Often underestimated, IR spectroscopy offers excellent time resolution combined with high structural sensitivity. Advanced techniques promise a breakthrough for time-resolved structural characterization of proteins and for structure-function studies in functional proteomics.
By Aihua Xie and Wouter D. Hoff
In the quest to understand biology at the molecular level, scientists have identified millions of genes through DNA sequencing, and determined the structures of thousands of proteins using x-ray and nuclear magnetic resonance (NMR) techniques. At this point, though, we have in-depth understanding of just a handful of proteins–and the desire to understand how proteins perform their biological functions is growing.
Achieving a better understanding, though, requires that we have a proper tool with which to “watch” how nanometer-size proteins perform their biological functions in real time. An important challenge is to detect structural motions of proteins from picoseconds to seconds with high structural sensitivity. Without stimuli, proteins settle in their steady states. Once triggered by interactions with stimuli such as hormones, ligands, photons, and DNA, however, proteins spring into action. Their biological functions are carried out through a series of energy-driven structural transitions, from fast motions in the subpicosecond domain all the way to slow motions that require milliseconds, seconds, or longer.
Traditional x-ray crystallographic techniques are highly successful for three-dimensional structural determination of proteins in steady states. NMR techniques enable structural studies of steady state proteins in solution, and new solid-state NMR techniques are being developed for structural studies of membrane proteins. But while researchers have worked to develop time-resolved x-ray crystallography for light-activated proteins, their successes have been limited. Advanced infrared (IR) spectroscopy, combined with vibrational structural markers, is now emerging as a promising technique to provide insight into time-resolved structural biology of proteins.1
Understanding structural details
Infrared spectroscopy, a standard technique for structural identification of small chemical compounds, provides rich structural information based on the fact that each chemical structure exhibits a unique set of vibrational frequencies. In biophysics, the structural information of interest is very different from that of small chemical compounds. The primary structure of a protein (its amino-acid sequence) is usually determined from its DNA sequence–and the 3-D structure of its initial state is determined by x-ray or NMR techniques. The structural information that is missing for structure-function studies of proteins is functionally important structural motions in real time.
Infrared spectroscopy is highly sensitive to electron transfer, proton transfer, hydrogen bonding interactions, and chromophore isomerization–all of which are functionally important in many proteins. It also probes the secondary structure of proteins using Amide I signals. We will discuss two examples:
Proton transfer. Proton transfer is a molecular event common in protein function, including biological energy transduction, biological signaling and regulation, and catalytic reactions. X-ray crystallography is not sensitive to proton location due to protons’ weak electron density. Most crystal structures of proteins lack structural information on hydrogen atoms. In contrast, infrared spectroscopy is highly sensitive to the location of protons. For example, a protonated carboxylic group (COOH) of Asp or Glu exhibits a signature C=O stretching frequency in the range of 1700–1760 cm-1, while that of deprotonated carboxylate (COO-) group display dramatically different vibrational frequencies: ~1400 cm-1 for symmetric and 1550 cm-1 for asymmetric vibrations of COO-. Thus, infrared spectroscopy is an ideal technique for time-resolved structural studies of proton transfer reactions.
Hydrogen bonding interactions. Hydrogen bonding interactions are essential for protein structure and folding. The energy difference between a folded and unfolded protein is typically ~40 kJ/mol, while the stabilizing energy of hydrogen bonds is around 10 to 20 kJ/mol per hydrogen bond. Thus, disruption of a single hydrogen bond can significantly alter the stability of a protein, demonstrating the structural and functional importance of hydrogen bonding interactions. Because hydrogen bonding interactions change the bond strength, infrared spectroscopy is highly sensitive to hydrogen bonding interactions. As we have demonstrated,5 infrared spectroscopy can be used to identify the number, type, and strength of hydrogen bonding interactions of carboxylic groups in proteins.
A particularly attractive aspect of infrared for protein function study is its ability to measure spectral differences between different functional states of a protein. While a protein exhibits a large number of vibrational modes, only a small fraction is altered during functional transitions in a protein. The modes can be selectively detected by measuring the protein in its initial resting state and subsequently during a specific stage of its functional cycle. The resulting infrared difference spectrum reveals vibrational modes that are altered in the initial state of the protein as negative signals, and modes that are new to the functional intermediate as positive signals. This greatly reduces the overlap problem and selectively identifies functionally relevant modes.
Despite its great potential, infrared spectroscopy is underutilized in biology. In part, this is due to misconceptions regarding infrared spectroscopic techniques. Many researchers do not know that infrared spectroscopy can be time-resolved. Others think infrared spectroscopy is a low-resolution technique. In part, this is due to a data interpretation limitation: an inability to extract structural information from time-resolved infrared spectroscopy.
Infrared spectroscopic techniques offer excellent time resolution and an exceedingly broad time window. For light-activated proteins, the entire time window of protein structural motions can be probed using three different types of advanced infrared spectroscopic techniques. For ultrafast structural dynamics from subpicoseconds to a few nanoseconds, Ti:sapphire-laser-based visible pump infrared probe systems are available. This technique has been applied to study the photochemical reactions and structural transitions of early intermediates of bacteriorhodopsin, photoactive yellow protein, and myoglobin.
Many functionally important motions take place in the time window from a few nanoseconds to a few milliseconds, and can be detected using time-resolved step-scan Fourier-transform infrared (FTIR) spectroscopy. The FTIR approach is technically challenging because it is sensitive to vibration and requires long measurement times for data averaging. Only a limited number of studies have been made using step-scan FTIR spectroscopy.
For detecting slower motions–from a few milliseconds to seconds–time-resolved rapid-scan FTIR spectroscopy is ideal. The use of standard rapid-scan FTIR spectroscopy is an experimentally accessible approach that allows a time resolution of 60 ms at 4.5 cm-1 resolution. Our group has used quadruple splitting to achieve 8 ms time resolution. While many time-resolved FTIR difference spectroscopy experiments use light-triggered reactions in proteins such as photoreceptors and photosynthetic pigments, the time scale of rapid-scan FTIR is also suitable for proteins that are not light-triggered based on rapid mixing applications.
Markers for structural interpretation
The time-resolved FTIR spectral differences generated during protein function often contain many vibrational bands. Structural analysis of these bands requires two essential steps: first, assignment of the band to a specific group in the protein; and second, interpretation of the signals of the assigned vibrational mode in terms of specific structural changes.
Assignment is often performed by using site-directed mutants and homogeneous or side-chain specific isotopic labeling.2 FTIR spectroscopy allows the detection of (de)protonation of a single acidic side chain during protein function with very high accuracy.3, 4 In addition, our research group has implemented an approach that allows highly specific structural conclusions to be drawn based on assigned signal differences. The method involves performing high-level vibrational calculations to determine the effects of various possible structural changes on the vibrational spectrum of the assigned group. Vibrational modes are then selected that are sensitive to only a single structural effect. These vibrational structural marker bands can serve as reliable tools to determine which structural changes affect the assigned group in the protein. We have used this approach to develop a generally applicable structural interpretation of the frequency of the C=O stretching mode of COOH groups in terms of their hydrogen bonding interactions.5 Further development of this approach to additional vibrational modes and amino-acid side chains promises to offer a generally applicable set of high-resolution tools for time-resolved structure determination during protein function.
Unraveling receptor activation
Our FTIR-based analysis of the mechanism of receptor activation in photoactive yellow protein (PYP) illustrates the power of infrared spectroscopy (see Fig. 1). PYP is a blue-light receptor, derived from the bacterium halorhodospira halophila, which controls the swimming behavior of the cell in response to light stimuli. The light-induced activation mechanism of PYP is triggered by photoisomerization of its chromophore. We used a single infrared band (the carboxylic stretching mode of Glu46) out of more than 3500 vibrational modes in PYP to determine the correct chromophore photoisomerization, the proton donor for chromophore protonation, and the hydrogen bonding interactions of Glu46.3,4 We found that proton transfer step from active site residue Glu46 to the light-sensitive chromophore in PYP is a key factor in driving the large protein conformational changes that activate the receptor. In this mechanism the deprotonation of Glu46 in a hydrophobic pocket provides the electrostatic driving force for the protein conformational changes. This “electrostatic epicenter model” for driving a functionally important “protein quake” may also operate in other signaling proteins.
Prospects for functional proteomics
A number of exciting developments promise to further increase the applicability of protein FTIR spectroscopy.
Until recently, FTIR spectroscopy required high protein concentrations (~5 mM), and at least a 3-microliter sample per measurement. This requires high protein solubility and a good protein overexpression system. However, recent techniques using surface enhancement have proved promising for increasing sensitivity by up to ~100-fold.6 This advance greatly expands the number of proteins that can be studied by FTIR difference spectroscopy.
Most functional difference FTIR spectroscopy has been performed using light-sensitive proteins, because of their experimental accessibility. In addition, the use of laser pulses allows the application of step-scan FTIR spectroscopy with up to nanosecond time resolution. Recently, the application of microfluidic devices that allow rapid mixing with a time resolution of a few hundred microseconds have opened time-resolved FTIR spectroscopic studies to reactions that are initiated by chemical stimuli, including protein folding and ligand binding.7
An exciting recent development is the application of focal-plane-array detectors in FTIR experiments.8 In combination with microfluidic devices this can provide time resolution or the simultaneous measurements of a large number of different protein samples. Further development of this approach promises to significantly increase the throughput of functional FTIR difference spectroscopy measurements. This may prove to be a powerful platform for proteomic-scale studies using FTIR spectroscopy–and its use could have substantial impact on numerous applications including medicine, agriculture, and biotechnology.
- R. Vogel, F. Siebert, Curr. Opin. Struct. Biol. 4, p. 518 (2000).
- B. Warscheid, et. al, Vibrational spectroscopy 48, p. 28 (2008).
- A. Xie et. al., Biochemistry 40, p. 1510 (2001).
- A. Xie et. al. Biochemistry 35, p. 14671 (1996).
- B.N. Nie et. al., A Biophys. J. 88, p. 2833 (2005).
- K. Ataka, et. al., J. Am. Chem. Soc. 126, p. 16199 (2004).
- E. Kauffmann et. al., Proc. Natl. Acad. Sci. USA 98, p. 6646.
- N. Kaun et. al. Applied Spectroscopy 60, p. 1273 (2006).
Aihua Xie is professor of physics, and Wouter D. Hoff is associate professor in microbiology and molecular genetics, at Oklahoma State University, Stillwater, OK; http://osu.okstate.edu. Contact Xie at firstname.lastname@example.org.