Selective Isotope Labelling of Leucine Residues by Using α-Ketoacid Precursor Compounds

Full text of DOI:10.1002/cbic.201200737

CHEMBIOCHEM
COMMUNICATIONS
DOI: 10.1002/cbic.201200737
Selective Isotope Labelling of Leucine Residues by Using
a-Ketoacid Precursor Compounds
Roman J. Lichtenecker,*[a] Nicolas Coudevylle,[b] Robert Konrat,[b] and Walther Schmid[a]
High-resolution NMR spectroscopy and X-ray analysis are the
two most important methods for providing atomic-level three-
dimensional structural information about large biomolecules.
In addition, NMR experiments can be applied to characterise
protein dynamics involved in different processes such as pro-
tein folding, enzymatic activity and molecular interaction.[1] Ali-
phatic side-chain residues are considered to be especially val-
uable sources of structural and dynamic information, due to
their frequent occurrence in the hydrophobic cores of globular
proteins and their unique relaxation properties.[2] Highly ad-
vanced NMR experiments, as well as diverse stable isotope la-
belling techniques, have been developed in order to maximise
the number of attainable structural and dynamic parameters
while reducing spectra complexity.[3]
Scheme 1. Metabolic intermediates—originating from pyruvate—in the bio-
synthesis of valine and leucine in E. coli.
The methyl resonances of Ile, Val, Leu and Ala all appear in
a narrow spectral range, however, and signal overlapping af-
fects the investigation of large protein complexes. Metabolic
precursor compounds—such as pyruvate,[4] a-ketoacid[5] or
acetolactate derivatives[6]—have been efficiently applied in
cell-based protein expression systems to incorporate stable iso-
topes at well-defined positions in the target macromolecules.
Although selective isoleucine labelling can be achieved by
adding a-ketobutyrate[5a] or acetohydroxybutyrate[7] to the
growth medium, valine and leucine labelling has been de-
scribed only in the case of simultaneous labelling, with a-keto-
isovalerate[5b] or acetolactate[6] as common precursor com-
pounds (Scheme 1). Various methods for achieving residue-
type-specific editing of the methyl spectra, based on residue-
discriminating NMR experiments, together with convenient la-
belling techniques, have been published.[8] However, the so far
unprecedented separation of leucine and valine labelling by
using independent precursor compounds would offer a more
general approach for achieving diverse isotope patterns. Here
we introduce a new methodology for labelling leucine by
using a-ketoisocaproate as a direct metabolic precursor with-
out interfering with the valine metabolic pathway (Scheme 1).
To explore the application of 2-ketoisocaproate as a selective
precursor for leucine residues we synthesised 1-13C-labelled
compounds 4 and 5 (Scheme 2). Firstly, the low-cost source of
label [1-13C]glycine (1) was converted into the corresponding
hydantoin.[9] In the next step, [1-13C]hydantoin was subjected
to condensation either with acetone or with isobutyraldehyde
in the presence of ethanolamine to give compound 2 or 3, re-
spectively. Finally, opening of the hydantoin ring under basic
[a] Dr. R. J. Lichtenecker, Prof. Dr. W. Schmid
Institute of Organic Chemistry, University of Vienna
Wꢀhringerstrasse 38, 1090 Vienna (Austria)
E-mail: roman.lichtenecker@univie.ac.at
[b] Dr. N. Coudevylle, Prof. Dr. R. Konrat
Department of Structural and Computational Biology
Max F. Perutz Laboratories, University of Vienna
Dr. Bohr-Gasse 9, 1030 Vienna (Austria)
Scheme 2. Conditions: a) KOCN, H2O, 1008C, 2 h; then HCl, 1008C, 1 h, 95%;
b) ethanolamine, acetone, H2O, 708C, 20 h, 75%; c) ethanolamine, isobutyral-
dehyde, H2O, 1108C, 9 h, 72%; d) 20% NaOH, 1008C, 5 h, 85%; e) protein ex-
pression in BL21-pLysS cells in 15NH4Cl-supplemented medium.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.201200737.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2013, 14, 818 – 821 818

CHEMBIOCHEM
COMMUNICATIONS
www.chembiochem.org
conditions yielded the target precursor 4 or 5. As a proof of
concept, His-tag–GB1 fusion proteins (109 residues) were syn-
thesised by using an overexpressing E. coli strain in minimal
medium containing unlabelled glucose, 15NH4Cl and 4 or 5
(100 mgLÀ1).
13C incorporation was analysed by measuring 13CO,15N,1HN
correlations in both the Val- and the Leu-labelled proteins
(Figure 1). In each case, the number of signals corresponded to
the relevant number of residues in the protein sequence and
Figure 2. Influence of the concentration of precursor 5 in the growth
medium on the level of protein isotope incorporation. The intensity ratios of
the Leu residue signals in the corresponding HNCO and HSQC spectra are
normalised to a uniformly labelled sample obtained by protein expression in
the presence of 13C6-glucose as sole carbon source. Each data point repre-
sents the average resulting from three Leu residues in the His-tag–GB1 pro-
tein sequence.
1
Figure 1. Overlay of H,15N HSQC (black resonances) and 2D HNCO (red and
blue resonances) of His-tag–GB1 fusion proteins produced from precursors
4 (red; U-[15N] Val [1-13C]) and 5 (blue; U-[15N] Leu [1-13C]).
could be assigned accordingly. This result demonstrates effi-
cient transformation of precursors 4 and 5 into the corre-
sponding residues without any observable metabolic isotope
scrambling. To correlate the amount of precursor compound in
the growth medium with the observed protein isotope incor-
poration level, His-tag–GB1 (66 amino acids) was expressed in
the presence of different concentrations of compound 5. The
HNCO/HSQC signal intensity ratio observed for Leu residues
was used as a reporter of 13C incorporation (Figure 2). The
maximum level of labelling was achieved by using
>150 mgLÀ1 of precursor in the minimal medium.
In order to provide a broadly applicable leucine precursor,
a route to labelled sodium [3-2H2, 4-2H, 5-13C, 5’-2H3]-a-ketoiso-
caproate (10) has been developed and is described in
Scheme 3. Stepwise methylation of diethyl malonate (6) gave
diethyl [13CH3/C2H3]dimethylmalonate (7).[10] Subsequent sapo-
nification and decarboxylation yielded labelled isobutyric acid
(8), which was further reduced to the corresponding alcohol
with LiAlH4.[11] Oxidation to isobutyraldehyde under Pfitzner–
Moffat conditions proved to be superior to conventional
Swern oxidation in terms of reaction reproducibility. The prod-
uct was directly frozen into a cooling trap in THF at 90 mbar
and subsequently converted into 5-isobutylidenehydantoin (9).
Performing this reaction step in the presence of D2O allowed
Scheme 3. Conditions: a) NaOEt, ethanol, CD3I, reflux, 2 h, 72%; b) NaOEt,
ethanol, 13CH3I, reflux, 80%; c) KOH, water/ethanol (1:2), 1008C, 4 h, 92%;
d) H2O, microwave irradiation, 1808C, 15 min, 94%; e) LiAlH4, diethyl ether,
6 h, 81%; f) dicyclohexylcarbodiimide, DMSO, pyridine, CF3CO2H, THF, RT;
then hydantoin, ethanolamine, D2O, 1108C, 15 h, 55%; g) 20% NaOD, 1008C,
5 h, 84%; h) protein expression in BL21-pLysS cells in 15NH4Cl supplemented
medium.
target compound 10. The synthetic route offers high flexibility
for the creation of diverse 13C/2H patterns in sodium a-ketoiso-
caproate from commercially available isotope-enriched starting
materials. Labelled forms of diethyl malonate (6) could make
additional isotope patterns possible in future applications and
are synthetically accessible as [1,3-13C2][12] or [2-13C][13] deriva-
tives.
2
facile introduction of H at the C4 position of the molecule.
Treating 9 with dilute NaOD (prepared by slow addition of Na
to D2O) introduced deuterium in the a-position and gave the
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2013, 14, 818 – 821 819

CHEMBIOCHEM
COMMUNICATIONS
www.chembiochem.org
Precursor 10 was applied for selective 13C/2H-labelling of His-
tag–GB1. 13C HSQC of the purified protein showed six signals,
which could easily be assigned to the Leu residues in the pro-
tein sequence (Figure 3). Again, no significant interference
with any other amino acid metabolic pathway was observed.
We strongly anticipate that precursor 10 will be of special
value in various areas of protein NMR spectroscopy. First of all,
plications in the fields of NMR-based protein structure determi-
nation, the probing of fast- and slow-timescale side-chain dy-
namics, and solid-state NMR spectroscopy.
Experimental Section
Complete experimental details of the chemical synthesis and
chemical characterisation of the precursor compounds and syn-
thetic intermediates, as well as details concerning protein over-
expression and purification, are provided in the Supporting Infor-
mation.
Acknowledgements
This work was supported by grants from the Austrian Science
Foundation (FWF P20549, P22125 and W1221B03).
Keywords: cell-based protein expression · isotopic labeling ·
metabolic precursors · NMR spectroscopy · proteins
[1] a) M. Takeda, M. Kainosho in Protein NMR Spectroscopy: Principal Tech-
niques and Applications (Eds.: L.-Y. Lian, G. Roberts), Wiley, Hoboken,
2011, pp. 23–53; b) V. Tugarinov, P. M. Hwang, L. E. Kay, Annu. Rev. Bio-
chem. 2004, 73, 107–146; c) L. Banci, I. Bertini, C. Luchinat, M. Mori,
Prog. Nucl. Magn. Reson. Spectrosc. 2010, 56, 247–266; d) M. P. Foster,
C. A. McElroy, C. D. Amero, Biochemistry 2007, 46, 331–340; e) V. Tugari-
nov, V. Kanelis, L. E. Kay, Nat. Protoc. 2006, 1, 749–754.
[2] a) V. Tugarinov, L. E. Kay, ChemBioChem 2005, 6, 1567–1577; b) V. Tugar-
inov, L. E. Kay, J. Am. Chem. Soc. 2003, 125, 13868–13878; c) H. van
Melckebeke, J.-P. Simorre, B. Brutscher, J. Am. Chem. Soc. 2004, 126,
9584–9591; d) A. M. Ruschak, L. E. Kay, J. Biomol. NMR 2010, 46, 75–87;
e) R. Lichtenecker, M. L. Ludwiczek, W. Schmid, R. Konrat, J. Am. Chem.
Soc. 2004, 126, 5348–5349; f) J. D. Gross, V. M. Gelev, G. Wagner, J.
Biomol. NMR 2003, 25, 235–242; g) P. J. Hajduk, D. J. Augeri, J. Mack, R.
Mendoza, J. Yang, S. F. Betz, S. W. Fesik, J. Am. Chem. Soc. 2000, 122,
7898–7904.
[3] a) G. Wagner, J. Biomol. NMR 2010, 46, 1–2; b) L.-Y. Lian, D. A. Middle-
ton, Prog. Nucl. Magn. Reson. Spectrosc. 2001, 39, 171–190; c) R. M.
Rasia, B. Brutscher, M. J. Plevin, ChemBioChem 2012, 13, 732–739;
d) C. G. Hoogstraten, J. E. Johnson, Jr., Concepts Magn. Reson. Part A
2008, 32A, 34–55; e) S. Ohki, M. Kainosho, Prog. Nucl. Magn. Reson.
Spectrosc. 2008, 53, 208–226.
[4] C. Guo, C. Geng, V. Tugarinov, J. Biomol. NMR 2009, 44, 167–173.
[5] a) K. H. Gardner, L. E. Kay, J. Am. Chem. Soc. 1997, 119, 7599–7600;
b) N. K. Goto, K. H. Gardner, G. A. Mueller, K. C. Willis, L. E. Kay, J. Biomol.
NMR 1999, 13, 369–374.
[6] P. Gans, O. Hamelin, R. Sounier, I. Ayala, M. A. Dura, C. D. Amero, M.
Noirdlerc-Savoye, B. Franzetti, M. J. Plevin, J. Boisbouvier, Angew. Chem.
2010, 122, 2002–2006; Angew. Chem. Int. Ed. 2010, 49, 1958–1962.
[7] a) A. M. Ruschak, A. Velyvis, L. E. Kay, J. Biomol. NMR 2010, 48, 129–135;
b) I. Ayala, O. Hamelin, C. Amero, O. Pessey, M. J. Plevin, P. Gans, J. Bois-
bouvier, Chem. Commun. 2012, 48, 1434–1436.
[8] a) H. Senn, B. Werner, B. A. Messerle, C. Weber, R. Traber, K. Wꢁthrich,
FEBS Lett. 1989, 249, 113–118; b) W. Hu, A. T. Namanja, S. Wong, Y.
Chen, J. Biomol. NMR 2012, 53, 113–124; c) G. Jaipuria, N. P. Lobo, D.
Shet, H. S. Atreya, J. Biomol. NMR 2012, 54, 33–42; d) R. P. Barnwal, H. S.
Atreya, K. V. R. Chary, J. Biomol. NMR 2008, 42, 149–154; e) C. Guo, V.
Tugarinov, J. Biomol. NMR 2010, 46, 127–133; f) B. Krishnarjuna, G. Jai-
puria, A. Thakur, P. D’Silva, H. S. Atreya, J. Biomol. NMR 2011, 49, 39–51;
g) G. Lv, H. K. Faßhuber, A. Loquet, J.-P. Demers, V. Vijayan, K. Giller, S.
Becker, A. Lange, J. Magn. Reson. 2013, 228, 45–49; h) I. Ayala, R. Souni-
er, N. Usꢂ, P. Gans, J. Boisbouvier, J. Biomol. NMR 2009, 43, 111–119.
[9] E. Ware, Chem. Rev. 1950, 46, 403–470.
Figure 3. 1H,13C HSQC of U-[15N] Leu-[13CH3, C2H3] His-tag–GB1, selectively
labelled at leucine residues by using precursor 10.
the dissociation of leucine labelling from valine labelling allevi-
ates spectral crowding in the methyl resonance region, and
this should be beneficial for the extraction of distance and
angle restraints, especially in larger protein systems. A proto-
typical 13C,1H NOESY-HSQC spectrum obtained with the isopro-
pyl 13CH3/C2H3 pattern of compound 10 is given in the Sup-
porting Information. The labelling scheme of 10 additionally
qualifies for implementations in methyl-TROSY experiments,
the value of which for NMR spectroscopy of large proteins has
been shown in numerous examples.[14] The methyl-TROSY
effect is substantially increased by branched side chain 13CH3 /
C2H3 labelling schemes, due to optimised relaxation rates.[15]
Moreover, an isolated 13CH3 group in a highly deuterated back-
ground is also regarded as a sensitive reporter of protein dy-
namics and folding, and the presented precursor should fur-
ther contribute to increasing the size of systems that can be
studied.[16] Another important future field of application in-
cludes solid-state NMR spectroscopy. The flexible route to dif-
ferent labelling patterns of precursor 10 allows for the synthe-
sis of isolated 13C carbon sites, so that homonuclear dipole–
dipole couplings are minimised to gain additional sensitivity in
13C detection (e.g., in 3D NCACX or NCOCX experiments).[17]
In conclusion, we have demonstrated selective and exclusive
leucine labelling by usign a-ketoisocaproate as a direct meta-
bolic precursor in cell-based protein expression systems. This
compound can be prepared with different labelling patterns
and can be tailored to meet the requirements of numerous ap-
[10] M. Long, D. W. Thornthwaite, S. H. Rogers, G. Bonzi, F. R. Liyens, S. P.
Rannard, Chem. Commun. 2009, 6406–6408.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2013, 14, 818 – 821 820

CHEMBIOCHEM
COMMUNICATIONS
www.chembiochem.org
[11] A. H. G. Siebum, W. S. Woo, J. Lugtenburg, Eur. J. Org. Chem. 2003,
4664–4678.
[16] D. Sheppard, R. Sprangers, V. Tugarinov, Prog. Nucl. Magn. Reson. Spec-
trosc. 2010, 56, 1–45.
[12] H. Thies, W. Franke, H. Schwarz, Synthesis 1982, 7, 587–588.
[13] M. Lang, S. Lang-Fugmann, W. Steglich, Org. Synth. 2002, 78, 113.
[14] a) V. Tugarinov, P. M. Hwang, S. E. Ollerenshaw, L. E. Kay, J. Am. Chem.
Soc. 2003, 125, 10420–10428; b) L. E. Kay, J. Magn. Reson. 2011, 210,
159–170.
[17] V. A. Higman, J. Flinders, M. Hiller, S. Jehle, S. Markovic, S. Fiedler, B.-J.
van Rossum, H. Oschkinat, J. Biomol. NMR 2009, 44, 245–260.
Received: November 26, 2012
[15] V. Tugarinov, L. E. Kay, J. Biomol. NMR 2004, 28, 165–172.
Published online on April 5, 2013
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2013, 14, 818 – 821 821