase, kenesin and cytoplasmic dynein. Much of this is described on the general page on molecular motors above.
The other general approach, which really gets down to the single molecule (single fluorophore) level is some kind of small volume fluorescence method. Scanning Near Field Microscopy and Scanning (or stationary) Confocal Microscopy are the two most popular. Scanning Near Field Microscopy involves taking a fiber optic and stretching the tip out to a very narrow end (less than 100 nm across). If this tip is moved up right next to the sample of interest (within tens of nanometers), one can actually beat the diffraction limit of light and see areas smaller in dimension than the wavelength of light. On the other end of the fiber is a very sensitive detector (usually a photomultiplier tube or an avalanche photodiode) which counts the photons emitted into the fiber by the sample. The problem with scanning near field microscopy is that it is hard to use in solution, which is where most biology occurs. Thus, people often use confocal microscopy instead. This approach takes advantage of the fact that if you focus on one depth of field in the microscope, only a particular point at that depth will map to a particular position in a particular plane in the image field. If you put a pin-hole at that point, you limit the volume of view in all three dimensions. This is done both on surfaces and in solution. The volumes of view are determined by the wavelength of light. Typically, they can be limited to about a femtoliter. One fluorescent molecule in this femtoliter can be observed on millisecond and longer time scales. These kinds of techniques allow one to look in detail at dynamics of individual molecules or complexes. Others have used these techniques to look at the fluorescence excitation spectra of individual molecules (a plot of the amount of fluorescence you get by exciting at different excitation wavelengths -- closely related to an absorbance spectrum).
Scanning Probe Methods. People have also visualized individual molecules and complexes with various kinds of scanning probe microscopy. The most popular in the biological community these days is called atomic force microscopy (AFM) and allows one to look at molecular surfaces with nanometer resolution. Basically, you take a very, very narrow tip and drag it along the surface. Using a laser, you watch it go up and down as it crosses over the molecules on the surface. It is very sensitive. People have visualized atoms this way. If you want to learn more about it look up Professor Stuart Lindsay in the Physics Department. He is a pioneer in this field. You can also manipulate molecules using this methodology. These days it is quite the thing to attach the end of something like a protein molecule to the AFM tip and then attach the other end either to a surface or to a big bead. In the case of the surface you have one end of the molecule or pair of molecules attached to the surface and the other end attached to the tip. Let's say it is a protein. Now pull. How much force does it take to unfold the protein? How much force does it take to break a bond? These kinds of direct measurements have been very interesting and informative. Another variation is to attach one end of the molecule to a small bead of some kind (say a cellulose bead). Now you use something called optical trapping or laser tweezers to pull on the bead. This can provide a more calibrated force test. Optical trapping is a phenomenon that occurs at the narrowest part of very well focused beams of laser light. Basically, as a photon of light goes through a particle, it is refracted, causing a force on the particle due to the momentum change of the photon. People have not only measured energies involved in protein unfolding but also forces associated with the molecular motors we discussed above. One can ask, how much light is required to hold the bead in place. This gives a measure of how hard the bead is being pulled on by the molecular motor.