De la reconnaissance protéine-glucides. Une activité lectine avec une spécificité mannose/glucose a été détectée dans les graines de Platypodium elegans, une légumineuse de la sous-tribu Dalbergiae. Le gène de la lectine PELa a été cloné. Son produit est une protéine de 261 acides aminés appartenant à la famille des lectines de légumineuses et présentant des similarités avec l'agglutinine de Pterocarpus angolensis (PAL). La lectine recombinante a été exprimée dans E. coli et renaturée à partir des corps d'inclusion. L'analyse de la spécificité par Glycan Array montre une préférence très rare pour des N-glycanes de type complexe avec des branches disymmétriques. Une branche courte composée d'un résidu de mannose est préférée sur le bras 1-6 des N-glycanes, tandis que l'extension par les résidus GlcNAc et Gal et favorable sur le bras 1-3. Les affinités ont été mesurées par microcalorimétrie de titration en utilisant des heptasaccharides liés à une asparagine et obtenus par une méthode semi-enzymatique. Une très forte affinité de 5 uM a été obtenue pour deux ligands symétriques et disymmétriques. Les structures cristallographiques de PELa complexé avec le trimannose branché et l'heptasaccharide-Asn symétrique de type complexe ont été résolues respectivement à 2,1 et 1,65 Å de résolution. La lectine adopte l'organisation dimérique canonique des lectines de légumineuses. Le trimannose ponte les sites de liaison de deux dimères voisins, résultant en la formation de chaînes infinies dans le cristal. L'heptasaccharide-Asn se lie par le mannose du bras 1-6 dans le site principal de liaison et de nombreux contacts supplémentaires sont établis avec les autres résidus glucidiques. Le GlcNAc du bras 1-3 interagit avec la surface de la protéine dans une conformation contrainte qui peut expliquer la plus grande affinité que l'on observe sur les puces pour les oligosaccharides avec des bras 1-3 courts qui ne contiennent pas ce monosaccharide. / Lectin activity with specificity for mannose and glucose has been detected in the seed of Platypodium elegans, a legume from the Dalbergiae tribe. The gene of the lectin PELa has been cloned and the resulting 261 amino acid protein belongs to the legume lectin family with similarity with Pterocarpus angolensis agglutinin (PAL) from the same tribe. The recombinant lectin has been expressed in E. coli and refolded from inclusion bodies. Analysis of specificity by Glycan Array evidenced a very unusual preference for complex type N-glycans with asymmetrical branches. A short branch consisting of one mannose residue is preferred on the 6- arm of the N-glycan, while extension by GlcNAc, Gal and NeuAc are favorable on the 3-arm. Affinities have been obtained by microcalorimetry using symmetrical and asymmetrical Asn- linked heptasaccharide prepared by semi-enzymatic method. Strong affinity of 5 µM was obtained for both ligands. Crystal structures of PELa complexed with branched trimannose and symmetrical complex type Asn-linked heptasaccharide have been solved at 2.1 and 1.65 Å resolution respectively. The lectin adopts the canonical dimeric organization of legume lectins. The trimannose bridges the binding sites of two neighbouring dimers, resulting in the formation of infinite chains in the crystal. The Asn-linked heptasaccharide binds with the 6-arm in the primary binding site and extensive additional contacts on both arms. The GlcNAc on the 3-arm is bound in a constrained conformation that may rationalize the higher affinity that is observed on chips for oligosaccharide with shorter 3-arm that do not present this monosaccharide.

The thesis entitled “Folding studies on Peanut Agglutinin: A lectin with an unusual quaternary structure” deals with the several aspects of the folding of the tetrameric legume lectin Peanut Agglutinin (PNA). PNA is a well studied legume lectin and several interesting observations regarding its unfolding have been published from our laboratory. The present thesis is an extension of the same work to enrich our knowledge about the folding behaviour of PNA. The thesis describes both experimental as well as theoretical insight on unfolding of PNA. Chapter 1 is a general discussion on lectins. Lectins are carbohydrate binding proteins of non immune source. Lectins are generally found in all type of organisms- plants, animals as well as micro-organisms. Among the plant lectins “legume lectin” is a very well studied system. Legume lectins share a general tertiary structural fold; “jelly roll fold” while they vary in their quaternary structure. Thus they can be considered as “natural mutants” in the context of quaternary structure. The origin of the lectins, structure and sugar specificity have been discussed with emphasis on legume lectin family. Chapter 2 describes the thermodynamics related to the urea induced denaturation of PNA. PNA shows a very interesting unfolding profile, populating one molten globule like intermediate during thermal as well as chaotrope induced denaturation. The molten globule like intermediate loses most of its tertiary structure but retains sufficient secondary structure. Surprisingly, the molten globule like state retains its carbohydrate binding specificity like the native PNA. A model has been developed to fit the chaotrope induced three state denaturation profile of PNA. The model considers the tetramer to dissociate to monomeric intermediate, which in turn dissociates to complete denatured state. All the relevant thermodynamic parameters (∆G, ∆Cp, Tg) associated in the denaturation process have been extracted. The tetramer is found to be ~30 kcal/Mole more stable compared to the intermediate and the intermediate is ~8 kcal/Mole more stable compared to the denatured. The denaturation process has been followed by the changes in hydrodynamic radii by dynamic light scattering (DLS). The profile of change in hydrodynamic radius and the % intensity clearly identify the generation of two species simultaneously. The analysis shows that the intermediate is ~40 % unfolded in nature. Thus this chapter deals with the detailed study of thermodynamics and dynamic light scattering study of the urea induced denaturation of PNA. Chapter 3 deals with the effect of 2, 2, 2 - trifluoroethanol (TFE) on the structure of PNA at two different pH. TFE is a well known co-solvent and is widely used to induce α- helical structure in a protein. The secondary structures induced by TFE are assumed to reflect conformations that prevail during early stages of protein folding. Thus it was quite interesting to notice the structural changes induced by TFE. The effect of TFE has been studied at two different pH- neutral pH of 7.4 and acidic pH 2.5. The  structure of the protein is accentuated in the presence of TFE at low concentration at both the pH. TFE induces α-helical structure from 40 % (v/v) concentration onwards at both the pH. TFE at 15 % concentration induces a molten globule like structure at low pH. The quenching of acrylamide suggests that the protein at low pH and 15 % TFE concentration has a more compact structure compared to the protein at low pH in absence of TFE as well as 6M guanidine hydrochloride (GdnHCl). Further studies of hydrodynamic radii by dynamic light scattering (DLS) also reveal that the protein undergoes some kind of compaction in presence of 15 % TFE at low pH. The induction of this type of molten globule like state at neutral pH has not been observed. Chapter 4 describes the molecular dynamics simulation of deoligomerization of PNA. The native PNA (PDB code 2PEL), excluding any ligand and metal ions has been simulated at 300 K, 400 K, 500 K and 600 K for 500 ps. The overall destabilisation has been followed by root mean square deviation (RMSD), the radius of gyration (Rg) and the solvent accessible surface area (ASA), while the atomistic details are revealed by residue wise RMSD (RRMS), hydrogen bonds and cluster analysis. The protein shows a quite a dramatic change in RMSD and radius of gyration profile at 600 K. RRMS shows that the residues belonging to the loops, mainly in the metal binding site show quite high flexibility. The relative change in average accessible surface area reveals that the primary core of the protein is exposed at 600 K while it is well buried till 500 K. The hydrogen bond analysis clearly shows that with increase in temperature number of hydrogen bonds starts decreasing. Mainly the hydrogen bonds involving side chain interactions are broken. Surprisingly, not all the monomers behave similarly. Monomers C and D are more perturbed compared to monomers A and B. The asymmetry in the interfaces of the monomers may be the key reason for it. The change in the interfaces has been probed by hydrogen bond analysis and cluster analysis. The GSIV type interfaces (A-D and B-C) have been found out to be the most dynamic in nature compared to the other two interfaces. Thus, this chapter reveals the early stage of unfolding of PNA, where perturbation in secondary and tertiary structural level is quite prominent but the interfaces are still holding weakly and are not completely dissociated. Chapter 5 is the continuation of the molecular dynamics simulation on unfolding of PNA, where the effect of metal ions has been illustrated. The monomeric PNA has been simulated to compare its dynamics with the tetramer. The metal binding loop (125-135) becomes unstable and opens up for the monomer even at 300 K after 800 ps. The monomer at 600 K is completely disorganized. The instability of the metal binding loop of the monomer triggers the urge to study the simulation in presence of metal ions (Ca2+ and Mn2+). The monomer bound with metal ions shows steady fluctuation at 300 K. Binding of metal ions seems to bring stability even at 600 K. Surprisingly binding of metal ions to the metal binding site not only stabilises the metal binding loop but also stabilises residues at back beta sheet which are involved in oligomerization. Hence, another simulation of the tetramer at 600 K bound with metal ions has been done. It has been shown that binding of metal ions increases the stability of the protein without altering the denaturation pathway. Appendix A describes a completely different study from PNA. The initial spectral and kinetic characterization of 7, 8- Diaminopelargonic acid Synthase (DAPA Synthase) has been done from Mycobacterium tuberculosis. The DAPA Synthase gene has been cloned earlier in our laboratory and the same has been used for further studies. This is a well known pyridoxal-5′ phosphate (PLP) dependent enzyme, which converts 8- Amino-7-oxopelargonic Acid (KAPA) to 7, 8-Diaminopelargonic Acid (DAPA) in the second step of biotin biosynthesis. DAPA Synthase uses S-adenosylmethionine (SAM) and KAPA as substrate. The first half of the enzymatic reaction has been followed spectroscopically, both by steady state and stopped flow. The enzyme seems to undergo change in conformation as evident from fluorescence and circular dichroism study. The Km value has been determined using bioassay technique. The detailed characterization of the enzyme has been described in this chapter.

Das Lektin aus <i>Pisum sativum</i>, der Gartenerbse, ist Teil der Familie der Leguminosenlektine. Diese Proteine haben untereinander eine hohe Sequenzhomologie, und die Struktur ihrer Monomere, ein all-ß-Motiv, ist hoch konserviert. Dagegen gibt es innerhalb der Familie eine große Vielfalt an unterschiedlichen Quartärstrukturen, die Gegenstand kristallographischer und theoretischer Arbeiten waren. Das Erbsenlektin ist ein dimeres Leguminosenlektin mit einer Besonderheit in seiner Struktur: Nach der Faltung in der Zelle wird aus einem Loop eine kurze Aminosäuresequenz herausgeschnitten, so dass sich in jeder Untereinheit zwei unabhängige Polypeptidketten befinden. Beide Ketten sind aber stark miteinander verschränkt und bilden eine gemeinsame strukturelle Domäne. Wie alle Lektine bindet Erbsenlektin komplexe Oligosaccharide, doch sind seine physiologische Rolle und der natürliche Ligand unbekannt. In dieser Arbeit wurden Versuche zur Entwicklung eines Funktionstests für Erbsenlektin durchgeführt und seine Faltung, Stabilität und Monomer-Dimer-Gleichgewicht charakterisiert. Um die spezifische Rolle der Prozessierung für Stabilität und Faltung zu untersuchen, wurde ein unprozessiertes Konstrukt in <i>E. coli</i> exprimiert und mit der prozessierten Form verglichen. <br /> <br /> Beide Proteine zeigen die gleiche kinetische Stabilität gegenüber chemischer Denaturierung. Sie denaturieren extrem langsam, weil nur die isolierten Untereinheiten entfalten können und das Monomer-Dimer-Gleichgewicht bei mittleren Konzentrationen an Denaturierungsmittel auf der Seite der Dimere liegt. Durch die extrem langsame Entfaltung zeigen beide Proteine eine apparente Hysterese im Gleichgewichtsübergang, und es ist nicht möglich, die thermodynamische Stabilität zu bestimmen. Die Stabilität und die Geschwindigkeit der Assoziation und Dissoziation in die prozessierten bzw. nichtprozessierten Untereinheiten sind für beide Proteine gleich. Darüber hinaus konnte gezeigt werden, dass auch unter nicht-denaturierenden Bedingungen die Untereinheiten zwischen den Dimeren ausgetauscht werden.<br /> <br /> Die Renaturierung der unprozessierten Variante ist unter stark nativen Bedingungen zu 100 % möglich. Das prozessierte Protein dagegen renaturiert nur zu etwa 50 %, und durch die Prozessierung ist die Faltung stark verlangsamt, der Faltungsprozess ist erst nach mehreren Tagen abgeschlossen. Im Laufe der Renaturierung wird ein Intermediat populiert, in dem die längere der beiden Polypeptidketten ein Homodimer mit nativähnlicher Untereinheitenkontaktfläche bildet. Der geschwindigkeitsbestimmende Schritt der Renaturierung ist die Assoziation der entfalteten kürzeren Kette mit diesem Dimer. / The lectin from <i>Pisum sativum</i> (garden pea) is a member of the family of legume lectins. These proteins share a high sequence homology, and the structure of their monomers, an all-ß-motif, is highly conserved. Their quaternary structures, however, show a great diversity which has been subject to cristallographic and theoretical studies. Pea lectin is a dimeric legume lectin with a special structural feature: After folding is completed in the cell, a short amino acid sequence is cut out of a loop, resulting in two independent polypeptide chains in each subunit. Both chains are closely intertwined and form one contiguous structural domain. Like all lectins, pea lectin binds to complex oligosaccharides, but its physiological role and its natural ligand are unknown. In this study, experiments to establish a functional assay for pea lectin have been conducted, and its folding, stability and monomer-dimer-equilibrium have been characterized. To investigate the specific role of the processing for stability and folding, an unprocessed construct was expressed in <i>E. coli</i> and compared to the processed form.<br /> <br /> Both proteins have the same kinetic stability against chemical denaturant. They denature extremely slowly, because only the isolated subunits can unfold, and the monomer-dimer-equilibrium favors the dimer at moderate concentrations of denaturant. Due to the slow unfolding, both proteins exhibit an apparent hysteresis in the denaturation transition. Therefore it has not been possible to determine their thermodynamic stability. For both proteins, the stability and the rates of association and dissociation into processed or unprocessed subunits, respectively, are equal. Furthermore it could be shown that even under non-denaturing conditions the subunits are exchanged between dimers.<br /> <br /> Renaturation of the unprocessed variants is possible under strongly native conditions with 100 % yield. The processed protein, however, can be renatured with yields of about 50 %, and its refolding is strongly decelerated. The folding process is finished only after several days. During renaturation, an intermediate is populated, in which the longer of the two polypeptide chains forms a homodimer with a native-like subunit interface. The rate limiting step of renaturation is the association of the unfolded short chain with this dimer.