ACIDS PRECURSORS

Technical field of the invention

The invention relates to the field of biochemistry, and more precisely to protein synthesis. More precisely, it relates to amino-acids precursors, to a process for preparing said precursors, and to their use for the cell-free synthesis of isotopic labelled proteins. These proteins are useful for studies of proteins by Nuclear Magnetic Resonance spectroscopy.

Prior art

Nuclear Magnetic Resonance (NMR) can be used to identify and validate therapeutic targets. However, standard NMR techniques based on the uniform enrichment of ¹³ C and ¹⁵ N proteins have a rather low sensitivity and allow to study only proteins with a molecular size less than 30 kDa.

This is due, inter alia, to hydrogen ( ¹ H) which is widely present in proteins. In fact, due to its high gyromagnetic ratio, ¹ H spins give rise to the most intense NMR signal compared to the other observable nuclei by NMR in proteins (i.e. ¹³ C and ¹⁵ N). However, the high density of ¹ H in the protein leads to numerous dipolar interactions between them. These interactions shorten the lifetime of the NMR signals. In other words, the larger the size of the protein, the higher its content of hydrogen atoms, leading to the rapid loss of the NMR signal. This limitation is quite critical because numerous value-added proteins, such as proteins involved in pathological processes, have a molecular size greater than 100 kDa. Thus, in order to study large proteins, various tools have been developed including, in particular, specific isotopic labelling.

Recent advances in the specific isotopic labeling of methyl groups have pushed forward the limits of NMR and have made it possible to study proteins of more than 100 kDa and up to 1 MDa. These advances are based on the reduction of the number of ¹ H atoms in the target protein. This is achieved by expressing the target protein in a deuterated culture medium resulting in a protein with deuterium ( ² H) incorporation in most of ¹ H atom positions. More precisely, in this approach only certain strategic sites are kept with ¹ H atoms, such as methyl groups and aromatic residues, by using their corresponding amino- acids or precursors bearing a specific protonation. In order to study these strategic sites in proteins, they are additionally isotopically labelled, in particular with ¹³ C in case of methyl groups and aromatic residues. This isotopic labeling and selective protonation strategy improves the sensitivity and the resolution of protein NMR analysis. Regarding hydrogen isotopes, ¹ H is NMR active, while ² H (also abbreviated as "D") is NMR inactive. Thus, any protein that contains ¹ H can be rendered "invisible" in the hydrogen NMR spectrum by replacing all the ¹ H by ² H.

Regarding CH ₃ , there are six naturally occurring amino-acids carrying a methyl group: alanine (Ala), isoleucine (lie), leucine (Leu), methionine (Met), threonine (Thr) and valine (Val). The amino acids lie, Leu and Val each carry two methyl groups. Methyl groups are the main constituents of the hydrophobic cores of proteins where catalytic sites are located. Furthermore, in supramolecular systems, CH ₃ containing amino-acids can represent up to 40% by number of the total of amino-acids present in proteins and up to 50% of their catalytic pockets (hydrophobic cores). So, the specific labeling of methyl groups in proteins may be sufficient to answer some biological questions such as the interaction of the protein of interest with drug candidates. In addition, methyl groups carry three hydrogen atoms which revolve around a symmetrical axis. This characteristic gives the NMR signal an intensity three times greater than that of the other ¹ H atoms (i.e. CH or NH).

The use of ¹³ C labelled methyl groups ( ¹³ C ¹ H ₃ ) was introduced in 1997 by Kay and al. (see Gardner & Kay, "The use of ² H, ¹³ C, ¹⁵ N multidimensional NMR to study the structure and dynamics of proteins", Ann. Rev. of Biophysics and Biomolecular Structure (1998) vol. 27 p 357-406) who proposed a protocol to specifically label the methyl group ( ¹³ CH ₃ ) in the deltal position of isoleucine (lle-51 ) using its precursor ( 2-ketobutyrate) and then use this specific labelled isoleucine precursor to express a target protein in bacteria that grow in a completely deuterated culture medium. The specific isotopic labeling ( ¹³ C ¹ H ₃ ) of the methyl groups of the amino-acids present in the target protein can be achieved by the addition of the amino-acid labelled with ¹³ C ¹ H ₃ , or by the addition of a suitably labelled precursor of the amino acid.

The specifically labelled ( ¹³ C ¹ H ₃ ) amino-acid is added to the fully deuterated culture medium before the induction of expression of the protein. Thereby, the ¹³ C ¹ H ₃ residue is incorporated into the synthesized protein. This approach has been applied to residues Ala, Thr and Met, but not to lie, Leu and Val- ¹³ C ¹ H ₃ because their very high cost, as described in Table 1 of the review of Kerfah et al., "Methyl-specific isotopic labelling : a molecular tool box for solution NMR studies of large proteins", Current Opinion in Structural Biology (2015) vol. 32 p1 13-122. In another approach to the synthesis of ¹³ C ¹ H ₃ labelled amino-acids, their labelled precursors are added prior to the induction of the expression of the protein and are internalized by the bacterium, and then metabolically converted into amino-acids carrying the targeted labeling. Residues lie (Gardner & Kay, "Production and incorporation of ¹⁵ N, ¹³ C, ² H (1H- 51 Methyl) Isoleucine into proteins for multidimensional NMR studies", J. Am. Chem. Soc. (1997) vol. 1 19 p 7599-7600), Leu and Val have been labelled in this way; this approach is fairly affordable (see Table 1 of the review of Kerfah et al., previously cited).

Specific ¹³ C labeling of aromatic residues (i.e. tyrosine, phenylalanine and tryptophan) has also been used as a complementary strategy to obtain more versatile NMR probes. Indeed, they may represent up to 15% in number of the amino-acid protein sequence (S. Miller et al., "Interior and Surface of Monomeric Proteins", J. Mol. Biol. (1987) vol.196 p 641 -656) and affect considerably the spatial packing and dynamics within the protein hydrophobic core (see Miyanoiri et al., "Alternative SAIL-Trp for robust aromatic signal assignment and determination of the conformation by intra-residue NOEs", J. Biomol. NMR (201 1 ) vol.51 p 425-435).

Their labeling, in protein produced in E. coli, relies on the same approaches as the methyl containing residues, using either the ¹³ C labelled amino-acid or its precursor (see Wang et al., "A Method for Assigning Phenylalanines in Proteins", J. Am. Chem. Soc. (1999) vol. 121 p 161 1 -1612; Raap et al., "Synthesis of Isotopically Labelled L-Phenylalanine and L- Tyrosine", Eur. J. Org. Chem. (1999) p 2609-2621 ; Rajesh et al., "A novel method for the biosynthesis of deuterated proteins with selective protonation at the aromatic rings of Phe, Tyr and Trp", Journal of Biomolecular NMR (2003) vol.27 p 81-86; Rasia et al., "Selective Isotopic Unlabeling of Proteins Using Metabolic Precursors: Application to NMR Assignment of Intrinsically Disordered Proteins", ChemBioChem (2012) vol.13 p 732-739; Kasinath et al., "A ¹³ C Labeling Strategy Reveals a Range of Aromatic Side Chain Motion in Calmodulin", J. Am. Chem. Soc. (2013) vol.135 p 9560-9563 or Lichtenecker et al., "Selective Isotope Labelling of Leucine Residues by Using a-Ketoacid Precursor

Compounds", ChemBioChem (2013) vol.14 p818-821 ).

To date, the technology for selective isotopic labeling of methyl groups and aromatic residues using precursors is mainly operational for bacterial systems, in particular E. coli, or for yeast systems. However, many challenging proteins such as membrane proteins, are difficult, or even impossible, to be expressed in bacterial systems.

Therefore, the production of these labelled proteins requires, according to their nature and their use, other expression systems, such as cell-free systems. Cell-free synthesis is a cost-effective approach to produce proteins. Furthermore, cell-free synthesis is required when the target protein is, inter alia, toxic to the cellular expression system, susceptible to proteolytic degradation in vivo or requires additional partners including co-chaperons and co-factors to fold properly. The so-called "cell-free" approach consists in an in vitro synthesis of the target protein by exploiting the protein machinery of transcription and translation (e.g. ribosomes, transcription factors, etc.) extracted from eukaryotic or prokaryotic cells. The cell extracts containing the enzymes are recovered after removal of the DNA / RNA of the original organism, then supplemented with amino-acids and other essential compounds such as energy sources, and finally used for the expression of the target protein.

Due to the absence of a biological membrane, the cell-free environment is an open system, offering the possibility of adding at any time compounds such as cofactors, ligands and stabilizers in order to improve the synthesis of proteins.

Cell-free systems have already been used for obtaining isotopically labelled proteins.

Some labelled proteins have been obtained by the Stereo-Array Isotope Labeling (SAIL) technology (see. Kainosho et al., "Optimal isotope labelling for NMR protein structure determinations", Nature (2006) vol. 440 p.52-57). EP 1 457 482 describes a labelled protein obtained by the SAIL technology from cell-free protein expression using chemically and enzymatically synthesized stable isotopically labelled amino-acids in which all hydrogen atoms except one of them in a methyl group are deuterated, and all carbon and nitrogen atoms are replaced respectively by ¹³ C and ¹⁵ N. This labeling allows to reduce the content of the proteins in ¹ H by a factor of 2 and improves the intensity and the resolution of the NMR signals of a relatively large protein of 50 kDa. However, its use is limited by the high price of SAIL amino-acids.

Linser et al. describe the cell-free expression of a protein perdeuterated and isotopically labelled on methyl groups of isoleucine, valine or/and leucine (see Linser et al., "Selective methyl labeling of eukaryotic membrane proteins using cell-free expression", J. Am. Chem.

Soc. (2014) vol.136, p 1 1308-1 1310). This publication describes a protocol to label the methyl groups of lie, Leu, and Val of a model protein; lie is specifically labelled in the 51 position. After the production in E. coli of an easy to express and labelled protein on lie, Leu and Val using standard protocols, this labelled protein is then purified and hydrolyzed to generate the labelled amino-acids in order to obtain specifically labelled amino-acids.

This pool of amino-acids is then added to a cell extract in order to express the protein of interest marked on the lie in 51 position, Leu and Val. This approach represents an interesting solution for the labeling of the methyl groups of proteins expressed in Cell-free media. However, it implies a step of producing proteins in bacteria, followed by an inefficient hydrolysis with a yield around 20%, a purification step of amino-acids and then a step of cell-free synthesis. This method remains complicated, time-consuming, expensive and difficult to apply on a large scale. The objective of the present invention is to at least partially overcome these disadvantages and to propose a simple way to express specifically labelled proteins, especially specifically labelled proteins which can be difficult, or even impossible, to be expressed in bacterial systems, and in particular (but not exclusively) of size≥100 kDa.

Another objective of the invention is to propose an extract for use in a cell-free synthesis comprising at least one partially deuterated and isotopic labelled precursor of amino-acids, which ensures a reliable elaboration of partially deuterated and isotopic labelled protein, while being relatively easy to prepare.

Another objective of the invention is to propose a kit for use in a cell-free protein synthesis comprising this extract, which is simple to use.

Subject matter

According to the invention, the problems are solved by a process of cell-free synthesis of a target protein, comprising at least the following steps:

(a) providing at least one precursor of an amino-acid, a cell-free extract comprising means for transforming said precursor of an amino-acid into the corresponding amino-acid, at least one tRNA for integrating said amino- acid into the corresponding target protein,

in which the at least one precursor of amino-acids is an alpha-keto acid, with the proviso that 2-ketobutyric acid is excluded,

(b) mixing the compounds provided in step (a) to obtain a mixture, (c) adding to said mixture obtained in step (b) a DNA or a mRNA coding to said target protein in order to produce said target protein.

Said tRNA is selected or designed as a means for producing said target protein (said target protein being called the protein that "corresponds" to said tRNA). In particular, said tRNA is selected or designed as a means for integrating said amino-acid, isotopically labelled or not, into the corresponding target protein.

Advantageously, said means for transforming said precursor of an amino-acid into the corresponding amino-acid comprises a branched-chain aminotransferase.

Advantageously, after the step (c) said target protein is isolated and purified. Advantageously, the at least one precursor of amino-acid is at least partially isotopically labelled allowing to produce the corresponding labelled target protein.

Advantageously, the at least one precursor of amino-acid comprises at least one atom isotopically labelled with a stable isotope ¹³ C and / or D, wherein the isotopic enrichment rate is higher than 80% by number.

Advantageously, the amino-acids obtained in step (b) are naturally occurring or non- naturally occurring amino-acids.

Advantageously, in step (a), the cell-free extract is obtained from biological matter selected from the group formed by: E coli (S12 to S100, preferably S30), wheat germ, insect cells, rabbit reticulocyte, HeLa cell and Chinese hamster ovary (CHO) cell line. According to one embodiment of the invention, in step (a), said precursor of amino-acid is specifically labelled with ¹³ C on at least one methyl group or one atomic position of aromatic ring.

According to one embodiment of the invention, in step (a), said precursor of amino-acid comprises at least one, and especially one methyl group which is labelled either ¹³ CH ₃ , ¹³ CHD ₂ , or ¹³ CH ₂ D with an isotopic enrichment rate for ¹³ C and/or D higher than 80% by number.

According to one embodiment of the invention, in step a), said precursor of amino-acid is perdeuterated and comprises one methyl group which is labelled either ¹³ CH ₃ , ¹³ CHD ₂ , or

¹³ CH ₂ D with an isotopic enrichment rate for ¹³ C and/or D higher than 80% by number.

According to one embodiment of the invention, in step a), said precursor of amino-acids comprises one or more compounds chosen from :

o 3 - Methyl - 2-oxopentanoate (MOP),

o 3 - Methyl - 2 - oxobutanoic acid (KIV),

o 4 - Methyl - 2 - oxopentanoic acid (4MOP),

o 2 - oxo - 3 - phenyl propanoate (PP),

o 4 - hydroxy phenyl pyruvate (HPP). Preferably, in step (a), said at least one precursor of amino-acid has at least one atom isotopically labelled with a stable isotope ¹³ C and / or D, and is characterized by an enrichment ratio higher than 80% by number, and is selected from the group formed by: o 3 - Methyl - 2-oxopentanoate (MOP),

o 3 - Methyl - 2 - oxobutanoic acid (KIV),

o 4 - Methyl - 2 - oxopentanoic acid (4MOP),

o 2 - oxo - 3 - phenyl propanoate (PP),

o 4 - hydroxy phenyl pyruvate (HPP).

Preferably, in step (a), said at least one precursor of amino-acid is selected from the group formed by:

o (S) 3-( ¹³ CH ₃ )2-oxopentanoate,

o (S) 3-(CD ₃ )- 4,4-(D ₂ )- 4-( ¹³ CH ₃ ) - 2-oxopentanoate (MOP- 51 ),

o (S) 1 ,2,3,4,5-( ¹³ C ₅ )-3-( ¹³ CD ₃ )-4,4-(D ₂ )- 2-oxopentanoate,

o (S) 1 ,2,3,4,5-( ¹³ C ₅ )- 3-(CD ₃ )-, 4,4-(D ₂ )-2-oxopentanoate,

o (S) 3,4,4 -(D ₃ )- 3,4-( ¹³ CH ₃ )-2-oxopentanoate,

o (S) 3,4,4 -(D ₃ )- U-( ¹³ C)-2-oxopentanoate,

o (S) 3,4,4,5,5,5-(D ₆ )-3-( ¹³ CH ₃ ) - 2-oxopentanoate (MOP- γ2),

o (S) 1 ,2,3,4,5-( ¹³ C ₅ )- 3,4,4,5,5,5-(D ₆ )-3-( ¹³ CH ₃ )-2-oxopentanoate,

o (S) 1 ,2,3-( ¹³ C ₂ )-3,4,4,5,5,5-(D ₆ )- 3-( ¹³ CH ₃ )- 2-oxopentanoate,

o (S) 5-( ¹³ C)-2-oxopentanoate,

o (S) 3-(CD ₃ ) - 3(D - 3-( ¹³ CH ₃ ) - 2-oxobutanoic acid (KIV - ProS),

o (R) 3-(CD ₃ ) - 3(D - 3-( ¹³ CH ₃ ) - 2-oxobutanoic acid (KIV - ProR),

o (S) 4-(CD ₃ ) - 4(D - 4-( ¹³ CH ₃ ) - 2-oxopentanoic acid (4MOP - ProS),

o (R) 4-(13CH ₃ ) - 4(D - 4-(CD ₃ ) - 2-oxopentanoic acid (4MOP- ProR),

o U-( ¹³ C)-2-Oxo-3-phenylpropanoate,

o 3-phenyl-3,3-(D ₂ )-2-oxopropanoate,

o 3-(2,3,5,6-D ₄ -4-( ¹³ C,H))phenyl-3,3-(D ₂ )-2-oxopropanoate,

o 3-(2,4,6-D ₃ -3,5-( ¹³ C,H)2)-phenyl-3,3-(D ₂ )-2-oxopropanoate,

o 3-(3,4,5-D ₃ -2,6-( ¹³ C,H)2)-phenyl-3,3-(D ₂ )-2-oxopropanoate,

o U-( ¹³ C)-2-Oxo-3-(4-hydroxy)phenylpropanoate,

o (4-hydroxy)phenyl-3,3-(D ₂ )-2-oxopropanoate,

o 3-(2,6-D ₂ -3,5-( ¹³ C,H)2-4-hydroxy)-phenyl-3,3-(D ₂ )-2-oxopropanoate,

o 3-(3,5-D ₂ -2,6-( ¹³ C,H)2-4-hydroxy)-phenyl-3,3-(D ₂ )-2-oxopropanoate. Another subject-matter of the invention is a supplemented extract for use in a cell-free synthesis according to the invention, made in a process comprising at least the following steps:

(a) providing at least one precursor of amino-acid, a cell-free extract comprising means for transforming said precursor of amino-acid into corresponding amino-acid, at least one tRNA for integrating said amino- acid into corresponding target protein,

in which the at least one precursor of amino-acids is an alpha-keto acid, with the proviso that 2-ketobutyric acid is excluded,

(b) mixing the compounds provided in step (a) resulting in a mixture.

According to the invention, the cell-free extract is supplemented with at least one precursor. Preferably, in a supplemented extract, said means for transforming said precursor of an amino-acid into the corresponding amino-acid comprises a branched-chain aminotransferase.

Another subject-matter of the invention is a kit for producing by cell-free synthesis a target protein, said kit comprising :

- at least one precursor of amino-acid in which the at least one precursor of amino- acid is an alpha-keto acid, with the proviso that 2-ketobutyric acid is excluded,

- a cell-free extract comprising means for transforming said precursor of amino-acid into corresponding amino-acid,

- and at least one tRNA for integrating said amino-acid into corresponding target protein.

Another subject-matter of the invention is a kit for producing by cell-free synthesis a target protein as described above and wherein said precursor of amino-acids are chosen from labelled:

o (S) 3-( ¹³ CH ₃ )2-oxopentanoate,

o (S) 3-(CD ₃ )- 4,4-(D ₂ )- 4-( ¹³ CH ₃ ) - 2-oxopentanoate (MOP- 51 ),

o (S) 1 ,2,3,4,5-( ¹³ C ₅ )-3-( ¹³ CD ₃ )-4,4-(D ₂ )- 2-oxopentanoate,

o (S) 1 ,2,3,4,5-( ¹³ C ₅ )- 3-(CD ₃ )-, 4,4-(D ₂ )-2-oxopentanoate,

o (S) 3,4,4 -(D ₃ )- 3,4-( ¹³ CH ₃ )-2-oxopentanoate,

o (S) 3,4,4 -(D ₃ )- U-( ¹³ C)-2-oxopentanoate,

o (S) 3,4,4,5,5,5-(D ₆ )-3-( ¹³ CH ₃ ) - 2-oxopentanoate (MOP- γ2), o (S) 1 ,2,3,4,5-( ¹³ C ₅ )- 3,4,4,5,5,5-(D ₆ )-3-( ¹³ CH ₃ )-2-oxopentanoate, o (S) 1 ,2,3-( ¹³ C ₂ )-3,4,4,5,5,5-(D ₆ )- 3-( ¹³ CH ₃ )- 2-oxopentanoate,

o (S) 5-( ¹³ C)-2-oxopentanoate,

o (S) 3-(CD ₃ ) - 3(D - 3-( ¹³ CH ₃ ) - 2-oxobutanoic acid (KIV - ProS),

o (R) 3-(CD ₃ ) - 3(D - 3-( ¹³ CH ₃ ) - 2-oxobutanoic acid (KIV - ProR),

o (S) 4-(CD ₃ ) - 4(D - 4-( ¹³ CH ₃ ) - 2-oxopentanoic acid (4MOP - ProS),

o (R) 4-(13CH ₃ ) - 4(D - 4-(CD ₃ ) - 2-oxopentanoic acid (4MOP- ProR),

o U-( ¹³ C)-2-Oxo-3-phenylpropanoate,

o 3-phenyl-3,3-(D ₂ )-2-oxopropanoate,

o 3-(2,3,5,6-D ₄ -4-( ¹³ C,H))phenyl-3,3-(D ₂ )-2-oxopropanoate,

o 3-(2,4,6-D ₃ -3,5-( ¹³ C,H)2)-phenyl-3,3-(D ₂ )-2-oxopropanoate,

o 3-(3,4,5-D ₃ -2,6-( ¹³ C,H)2)-phenyl-3,3-(D ₂ )-2-oxopropanoate,

o U-( ¹³ C)-2-Oxo-3-(4-hydroxy)phenylpropanoate,

o (4-hydroxy)phenyl-3,3-(D ₂ )-2-oxopropanoate,

o 3-(2,6-D ₂ -3,5-( ¹³ C,H)2-4-hydroxy)-phenyl-3,3-(D ₂ )-2-oxopropanoate,

o 3-(3,5-D ₂ -2,6-( ¹³ C,H)2-4-hydroxy)-phenyl-3,3-(D ₂ )-2-oxopropanoate.

Another subject-matter of the invention is a method for analysing a target protein by NMR spectroscopy comprising the step of obtaining said target protein according to the process described above.

Description of the figures

Figure 1 shows a metabolic pathway of valine and Isoleucine biosynthesis in E.coli, in particular, the incorporation pathway of HMOB (2-Hydroxy-2-Methyl-3-OxoButanoate or 2- acetolactate) and AHB (2-Aceto-2-hydroxybutanoate) into the valine and isoleucine residues in E. coli. The enzymes responsible for catalyzing each reaction are indicated by their EC (Enzyme Commission) number, wherein EC 1 .1 .1.86 represents ketol-acid reductoisomerase (hereinafter referred to as KARI), EC 4.2.1 .9 represents dihydroxy-acid dehydratase (hereinafter referred to as DHAD) and EC 2.6.1.42 represents branched- chain amino-acid aminotransferase (BCAT).

Figure 2 shows fluorescence results on a cell-free Green Fluorescent Protein (GFP) synthesis using precursors. GFP was used as a model protein to confirm the ability of the cell-free system to synthesize proteins from different kind of amino-acids precursors.

Figure 2 (A) shows fluorescence results of a cell-free GFP synthesis using commercial precursors, i.e. starting precursors in the present invention, usually used as precursors in E. coli. We can see bars representing fluorescence intensity of GFP obtained by a cell- free synthesis (determined by fluorescence measurements) without amino-acids (lane 1 ); with a mix of 20 amino-acids (lane 2); 19 amino-acids without lie (lane 3); 19 amino-acids without lie and with commercial oketobutyric acid hereinafter referred to as a-KB (lane 4); 19 amino-acids without Val (lane 5); 19 amino-acids without Val and with commercial HMOB (lane 6).

After cell-free GFP synthesis, the fluorescence intensity of the GFP obtained is measured and compared to the fluorescence intensity of GFP obtained by cell-free synthesis from a mix of 20 amino-acids. This last measure corresponds to a percentage of fluorescence of 100% in the figure 2 (see figure 2 (A), lane 2).

As shown in Figure 2 (A) fluorescence intensity of GFP obtained from a mix of 19 amino- acids without Val and with commercial HMOB (line 6) or from a mix of 19 amino-acids without lie and with commercial a-KB (line 4) was close to the baseline, indicating that the GFP protein is not produced and consequently that these precursors are not converted to amino-acids by the cell-free system. In other words, the addition of a-ketobutyric acid (a- KB) or 2-acetolactate (HMOB), precursors commonly used in bacteria for the regio- and stereo-specific labeling of lie and Val residues, does not permit the in vitro synthesis of the GFP protein. Figure 2 (B) shows the expression of GFP by cell-free synthesis from amino- acid precursors according to the invention, i.e. ketoacids precursors. We can see bars representing fluorescence intensity of GFP obtained by a cell-free synthesis (determined by fluorescence measurements) without amino-acid (lane 1 ); with a mix of: 20 amino- acids (lane 2); 19 amino-acids without lie (lane 3); 19 amino-acids without lie and with MOP-51 , i.e. labelled (S)-3-Methyl-2-oxopentanoic acid (lane 4); 19 amino-acids without Val (lane 5); 19 amino-acids without Val and with KlV-ProS, i.e. labelled 2- KetolsoValerate or labelled 3-Methyl-2-oxobutanoic acid (lane 6). Fluorescence of GFP synthetized by addition of 20 amino-acids mix was arbitrary determined as 100 %.

When the starting precursors presented in Figure 2(A) were replaced by amino-acid precursors according to the invention, such as KlV-ProS and MOP, which are downstream in the isoleucine and valine biosynthesis pathway and represent transaminases substrates, GFP was produced at a level close to 100% (level corresponding to the synthesis of GFP with 20 amino-acids added).

Figure 3 shows a SDS-PAGE gel of the purification of KARI enzyme on Ni-NTA resin; SDS-PAGE means Sodium DodecylSulfate containing PolyAcrylamide Gel Electrophoresis. Cells expressing KARI were lysed using sonicator and centrifuged for 30 min at 20 000 rpm. Insoluble (P) and soluble (SN) fractions were analysed on SDS-PAGE gel. 7 mL of Ni-NTA resin were equilibrated in loading buffer (50 mM Hepes pH 8 10% glycerol 0,1 % Tween 20). Supernatant was passed through the column and flowthrough (FT) analysed. Resin was washed with 20 mM imidazole (W20) and KARI was eluted with 500 mM imidazole (Elutions).

Figure 4 shows a SDS-PAGE gel of the purification of DHAD enzyme on a 5 mL His-Trap column.

Figure 4 (A) shows a chromatogram of the affinity purification. Figure 4 (B) shows a SDS- PAGE gel of the lysis and purification process.

Cells were lysed using a sonicator and centrifuged for 30 min at 20 000 rpm. Insoluble (P) and soluble (SN) fractions were analysed on SDS-PAGE gel. A 5-mL his-trap column (GE Healthcare) was equilibrated in loading buffer. Supernatant was passed through the column and flowthrough (FT) analysed. Resin was washed with 20 mM imidazole (W20) and DHAD was eluted with an imidazole gradient ranging from 20 to 500 mM (Elutions).

Figure 5 shows a kinetic curve representing the synthesis of stereospecifically labelled KlV-ProS and ΜΟΡ-γ2.

Synthesis of stereospecifically labelled KlV-ProS (A) or ΜΟΡ-γ2 (B) was monitored by NMR during 3 h at 37°C on a 600 MHz NMR spectrometer equipped with a cryogenic probe using a series of 1 d experiments. Intensities of the methyl groups of HMOB (A ·), KIV (A ■), AHB-51 , i.e. labelled (S) 2-Aceto-2-HydroxyButanoate or 2-hydroxy-2-(2'- ( ¹³ C),1 '-(D ₂ )) ethyl-4-(D ₃ )-3-oxobutanoate (B ·) and ΜΟΡ-γ2, i.e. labelled (S) 3-Methyl-2- OxoPentanoate or (S) 3,4,4,5,5,5-(D6)-3-( ¹³ CH ₃ )-2-oxopentanoate (B■) were indicated as a function of time.

Figure 6 shows bars representing fluorescence intensity of GFP obtained by a cell-free synthesis (fluorescence) using precursors of aromatic residues.

We can see a cell-free GFP synthesis (determined by fluorescence measurements) with a mix of: 19 amino-acids without Phe (lane 1 ); 19 amino-acids without Tyr (lane 2); 18 amino-acids without Phe and Tyr (lane 3); 19 amino-acids with phenylpyruvate (PP) (lane 4) ; 19 amino-acids with 4-hydroxyphenylpyruvate (HPP) (lane 5); 18 amino-acids with PP and HPP (lane 6) or 20 amino-acids (lane 7). Fluorescence of GFP synthetized by addition of 20 AA mix was arbitrary determined as 100 %.

The inventors tested the incorporation of HPP and PP which are respectively the amino- acids precursors of tyrosine (Tyr) and phenylalanine (Phe) in the cell-free synthesis of GFP. Cell-free GFP production was attempted in different conditions. As expected, without Phe, Tyr or both of them, GFP protein was not produced by a cell-free synthesis process. When the adequate precursor was then added to replace the missing aromatic amino-acid, the synthesis was restored to about 80%, proving that these precursors are effectively incorporated by the cell-free system. Figure 7 shows a SDS-PAGE gel of the purification of H23 produced by cell-free using labelled amino-acid precursors according to the invention.

H23 was synthetized by cell-free system using KlV-ProS, MOP-51 or ΜΟΡ-γ2 according to the protocol described later. Proteins were purified on Ni-NTA resin and different fractions (SN: supernatant; FT: flowthrough; E: Elution) were analysed on SDS-PAGE gel. Figure 7 (A) shows a H23 protein labelled on Val proS, Figure 7 (B) shows a H23 protein labelled on lle-51 , and Figure 7 (C) shows a H23 protein labelled on Ile-y2. This SDS- PAGE gel of the purification of H23 produced by cell-free using labelled amino-acid precursors according to the invention allows to verify that ¹³ CH ₃ -labelled amino-acid precursors according to the invention were converted to specifically labelled Val ProS, lle- δ 1 or Ile-y2 methyl groups of H23, a 17-kDa protein, without scrambling.

Figure 8 shows a comparison of 2D methyl-TROSY (Transverse relaxation optimized spectroscopy) NMR spectra recorded on ¹³ CH ₃ specifically methyl-labelled H23 samples.The 2D methyl-TROSY NMR spectra were recorded at 30 °C in D ₂ 0 buffer (0,1 mM H23 protein in 20 mM Tris 50 mM NaCI, pH 7,1 ) on an NMR spectrometer operating at a proton frequency of 600 MHz. H23 samples were ¹³ CH ₃ labelled on Ile-y2 (A), lle-51 (B) or Val ProS (C). A total of 9 peaks were observed for Ile-y2 (see Figure 8 (A)), 9 peaks for lle-51 (see Figure 8 (B)) and 13 peaks for Val ProS (see Figure 8 (C)) suggesting that these three amino-acid precursors according to the invention were correctly converted into ¹³ CH ₃ amino-acids in H23 protein.

Detailed description

In the present patent document the atom "D" means deuterium ( ² H), and when the symbol D is used for deuterium, the atom "H" means ¹ H. When isotopes are not relevant in a molecule or context, the atom "H" means hydrogen, with no isotopic specificity. The inventors have recognized that an alternative to isotopically labelled amino-acids could involve the use of precursors of the labelled amino-acids. The use of isotopically labelled precursors in a cell-free system for the selective labeling of methyl residues has not yet been reported in the literature. In the present invention, the cell-free synthesis of a protein is an in vitro protein synthesis using biological machinery of a cell extract in which a cell-free transcription system produces mRNA using DNA as the template, such as a plasmid, and in which mRNA information are translated into proteins. According to the invention, the cell-free synthesis of a protein can also be carried out by using biological machinery of a cell extract in which a cell-free translation system produces proteins in ribosome through reading of information of mRNA used as the template.

In order to allow the expression of only a specific protein, the DNA of host cell is removed from this cell extract. Furthermore, this cell extract may contain buffer solutions, salts, RNase inhibitors, RNA polymerase in case where DNA is used as template.

For preparing the cell extract, any eukaryotic or prokaryotic cell extract containing factors required for protein synthesis such as ribosomes, factors for transcription and translation machinery can be used. Any of generally known cells can be used for the preparation of the cell extract. For example, E. coli strains (e.g. MRE, BL21 /DE3/C+/RIL/Rosetta...) from S12 to S100, especially S30, wheat germ, insect cells, wheat germ or insect cells can be used. The E. coli cell extract can be prepared from E. coli BL21 (DE3) cells in accordance with generally known methods (Apponyi et al., "Cell-free protein synthesis for analysis by NMR spectroscopy", Methods in Molecular Biology (2008) p.257-268 and specifically parts 2 & 3), or can be purchased (from companies such as Promega or Sigma-Aldrich).

This cell extract comprising biological machinery (i.e. comprising means for transforming precursor of an amino-acid into the corresponding amino-acid such as BCAT enzyme (EC 2.6.1 .42: branched chain aminotransferase) and lacking host cell DNA is hereinafter referred to as "cell-free extract".

In order to produce a target protein by a cell-free synthesis process according to the invention, at least one tRNA is added to the cell-free extract for integrating amino-acid obtained from amino-acid precursor into the corresponding target protein.

The present invention deals with cell-free synthesis of proteins from precursor of amino- acids (i.e. methyl-containing and aromatic residues), and in particular with cell-free synthesis of at least partially deuterated and at least partially ¹³ C labelled protein or biomolecular assembly from particular precursor of amino-acids.

Even though there are a few solutions for obtaining proteins bearing specifically labelled methyl groups from a cell-free system, in vitro synthesis of a protein has never been made from a regio- and stereospecific labelled precursor. Indeed, unlike bacterial system, HMOB (Val precursor), a-KB (lie precursor) and AH B (lie precursor) (see Figure 1 ) are not converted into amino-acids by the cell-free system (see Figure 2(A)), probably because some enzymes of the bacterial cell extracts such as DHAD (EC 4.2.1 .9: dihydroxy-acid dehydratase) enzyme are denatured. To solve this problem and be able to carry out specific methyl labeling in the cell-free system, the inventors have developed a protocol for the precursor incorporation, including synthesis of amino-acids precursors such as KlV-ProS and MOP-51 (or ΜΟΡ-γ2), and shown that these new regio- and stereospecifically labelled precursors are converted into valine and isoleucine respectively, enabling the synthesis of methyl labelled proteins using the cell-free system (see Figure 2 (B)).

Phenylpyruvate and 4-hydroxyphenylpyruvate, precursors of Phe and Tyr respectively, can be used for cell-free protein expression. These compounds can bear any type of specific protonation and isotopic labeling at any atomic position. 1 . Amino-acid Precursors synthesis

As starting precursors usually used in bacteria (HMOB and AHB) were not efficient in the cell-free system, the inventors have synthetized downstream precursors to restore their conversion into amino-acids in the cell-free extract. Thus, in order to obtain the precursor of amino-acids from starting precursors, the use of KARI (EC 1 .1 .1.86: ketol-acid reductoisomerase) and DHAD (EC 4.2.1.9: dihydroxy-acid dehydratase) is necessary. These enzymes act sequentially on the biosynthetic pathway of isoleucines and valines to convert precursors such as HMOB or AHB to transaminase substrates, i.e. precursors of amino-acids of the present invention such as KIV or MOP (see Figure 1 ). a. Recombinant KARI

KARI is a family of nicotinamide adenine dinucleotide phosphate (NADPH)-dependent oxidoreductases that is involved in the biosynthesis of the branched-chain amino-acids (see Figure 1 ). It catalyzes the second step in the branched chain amino-acid (BCAA) biosynthesis pathway, converting HMOB to (R)-DHIV ((R)-2,3-Dihydroxy-isovalerate) via a methyl shift coupled to a reduction with concomitant oxidation of a nicotinamide adenine dinucleotide cofactor. It also converts AHB into (R)-3-hydroxy-3-methyl-2-oxopentanoate, in the isoleucine biosynthesis pathway. It is necessary to dispose of a source of reducing power to ensure complete conversion of the acetolactates to dihydroxyacids. All kind of KARI can be used in the biosynthesis of the branched-chain amino-acids. Most of wild- type KARIs characterized and described in the literature have displayed a strong preference for nicotinamide adenine dinucleotide phosphate (NADPH) over nicotinamide adenine dinucleotide (NADH). Consequently, it is needed to have a source of reducing power able to regenerate the NADPH. Only a few NADPH regenerating systems are known. The most common uses glucose-phosphate and glucose-phosphate dehydrogenase. However, because NAD(H) is much less expensive than NADP(H), KARIs with the reversed cofactor preference are preferred for the in vitro conversion of acetolactate. Furthermore, the NADH regenerative system, based on formate dehydrogenase and the conversion of formate into C0 ₂ , allows synthesis without any residual side-products. Mutations in E.coli KARI gene enable a switch in the cofactor preference), but natural NADH-dependent KARIs, too, are known, and the inventors eventually prefer a NADH dependant, thermo- and solvent-stable KARI derived from the bacterium Meiothermus ruber. This specific KARI is described in Rei3e et al., « Identification and optimization of a novel thermo- and solvent stable ketol-acid reductoisomerase for cell free isobutanol biosynthesis», Biochimie (2015) vol.108, p.76-84. Recombinant KARI is expressed and purified on Ni-NTA resin (see Figure 3) and can be obtained after dialysis at a concentration of about 480 mg/L. This enzyme is stored at - 80 °C. KARI is also stable for few weeks at 4 °C. b. Recombinant DHAD

The DHAD enzyme is part of naturally occurring biosynthetic pathways producing valine, isoleucine, leucine and pantothenic acid (vitamin B5). It works downstream of KARI and catalyzes the conversion of DHIV to KIV and of 2,3-dihydroxy-3-methylpentanoate to MOP (see Figure 1 ). To be able to exhibit its catalytical function, the iron sulphur cluster in the core of the protein is crucial. Two families of DHAD are known, one with (4Fe-4S) cluster which are very sensitive to oxygen, inducing its degradation. For example, the activity of the E. coli DHAD is completely inhibited after a day. The second family is characterized by (2Fe-2S) clusters which are reported to be less sensitive to 0 ₂ (as described in US 2010/0 081 154). Consequently, the inventors chose to work with the Lactococcus lactis DHAD, a member of this last family but all kind of DHAD can be used.

Recombinant DHAD was expressed, purified (see Figure 4) and the final yield reached 300 mg/L. Recombinant DHAD is used extemporaneously when unfrozen. It was stable enough to carry out the conversion of DHIV to KIV in few hours, under aerobic conditions (see Figure 5). To improve stability, DHAD can be produced and purified in an inert atmosphere glovebox. c. Synthesis of precursor of amino-acids

Valine, leucine and isoleucine form the small group of branched-chain amino-acids (BCAAs) classified by their small branched hydrocarbon residues. A unique feature of

BCAA biosynthesis is that Val and lie are synthesized in two parallel pathways (see Figure 1 ). This is achieved with a single set of four enzymes, which catalyze the four reactions towards the formation of these amino-acids with different substrates. KARI and DHAD catalyze respectively the second and third steps of these parallel pathways.

KARI catalyzes an unusual two-step reaction including an alkyl migration to form the intermediates 3-hydroxy-3-methyl-2-oxobutyrate from HMOB or 3-hydroxy-3-methyl-2- oxopentanoate from AHB. This isomerization depends on Mg ²⁺ and is followed by the reduction of 3-hydroxy-3-methyl-2-oxobutyrate, respectively of 3-hydroxy-3-methyl-2- oxopentanoate to the final dihydroxyacid products DHIV (i.e. (R)-2,3-dihydroxy-3- methylbutanoate), and respectively (R)-2,3-dihydroxy-3-methylpentanoate. The reduction depends on both NADPH (NADH) and Mg ²⁺ . DHAD then catalyzes the dehydration of DHIV, to the ketoacid KIV, and respectively catalyzes the dehydration of (R)-2,3- dihydroxy-3-methylpentanoate to the ketoacid MOP.

As an example, KIV synthesis is presented in Figure 1. HMOB is used as a substrate of KARI and HMOB is converted to (R)-DHIV via a methyl shift coupled to a reduction with oxidation of NADH (see Figure 1 ). (R)-DHIV is then converted by DHAD to KIV, precursor of valine in the present invention.

In order to optimise the synthesis of precursors of amino-acids, it is possible to use, for NADH regeneration, formate dehydrogenase (FDH) isolated from Candida boidinii to convert formate and NAD into carbon dioxide and NADH.

As a buffer solution, a buffer agent such as Hepes can be used, but the invention is not limited to this specific buffer.

Reactions are carried out in D ₂ 0 in order to introduce a deuterium atom in the beta- position of the KIV during the dehydration reaction catalyzed by DHAD.

An optimised composition of materials used for the synthesis of amino-acids precursor from starting precursor is presented in table 1 . This composition can be used to synthesize KIV from HMOB (see table 1 ) or MOP from AHB.

Table 1

Composition of materials used for the synthesis of amino-acids precursors such as KIV, precursor of Valine, from HMOB

KARI is derived from a thermophilic bacterium which grows at a temperature comprised between 35°C and 70°C, with an optimum growth temperature at 60°C. DHAD is derived from a mesophilic strain. Moreover, acetolactate has a poor stability, i.e. is stable for a few hours above 37°C. According to an advantageous embodiment of the invention, the temperature of the synthesis is comprised between 30°C and 40°C. The inventors have found that 37°C is a good temperature, which is around the optimal temperature for DHAD and the maximal temperature for acetolactate. The pH value may vary between 7 and 8 so that the KARI and DHAD enzymes are active, especially KARI, with an optimum around 7. NAD or NADH can be indifferently added to the medium provided that there is an efficient regeneration system.

After the addition, in any order, of all the components presents in the composition listed in table 1 , the reaction mixture is maintained at 37 °C for about 3 hours under stirring. Once set up, kinetics studies on conversion of HMOB-ProS are performed on a 600 MHz NMR spectrometer and shown in Figure 5A. HMOB-ProS, prepared in buffer according to table 1 , was followed for 3 hours at 37 °C. After 2 h, more than 90 % of HMOB-ProS disappeared (see · in Figure 5(A)) in favor of KlV-ProS (see■ in Figure 5A).

This protocol is also used to convert labelled AHB-51 to corresponding MOP- 51 or labelled ΑΗΒ-γ2 to corresponding ΜΟΡ-γ2 as shown in Figure 5(B) and in table 2.

After the synthesis of amino-acid precursor, a purification step can be carried out. The reactional medium containing amino-acid precursor undergo a heat shock at 80 °C during 10 min in order to precipitate FDH, KARI and DHAD enzymes. After centrifugation of the reactional medium containing the precipitate of FDH, KARI and DHAD, amino-acid precursor purified are obtained.

Table 2

Use of the described protocol for the synthesis of amino-acid precursors from starting

precursor

Starting precursor amino-acids precursor Amino-acids

HMOB-ProS KlV-ProS precursor of Valine

AHB-51 MOP- 51 precursor of Isoleucine

ΑΗΒ-γ2 ΜΟΡ-γ2 precursor of Isoleucine

4MOP - ProS

AHB-51 ((S) 4-(CD3) - 4(D1 ) - 4-(13CH3) - 2- precursor of Leucine oxopentanoic acid)

4MOP - ProR

ΑΗΒ-γ2 ((R) 4-(CD3) - 4(D1 ) - 4-(13CH3) - 2- precursor of Leucine oxopentanoic acid) This results are applicable to HMOB and AHB bearing any labelling pattern including different types of isotope enrichments (e.g. uniformly ¹³ C or not, whith ¹³ CH ² H ₂ or ¹³ CH ₂ ² H isotopomers etc.). The targeted isotope enrichment in number ( ¹³ C or ² H) is in all cases higher than 80% by number.

According to the invention, the amino-acid precursors can be used as an acid or as the corresponding anion of this acid (said precursor being used for instance as a salt).

2. Cell-free synthesis of protein from precursor of amino-acids

Proteins produced by the present invention may be any known and/or novel protein. DNA is used as the nucleic acid coding for target proteins, and can be extracted from eukaryotic or prokaryotic or can be used under plasmid form or template.

In order to produce a protein by a cell-free synthesis process according to the invention, a cell-free extract is used as described above and to which we can add ATP (0,5 to 5 mM), CTP (0,5 to 5 mM), GTP (0,5 to 5 mM), UTP (0,5 to 5 mM), buffer solutions, salts, antibacterial agents, RNA polymerase such as T7 RNA polymerase, tRNA and amino- acids with the exception of those amino-acids for which one or more amino-acid precursors will be added instead.

According to the invention, some of the amino-acid precursors used in cell-free synthesis of isotopic labelled proteins are presented in table 3.

According to the invention, one or more of the amino-acid precursors and the enzymes allowing to transform said precursors into corresponding amino-acid are added to the cell- free extract. The amino-acid precursors can be added to the cell-free extract by the addition of the reaction mixture that has been used for the synthesis of precursor as presented above in § 1.c.

According to the invention, the reaction mixture comprising cell-free extract, amino-acids, at least one amino-acid precursor, and DNA coding for target protein, is incubated at a temperature comprised between 20 and 40 °C, preferably at 30°C, for a time comprised between 1 hour and 24 hours, preferably for 3.5 hours, in order to allow the production of the target protein. .

After the reaction, the target protein can be isolated and purified by known methods. Depending on the properties of the proteins, purification can be carried out by a single method or by a combination of two or more methods. For example, anion or cation exchange chromatography, affinity chromatography, gel filtration chromatography, and/or HPLC, can be used. Table 3 below shows amino-acid precursors according to the invention. The formulae (I), (II), III) and (IV) identify the precursors.

(formula I)

(formula II)

(formula III)

(formula IV)

Table 3: Amino-acid precursors according to the invention

Examples

The present invention is not intended to be limited to the specific embodiments in the examples given below and/or in the attached drawings. These examples only illustrate the present invention, and in no way limit its legal scope.

1 . Production of KARI and DHAD enzymes

1 .1 . Recombinant KARI

KARI (Meiothermus ruber) was cloned in pET21 b plasmid with a C-terminal his-tag (AmpR).

PET21 b-KARI transformed BL21 (DE3) cells were grown in Terrific Broth (TB) medium supplemented with MgS0 ₄ (5 mM). When OD ₆₀₀ reached 2.5, IPTG (0.5 mM) was added and cells were harvested after 3 h at 37 °C (OD=9; centrifugation 7 000 g, 20 min, 4 °C). Cell pellets were frozen and stored at -20 °C until use.

Cells were lyzed by sonication in 50 mM Hepes pH 8 10 % glycerol 0.1 % Tween 20 and DNAse (50 μg mL final). After centrifugation (30 min; 40 000 g), supernatant was loaded on a 7 mL Ni-NTA resin (Qiagen) equilibrated with loading buffer (50 mM Hepes pH8 10 % glycerol 0.1 % Tween 20). After substantial washing with said loading buffer + 20 mM imidazole, KARI was eluted with a 500 mM imidazole step and fractions were analysed on SDS-PAGE gel. Fractions containing KARI were pooled and imidazole was removed by dialysis or desalting column (HiPrep 26/10 desalting) against 50 mM Hepes pH8 10 % glycerol 0.1 % Tween 20. Enzyme was concentrated up to 15 mg/mL, flash frozen and stored at -80 °C.

To perform the synthesis of precursors in fully deuterated buffer in order to prevent introduction of unwanted protons, KARI was buffer-exchanged against D ₂ 0 buffer using an amicon-15with a cut off 10 kDa (centrifugal filter).

1 .2. Recombinant DHAD

DHAD (Lactococcus lactis) was cloned in pET28a plasmid with a N-terminal his-tag

(KanR).

PET28a-DHAD transformed BL21 (DE3) cells were grown in TB medium supplemented with MgS04 (5 mM). When OD ₆₀₀ reached 2.5, IPTG (0.5 mM) was added and cells were harvested after an overnight induction at 20 °C (OD=14 ; centrifugation 7 000 g 20 min 4 °C). Cell pellets were frozen and stored at -20 °C until use.

Cells were lyzed by sonication in 50 mM Tris pH8 10 mM MgCI ₂ (TM8). After centrifugation (30 min; 40 000 g), supernatant was loaded on a 5 mL his-Trap (GE Healthcare) equilibrated with loading buffer (TM8). After substantial washing with loading buffer + 20 mM imidazole, DHAD was eluted with an imidazole gradient ranging from 20 mM to 500 mM and fractions were analysed on SDS-Page gel. Fractions containing DHAD were pooled and imidazole was removed by dialysis or gel filtration (HiLoad Superdex 200 26/60) against TM8. Enzyme was concentrated up to 15 mg/mL, incubated 10 min at 37 °C, centrifuged (10 min 12 000 rpm), aliquoted, flash frozen and stored at - 80 °C.

2. Synthesis of regio- and stereospecifically labelled amino-acids precursors Labelled amino-acids precursors presented below were elaborate from enzymatic synthesis because this method is robust, efficient and cost-effective.

The required starting compounds for the targeted labelled amino-acids precursors synthesis i.e. (S)-2-Hydroxy-2-[ ¹³ C]methyl-3-oxo-4,4,4-tri-[ ² H]butanoate (HMOB-proS), (S)-2-Hydroxy-2,2,2-tri-[ ² H ₃ ]methyl-3-oxo-4-[ ¹³ C]butanoate (HMOB-proR), 2-hydroxy-2- ethyl-3-oxo-butanoate either ¹³ CH ₃ -labelled on the 4-methyl or on the 1 '-methyl and fully deuterated on others positions (ΑΗΒ-δ1 or ΑΗΒ-γ2, respectively precursors of labelled δ 1 or j2 isoleucine) were obtained from NMR-Bio. These starting compounds comprising specific labelling can be made by several chemical methods such as those described in WO 201 1/083 356.

2.1 . Synthesis of stereospecifically labelled 2-oxoisovalerate (KlV-ProS)

KlV-ProS was produced in 50 mM Hepes at pH 7, 10 mM MgCI ₂ containing 15 mM (S)-2- Hydroxy-2-[ ¹³ C]methyl-3-oxo-4,4,4-tri-[ ² H]butanoate, i.e. HMOB-proS, 100 mM Sodium Formate, 1 .2 U/mL Formate deshydrogenase, 0,9 mM NAD, 3.4 μΜ (i.e. 1 : 5 000 (molar))

KARI and 2 μ Μ (i.e. 1 : 10 000 (molar)) DHAD as described in table 1 . The reaction was incubated for 2 h at 37°C under agitation (see Figure 1 ). Enzymes were precipitated by heat shock for 10 min at 80°C and removed by centrifugation for 30 min at 12 000 rpm. Supernatant was analysed by NMR on a Bruker Avance spectrometer operating at a proton frequency of 600 MHz and equipped with a cryogenic triple resonance probe head.

For kinetics studies, production of KlV-ProS was monitored by NMR on the same spectrometer by a series of I d-experiments.

2.2. Synthesis of regioselectively labelled MOP

Same conditions were applied to produce labelled MOP-51 or ΜΟΡ-γ2 from AHB-51 or ΑΗΒ-γ2 (see Figure 1 ). MOP was produced in 50 mM Hepes at pH 7, 10 mM MgCI ₂ containing 15 mM AHB, 100 mM Sodium Formate, 1 .2 U/mL Formate deshydrogenase, 0.9 mM NAD, 3.4 μΜ (i.e. 1 : 5 000 (molar) KARI and 2 μΜ (i.e. 1 : 10 000 (molar)) DHAD as described in table 1 . The reaction was incubated for 2 h at 37°C under agitation (see Figure 1 ). Enzymes were precipitated by heat shock for 10 min at 80°C and removed by centrifugation for 30 min at 12 000 rpm. Supernatant was analysed by NMR on a Bruker Avance spectrometer operating at 600 MHz and equipped with a triple resonance probe head. For kinetics studies, production of ΜΟΡ-γ2 was monitored by NMR on the same spectrometer by a series of I d-experiments (see Figure 5(B)).

3. Cell-free synthesis of GFP isotopically labelled protein from amino-acids precursor GFP was synthesized in vitro.

For GFP protein cell-free synthesis, a mixture is prepared, for a total volume of 50 μΙ_, containing 55 mM HEPES/KOH at pH 7.5, 3.4 mM dithiothreitol, 1 ,2 mM Adenosine triphosphate (ATP), 0.8 mM Cytidine triphosphate (CTP), 0.8 mM Guanosine triphosphate (GTP) and 0.8 mM Uridine triphosphate (UTP), 0.64 mM 3',5'-cyclic AMP, 68 μΜ folinic acid, 27.5 mM ammonium acetate, 2 mM spermidine, 208 mM potassium glutamate, 80 mM creatine phosphate, 250 μg mL creatine kinase, a mixture of 1 mM of each of amino- acids or if one amino-acids is exempt 3 mM of the corresponding amino-acids precursor is added, 14 mM magnesium acetate, 175 μg mL E. coli total tRNA, 20 μΙ_ of S30 extract (A254 nm > 250 as presented above), 16 μg mL of supercoiled plasmid DNA of GFP, and 50 μg mL of T7 RNA Polymerase. When the starting precursors presented in Figure 2(A) were replaced by amino-acids precursors according to the invention (see Figure 2 (B)), such as KlV-ProS and MOP, which are downstream in the isoleucine and valine biosynthesis pathway and represent transaminases substrates, GFP was produced at a level close to 100% (level corresponding to the synthesis of GFP with 20 amino-acids added).

Similarly, HPP and PP which are respectively the precursors of tyrosine and phenylalanine can be used in the cell-free synthesis of protein, as shown with GFP (see Figure 6).

4. Cell-free synthesis of H23 isotopically labelled protein from amino-acids precursor

H23 isotopic labelled protein was prepared by cell-free synthesis from amino-acids precursor in order to confirm the ¹³ CH ₃ -precursor incorporation by NMR.

Cell-free preparation, i.e. S30 extract was prepared in flasks from E. coli BL21 (DE3) cells as described by the procedure of Apponyi (Apponyi et al. , "Cell-free protein synthesis for analysis by NMR spectroscopy", Methods in Molecular Biology (2008) p.257-268 and specifically parts 2 & 3). Proteins, sub-cloned in pIVEX 2.3d or 2.4d (Roche Applied Science), were synthetized by cell-free E. coli coupled transcription-translation system.

H23 protein (17 kDa) was synthesized in vitro.

First, a mixture of each of 19 L-amino-acids (which can be deuterated or protonated) at 1 mM, i.e. 20 amino-acids except Valine was prepared. Stereospecifically labelled amino- acids precursor KlV-ProS (precursor of valine) was added at 3 mM in the said amino- acids mixture without Valine.

For H23 protein cell-free synthesis, a mixture is prepared, for a total volume of 3 mL, containing 55 mM HEPES/KOH at pH 7.5, 3.4 mM dithiothreitol, 1 .2 mM Adenosine triphosphate (ATP), 0.8 mM Cytidine triphosphate (CTP), 0.8 mM Guanosine triphosphate (GTP) and 0.8 mM Uridine triphosphate (UTP), 0.64 mM 3',5'-cyclic AMP, 68 μΜ folinic acid, 27.5 mM ammonium acetate, 2 mM spermidine, 208 mM potassium glutamate, 80 mM creatine phosphate, 250 μg mL creatine kinase, a mixture of 1 mM of each of 19 L- amino-acids (i.e. 20 amino-acids except Valine) and 3 mM of KlV-ProS (precursor of valine), 14 mM magnesium acetate, 175 μg mL E. coli total tRNA, 1.2 mL of S30 extract (A254 nm >250), 16 μg/mL of supercoiled plasmid DNA of H23, and 50 μg/mL of T7 RNA Polymerase. Samples were then incubated for 3h30 at 30 °C and the H23 protein was purified. Briefly, samples were diluted in loading buffer comprising 20 mM Tris, 100 mM NaCI and 5 mM Imidazole at pH 7.1 and loaded on Ni-NTA resin (1 mL). Resin was washed with loading buffer comprising 20 mM Tris, 100 mM NaCI and 25 mM Imidazole and protein was eluted with loading buffer comprising 20 mM Tris, 100 mM NaCI and 305 mM Imidazole. Protein was buffer-exchanged against 20 mM Tris 50 mM NaCI pH 7.1 prepared in D ₂ 0 and concentrated to 0.1 mM on a membrane for concentration vivaspin with a cut off 5 kDa.

H23 labelled with Val proS was analysed on SDS-PAGE gel (see Figure 7 (A)).

H23 protein was also synthetized by cell-free system using MOP-51 or ΜΟΡ-γ2 (see Figure 7 (B) and Figure 7(C)). These SDS-PAGE gel of the purification of H23 produced by cell-free using labelled amino-acids precursor according to the invention allow to verify that ¹³ CH ₃ -labelled amino-acids precursor according to the invention were converted to specifically labelled Val ProS, lie- δ 1 or Ile-y2 methyl groups of H23, a 17-kDa protein, without scrambling.

Methyl-TROSY (Transverse Relaxation Optimised Spectroscopy) spectra of labelled H23 were recorded at 30 °C on a Bruker advance 600 MHz NMR spectrometer equipped with a cryogenic probe during 10 min each (cf. Figure 8). These NMR spectra suggest that the amino-acids precursors according to the invention were correctly converted into ¹³ CH ₃ amino-acids in H23 protein.

The use of deuterated amino-acids along with the described precursors in the invention results in the production of a uniformly deuterated protein with the specific protonation and isotopic labeling of ¹³ CH ₃ -(lle-δΐ and/or Ile-y2 and/or Val-proR and/or Val-proS).