Le gène de l’entérotoxine thermostable b (estB) d’Escherichia coli a été fusionné au gène de la protéine liant le maltose (malE) dans le vecteur pMAL-p via PCR. Par la suite, deux constructions plasmidiques ont été realisées à partir de ce nouveau vecteur, nommé pMAL-STb. Dans un premier temps, un marqueur hexahistidine (His6) a été ajouté entre malE et estB, et dans un deuxième temps, un marqueur décahistidine (His10) a été placé en amont de malE. La séquence signal de la protéine liant le maltose (MBP) dirige l’exportation de la protéine de fusion du cytoplasme vers le périplasme, où l’entérotoxine STb acquière sa conformation active. MBP est également reconnue pour améliorer le rendement et la solubilité de la protéine passagère tandis que le marqueur histidine, connu comme étant le meilleur marqueur d’affinité pour la purification protéique, facilite sa purification jusqu’à homogénéité. De plus, les gènes fusionnés sont sous le contrôle du promoteur tac (Ptac), un promoteur fort et inductible. Suite à l’induction par l’IPTG, la souche recombinante exprime une protéine d’environ 48 kDa, qui est facilement identifiable par électrophorèse à partir du surnageant obtenu via choc osmotique. Une séquence encodant un site de clivage spécifique au facteur Xa est présente dans le plasmide afin de séparer les marqueurs MBP et histidine de STb. Le clivage de la protéine de fusion avec le facteur Xa libère MBP (42 kDa) attachée au marqueur histidine et un polypeptide de 5.2 kDa, correspondant au poids moléculaire de STb mature. Avec cette méthode, nous visons à obtenir une méthode plus efficace pour la production et la purification de STb. / The heat-stable enterotoxin b gene (estB) of Escherichia coli was fused to the gene for maltose-binding protein (malE) into the pMAL-p vector using PCR. Afterward, two plasmid constructs were realized from this new vector, named pMAL-STb. Firstly, a hexahistidine tag (His6) was added between malE and estB and secondly, a decahistidine tag (His10) was placed upstream of malE. The signal sequence of maltose-binding protein (MBP) directs the export of the fusion protein from the cytoplasm to the periplasm, where the enterotoxin STb acquires its active conformation. MBP is also known to improve the yield and solubility of the passenger protein while the histidine tag, viewed as the best affinity tag for protein purification, facilitates its purification to homogeneity. Furthermore, the fused genes are controlled by the tac promoter (Ptac), a strong inducible promoter. Following IPTG induction, the recombinant strain expressed a protein of approximately 48 kDa, which is easily identified from osmotic shock fluid following electrophoresis. A sequence encoding a factor Xa cleavage site is present in the plasmid to separate MBP and histidine tags from STb. The cleavage of the fusion protein with factor Xa generates the maltose-binding protein (42 kDa) attached to the histidine tag and a polypeptide of 5.2 kDa, corresponding to the molecular mass of mature STb. With this method, we aim at obtaining a more efficient way to produce and purify STb.

Chapter 1gives a general introduction to the CXXC motif found in natural proteins. It then reviews the studies where disulphides were engineered in various proteins. The various strategies developed to engineer metal binding activity and redox activity are described. The objectives behind engineering the CXXC motif into a protein, such as imparting it novel metal-binding and redox activities, are discussed next. Alternative strategies which achieve the same objectives are described as well. This chapter then introduces the model proteins used in the course of this thesis: maltose-binding protein (MBP) and E. coli. Thioredoxin (Trx). This chapter also briefly discusses the role of signal peptide in protein export. Chapter 2describes the experimental studies and their results in which we introduced the widely occurring cysteine motif CXXC into the maltose binding protein (one-at-a-time, in five alpha-helices, at the N-termini) to test three hypotheses: 1) Does a disulphide bond form at the N-terminus? 2) Does the protein acquire any oxido-reductase activity? 3) Does it acquire new metal-binding properties? The results confirmed: 1) Each cysteine pair forms a stable intrahelical disulphide bond under non-reducing conditions. 2) The five mutant proteins acquire considerable oxidoreductase activity, tested by the insulin aggregation assay. 3) The mutants acquire novel metal-binding properties for Ni2+, Cd2+, and Zn2+ upon reduction. Further, introducing the CXXC motif neither destabilizes the protein nor affects its global structure. Our results demonstrated that introduction of CXXC motifs can be used to probe alpha-helix start sites and to introduce oxidoreductase and metal binding functionality into proteins. Chapter 3describes further experimentson a few of the metal ion binding mutants discussed in the previous chapter. We explore the effect and usefulness of reducing agents (DTT and TCEP) on the binding of metal salts to the CXXC mutants. We also studied the explore of metal salts on the thermal stability of the mutants and show that metal ions bind to the CXXC motif even when the protein is in the unfolded state. The chapter describes the use of an immobilized metal affinity chromatography (IMAC) based method for the purification of MBP mutants.Yields ranging from 60-85% were obtained for thethree MBP mutants. The cysteines were located at different positions in thesethree MBP mutants (MBP 42-45 Cys, MBP 128-131 Cys, and MBP 359-359 Cys mutants). The yields for wild-type MBP, a single cysteine mutant (MBP S211C), a double cysteine mutant (MBP 230, 30) were all below 15%. Chapter 3 also reports a new crystal structure of the MBP356-359 mutant in ligand bound form:it crystallizes as an intermolecular dimer, bonded by two disulfides formed by the cysteines of the CXXC motif. Chapter 4describes the effects of inserting signal peptide sequences on protein folding and expression. We fused the malE and pelB signal sequences at the N-terminus of the model protein thioredoxin and observed that the wild-type and pelB fusion constructs are soluble when expressed, but the malE construct was targeted to inclusion bodies. Nonetheless, it could be refolded in vitro to yield a monomeric product with a secondary structure identical to the wild-type thioredoxin. This chapter also details the thermodynamic stability, aggregation propensity and activity of the purified recombinant proteins in comparison with the wild-type thioredoxin. The presence of the signal sequences reduces the thermodynamic stability and activity of the recombinants and increases their aggregation propensity, with malE having much larger effects than pelB. These studies show that besides acting as address labels, different signal sequences affect protein stability and aggregation differently. Chapter 5describes three different strategies to label a protein at different sites with cysteine-specific fluorophores using MBP as the model. The first strategy exploits the differential accessibility of residues within MBP in its maltose-bound and maltose-free states. The second strategy involves insertion of a 14-amino-acid loop called V3 from the HIV gp120 protein into MBP; anti-V3 antibodies shield the cysteine residue present inside the inserted loop, while we label another cysteine present outside the loop. In the third strategy, we introduce a third cysteine residue onto the background of the MBP mutant already containing a disulphide bridge at the N-terminus of one of its helices (discussed in Chapter 2). We label the third, free cysteine while the cysteines involved in the disulphide bridge remain protected. We observed successful differential labelling using the first strategy and also observed FRET between the fluorophore labels. Similarly, after trying the second strategy we could individually label all the mutants except one. The third strategy based on the triple-cysteine mutant was not successful because the fluorophore we chose (DBM) did not show site specificity and instead labelled all three cysteines. In addition, the triple-cysteine mutant did not even show disulphide-bridge formation.We showed that indeed the V3 loop inserted in MBP binds anti-V3 antibodies and we could individually label all the mutants expect D41C. The third strategy was not successful because unfortunately in the triple cysteine mutant, the fluorophore we chose (DBM) did not show site specificity and labeled all three cysteines. In addition, the disulfide bridge was not found to be present in the triple cysteine mutant. Chapter 6discusses the synthesis, characterization and binding of various maltolipids, (and their corresponding maltose-free controls) to MBP. The maltolipids were synthesised with varying linker lengths and anchor- & head-groups and then used to prepare liposomes and micelles. Although both liposomal and micellar forms could bind to MBP, only the micelles were screened subsequently for their ability to bind to MBP. The binding was assessed using various techniques such as fluorescence spectroscopy, gel filtration and thermal stability assay. We screened the maltolipids and determined how their anchor group, linker length and charge on the head group influences the binding of MBP to micelles formed by these maltolipids.