Probing proteins in a flash
Technique offers faster timescales, consumes less sample
Protein dynamics can reveal much about the structure and function of a protein. But the techniques used to study them are often limited, requiring a significant amount of sample or offering insufficient temporal resolution. As a result, the functional dynamics of many proteins remain poorly characterized.
In an Analytical Chemistry paper published online Nov. 11, researchers with the University of Illinois at Urbana-Champaign, Los Alamos National Laboratory in New Mexico and Rensselaer Polytechnic Institute in Troy, N.Y., reported a technique called microfluidic flow-flash that addresses these limitations. It requires only a minimum of sample and covers a wide range of reaction times.
Researchers have used microfluidics with a flash-flow technique to investigate protein dynamics. The microfluidic flow cell helped reduce the timescales by decreasing the length scales, and the three-dimensional sheath stream (above) limited sample consumption by minimizing dispersion. Reprinted with permission of Analytical Chemistry.
The investigators generally study protein dynamics with a technique involving disassociation of CO from a protein after mixing with an oxygenated buffer, which leads to irreversible binding of oxygen and other reactions. But this method presents challenges. “One problem we often encounter is that protein samples are scarce and hard to make,” said R. Brian Dyer, a researcher with Los Alamos National Laboratory and one of the authors of the paper. Microfluidics minimizes sample dispersion and consumption by using a three-dimensional sheath flow around the sample stream, for example, therefore reducing the amount of sample needed.
Another problem is how to initiate the reaction on a short timescale. For decades, people have used stop-flow techniques involving simple diffusion, Dyer said. “These are great if what you’re interested in is what happens on a timescale of 10 milliseconds to a second; a lot of things happen on that timescale. But a lot more fundamental processes in biology happen faster than that.”
Microfluidics offers a reliable means to initiate and measure these reactions on a short timescale. It is possible to achieve a shorter timescale by reducing the length scales of mixing, Dyer explained. A micron-length scale corresponds with a microsecond diffusion timescale, for example. The short path lengths of microfluidic flow cells therefore allow the researchers to increase the temporal resolution to where they can measure events occurring on the faster timescales.
Others have sought to increase the temporal resolution by shrinking the dimensions. In most cases, though, they have used an approach that might be called turbulent flow mixing -- in which mixing is achieved by creating turbulence between two streams. Dyer and colleagues instead used laminar flow, in which they bring the two streams together in such a way that they do not physically mix. Rather, the method takes advantage of diffusion from one stream to the other.
Laminar flow enables this by creating a sheath flow, where the analyte stream is in the center of a buffer stream. “Having the analyte only in the middle of the channel eliminates most of the effects of dispersion, thus providing more accurate data,” said Paul J.A. Kenis, a researcher at the University of Illinois at Champaign-Urbana and another of the paper’s authors. “Only in a laminar flow regime can one create such a sheath flow.”
The researchers further sped up the timescale by combining microfluidics with laser-induced dissociation of CO from the protein in question, instead of relying on diffusion alone to initiate the reaction. This technique -- known as flow-flash -- has been in use for decades, Dyer said, but to his team’s knowledge no one has miniaturized it for use with microfluidics.
Testing the technique
To demonstrate this microfluidic flow-flash technique, the researchers used it to probe the kinetics of CO recombination or oxygen binding to myoglobin after laser-induced dissociation of CO from horse heart myoglobin. The myoglobin-CO buffer in the flow cell was surrounded by a sheath flow buffer saturated with either CO or oxygen. They dissociated the CO from the myoglobin using the 514.5-nm laser line focus of a Spectra-Physics continuous-wave argon-ion laser and monitored the resulting kinetics using ultraviolet/visible absorbance spectral imaging.
This is a microfluidic flow-flash time-resolved difference visible absorbance image of the reaction of cytochrome c oxidase with oxygen, following photolysis of carbon monoxide. The flow direction is from top to bottom of the image, and the laser photolysis beam is focused at about 1.5 ms on the vertical axis. The spectral changes represent the binding of oxygen to the binuclear active site and reduction to water.
The detection system was based on a Carl Zeiss inverted microscope outfitted with a 10× or 40× air objective. Either a halogen or a xenon arc lamp illuminated the sample; a broadband grating in a 0.25-m imaging spectrograph dispersed the white light, which was then imaged by a thermoelectrically cooled CCD array detector made by Princeton Instruments of Trenton, N.J.
The work showed, first, that the experiments can be completed with considerably less sample consumption. A spectral image containing 1340 spectral pixels and 400 time points required only about 69 pmol of protein sample, based on a 2-MHz digitization rate of the CCD array, 0.1-s integration time, 750 μM protein concentration and a 15 μL/min flow rate. The sample consumption could be reduced even further, the researchers noted, by digitizing a smaller region of the CCD array detector.
In contrast, “flash-flow” experiments using a microcuvette would require approximately 256 nmol of sample to obtain spectra at the same number of time points with a protein concentration of 8 μM. The significantly reduced sample consumption of the microfluidic “flash-flow” technique bolsters the study of proteins -- and mutants, in particular -- that are otherwise difficult or expensive to investigate.
In addition, the improved timescale of the measurements provided new insights into the reaction probed in the study. It turns out that the initial process in most enzyme reactions (encounter complex formation) is “much more complex than anyone ever thought it would be,” Dyer said.
The researchers continue to develop the technique. First, they are working to determine the best achievable temporal resolution, focusing on two known problems currently limiting the resolution. In addition, they are exploring use of Fourier transform infrared (FTIR) spectroscopy in place of ultraviolet/visible absorbance spectral imaging. The connection between spectrum and structure is much closer with FTIR, Dyer said.
“If you want to see how a specific amino acid gets perturbed during the reaction, you could look at the infrared absorption that corresponds to the amino acid and actually follow it as the reaction proceeds. That’s where we’re going with the technique.”
MORE FROM PHOTONICS MEDIA