Biomolecular NMR: Isotope Labeling Methods for Protein Dynamics Studies
Liquid-phase nuclear magnetic resonance spectroscopy has long relied on uniform and stable isotopic enrichment of 13C and 15N to reduce resonance overlap and allow for multi-distance and multi-angle confinement on as many atomic sites as possible, in order to facilitate the calculation of the best three-dimensional structural model.1 In recent years, the optimization of these labeling techniques has increased the range of modifiable protein sizes, improved the quality of three-dimensional structures, and simplified the process of analyzing experimental data.2 At the same time, advances in the field of protein dynamics have also benefited from advances in isotope labeling, which have allowed researchers to study the properties of protein motion in larger volumes over a wider time scale, while more accurately characterizing protein trajectories. In many ways, advances in kinetic isotope labeling mirror advances in techniques used in structural studies. However, spin-relaxation experiments designed to study protein dynamics have unique requirements for residue labeling, requiring the rapid development of isotopic enrichment techniques for better application to these advanced research topics.
The Need for Isolated Spin Systems in Dynamics
olution NMR is a powerful method to characterize the motion of proteins over a wide range of time scales by measuring the relaxation rates of desired nuclei. The design of these relaxation experiments, as well as the analysis and interpretation of the data, is greatly simplified if the position of the target protein is treated as an isolated spin pair. In this case, the pulse sequence design does not need to consider and manipulate multiple undesired coherent paths, and the resulting relaxation rate can be measured directly from the mono-exponential decay profile of the peak intensity. However, the presence of multiple large single-bond couplings complicates experimental results through multi-exponential relaxation paths and signal noise degradation. Because of this, most of the conventional labeling methods are provided by providing a way to label the isotopes of different isolated spin pairs in proteins so that one-bond scalar (J) coupling does not pose a problem in the analysis of experimental data.
15N-labeling
To date, most kinetic studies have been achieved through the enrichment of a single 15N. 15N is a good research direction, and the necessary nitrogen for cell growth can be controlled by readily available 15N-enriched trace or nitrogen-enriched growth media, making sample preparation easy. A single 15N label constitutes an isolated spin system (1H-15N), making it well suited for relaxation experiments. Each 15N position is separated from another 15N atom by at least two bonds, both in the main chain and in the side chain of the protein. Therefore, without 1JNN couplings that lead to complex multi-exponential relaxation behaviors, such couplings would be difficult to measure precisely and would affect the interpretation of the associated motion.
However, the enrichment of 15N alone does not provide a complete picture of protein movement. Nitrogen only accounts for 1/3 of the protein backbone, and only 6 of the 20 amino acids contain nitrogen in their side chains. Therefore, except for a few selectable positions (Asn, Gln, His, Trp, Lys and Arg), 15N relaxation experiments do not allow large amounts of dynamic coverage in the range to avoid obscuring the full picture of the motion of the main and side chains, especially the amide relaxation may be relatively insensitive to the motion of the hydrophobic core of the protein. It was also noted that by monitoring the relaxation of the amide positions, certain motor behaviors of the protein backbone could not be detected. Nonetheless, the ease of 15N protein labeling, the accumulation of robust existing experiments, and the sensitivity of nitrogen elements to structural, electrostatic, and hydrogen-bonding effects make 15N an important part of modern kinetic studies.
13C-labeling
On the face of it, the relaxation experiments of carbon provide many additional opportunities for molecular dynamics studies using NMR spectroscopy. The abundance of carbons in each amino acid provides more probes for the study of enzyme kinetics. These carbons are contained in the main and side chains and can provide dynamic information about the entire protein. Methyl residues are often buried in the hydrophobic core and are particularly suitable for providing information on the dynamic processes of protein folding and stability. The chemical shifts of 13C, especially Cα, are more dependent on protein secondary structure than nitrogen in amides, which makes chemical shift changes obtained from certain spin relaxation experiments easier to understand.
Unfortunately, the ideal carbon labeling method is not as straightforward as the nitrogen labeling method. The large number of carbon elements in proteins can cover most positions in protein dynamics research, which is also the biggest obstacle to its research; the large number of carbon atoms means that almost all carbon atoms are adjacent to another carbon atom. Thus, uniformly labeled 13C protein samples result in a large number of 13C-13C couplings at many residues, making data that should directly explain relaxation behavior intractable in many cases. The only position that is not affected by the 1JCC coupling described above is the methyl group on methionine, which is separated from the rest of the protein by a sulfur atom. Therefore, the study of kinetics in a unified 13C-labeled protein is very limited. While the relaxation data for methionine can be very useful, it does not provide a more complete level of dynamic coverage from carbon labeling schemes. Because of this, many isotopic labeling methods have been developed that utilize known bacterial metabolic pathways.
One method used to isolate 13C markers is to use 15% 13C-acetate for partial labeling, leaving the rest as 12C.3 In this way, the 13C marker can be diluted to a certain level to make the relaxation experiment feasible, but the signal-to-noise ratio will also decrease due to the reduction of the proportion of labeled protein.
Using [3-13C] pyruvate as the sole carbon source,13C labeling of Leuδ, Valγ, and Ileγ can be achieved at >90% incorporation levels.4 What's more, the isotopic labeling largely did not disrupt the directly bonded carbons, allowing us to perform relaxation measurements on these residues. Therefore, good single-noise and single-exponential relaxation behaviors can be observed at the positions of these methyl groups. Like methionine, these residues allow us to understand the trajectories of proteins in the hydrophobic core.
The use of α-keto acids also provides a cost-effective method for producing 13C-methyl-labeled amino acids that are also compatible with high levels of deuterated carbons at other positions.5,6 The use of deuterated methods allows kinetic studies to be performed on larger protein systems than other methods. Another benefit of this approach is the highest order of 13C markers, which minimizes the problems described above due to 13C-13C conjugation.
The use of 1- or 2-position labeled glycerol in auxotrophic cell lines allows for the incorporation of alternating labels on most carbons throughout the protein side chain. Typically, two protein samples are required to cover atomic positions as completely as possible.7 Aromatic residues can complement the data obtained for methyl groups, as they are often also present in the hydrophobic core. The specific labelling in these residues is particularly important given the strong J-coupling present and the small range of chemical shifts. Early studies showed that growth on glycerol at the [2-13C] position would result in isotopic enrichment of alternating carbons in most amino acids, including isolated aromatic carbons in Phe, Tyr, and Trp. Growth on glycerol at the [1,3-13C] position will occur with the opposite labeling pattern. An alternative to this is to use [1-13C]-glucose as the sole carbon source, and aromatic rings are labeled at the positions of Pheδ, Tyrδ, Hisδ2/ε1 and Trpδ1/ε3.8 The [1-13C]-glucose labeling method is very practical and efficient because it is not only more economical, but also enables higher protein yields.
Recent studies have shown that [1-13C]-glucose also leads to the enrichment of methyl residues in Ala, Val, Leu, Met and Ile up to about 45%, these residues are separated from other 13C-labeled atoms by two or more bonds.9 It was also found that expression using [2-13C]-glucose as the sole source of carbon was shown to result in an enrichment rate of 20-45% at the Cα site, not labeling the C' site, but only labeling Leu, Val and Cβ site of Ile. This labeling allows CPMG relaxation experiments to be performed at the Cα position, providing additional data to complement the conclusions drawn from common 15N CPMG experiments.
Finally, isotopic labeling can be precisely incorporated by introducing the desired marker amino acid directly into the growth medium.10,11 Often, to avoid label disturbance or dilution, the desired amino acid is included in a mixture containing all other unlabeled amino acids, but the synthesis can be time-consuming given the position of the labeled isotope. This technique has proven useful in kinetic and structural studies and has recently been used for NMR structure determination of bulk proteins.2
Future Prospects
As more and more bacterial metabolic pathways are exploited to provide specific isotopic labels, by developing these techniques for solution NMR relaxation experiments, in-depth studies of the dynamical behavior of proteins across various residue types can be achieved.
References
1. Muchmore DC, McIntosh LP, Russell CB, Anderson DE, Dahlquist FW. 1989. [3] Expression and nitrogen-15 labeling of proteins for proton and nitrogen-15 nuclear magnetic resonance.44-73. https://doi.org/10.1016/0076-6879(89)77005-1
2. Kainosho M, Torizawa T, Iwashita Y, Terauchi T, Mei Ono A, Güntert P. 2006. Optimal isotope labelling for NMR protein structure determinations. Nature. 440(7080):52-57. https://doi.org/10.1038/nature04525
3. Wand A, Bieber R, Urbauer J, Mcevoy R, Gan Z. 1995. Carbon Relaxation in Randomly Fractionally 13C-Enriched Proteins. Journal of Magnetic Resonance, Series B. 108(2):173-175. https://doi.org/10.1006/jmrb.1995.1119
4. Lee AL, Urbauer JL, Wand AJ. 1997. 9(4):437-440. https://doi.org/10.1023/a:1018311013338
5. Gardner KH, Kay LE. 1997. Production and Incorporation of15N,13C,2H (1H-?1 Methyl) Isoleucine into Proteins for Multidimensional NMR Studies. J. Am. Chem. Soc.. 119(32):7599-7600. https://doi.org/10.1021/ja9706514
6. Goto NK, Gardner KH, Mueller GA, Willis RC, Kay LE. 1999. 13(4):369-374. https://doi.org/10.1023/a:1008393201236
7. LeMaster DM, Kushlan DM. 1996. Dynamical Mapping ofE. coliThioredoxin via13C NMR Relaxation Analysis. J. Am. Chem. Soc.. 118(39):9255-9264. https://doi.org/10.1021/ja960877r
8. Teilum K, Brath U, Lundström P, Akke M. 2006. Biosynthetic13C Labeling of Aromatic Side Chains in Proteins for NMR Relaxation Measurements. J. Am. Chem. Soc.. 128(8):2506-2507. https://doi.org/10.1021/ja055660o
9. Lundström P, Teilum K, Carstensen T, Bezsonova I, Wiesner S, Hansen DF, Religa TL, Akke M, Kay LE. 2007. Fractional 13C enrichment of isolated carbons using [1-13C]- or [2-13C]-glucose facilitates the accurate measurement of dynamics at backbone C? and side-chain methyl positions in proteins. J Biomol NMR. 38(3):199-212. https://doi.org/10.1007/s10858-007-9158-6
10. LeMaste D, Cronan J. 1982. Biosynthetic production of 13C-labeled amino acids with site-specific enrichment. Journal of Biological Chemistry. 257(3):1224-30.
11. LeMaster DM, Richards FM. 1982. Preparative-scale isolation of isotopically labeled amino acids. Analytical Biochemistry. 122(2):238-247. https://doi.org/10.1016/0003-2697(82)90275-5