Biomolecular NMR: α-Ketoacids as Biosynthetic Precursors for Protein NMR

By: Vitali Tugarinov, Ph.D., and L.E. Kay, Ph.D., ISOTEC® Stable Isotopes 2008 - 2010 Catalog

The state-of-the-art NMR methodology for the studies of large proteins requires full deuteration of all non-exchangeable sites in the polypeptide chain. Unfortunately, full deuteration precludes the use of NOEs for structure determination. Since methyl groups give the most intense signals in NMR spectra of proteins due to their favorable relaxation properties and are frequently located in the hydrophobic cores of protein structures, it was suggested to protonate methyls selectively in otherwise fully deuterated proteins1. Selective protonation of methyl groups allows to retain protons in the key elements of protein structure and to measure methyl-methyl and methyl-HN NOEs with high sensitivity and resolution2 for determination of global folds of larger proteins1, 3-7

The development of strategies for selective methyl protonation of deuterated large proteins has a long history. It has been recognized that α-ketoacids may serve as biosynthetic precursors for incorporation of several methyl moieties with desired isotopic labeling patterns into proteins overexpressed in minimal media. Initially, [U-13C] pyruvate was used as a primary carbon source for production of protonated Alaβ, Ileγ, Valγ and Leuδ methyls in highly deuterated protein molecules1. However, this approach leads to extensive generation of methyl isotopomers of other than 13CH3 variety (13CHD213CH2D and possibly 13CD3) in protein samples. Due to isotope shifts, both 13C and 1H in each of these methyl variants resonate at different frequencies degrading resolution and sensitivity in NMR spectra. Although each of the isotopomers may be easily selected for spectroscopically, such losses in sensitivity are detrimental to NMR studies of high-molecular-weight (> ca. 30 kDa) proteins. Later, selective protonation of isoleucine δ1 methyls was suggested through addition of α-ketobutyrate (2-ketobutyric acid) (obtained by enzymatic reaction from threonine) to the growth medium8. It was also shown that the addition of α-ketoisovalerate (2-keto-3- methylbutyric acid) leads to efficient production of methyl-protonated Val and Leu residues9. Despite that the number of methyl-containing sites is restricted to Ileδ1, Valγ, and Leuδ in this case, the use of α-ketobutyrate and α-ketoisovalerate in the growth media does not lead to production of methyl isotopomers and is, generally, a robust and effective technique for selective protonation of methyl groups9. Sodium salts of α-ketobutyric and α-ketoisovaleric acids can be added to the D2O-based minimal growth medium in amounts of ~50 mg and ~80 mg/L, respectively, ca. 1 hour prior to induction of protein overexpression. These keto-acids can be obtained in the α-Ketoacids as Biosynthetic Precursors for Protein NMR Vitali Tugarinov, Ph.D., and L.E. Kay, Ph.D. Departments of Medical Genetics, Biochemistry, and Chemistry, The University of Toronto, Toronto, Ontario M5S 1A8, Canada protonated form at position 3 and quantitatively exchanged to 3-2H at high pH in D2O (pH 12.5, ~2-3 hrs. for α-ketoisovalerate and pH 10.5, ~12-14 hrs. for α-ketobutyrate) prior to use. In addition, α-ketobutyrate (the source of δ1-methyl protonated Ile) and α-ketoisovalerate (the source of methyl protonated Val and Leu) may be used separately or together without scrambling between Ile and Val,Leu labels in the course of protein production. 

Building upon the efficiency and robustness of selective methyl protonation in deuterated protein structures through the use of α-ketoacids, the Abbott group developed synthetic methods for production of α-ketobutyric and α-ketoisovaleric acids with 13C enrichment exclusively at methyl positions10. The main purpose of this approach was to eliminate 13C-13C scalar couplings that would allow recording very sensitive and highly resolved 1H-13C methyl correlation maps for high throughput screening of ligands to larger proteins. This group demonstrated that the screening methods based on 1H-13C spectroscopy of selectively [13C,1H]-labeled methyl groups of Ile,Leu and Val can be at least 3-5 times more sensitive than those based on 1H-15N TROSY spectroscopy10. An alternative and more cost-effective synthetic strategy for production of these ketoacids was proposed later by Gross et al. and used for the studies of protein-ligand interactions11. Recently, Konrat and co-workers have reported a versatile synthetic procedure that allows to incorporate 13CH3/12CD3 (one methyl group - 13CH3, the other - 12CD3) isotope labels into isopropyl moieties of Val and Leu residues with high efficiency12. Indeed, the primary advantage of using α-ketobutyric and α-ketoisovaleric precursors is in the possibility of incorporation of practically any desired labeling pattern into the side-chains of Ile,Leu,Val residues as well as into methyl moieties themselves. For example, using appropriate precursors the side chains of these residues may be uniformly 13C-labeled or, alternatively, the 13C labels can be restricted to methyl groups. The isopropyl moieties of (13CH3/13CH3) or (13CH3/12CD3) type can be produced in Val/Leu side-chains. Likewise, methyl groups of 13CH313CHD2, and 13CH2D variety can be incorporated into Ileδ1, Valγ, and Leuδ sites. Such isotope labeling possibilities in methyl-bearing side-chains open new perspectives for NMR of large protein molecules. 

Recently, Tugarinov & Kay showed that NMR assignments can be obtained for Leu and Val groups of very large proteins using ‘linearization’ of the corresponding spin-systems through the use Content originally published in ISOTEC® Stable Isotopes 2008–2010 Catalog. α-ketoisovalerate non-stereospecifically labeled with 13C in only one of its two methyl positions13. The same laboratory showed that factors of 2-to-3 in sensitivity can be achieved through the use of HMQC as opposed to HSQC experiments on protonated Ile δ1 methyls (obtained using α-ketobutyrate 13C-enriched only in the methyl position) in a highly deuterated protein environment14. The Methyl-TROSY methodology opened new perspectives for NMR studies of methyl groups attached to very large proteins and macromolecular complexes. However, as with 1H-15N TROSY, Methyl-TROSY requires extensive deuteration of the protein and a modification of standard methyl labeling strategies. The labeling strategy that maximizes the dipolar TROSY effect in Leu and Val side chains involves the 13CH3/12CD3 labeling of isopropyl groups15. Using the samples of a 723-residue monomeric protein Malate Synthase G (MSG, 82 kDa), selectively protonated at Ileδ1 (13CH3), and Valγ/Leuδ sites (13CH3/12CD3), the global fold of this protein could be derived very recently exclusively from solution NMR data16-18. Finally, the utility of 13CHD2 and 13CH2D methyl isotopomers incorporated into the side chains of Ile/Val/Leu amino acids has been established recently for 2H and 13C relaxation studies in high-molecular-weight proteins in the 100 kDa range19

Applications of the labeling methodology and the methyl-TROSY approach have been published in studies of a number of supramolecular systems by the Kay group. Applications include the allosteric enzyme aspartate transcarbamoylsae (306 kDa)20, the ClpP protease (300 kDa)21 and the 20S core particle proteasome (670 kDa)22. A survey of the approaches used and some of the applications have recently been published as well23

Please visit our α-Ketoacids webpage for a complete list of products.


Materials List



  1. Rosen, M. K., Gardner, K. H., Willis, R. C., Parris, W. E., Pawson, T. and Kay, L. E. (1996) J. Mol. Biol263, 627-636. 
  2. Zwahlen, C., Gardner, K. H., Sarma, S. P., Horita, D. A., Byrd, R. A. and Kay, L. E. (1998) J. Am. Chem. Soc. 120, 7617-7625. 
  3. Gardner, K. H., Rosen, M. K. and Kay, L. E. (1997) Biochemistry36, 1389-1401. 
  4. Metzler, W. J., Wittekind, M., Goldfarb, V., Mueller, L. and Farmer, B. T. (1996)J. Am. Chem. Soc118, 6800-6801. 
  5. Gardner, K. H. and Kay, L. E. (1998) Annu. Rev. Biophys. Biomol. Struct.27, 357- 406. 
  6. Mueller, G. A., Choy, W. Y., Yang, D., Forman-Kay, J. D., Venters, R. A. and Kay, L. E. (2000) J. Mol. Biol300, 197-212. 
  7. Venters, R. A., Metzler, W. J., Spicer, L. D., Mueller, L. and Farmer , B. T. (1995)J. Am. Chem. Soc117, 9592-9593. 
  8. Gardner, K. H. and Kay, L. E. (1997) J. Am. Chem. Soc.119, 7599-7600. 
  9. Goto, N. K., Gardner, K. H., Mueller, G. A., Willis, R. C. and Kay, L. E. (1999) J. Biomol. NMR 13, 369-374. 
  10. Hajduk, P. J., Augeri, D. J., Mack, J., Mendoza, R., Yang, J. G., Betz, S. F. and Fesik, S. W. (2000) J. Am. Chem. Soc122, 7898-7904. 
  11. Gross, J. D., Gelev, V. M. and Wagner, G. (2003) J. Biomol. NMR 25, 235-242. 
  12. Lichtenecker, R., Ludwiczek, M.L., Schmid, W. and Konrat, R. (2004) J. Am. Chem. Soc126, 5348-5349. 
  13. Tugarinov, V. and Kay, L.E. (2003) J. Am. Chem. Soc.125, 5701-5706. 
  14. Tugarinov, V. Hwang, P.M., Ollerenshaw, J.E. and Kay, L.E. (2003) J. Am. Chem. Soc125, 10420-10428.
  15. Tugarinov, V. and Kay, L.E. (2004) J. Biomol. NMR 28, 165-172. 
  16. Tugarinov, V. and Kay, L.E. (2004) J. Am. Chem. Soc.126, 9827-9836. 
  17. Tugarinov, V., Kay, L.E., Ibraghimov, I. and Orekhov, V. Yu. (2005) J. Am. Chem. Soc127, 2767-2675.
  18. Tugarinov, V., Choy, W.Y., Orekhov, V.Y. and Kay, L.E. (2005) Proc. Natl. Acad. Sci. USA 102, 622-627. 
  19. Tugarinov, V., Hwang, P.M. and Kay, L.E. (2004) Annu. Rev. Biochem73, 107-146. 
  20. Velyvis, A., Yang, Y.R., Schachman, H.K. and Kay, L.E. (2007). Proc. Natl. Acad. Sci. USA 104, 8815-20. 
  21. Sprangers, R., Gribuin, A., Hwang, P.M., Houry, W.A. and Kay, L.E. (2005). Proc. Natl. Acad. Sci. USA 102, 16678-83. 
  22. Sprangers, R. and Kay, L.E. (2007) Nature, 445, 618-22. 
  23. Sprangers, R., Velyvis, A. and Kay, L.E. (2007). Nature Methods 4, 697-703.


Related Links