System for cyclisation of recombinant proteins for their expression and purification

Abstract

The invention relates to a novel method for cyclising a protein of interest using a protein tag consisting of a specific peptide (NUFIP1 or TAH1) and its receptor protein (ZNHIT3 or PIH1, respectively), the specific peptide and the receptor protein being fused respectively to the N- and C-termini of the protein of interest, and to the resulting cyclised fusion protein, thereby making it possible to increase thermostability and / or to assist in the expression of recombinant proteins and / or to assist in the purification of recombinant proteins of interest rendered more soluble for an improved production yield.

Description

[0001] SYSTEM FOR CYCLING RECOMBINATING PROTEINS FOR THEIR EXPRESSION AND PURIFICATION

[0002] DESCRIPTION

[0003] Technical field of the invention

[0004] The invention relates to a novel method for cyclizing a protein of interest using a protein tag consisting of a specific peptide (NUFIP1 or TAH1) and its protein receptor (ZNHIT3 or PIH1, respectively) attached to the N- and C-terminal ends of said protein of interest, and the resulting cyclized fusion protein. This allows for increased thermostability and / or facilitates the expression of recombinant proteins and / or aids in the purification of recombinant proteins of interest, making them more soluble for improved production yields.

[0005] In the description below, references in brackets ([ ]) refer to the list of references at the end of the text.

[0006] State of the art

[0007] For several decades, industrial demand for recombinant proteins has been growing. These proteins are generally active biomolecules produced by host microorganisms (such as bacteria or yeast) that have been genetically optimized to overproduce a target protein. This is how, for example, insulin and its derivatives, erythropoietin (or EPO), lipolytic enzymes (or lipases) capable of digesting fats, and, more recently, PET hydrolases, which have the property of digesting plastic, are synthesized. The areas of application are therefore numerous and promising. For instance, biocatalysis, a phenomenon made possible by the enzymatic properties of certain recombinant proteins, is an environmentally compatible industrial process, as many of the enzymatic reactions take place in water and at temperatures below 100°C.New diagnostic and research tools based on the use of lightweight antibodies, also known as "nanobodies," are also being developed. These small, recombinantly produced proteins have very strong binding capabilities to specific and highly varied ligands, such as toxins or viral proteins like those of SARS-CoV-2. While the rise of applied biology has made the large-scale production of recombinant proteins possible, numerous obstacles remain, from the discovery of a promising target to its intensive use. These include low production yields, poor solubility, and the aggregation of the protein of interest. An additional point of concern is the stability of the target protein in response to temperature, mechanical agitation, and over time.Indeed, some enzymes, even when successfully purified and used at moderate temperatures, quickly become deactivated due to their intrinsic instability.

[0008] Since recombinant protein production processes are expensive, solutions improving the points mentioned above had to be considered to facilitate industrial applications.

[0009] Improving protein stability is a subject of much research in the literature. One commonly used approach is mutagenesis. This involves modifying the amino acid sequence of the target protein, either randomly or in a targeted manner. For example, it has been shown that a high content of charged residues such as lysine, arginine, or glutamate is correlated with greater temperature resistance (Kumar et al., 2000; Yokota et al., 2006; Strickler et al., 2006) [1-3]. However, this approach can alter the catalytic or binding capacities of the mutated protein. Indeed, to preserve the desirable properties of the native protein, it is necessary to spare the residues of its active site (or in its immediate vicinity) as well as the residues that give the protein its three-dimensional shape.

[0010] Another method has proven effective in increasing temperature stability and decreasing the propensity for aggregation. This involves cyclizing the protein by capturing its ends. Indeed, the regions at the ends of proteins are areas of high entropy, prone to disorganization and thus capable of destabilizing the three-dimensional folding of the protein, which is essential for its function. Several cyclization methods are possible (Hayes et al., 2021) [4]. One approach relies on chemical methods (activation or intrinsic reactivity of specific amino acids) that result in the formation of covalent bonds between the ends. A second approach consists of using an enzyme capable of taking up specific amino acid sequences implanted at each end of the protein, bringing them closer together in space, and forming a covalent bond.This is the mode of operation of subtiligase, sortase, and butelase-1. Finally, a third approach involves protein tags. This is the case with inteins, whose mode of action is similar to that of splicing observed with nucleic acids. Recently, protein / peptide pairs have also been developed which, in addition to specific mutual recognition, establish a covalent isopeptide bridge between them (Spy-ring, Snoop-ring, etc.) (Sun et al., 2019; Buldun et al., 2018; Wang et al., 2016) [5-7]. This property has been exploited to clamp the ends of phytase-C, an enzyme used by the food industry to release phosphate from phytic acid and thus improve the digestibility of animal feed.This cyclization has increased the enzyme's thermal resistance and facilitated its use in an industrial process requiring a sterilization step performed between 75 and 90°C. In some cases, such as that of ammonia lyase, cyclization also improves thermal stability. Some PET hydrolases have undergone cyclization using protein tag systems, which has increased their catalytic activity and / or stability (Hayes et al., 2023) [8].

[0011] Description of the invention

[0012] In this context, the inventors have developed a system capable of increasing the overexpression and / or stability of a recombinant protein and / or aiding its purification by integrating adapters, which have the capacity to clamp the mobile ends of the target proteins. This cyclization process thus makes it possible to increase the production yield of a recombinant protein, for example, an enzyme (e.g., PET hydrolase).

[0013] The new system proposed by the Inventors is based on a protein tag formed by a specific protein / peptide pair. This is a highly specific, heat-resistant, and high-affinity complex between a peptide and its protein receptor. When the peptide and its protein receptor are fused to the ends of a target protein of interest, they interact to form a clamp, resulting in the cyclization of the protein of interest. This allows for increased expression and / or thermostability and / or facilitates the purification of the protein of interest, making it more soluble for improved production yields.

[0014] To do this, the two protein / peptide pairs ZNHIT3 / NUFIP1 and PIH1 / TAH1, which have already been the subject of many fundamental studies (Chagot et al., 2022; Henri et al., 2018; Quinternet et al., 2016; Quinternet et al., 2015; Quinternet et al., 2015; Rothé et al., 2014) [9-14], were selected and respectively named ZN-ring and TP-ring. Interestingly, it has been shown that the ZNHIT3 / NUFIP1 complexes of Homo sapiens, Saccharomyces cerevisiae and Plasmodium falciparum adopt very similar three-dimensional folds despite a low percentage of identity between the protein sequences of these three species (17% minimum) (Chagot et al., 2022; Quinternet et al., 2016) [9,12],

[0015] The general principle of the cyclization system according to the invention is illustrated in Figure 1. ZNHIT3 and PIH1 (also known as TRIP3 or HIT1 and PID1D1, respectively) are both globular protein domains of approximately 80 residues. NUFIP1 and TAH1 (also known as RSA1 and RPAP3) are linear peptides of approximately 20 residues that exhibit three-dimensional folding upon contact with their partner (protein receptor). The ZNHIT3 / NUFIP1 and PIH1 / TAH1 complexes formed are very stable and have a denaturation temperature of up to 90°C (Chagot et al., 2022) [9]. Below this temperature, the association between the two partners remains extremely strong due to the nature of the interactions established between the peptide and its protein domain.Indeed, strong hydrophobic and electrostatic interactions significantly slow the association / dissociation equilibrium of these complexes, which remain associated even at very low concentrations. Unlike Spy-ring or Snoop-ring systems, these complexes do not require the formation of isopeptide bridges to be strong (Wang et al., 2016) [7]. This strong association at low concentrations is an advantage, as minimizing enzyme concentration is a parameter considered in calculating the profitability of an industrial application.

[0016] Another advantage of the ZN-ring and TP-ring systems according to the invention lies in the components ZNHIT3 and PIH1 which can only be soluble in the presence of their respective partners NUFIP1 and TAH1 (Quinternet et al., 2016; Rothé et al., 2014) [11,14]. In other words, when ZNHIT3 and PIH1 are not respectively complexed with NUFIP1 and TAH1, these protein domains are totally insoluble. Interestingly, the ZNHIT3 / NUFIP1 and TAH1 / PIH1 complexes are highly soluble at high concentrations (Chagot et al., 2022; Henri et al., 2018; Quinternet et al., 2016; Quinternet et al., 2015; Quinternet et al., 2015; Rothé et al., 2014) [9-14]. Thus, when TP-ring and ZN-ring systems are used as protein tags, a natural sorting of the resulting cyclized proteins occurs during the first purification step. This step consists of separating soluble proteins from insoluble proteins.Thus, in the case of an intrinsically soluble protein of interest fused to the TP-ring and ZN-ring system, but for which the clamp has not formed, the ZNHIT3 and PIH domains not complexed with NUFIP1 and TAH1 will drive all protein fusion into the insoluble fraction. Conversely, when the clamp is formed and cyclization occurs, protein fusion will be driven into the soluble fraction.

[0017] Conversely, the protein and peptide components of Spy-ring and Snoop-ring systems are soluble, independently of each other. This precludes drastic sorting based on the solubility of cyclized or non-cyclized proteins.

[0018] The use of ZN-ring and TP-ring systems makes them all the more attractive because they can be easily adapted, as the ZNHIT3 / NUFIP1 and PIH1 / TAH1 pairs are naturally found in various eukaryotic organisms. Here, for the ZNHIT3 / NUFIP1 complex, the (protein) sequences of Plasmodium falciparum were used, but the inventors have already shown that the human and yeast versions of this complex are resistant to temperatures above 70°C (Chagot et al., 2022) [9]. For the PIH1 / TAH1 complex, the protein sequences of Saccharomyces cerevisiae were used (Quinternet et al., 2015)

[0012] .

[0019] The present invention therefore relates to a cyclized fusion protein comprising a cyclization system consisting of a peptide and its protein receptor respectively fused to the N- and C-terminal ends of a protein of interest, where said peptide is the peptide NUFIP1 or TAH1 and said protein receptor is respectively the protein ZNHIT3 or PIH1.

[0020] The present invention also relates to the use of a cyclized fusion protein according to the invention for the expression and / or purification of a protein of interest.

[0021] The present invention also relates to an in vitro process for the expression and purification of a protein of interest, characterized in that it comprises the following steps:

[0022] a) fusion of a cyclization system consisting of a peptide and its protein receptor respectively fused to the N- and C-terminal ends of a protein of interest;

[0023] b) expression in a host cell of the cyclized fusion protein obtained in step a);

[0024] c) purification of the cyclized fusion protein; and

[0025] d) optionally isolation of the protein of interest from the cyclization system; where said peptide is the NUFIP1 or TAH1 peptide and said protein receptor is respectively the ZNHIT3 or PIH1 protein.

[0026] According to a particular embodiment of the present invention, the ZNHIT3 / NUFIP1 cyclization system is that of P. falciparum, H. sapiens, S. cerevisiae and / or the PIH1 / TAH1 cyclization system is that of S. cerevisiae. In particular, the NUFIP1 peptide has the sequence SEQ ID NO: 4 or a sequence having at least 17% identity with SEQ ID NO: 4, the ZNHIT3 protein receptor has the sequence SEQ ID NO: 3 or a sequence having at least 18% identity with SEQ ID NO: 3, the TAH1 peptide has the sequence SEQ ID NO: 6, and / or the PIH1 protein receptor has the sequence SEQ ID NO: 5.

[0027] According to a particular embodiment of the present invention, the host cell is a non-human eukaryotic host cell or a prokaryotic cell, for example chosen from Saccharomyces cerevisiae, Escherichia coli or Pichia pastoris.

[0028] Brief description of the figures

[0029] Figure 1 illustrates the general principle of the cyclization system according to the invention. "POI" stands for "protein of interest".

[0030] Figure 2 shows the protein sequence of a helix cyclized by the TP-ring (A) and ZN-ring (B) systems. In this case, cloning sites for the restriction enzymes Xhol and Sacl, as well as polyglycine linker regions (GGGGS) (SEQ ID NO: 1), were introduced on either side of the region to be cyclized. The TP-ring system consists of regions 93–111 and 262–344 of the S. cerevisiae TAH1 and PIH1 proteins, respectively. The ZN-ring system consists of regions 265–332 and 817–841 of the P. falciparum ZNHIT3 and NUFIP1 proteins, respectively. (C) Purification follow-up of the constructs shown in (A) and (B) inserted into a pnEA plasmid and expressed in an E. coli BL21(DE3) strain. A 6xHIS label is added to the N-terminus of the constructs.Cso, SSo, Fix, Elu, and kDa refer respectively to the sonication pellet (insoluble fraction), the sonication supernatant (soluble fraction), the fixation on TALON beads, the eluate after imidazole addition to the beads, and the size marker. (D) Half-denaturation temperature curves for the helix with sequence EAAAKEAAAK (SEQ ID NO: 2) cyclized by the ZN-ring and TP-ring systems obtained by differential scanning calorimetry.

[0031] Figure 3 shows, as an example, (A) the monitoring of the purification process of CRD-gal3 inserted into the TP-ring system and expressed in an Escherichia coli BL21(DE3) strain using a denaturing polyacrylamide gel. Cso, SSo, Fix, Elu, and kDa refer, respectively, to the sonication pellet (insoluble fraction), the sonication supernatant (soluble fraction), fixation onto lactose-sepharose beads, the eluate after the addition of D-lactose to the beads, and the size marker. (B) Half-denaturation temperature curves for wild-type and TP-ring CRD-gal3, obtained by differential scanning calorimetry.

[0032] Figure 4 shows the purification follow-up using denaturing polyacrylamide gel of (A) Cel-CD cellulase, (B) Ideonella sakaiensis PETase (Is-PETAse), and (C) Carbios PETase (Carbios PETase), each inserted into TP-ring and ZN-ring systems. Is-PETase was purified in the uncyclized form (B). The constructs were expressed in an Escherichia coli BL21(DE3) strain. Cso, SSo, Fix, Elu, and kDa refer to the sonication pellet (insoluble fraction), the sonication supernatant (soluble fraction), fixation onto Lactose-Sepharose or TALON beads, the eluate after addition of D-Lactose or imidazole to the beads, and the size marker, respectively.

[0033] EXAMPLES

[0034] EXAMPLE 1: PROCESS FOR CYCLIZING PROTEINS OF INTEREST To perform cyclization, a plasmid DNA vector (e.g., pnEA, pET-20b, or pPICZ-A) is constructed such that the sequence of a protein or protein domain of interest is inserted between the DNA sequences of ZNHIT3 and NUFIP1 or TAH1 and PIH1. Preferably, ZNHIT3 or TAH1 are placed upstream of the protein or protein domain of interest, and NUFIP1 and PIH1, respectively, downstream. Optionally, the sequences of a short polyglycine segment (of the GGGGS type) (SEQ ID NO: 1) and / or a cloning enzyme restriction site may be placed downstream and / or upstream of the protein or protein domain of interest.

[0035] The sequences of the ZNHIT3 and NUFIP1 domains of P. falciparum and of TAH1 and PIH1 of S. cerevisiae respectively used in the ZN-ring and TP-ring systems are shown below:

[0036] pfZNHIT3(265-332):

[0037] DYDMLTEEQKKKLKEDHTLKILLKNNYVREVFKQFTLSNDKIGYLSHYINDPTIVQVID HIMKTIDDT (SEQ ID NO : 3)

[0038] pfNUFIPI (817-841) :DIYTYEKKLIKSIEYITKNKFFDDS (SEQ ID NO: 4)

[0039] scPIH 1(262-344):

[0040] EDVPEYEVKMKRFKGAAYKLRILIENKAPNSKPDRFSPSYNFAENILYINGKLSIPLPR DIWNAADIKIFHIRKERTLYIYI (SEQ ID NO : 5)

[0041] scTAH1(93-111):

[0042] SVQIPVVEVDELPEGYDRS (SEQ ID NO: 6)

[0043] The examples of the cyclization of the court segment hélicoïdal of the sequence EAAAKEAAAK (SEQ ID NO: 2) by the TP-ring and ZN-ring systems are indicated in Figures 2A and 2B. The options for poly-glycine segments and the location of the cloning of the figures.

[0044] The plasmid vector used is tailored to the host organism chosen for cyclization and may carry additional functions, such as protein tags to aid purification (histidine-tag or 6xHIS type). In a preferred mode, the host organism is an Escherichia coli BL21(DE3) strain.

[0045] Cyclization occurs spontaneously during the production of the plasmid-encoded construct by the host organism. When using an Escherichia coli BL21(DE3) strain, the bacteria must be transformed with the vector of interest. The bacteria are then cultured in a medium that supports their growth. Expression of the protein or protein domain of interest can be induced by adding sufficient amounts of IPTG (isopropyl-D-1-thiogalactopyranoside) when using a non-autoinducible culture medium. Purification of the cyclized protein can be performed using chromatographic methods (affinity, size exclusion, etc.) to be selected according to the biophysical properties of the expressed construct.

[0046] EXAMPLE 2: CYCLIZATION RESULTS OF PROTEINS OF INTEREST

[0047] I- Helical segment of sequence EAAAKEAAAK (SEQ ID NO: 2)

[0048] The sequences described in Figures 2A and 2B were inserted into a pnEA vector encoding an additional 6xHIS tag at the N-terminus of the construct. This vector was used to transform an E. coli BL21(DE3) strain. The bacteria were then cultured under shaking at 37°C in LB medium. When the absorbance of the medium (measured at 600 nm) reached 0.7T, the IPTG concentration was increased to 0.25 mM to induce plasmid expression of the cyclized protein. After 16 hours at 20°C, a cell pellet was recovered by centrifugation and then dissolved in a lysis buffer containing 25 mM HEPES, pH 7.5, 300 mM NaCl, and 10 mM imidazole. After centrifugation, the soluble fraction was placed in contact with a TALON resin to which the 6xHIS label was attached. Following three successive washes with lysis buffer, the proteins retained on the resin were eluted in a buffer containing 25 mM HEPES, pH 7.5, 300 mM NaCl and 300 mM Imidazole. The purification analysis was carried out on polyacrylamide gel under denaturing conditions and stained with Coomassie blue.

[0049] Figure 2C shows that the helix cyclized with the ZN-ring and TP-ring systems is present in the soluble fraction after lysis and centrifugation of the bacteria (SSo lane). Some of the constructs are found in the sonication pellet (Cso lane), corresponding to the constructs for which cyclization did not occur correctly. These are therefore efficiently removed at the beginning of the purification process. The cyclized constructs are efficiently retained on the resin (Fix. lane). Elution yields the cyclized EAAAKEAAAK helix (SEQ ID NO: 2) in solution (Elu. lane).

[0050] Figure 2D shows, by differential scanning calorimetry (DSC), that the half-denaturation temperature of the constructs is very high, namely 96 and 100°C for the EAAAKEAAAK helix (SEQ ID NO: 2), cyclized by the TP-ring and ZN-ring systems, respectively. DSC measurements were performed with protein concentrations of 30 pM in a 25 mM HEPES buffer, pH 7.5, 300 mM NaCl, using a VP-DSC instrument (MicroCai, Malvern, Ltd).

[0051] II- Human galectin-3 CRD domain (CRD-gal3)

[0052] The CRD-gal3 domain enables protein purification (International Application WO 2017 / 194888)

[0015] . It is a domain capable of binding D-lactose. Here, this domain was cyclized using the TP-ring system. In this case, the polyglycine segment and cloning site options were selected. The cyclized CRD-gal3 protein sequence was introduced into a pnEA vector.

[0053] TP-ring-CRD-gal3:

[0054] TAH1 (93-111 )-Linker-X / 7Q / -CRD-qal3(20-156)-Sac / -Linker-PIH1 (262-344) SVQIPVVEVDELPEGYDRSGGGGSLELIVPYNLPLPGGVVPKMLITILGTVKPNANRI ALDFQRGNDVAFHFNPRFNENNRRVIVCNTKLDNNWGREERQSVFPFESGKPFKI QVLVEPDHFKVAVNDAHLLQYNHRVKKLNEISKLGISGDIDLTSTSYTMIELGGGGSEDVPEYEVKMKRFKGAAYKLRILIENKAPNSKPDRFSPSYNFAENILYINGKLSIPLP RDIWNAADIKIFHIRKERTLYIYI (SEQ ID NO: 7)

[0055] The purification protocol is identical to that described for the cyclized EAAAKEAAAK helix (SEQ ID NO: 2), except that a lactose-sepharose resin was used. This resin retained CRD-gal3. Elution was performed using lysis buffer supplemented with 200 mM D-lactose. Figure 3A shows that CRD-gal3 cyclized with the TP-ring system is present in the soluble fraction after lysis and centrifugation of the bacteria (SSo lane). Some of the constructs are found in the sonication pellet (Cso lane), corresponding to constructs for which cyclization did not occur correctly. These are therefore efficiently removed at the beginning of the purification process. The cyclized constructs are effectively retained on the resin (Fix. lane), thus demonstrating the safety of the TP-ring on the function and three-dimensional folding of CRD-gal3. The elution is also very efficient and allows obtaining cyclized CRD-gal3 in solution (Elu. track).

[0056] Figure 3B shows, by differential scanning calorimetry (DSC), that the half-denaturation temperature of the CRD-gal3 domain was increased by approximately 7% when this protein was cyclized by the TP-ring method (increasing from 56 to 60°C). Cyclization thus increased the thermostability of CRD-gal3. DSC measurements were performed with 30 pM protein concentrations in a 25 mM HEPES buffer, pH 7.5, 300 mM NaCl, using a VP-DSC instrument (MicroCai, Malvern, Ltd).

[0057] Ill-Cellulase (Cel-CD)

[0058] This enzyme is capable of digesting cellulose, and the digestion products have potential for biofuel production. It has been cyclized using TP-ring and ZN-ring systems. In this case, the polyglycine segment and cloning site options were selected. The constructs shown below were introduced into a pnEA expression plasmid vector providing an additional 6xHIS tag at the N-terminus of cyclized Cel-CD. The sequences of cyclized Cel-CD are shown below:

[0059] TP-ring-Cel-CD:

[0060] TAH1 (93-111 )-Linker-X / 7Q / -Cel-CD(9-365)-Sac / -Linker-PIH1 (262-344)SVQIPVVEVDELPEGYDRSGGGGSLEDNFKHLLGNDNVKRPSEAGALQLQEVDGQ MTLVDQHGEKIQLRGMSTHGLQWFPEILNDNAYKALANDWESNMIRLAMYVGENG YASNPELIKSRVIKGIDLAIENDMYVIVDWHVHAPGDPRDPVYAGAEDFFRDIAALYP NNPHIIYELANEPSSNNNGGAGIPNNEEGWNAVKEYADPIVEMLRDSGNADDNIIIVG SPNWSQRPDLAADNPIDDHHTMYTVHFYTGSHAASTESYPPETPNSERGNVMSNT RYALENGVAVFATEWGTSQANGDGGPYFDEADVWIEFLNENNISWANWSLTNKNE VSGAFTPFELGKSNATSLDPGPDQVWVPEELSLSGEYVRARIKGVNYELGGGGSE DVPEYEVKMKRFKGAAYKLRILIENKAPNSKPDRFSPSYNFAENILYINGKLSIPLP RDIWNAADIKIFHIRKERTLYIYI (SEQ ID NO : 8)

[0061] ZN-ring-Cel-CD :

[0062] ZNHIT3(265-332)-Linker-X / 7Q / -Cel-CD(9-365)-Sac / -Linker-NUFIP1 (817-841 ) DYDMLTEEQKKKLKEDHTLKILLKNNYVREVFKQFTLSNDKIGYLSHYINDPTIVQVI DH I MKTI DDTGGGGSLEDNFKHLLGNDNVKRPSEAGALQLQEVDGQMTLVDQHGE KIQLRGMSTHGLQWFPEILNDNAYKALANDWESNMIRLAMYVGENGYASNPELIKS RVIKGIDLAIENDMYVIVDWHVHAPGDPRDPVYAGAEDFFRDIAALYPNNPHIIYELA NEPSSNNGGAGIPNNEEGWNAVKEYADPIVEMLRDSGNADDNIIIVGSPNWSQRP DLAADNPIDDHHTMYTVHFYTGSHAASTESYPPETPNSERGNVMSNTRYALENGV (SEQ ID NO: 9)

[0063] The purification protocol is identical to that described for CRD-gal3, except that a TALON resin was used. This resin allowed the retention of the 6xHIS tag introduced at the N-terminus of cyclized Cel-CD. Elution was performed using lysis buffer supplemented with 300 mM imidazole.

[0064] Figure 4 shows that recombinant Cel-CD constructs with the TP-ring or ZN-ring system (TP-ring or ZN-ring, respectively) are also very widely overexpressed and predominantly soluble. Again, the uncyclized constructs are efficiently eliminated in the insoluble fraction (Cso).

[0065] IV- PET hydrolases

[0066] Two PET hydrolases were used: one from the microorganism Ideonella sakaiensis (Is-PETase), and the other industrially exploited by CARBIOS (carbios-PETase). The purification protocol implemented was the same as that used for cyclized Cel-CD. In the case of Is-PETase, the polyglycine segment option was not selected. Conversely, this option was included for carbios-PETase. The cloning site option was present in both cases. The sequences of the cyclized proteins are shown below:

[0067] TP-ring-ls-PETase:

[0068] TAH1 (93-111 ) -X / îoZ-ls-PETase-Sac / -PIH1 (262-344) SVQIPVVEVDELPEGYDRSLEMTNPYARGPNPTAASLEASAGPFTVRSFTVSRPSG YGAGTVYYPTNAGGTVGAIAIVPGYTARQSSIKWWGPRLASHGFWITIDTNSTLDQ PSSRSSQQMAALRQVASLNGTSSSPIYGKVDTARMGVMGHSMGGGGSLISAANNP SLKAAAPQAPWDSSTNFSSVTVPTLIFACENDSIAPVNSSALPIYDSMSRNAKQFLEI NGGSHFCANSGNSNQALIGKKGVAWMKRFMDNDTRYSTFACENPNSTRVSDFRT ANCSELEDVPEYEVKMKRFKGAAYKLRILIENKAPNSKPDRFSPSYNFAENILYING KLSIPLPRDIVVNAADIKIFHIRKERTLYIYI (SEQ ID NO: 10)

[0069] ZN-ring-ls-PETase :

[0070] ZNHIT3(265-332)- X / ?o / -ls-PETase- Sac / -NUFIP1(817-841) DYDMLTEEQKKKLKEDHTLKILLKNNYVREVFKQFTLSNDKIGYLSHYINDPTIVQVI DH I MKTI DDTLEMTNPYARGPNPTAASLEASAGPFTVRSFTVSRPSGYGAGTVYYP TNAGGTVGAIAIVPGYTARQSSIKWWGPRLASHGFVVITIDTNSTLDQPSSRSSQQM AALRQVASLNGTSSSPIYGKVDTARMGVMGHSMGGGGSLISAANNPSLKAAAPQA PWDSSTNFSSVTVPTLIFACENDSIAPVNSSALPIYDSMSRNAKQFLEINGGSHFCA NSGNSNQALIGKKGVAWMKRFMDNDTRYSTFACENPNSTRVSDFRTANCSELDIY TYEKKLIKSIEYITKNKFFDDS (SEQ ID NO: 11)

[0071] TP-ring-carbios-PETase :

[0072] TAH1 (93-111 )-Linker-X / 7Q / -carbios-PETase-Sac / -Linker-PIH1 (262-344) SVQIPWEVDELPEGYDRSGGGGSLEMSNPYQRGPNPTRSALTADGPFSVATYTV SRLSVSGFGGGVIYYPTGTSLTFGGIAMSPGYTADASSLAWLGRRLASHGFWLVIN TNSRFDGPDSRASQLSAALNYLRTSSPSAVRARLDANRLAVAGHSMGGGGTLRIAE QNPSLKAAVPLTPWHTDKTFNTSVPVLIVGAEADTVAPVSQHAIPFYQNLPSTTPKV YVELCNASHIAPNSNNAAISVYTISWMKLWVDNDTRYRQFLCNVNDPALCDFRTNNRHCQELGGGGSEDVPEYEVKMKRFKGAAYKLRILIENKAPNSKPDRFSPSYNFAE NILYINGKLSIPLPRDIWNAADIKIFHIRKERTLYIYI (SEQ ID NO: 12)

[0073] ZN-ring-carbios-PETase :

[0074] ZNHIT3(265-332)-Linker-X / 7Q / -carbios-PETase-Sac / -Linker-NUFIP1 (817-841 ) DYDMLTEEQKKKLKEDHTLKILLKNNYVREVFKQFTLSNDKIGYLSHYINDPTIVQVI DH I MKTI DDTGGGGSLEMSNPYQRGPNPTRSALTADGPFSVATYTVSRLSVSGFG GGVIYYPTGTSLTFGGIAMSPGYTADASSLAWLGRRLASHGFWLVINTNSRFDGPD SRASQLSAALNYLRTSSPSAVRARLDANRLAVAGHSMGGGGTLRIAEQNPSLKAAV PLTPWHTDKTFNTSVPVLIVGAEADTVAPVSQHAIPFYQNLPSTTPKVYVELCNASHI APNSNNAAISVYTISWMKLWVDNDTRYRQFLCNVNDPALCDFRTNNRHCQELGGG GSDIYTYEKKLIKSIEYITKNKFFDDS (SEQ ID NO: 13)

[0075] Figure 4B shows that cyclization using the TP-ring and ZN-ring systems yielded significant quantities of soluble Is-PETase, whereas the latter is not identifiable when produced uncycled (as seen by comparing the intensity of the bands on the left-hand gel with those observed on the middle and right-hand gels). The inventors, as well as the literature (Fecker et al., 2018)

[0016] , have observed that this industrially intended enzyme can be completely insoluble, thus requiring complex renaturation protocols. Cyclization using the TP-ring and ZN-ring systems bypasses this requirement. Finally, Figure 4C shows that the cyclization systems according to the invention also allow for the purification of soluble CARBIOS-PETase in good proportions. References

[0076] [1] Kumar et al., Protein Eng., 13(3): 179-191, 2000

[0077] [2] Yokota et al., Science and Technology of Advanced Materials, 7(3): 255-262, 2006 [3] Strickler et al., Biochemistry, 45(9): 2761-2766, 2006

[0078] [4] Hayes et al., Org. Biomol. Chem., 19(18): 3983-4001, 2021

[0079] [5] Sun et al., N Biotechnol., 25: 28-36, 2019

[0080] [6] Buldun et al., J. Am. Chem. Soc., 40(8): 3008-3018, 2018

[0081] [7] Wang et al., Biotechnol. Biofuels, 31;9:79. doi: 10.1186 / s13068-016-0490-5, 2016 [8] Hayes et al., Scientific Reports, 13(1): 1267. doi: 10.1038 / s41598-023-27780-4, 2023

[0082] [9] Chagot et al., Biochemistry, 61(7): 479-493, 2022

[0083]

[0010] Henri et al., Structure, 26(9): 1196-1209.e8, 2018

[0084]

[0011] Quinternet et al., Structure, 24(10): 1693-1706, 2016

[0085]

[0012] Quinternet et al., J. Mol. Biol., 427(17): 2816-2839, 2015

[0086]

[0013] Quinternet et al., Biomol. NMR Assign., 2015

[0087]

[0014] Rothé et al., Nue. Ac. Research, 2014

[0088]

[0015] International Application WO 2017 / 194888

[0089]

[0016] Feckeretal., Biophys. J., 114(6): 1302-1312, 2018

Claims

DEMANDS 1) Cyclized fusion protein comprising a cyclization system consisting of a peptide and its protein receptor fused respectively to the N- and C-terminal ends of a protein of interest, where said peptide is the peptide NUFIP1 or TAH1 and said protein receptor is respectively the protein ZNHIT3 or PIH1. 2) Cyclized fusion protein according to claim 1, wherein the ZNHIT3 / NUFIP1 cyclization system is that of P. falciparum, H. sapiens or S. cerevisiae. 3) Cyclized fusion protein according to claim 1 or 2, wherein the PIH1 / TAH1 cyclization system is that of S. cerevisiae. 4) Fusion protein according to any one of claims 1 to 3, wherein the peptide NUFIP1 has the sequence SEQ ID NO: 4 or a sequence having at least 17% identity with SEQ ID NO: 4, the protein receptor ZNHIT3 has the sequence SEQ ID NO: 3 or a sequence having at least 18% identity with SEQ ID NO: 3, the peptide TAH1 has the sequence SEQ ID NO: 6, and / or the protein receptor PIH1 has the sequence SEQ ID NO:

5. 5) Use of a cyclized fusion protein according to any one of claims 1 to 4 for the expression and / or purification of a protein of interest. 6) An in vitro process for the expression and purification of a protein of interest, characterized in that it comprises the following steps: a) fusion of a cyclization system consisting of a peptide and its protein receptor respectively fused to the N- and C-terminal ends of a protein of interest; b) expression in a host cell of the cyclized fusion protein obtained in step a); c) purification of the cyclized fusion protein; and d) optionally isolation of the protein of interest from the cyclization system; where said peptide is the NUFIP1 or TAH1 peptide and said protein receptor is respectively the ZNHIT3 or PIH1 protein.7) Method according to claim 6, wherein the ZNHIT3 / NUFIP1 cyclization system is that of P. falciparum. 8) Method according to claim 6 or 7, wherein the PIH1 / TAH1 cyclization system is that of S. cerevisiae. 9) A method according to any one of claims 6 to 8, wherein the host cell is a non-human eukaryotic cell or a prokaryotic cell. 10) A method according to any one of claims 5 to 9, wherein the peptide NUFIP1 has the sequence SEQ ID NO: 4 or a sequence having at least 17% identity with SEQ ID NO: 4, the protein receptor ZNHIT3 has the sequence SEQ ID NO: 3 or a sequence having at least 18% identity with SEQ ID NO: 3, the peptide TAH1 has the sequence SEQ ID NO: 6, and / or the protein receptor PIH1 has the sequence SEQ ID NO: 5.

Images