Independent valine and leucine isotope labeling in Escherichia coli protein overexpression systems

Article abstract of DOI:10.1007/s10858-013-9786-y

The addition of labeled α-ketoisovalerate to the growth medium of a protein-expressing host organism has evolved into a versatile tool to achieve concomitant incorporation of specific isotopes into valine- and leucine-residues. The resulting target proteins represent excellent probes for protein NMR analysis. However, as the side-chain resonances of these residues emerge in a narrow spectral range, signal overlap represents a severe limitation in the case of high-molecular-weight NMR probes. We present a protocol to eliminate leucine labeling by supplying the medium with unlabeled α-ketoisocaproate. The resulting spectra of a model protein exclusively feature valine signals of increased intensity, confirming the method to be a first example of independent valine and leucine labeling employing α-ketoacid precursor compounds.

Full text of DOI:10.1007/s10858-013-9786-y

J Biomol NMR (2013) 57:205–209
DOI 10.1007/s10858-013-9786-y
COMMUNICATION
Independent valine and leucine isotope labeling in Escherichia coli
protein overexpression systems
•
Roman J. Lichtenecker Katharina Weinhaupl
•
¨
•
Lukas Reuther Julia Schorghuber
•
¨
Walther Schmid Robert Konrat
•
Received: 6 September 2013 / Accepted: 20 September 2013 / Published online: 28 September 2013
Ó Springer Science+Business Media Dordrecht 2013
Abstract The addition of labeled a-ketoisovalerate to the
growth medium of a protein-expressing host organism has
evolved into a versatile tool to achieve concomitant
incorporation of specific isotopes into valine- and leucine-
residues. The resulting target proteins represent excellent
probes for protein NMR analysis. However, as the side-
chain resonances of these residues emerge in a narrow
spectral range, signal overlap represents a severe limitation
in the case of high-molecular-weight NMR probes. We
present a protocol to eliminate leucine labeling by
supplying the medium with unlabeled a-ketoisocaproate.
The resulting spectra of a model protein exclusively feature
valine signals of increased intensity, confirming the method
to be a first example of independent valine and leucine
labeling employing a-ketoacid precursor compounds.
hydrophobic cores and on interaction surfaces of proteins
(Tugarinov and Kay 2005a). Progress in NMR methodology
led to a considerable shift of the upper size limit of NMR-
attainable protein structures, especially with the advent of
heteronuclear-edited NOESY techniques (Morshauser and
Zuiderweg 1999; Diercks et al. 1999) or the methyl trans-
verse relaxation optimized spectroscopy (methyl-TROSY;
Ollerenshaw et al. 2003). Concomitant to novel NMR-
techniques, robust protocols for selective labeling of iso-
leucines, valines, leucines, alanines and methionines using
cell-based expression systems have been designed and were
summarized previously (Ohki and Kainosho 2008; Ruschak
and Kay 2010). These methods have to fulfil the criteria of
providing defined isotope patterns at distinct atomic posi-
tions of the macromolecular assembly to suit the NMR
experiment applied. At the same time, these stable isotopes
should be implemented with reasonable synthetic and eco-
nomic efforts. A straightforward approach to generate
methyl labeled proteins is given by the use of an over-
expressing, plasmid transformed Escherichia coli strain.
The bacteria can be grown in minimal medium, supple-
mented with labeled amino acids or metabolic precursors
thereof (Hoogstraten and Johnson 2008). The application of
a-ketoacids in this context avoids the costly challenge of
isotope labeled amino acid synthesis and thus the intro-
duction of the stereocenter at the Ca position using synthetic
chemistry. This approach additionally enables facile intro-
duction of nitrogen-15 by using 15N-ammonium chloride as
a source of label. Combinations of 2-ketobutyrate and
2-ketoisovalerate as central metabolites in the biosynthesis
of isoleucine, leucine and valine have been shown to yield
appropriate isotope methyl patterns (often referred to as
ILV-labeling; Tugarinov et al. 2006; Sprangers et al. 2007).
Synthetic routes to access various labeled 2-ketobutyrates
and 2-ketoisovalerates have been published (Gardner and
Keywords Isotope labeling ꢀ Protein NMR ꢀ Protein
overexpression ꢀ Leucine ꢀ Valine
NMR correlations of side-chain methyl groups represent
valuable reporters for protein structure determination, due to
favorable relaxation properties and high abundance in
Electronic supplementary material The online version of this
article (doi:10.1007/s10858-013-9786-y) contains supplementary
material, which is available to authorized users.
¨
R. J. Lichtenecker (&) ꢀ J. Schorghuber ꢀ W. Schmid
Institute of Organic Chemistry, University of Vienna,
¨
Wahringerstr. 38, 1090 Vienna, Austria
e-mail: roman.lichtenecker@univie.ac.at
¨
K. Weinhaupl ꢀ L. Reuther ꢀ R. Konrat
Department of Structural and Computational Biology, Max F.
Perutz Laboratories, University of Vienna, Dr. Bohr-Gasse 9,
1030 Vienna, Austria
123

206
J Biomol NMR (2013) 57:205–209
O
O
O
COO-
OH
N
COO-
OH
N
N
COO-
N
O
a, b)
c)
COOtBu
O
O
COOH
COO
tBu
13CH3
COO-
COO-
(S)-2-Aceto-2-hydroxy-
butanoate
(S)-2-Acetolactate
3
1
2
2-Ketobutanoate
Pyruvate
O
d, e)
COONa
13CH3
O
O
COO-
4
COO-
COO-
O
H
H
O
N
O
O
N
2-Ketoisovalerate
(S)-3-Methyl-2-keto-
pentanoate
2-Ketoisocaproate
g)
f)
O
O
N
COONa
N
H
H
7
5
6
L-Valine
L-Isoleucine
L-Leucine
Fig. 2 Synthesis of sodium-(13CH3)-2-ketoisovalerate 4 and sodium-
2-ketoisocaproate 7. Conditions a tBuOH, methanesulfonyl chloride,
pyridine, THF, -5 °C - RT, 12 h, 64 %, b N,N0-dimethylhydrazine,
diethyl ether, 12 h, 73 %, c lithium diisopropylamide, -78 °C, 1 h,
then 13C-methyliodide, THF, 2 h, 83 %, d 1N HCl, THF, 3 h, 77 %,
e HClg, Et2O/CH2Cl2 1:1, 0 °C - RT, then 1 M NaOH until pH = 8,
lyophilize, 95 %, f ethanolamine, isobutyraldehyde, H2O, 110 °C,
9 h, 74 %, g NaOH (20 %), 100 °C, 5 h, 81 %
Fig. 1 Precursor compounds in ILV-labeling using E. coli expression
systems
Kay 1997; Hajduk et al. 2000; Gross et al. 2003; Lichte-
necker et al. 2004; Schedlbauer et al. 2008) and some of
them are commercially available. ILV-labeling has been
successfully applied to study very large proteins or protein
complexes, including the malate synthase G (82 kDa)
(Tugarinov and Kay 2005b), the aspartate transcarbamoyl-
ase (306 kDa) (Velyvis et al. 2007), the 20S core particle
proteasome (670 kDa) (Sprangers and Kay 2007) or the
translocase motor SecA (204 kDa) (Gelis et al. 2007).
Despite these impressive applications, extensive signal
overlap and restricted sensitivity remain the most profound
challenges in studies of high molecular weight NMR probes.
The limited resolution of complex spectra often results in
incomplete signal assignment and a significant loss of
structural restraints.
established 2-ketoisovalerate as a metabolic precursor in
the presence of unlabeled 2-ketoisocaproate to saturate the
metabolic branch of leucine biosynthesis. To validate our
hypothesis of independent isotope incorporation into ali-
phatic residues we used a His-Tag GB1 66 amino acid
protein as a model system and tested the appropriate
quantities of additives in the E. coli growth media with the
intention to maximize valine-, while at the same time
repressing leucine-labeling.
Sodium-(13CH3)-2-ketoisovalerate 4 was prepared sim-
ilar to the synthetic routes described previously (Fig. 2;
Hajduk et al. 2000; Lichtenecker et al. 2004). As a first
step, a-ketobutanoic acid 1 was converted to the corre-
sponding dimethylhydrazone-tert-butyl ester 2. Subsequent
methylation using iodo(13C)methane gave compound 3,
which was treated with 1N HCl to cleave the hydrazone to
the a-ketoester. Final hydrolysis was performed with gas-
eous HCl, followed by lyophilization from an aqueous
solution (pH = 8) to a white powder of product 4 in a total
yield of 28 % from 1 and 40 % from 13CH3I. Sodium
2-ketoisocaproate 7 was synthesized as reported (Lichte-
necker et al. 2013) via hydrolysis of isobutylidenehydan-
toin 6 in a total yield of 60 %.
Figure 1 gives an overview of the main metabolic
intermediates in ILV-labeling. Selective isoleucine labeling
can easily be realized using either 2-ketobutyrate (Goto
et al. 1999) or (S)-2-aceto-2-hydroxybutanoate (Ruschak
et al. 2010; Ayala et al. 2012) as unique precursor com-
pounds. On the other hand, leucines and valines are labeled
simultaneously in the presence of 2-ketoisovalerate (Goto
et al. 1999) or (S)-2-acetolactate (Gans et al. 2010). Sep-
aration of Val- and Leu-residue methyl resonances has
been realized by suitable NMR experiments which allow
for selective residue type editing (Hu et al. 2012). We
implied a more general approach based on the introduction
of 2-ketoisocaproate as a new precursor for selective leu-
cine labeling. This methodology was successfully applied
and resulted in selective labeling without interfering with
the valine metabolic pathway (Lichtenecker et al. 2013). In
the present study, we have further developed our approach
of independent ILV-labeling to exclusive incorporation of
isotopes into valine residues by applying the well-
His-tagged-GB1 protein was expressed and purified
according to a protocol described previously (Lichtenecker
et al. 2013). In brief, BL21-pLysS cells, containing the cor-
responding expression plasmid were grown in LB medium at
37 °C. Dilution in 15NH4Cl (1 g/L) and glucose (4 g/L)
supplemented M9 medium (250 mL) was followed by
addition of sodium-(13CH3)-2-ketoisovalerate 4 (100 mg/L)
123

J Biomol NMR (2013) 57:205–209
207
Fig. 3 a Relative 13C-HSQC signal intensities of valine—(V1–V8)
and leucine methyl groups (L1–L6) of His-tagged GB1 expressed in
presence of sodium-(13CH3)-2-ketoisovalerate 4 (100 mg/L) and
different concentrations of sodium-2-ketoisocaproate 7 (0–300 mg/
L). b Sequence of His-tag-GB1. c 13C-HSQC of His-tag-GB1
expressed in presence of sodium-(13CH3)-2-ketoisovalerate
4
(100 mg/L) showing valine (blue) and leucine (red) residues. d 13C-
HSQC of His-tag-GB1 expressed in presence of sodium-(13CH3)-2-
ketoisovalerate 4 (100 mg/L) and sodium-2-ketoisocaproate 7
(100 mg/L) exclusively showing valine residues
and sodium-2-ketoisocaproate 7 (0, 50, 100, 200 and
300 mg/L) 1 h prior to induction. Addition of Isopropyl b-D-
1-thiogalactopyranoside (IPTG) to a final concentration of
0.8 mM at an OD600 of 0.8 induced protein overexpression.
After 16 h of expression at 30 °C, the cells were collected by
centrifugation at 6,000 rpm for 15 min and suspended in ice-
cold PBS. Subsequently, the cells were passed through a
French press and the cell lysate was cleared by centrifugation
at 18,000 rpm for 30 min. Purification was performed by
loading the supernatant onto a Ni2? HiTrap 5 mL affinity
column, washed with PBS and eluted with high imidazole
buffer (0.5 M). The protein containing fractions were col-
lected, concentrated and further purified by gel filtration
over a Superdex75 column. 15N-HSQC and 13C-HSQC
experiments were measured at a protein concentration of
0.5 mM in 20 mM Tris-buffer (pH 7.4, 50 mM NaCl, sup-
plemented with 10 % 2H2O). In order to compare 13C
incorporation levels, the 13C-HSQC signal intensities were
divided by the sum of 10 signal intensities of the corre-
sponding 15N-HSQC spectrum.
Figure 3 shows the 13C-HSQC intensities of the 66 amino
acid model peptide His-tagged GB1, expressed in presence
of sodium-(13CH3)-2-ketoisovalerate 4 (100 mg/L) and
increasing amounts of sodium-2-ketoisocaproate 7 (50-
300 mg/L) relative to the control protein, which was grown
in a medium devoid of unlabeled compound 7 (Fig. 3a
black). The results depicted in Fig. 3a evidence decreasing
Leu-signal intensities as a function of the precursor 7
123

208
J Biomol NMR (2013) 57:205–209
Fig. 4 Approaches for Leu/
Val-labeling. a Concomitant
Leu/Val labeling, b selective
Leu labeling, c selective Val
labeling
concentration in the overexpression medium. Complete
extinction of the methyl signals resulting from the three
leucines present in the protein sequence was reached at
100 mg/L of sodium-2-ketoisocaproate 7 (Fig. 3c, d). Under
these conditions, a significant increase of the valine-methyl-
resonance intensities was observed, proving the transfor-
mation of sodium-(13CH3)-2-ketoisovalerate 4 to Val-resi-
dues to be more effective in presence of unlabeled sodium
2-ketoisocaproate 7 (Fig. 3a). Higher concentrations of
unlabeled 2-ketoisocaproate ([100 mg/L) led to a slight
decay of valine signal intensity. Therefore, the optimal ratio
of precursor compounds in the minimal medium was iden-
tified as a 1:1 mixture (100 mg/L each) since these condi-
tions resulted in complete deletion of leucine signals and a
maximum increase of valine signal intensities.
complete our efforts of separating valine and leucine
labeling, which are summarized in Fig. 4. Protein overex-
pression in presence of labeled 2-ketoisovalerates leads to
concomitant labeling of Val- and Leu-residues (Fig. 4a;
Goto et al. 1999; Lichtenecker et al. 2004). Labeled
2-ketoisocaproates can be used as selective leucine pre-
cursors without any cross-labeling to valine residues
(Fig. 4b; Lichtenecker et al. 2013). Finally, a combination
of labeled 2-ketoisovalerates and unlabeled 2-ketoisocap-
roates results in exclusive valine labeling (Fig. 4c). We
want to point out that in using our methodology different
isotope patterns for Val- and Leu-residues in the same
protein sample become available for the first time. We are
currently working on patterns that will allow for the dis-
crimination of valine-derived through-space connectivities
from those derived from leucines. The resulting spectra
simplification will increase the number of available dis-
tance restraints in the case of large proteins significantly.
When it comes to large proteins, a high deuteration
grade may be required in order to decrease transverse
relaxation rates and improve sensitivity. In this case, fully
2H-labeled 2-ketoisocaproate may be applied. This deu-
terated precursor is synthetically accessible via a route
similar to the one we described recently (Lichtenecker
et al. 2013). In summary, we could show that addition of
unlabeled sodium 2-ketoisocaproate 7 to an E. coli over-
expression medium efficiently saturates the leucine meta-
bolic pathway. The metabolic conversion of labeled
sodium 2-ketoisovalerate 4 to leucine was thus blocked,
leading to exclusive valine labeling from precursor 4. At
the same time, signal intensities derived from valine-
methyl groups were significantly increased. These results
Acknowledgments This work was supported by grants from the
Austrian Science foundation FWF P20549, P22125 and W1221B03.
References
Ayala I, Hamelin O, Amero C, Pessey O, Plevin MJ, Gans P,
Boisbouvier J (2012) An optimized isotope labelling strategy of
isoleucine-c2 methyl groups for solution NMR studies of high
molecular weight proteins. Chem Commun 48:1434–1436
123

J Biomol NMR (2013) 57:205–209
209
Diercks T, Coles M, Kessler H (1999) An efficient strategy for
assignment of cross-peaks in 3D heteronuclear NOESY exper-
iments. J Biomol NMR 15:177–180
Morshauser RC, Zuiderweg ERP (1999) High-resolution four-dimen-
sional HMQC-NOESY-HSQC spectroscopy. J Magn Reson
139:232–239
Gans P, Hamelin O, Sounier R, Ayala I, Asucion Dura M, Amero CD,
Noirclerc-Savoye M, Franzetti B, Plevin MJ, Boisbouvier J
(2010) Stereospecific isotopic labeling of methyl groups for
NMR spectroscopic studies of high-molecular-weight proteins.
Angew Int Ed 49:1958–1962
Ohki S, Kainosho M (2008) Stable isotope labeling methods for
protein NMR spectroscopy. Prog Nucl Magn Reson Spectrosc
53:208–226
Ollerenshaw JE, Tugarinov V, Kay LE (2003) Methyl TROSY:
explanation and experimental verification. Magn Reson Chem
41:843–852
Ruschak A, Kay LE (2010) Methyl groups as probes of supra-
molecular structure, dynamics and function. J Biomol NMR
46:75–87
Ruschak AM, Velyvis A, Kay LE (2010) A simple strategy for 13C,
1H labeling at the Ile-c2 methyl position in highly deuterated
proteins. J Biomol NMR 48:129–135
Gardner KH, Kay LE (1997) Production and incorporation of 15N,
2
13C, H (1H-d1 methyl) isoleucine into proteins for multidimen-
sional NMR studies. J Am Chem Soc 119:7599–7600
Gelis I, Bonvin AM, Keramisanou D, Koukaki M, Gouridis G,
Karamanou S, Economou A, Kalodimos CG (2007) Structural
basis for signal-sequence recognition by the translocase motor
SecA as determined by NMR. Cell 131:756–769
Goto NK, Gardner KH, Mueller GA, Willis RC, Kay LE (1999) A
robust and cost-effective method for the production of Val, Leu,
Ile (d1) methyl-protonated 15N-, 13C-, 2H-labelled proteins.
J Biomol NMR 13:369–374
Schedlbauer A, Auer R, Ledolter R, Tollinger M, Kloiber K,
Lichtenecker R, Ruedisser S, Hommel U, Schmid W, Konrat R
(2008) Direct methods and residue type specific isotope labeling
in NMR structure determination and model-driven sequential
assignment. J Biomol NMR 42:111–127
Sprangers R, Kay LE (2007) Quantitative dynamics and binding
studies of the 20S proteasome by NMR. Nature 445:618–622
Sprangers R, Velyvis A, Kay LE (2007) Solution NMR of supramo-
lecular complexes: providing new insights into function. Nat
Methods 4:697–703
Tugarinov V, Kay LE (2005a) Methyl groups as probes of structure
and dynamics in NMR studies of high-molecular-weight pro-
teins. ChemBioChem 6:1567–1577
Tugarinov V, Kay LE (2005b) Quantitative 13C and 2H NMR
relaxation studies of the 723-residue enzyme malate synthase G
Gross JD, Gelev VM, Wagner G (2003) A sensitive and robust
method for obtaining intermolecular NOEs between side chains
in large protein complexes. J Biomol NMR 25:235–242
Hajduk PJ, Augeri DJ, Mack J, Mendoza R, Yang JG, Betz SF, Fesik
SW (2000) NMR-based screening of proteins containing 13C-
labelled methyl groups. J Am Chem Soc 122:7898–7904
Hoogstraten CG, Johnson JE Jr (2008) Metabolic labeling: taking
advantage of bacterial pathways to prepare spectroscopically
useful isotope patterns in proteins and nucleic acids. Concepts
Magn Reson A 32A:34–55
Hu W, Namanja AT, Wong S, Chen Y (2012) Selective editing of Val
and Leu methyl groups in high molecular weight protein NMR.
J Biomol NMR 53:113–124
reveal
44:15970–15977
a
dynamic
binding
interface.
Biochemistry
Lichtenecker R, Ludwiczek ML, Schmid W, Konrat R (2004)
Simplification of protein NOESY spectra using bioorganic
precursor synthesis and NMR spectral editing. J Am Chem
Soc 126:5348–5349
Lichtenecker RJ, Coudevylle N, Konrat R, Schmid W (2013)
Selective isotope labelling of leucine residues by using
a-ketoacid precursor compounds. ChemBioChem 14:818–821
Tugarinov V, Kanelis V, Kay LE (2006) Isotope labeling strategies
for the study of high-molecular-weight proteins by solution
NMR spectroscopy. Nat Protoc 1:749–754
Velyvis A, Yang YR, Schachmann HK, Kay LE (2007) A solution
NMR study showing that active site ligands and nucleotides
directly perturb the allosteric equilibrium in aspartate transcar-
bamoylase. Proc Natl Acad Sci USA 104:8815–8820
123