Les protéines solubles sont généralement faiblement immunogènes, ce qui constitue unelimite pour le développement de vaccins sous unitaires à base de protéines. Mes travaux de thèseont eu pour objectif de décrypter certains mécanismes moléculaires et cellulaires qui contribuent àl’immunogénicité et d’en tirer partie pour développer des approches originales permettantd’améliorer la capacité des protéines à déclencher la réponse immunitaire. Pour cela, j’aiprincipalement utilisé le transactivateur transcriptionnel (Tat) du VIH-1. J’ai montré quel’oligomérisation de Tat permet à un mécanisme de collaboration B-TH-2 d’induire la réponseimmunitaire en absence d’adjuvant. J’ai identifié le déterminant minimal responsable de l’effet etmontré qu’il confère la propriété adjuvante à d’autres antigènes. J’ai ensuite montré que laprésentation aux cellules T restreinte aux CMH I et CMH II est accrue lorsque les protéines sontdotées de la capacité à lier des sucres sulfatés d’expression ubiquitaire: les héparanes sulfate. Cestravaux ont permis de définir de nouvelles approches pour améliorer l’immunogénicité de protéinessusceptibles d’être intégrées dans des préparations vaccinales. / Soluble proteins are usually poorly immunogenic, which is a limit to the development ofsubunit vaccines based on proteins. My thesis work aimed to decipher some molecular and cellularmechanisms that contribute to the immunogenicity and to exploit them to develop innovativeapproaches to improve the ability of proteins to trigger the immune response. For this purpose, Imainly used the transcriptional transactivator (Tat) of HIV-1. I showed that the oligomerization of Tatenables a B-TH-2 collaborative mechanism to induce the immune response in the absence ofadjuvant. I identified the minimum region determining the effect and showed that it confers the selfadjuvantingproperty to other antigens. In the second part of my work, I showed that the MHC I andMHC II restricted presentation to T cells is increased when the proteins are endowed with the abilityto bind ubiquitous sulfated polysaccharides: heparan sulfates. This work helped to define newapproaches to improve the immunogenicity of proteins that are likely to be included in vaccinepreparations.

<p>Fibroblast growth factors (FGFs) constitute a family (currently FGF-1 to FGF-23) of polypeptides that are essential in embryonal development and adult physiology, in animals from nematodes to humans. FGFs bind to four receptor tyrosine kinases, denoted FGFR-1 to FGFR-4. For proper function, the FGFs and their receptors depend on specific polysaccharide co-receptors, denoted heparan sulfate (HS). This thesis describes HS regulation of FGFR-1 signal transduction using blood vessel endothelial cells as a model.</p><p>We have determined HS structural features, necessary for FGF-2 induced FGFR-1 activation, using chemically modified heparin, which is structurally related to HS. Modified heparin, lacking sulfation at the 6-O position was inhibitory for FGFR-1 kinase activation and FGF-2 induced angiogenesis. Inhibition of blood vessel formation using modified heparin could be useful in treatment of diseases characterized by excess blood vessel formation. The critical role of HS sulfation for proper growth factor function was further underscored using an embryonal stem (ES) cell model. ES cells lacking expression of two isoforms of N-deacetyl N-sulfotransferase, NDST-1 and –2, failed to undergo embryonal development and to establish a vascular system. Exogenous heparin could not support development, but HS delivered from other ES cells allowed formation of primitive vessels and subsequent sprouting angiogenesis.</p><p>We have, furthermore, shown that the mechanism whereby HS supports FGF receptor activation is qualitative, as well as quantitative. Kinase activity could be induced by FGF-2 in the absence of HS, but this allowed only selected phosphorylation. In the presence of HS, the kinase activity was stabilized, allowing a broader spectrum of phosphorylation of sites on the FGF receptor itself as well as on cytoplasmic substrates. Finally, using selected microarrays, we have examined the potential regulation of enzymes in the HS biosynthesis pathway and of different proteoglycans to which HS is attached. Overall, we found no evidence for dramatic regulation on the transcriptional level, but could identify specific upregulation of HS proteoglycan syndecan-2, during blood vessel formation in vitro.</p><p>In conclusion, our studies demonstrate selective and complex regulation of HS synthesis and structure, essential in guiding growth factor function during health and disease.</p>

<p>Heparan sulfate (HS), a linear negatively charged polysaccharide located at the cell surface and in the extracellular matrix, interacts with, and thereby regulates the functions of numerous proteins. HS-protein interactions depend on the fine structure of HS, especially its sulfation pattern. This thesis aimed to understand how differently sulfated domains in HS are generated. Specifically, the substrate specificities of HS hexuronic acid 2-<i>O</i>-sulfotransferase (2OST) and HS glucosaminyl 6-<i>O</i>-sulfotransferases (6OSTs) were investigated. </p><p>Three different 6OSTs (6OST1-3) have been cloned and characterized. To study the mechanisms controlling 6-<i>O</i>-sulfation we incubated the recombinant purified 6-OST isoforms with different 6-<i>O</i>-desulfated poly- and oligosaccharide substrates and the active sulfate donor 3'-phosphoadenosine 5'-phospho[³⁵S]sulfate (³⁵S-labeled PAPS). All three enzymes catalyzed 6-<i>O</i>-sulfation of both <i>N</i>-acetylated (GlcNAc) as well as <i>N</i>-sulfated (GlcNS) glucosamines next to a nonreducing iduronic acid (IdoA) or glucuronic acid (GlcA). Similar specificities were demonstrated, although some differences in substrate preferences were noted.</p><p>To understand how pre-existing 2-<i>O</i>-sulfates affects 6-<i>O</i>-sulfation, 6OST2 and 6OST3 were incubated with pair-wise mixed octasaccharide substrates with different contents of 2-<i>O</i>-sulfates. The specificities for substrates with two or three 2-<i>O</i>-sulfates were higher compared to octasaccharides with no or one 2-<i>O</i>-sulfate indicating that 2-<i>O</i>-sulfate groups substantially promote the subsequent 6-<i>O</i>-sulfation. </p><p>Overexpression of the 6OSTs in a mammalian cell line resulted in increased 6-<i>O</i>-sulfation of -GlcA-GlcNS- and -GlcA-GlcNAc- sequences. The results were not isoform specific, but affected by the overexpression level. </p><p>The 2OST catalyzes 2-<i>O</i>-sulfation of both IdoA and GlcA residues, with high preference for IdoA units. To study how 2-<i>O</i>-sulfation of GlcA and IdoA is regulated, we incubated the enzyme with different substrates and ³⁵S-labeled PAPS. Our findings revealed that the 2OST almost exclusively sulfated IdoA also with a ratio of GlcA to IdoA of 99:1, suggesting that 2-<i>O</i>-sulfation of GlcA occurs before IdoA is formed.</p>

<p>Blood vessels transport blood with essential nutrients and oxygen to the cells in our body. In a healthy adult, formation of new vessels (angiogenesis) occurs only in case of tissue repair and growth. Physiological angiogenesis requires precise regulation of multiple signaling components, a process which is deregulated in a number of pathological conditions, such as cancer. This thesis is focused on the role of laminins, heparan sulfate proteoglycans (HSPGs) and vascular endothelial growth factor (VEGF)-A in regulation of vascular development and angiogenesis. As a model, we have used embryonic stem cells that differentiate to form blood vessels in a manner faithfully recapitulating the <i>in vivo</i> processes. </p><p>We show that the basement membrane (BM) protein laminin-111 promotes maturation of endothelial cells in the presence of fibroblast growth factor-2, a known endothelial cell mitogen. However, embryonic stem cells are able to differentiate into endothelial cells also in the absence of laminin deposition in the vascular BM. Sprouting angiogenesis, induced by VEGF-A, is also not strictly dependent on laminin deposition. On the other hand, in the absence of laminins, vessels are enlarged. These data suggest an important role for laminins in regulation of the vessel diameter.</p><p>We also show that HSPGs serve as coreceptors for VEGF-A to regulate vascular development. The mode of presentation of HSPGs, <i>in</i> <i>cis</i> (on the endothelial cell) or <i>in</i> <i>trans</i> (on an adjacent cell such as pericytes), is critical in regulation of VEGF receptor-2 activation and stimulation of vascular development. Binding of VEGF-A to HSPGs <i>in</i> <i>trans</i> leads to accumulation of activated VEGFR-2 in endothelial cells and to prolonged signaling. This demonstrates a potential role for HSPGs in regulation of receptor trafficking and signaling kinetics, with possible implications also for other HS-binding ligand/receptor systems.</p>

Fibroblast growth factors (FGFs) constitute a family (currently FGF-1 to FGF-23) of polypeptides that are essential in embryonal development and adult physiology, in animals from nematodes to humans. FGFs bind to four receptor tyrosine kinases, denoted FGFR-1 to FGFR-4. For proper function, the FGFs and their receptors depend on specific polysaccharide co-receptors, denoted heparan sulfate (HS). This thesis describes HS regulation of FGFR-1 signal transduction using blood vessel endothelial cells as a model. We have determined HS structural features, necessary for FGF-2 induced FGFR-1 activation, using chemically modified heparin, which is structurally related to HS. Modified heparin, lacking sulfation at the 6-O position was inhibitory for FGFR-1 kinase activation and FGF-2 induced angiogenesis. Inhibition of blood vessel formation using modified heparin could be useful in treatment of diseases characterized by excess blood vessel formation. The critical role of HS sulfation for proper growth factor function was further underscored using an embryonal stem (ES) cell model. ES cells lacking expression of two isoforms of N-deacetyl N-sulfotransferase, NDST-1 and –2, failed to undergo embryonal development and to establish a vascular system. Exogenous heparin could not support development, but HS delivered from other ES cells allowed formation of primitive vessels and subsequent sprouting angiogenesis. We have, furthermore, shown that the mechanism whereby HS supports FGF receptor activation is qualitative, as well as quantitative. Kinase activity could be induced by FGF-2 in the absence of HS, but this allowed only selected phosphorylation. In the presence of HS, the kinase activity was stabilized, allowing a broader spectrum of phosphorylation of sites on the FGF receptor itself as well as on cytoplasmic substrates. Finally, using selected microarrays, we have examined the potential regulation of enzymes in the HS biosynthesis pathway and of different proteoglycans to which HS is attached. Overall, we found no evidence for dramatic regulation on the transcriptional level, but could identify specific upregulation of HS proteoglycan syndecan-2, during blood vessel formation in vitro. In conclusion, our studies demonstrate selective and complex regulation of HS synthesis and structure, essential in guiding growth factor function during health and disease.

Heparan sulfate (HS), a linear negatively charged polysaccharide located at the cell surface and in the extracellular matrix, interacts with, and thereby regulates the functions of numerous proteins. HS-protein interactions depend on the fine structure of HS, especially its sulfation pattern. This thesis aimed to understand how differently sulfated domains in HS are generated. Specifically, the substrate specificities of HS hexuronic acid 2-O-sulfotransferase (2OST) and HS glucosaminyl 6-O-sulfotransferases (6OSTs) were investigated. Three different 6OSTs (6OST1-3) have been cloned and characterized. To study the mechanisms controlling 6-O-sulfation we incubated the recombinant purified 6-OST isoforms with different 6-O-desulfated poly- and oligosaccharide substrates and the active sulfate donor 3'-phosphoadenosine 5'-phospho[35S]sulfate (35S-labeled PAPS). All three enzymes catalyzed 6-O-sulfation of both N-acetylated (GlcNAc) as well as N-sulfated (GlcNS) glucosamines next to a nonreducing iduronic acid (IdoA) or glucuronic acid (GlcA). Similar specificities were demonstrated, although some differences in substrate preferences were noted. To understand how pre-existing 2-O-sulfates affects 6-O-sulfation, 6OST2 and 6OST3 were incubated with pair-wise mixed octasaccharide substrates with different contents of 2-O-sulfates. The specificities for substrates with two or three 2-O-sulfates were higher compared to octasaccharides with no or one 2-O-sulfate indicating that 2-O-sulfate groups substantially promote the subsequent 6-O-sulfation. Overexpression of the 6OSTs in a mammalian cell line resulted in increased 6-O-sulfation of -GlcA-GlcNS- and -GlcA-GlcNAc- sequences. The results were not isoform specific, but affected by the overexpression level. The 2OST catalyzes 2-O-sulfation of both IdoA and GlcA residues, with high preference for IdoA units. To study how 2-O-sulfation of GlcA and IdoA is regulated, we incubated the enzyme with different substrates and 35S-labeled PAPS. Our findings revealed that the 2OST almost exclusively sulfated IdoA also with a ratio of GlcA to IdoA of 99:1, suggesting that 2-O-sulfation of GlcA occurs before IdoA is formed.

Blood vessels transport blood with essential nutrients and oxygen to the cells in our body. In a healthy adult, formation of new vessels (angiogenesis) occurs only in case of tissue repair and growth. Physiological angiogenesis requires precise regulation of multiple signaling components, a process which is deregulated in a number of pathological conditions, such as cancer. This thesis is focused on the role of laminins, heparan sulfate proteoglycans (HSPGs) and vascular endothelial growth factor (VEGF)-A in regulation of vascular development and angiogenesis. As a model, we have used embryonic stem cells that differentiate to form blood vessels in a manner faithfully recapitulating the in vivo processes. We show that the basement membrane (BM) protein laminin-111 promotes maturation of endothelial cells in the presence of fibroblast growth factor-2, a known endothelial cell mitogen. However, embryonic stem cells are able to differentiate into endothelial cells also in the absence of laminin deposition in the vascular BM. Sprouting angiogenesis, induced by VEGF-A, is also not strictly dependent on laminin deposition. On the other hand, in the absence of laminins, vessels are enlarged. These data suggest an important role for laminins in regulation of the vessel diameter. We also show that HSPGs serve as coreceptors for VEGF-A to regulate vascular development. The mode of presentation of HSPGs, in cis (on the endothelial cell) or in trans (on an adjacent cell such as pericytes), is critical in regulation of VEGF receptor-2 activation and stimulation of vascular development. Binding of VEGF-A to HSPGs in trans leads to accumulation of activated VEGFR-2 in endothelial cells and to prolonged signaling. This demonstrates a potential role for HSPGs in regulation of receptor trafficking and signaling kinetics, with possible implications also for other HS-binding ligand/receptor systems.

<p>Extracellular GRP94 can elicit both innate and adaptive immune responses by interacting with endocytic and signaling receptors on professional antigen presenting cells (pAPCs). CD91 was the first receptor proposed to facilitate GRP94-mediated immune responses. Using a GRP94 affinity matrix, a CD91 fragment was isolated from the detergent-solubilized membranes of a pAPC cell line. It was then demonstrated that CD91 ligands could inhibit GRP94-mediated peptide cross-presentation, suggesting that CD91 played a critical role in this process. While these studies implied that CD91 could function as a GRP94 endocytic receptor, later works suggested that CD91 may not recognize GRP94 at the cell surface. These opposing observations have lead to a significant controversy surrounding the identity of CD91 as an endocytic receptor for GRP94. Because the ability of CD91 to directly mediate GRP94 surface binding and uptake has not been established, the studies included in this dissertation have focused on evaluating the ability of CD91 to facilitate three processes that are necessary for GRP94-mediated peptide cross-presentation: surface binding, internalization, and processing.</p><p>These studies utilized a recombinantly-expressed N-terminal domain of GRP94 (GRP94.NTD), which was previously shown to have nearly identical biological activity to full length GRP94. The ability of CD91 to directly bind and internalize GRP94.NTD was examined using murine embryonic fibroblast (MEF) cell lines whose expression of CD91 was either reduced via siRNA, or eliminated by genetic disruption of the CD91 locus. Binding competition experiments were also conducted. Together, these studies reveal that CD91 does not directly interact with GRP94 at the cell surface. The ability of CD91 to directly facilitate GRP94 internalization was examined using various internalization and internalization competition assays. These studies demonstrated that GRP94.NTD and the CD91 ligand RAP were internalized through spatially and kinetically distinct pathways, that CD91 was not necessary for GRP94.NTD internalization, and that RAP did not inhibit GRP94 endocytosis. Together, these studies strongly suggest that CD91 does not directly facilitate GRP94 internalization. When these studies were extended to DC2.4 mouse dendritic cells, the CD91 ligand RAP reduced GRP94.NTD internalization/process by ~15%. This suggests that CD91 may indirectly facilitate GRP94 internalization in pAPC cell lines. Lastly, cross-presentation studies were utilized to examine the ability of various CD91 ligands to influence GRP94.NTD-mediated peptide cross-presentation through a post-uptake mechanism using the DC2.4/OT-1 system. Although it was discovered that DC2.4 cells can internalize and process GRP94.NTD/peptide complexes through fluid-phase endocytosis, CD91 ligands did not significantly inhibit GRP94-mediated peptide cross-presentation by DC2.4 cells. These studies demonstrate that CD91 does not play a primary role in GRP94-mediated peptide cross-presentation.</p><p>In the course of these studies, cell surface heparan sulfate proteoglycans (HSPGs) were identified as novel cell surface binding sites for GRP94.NTD on MEF cells. This conclusion was established using three distinct experimental approaches. GRP94.NTD surface binding was significantly decreased following heparin pre-treatment, following incubation with the sulfation inhibitor sodium chlorate, and following digestion with extracellular heparinase II. Conversely, these treatments did not significantly influence GRP94.NTD binding to RAW264.7 mouse macrophage-like cells. This suggested that GRP94.NTD-HSPG cell surface interactions may require the expression of a specific type of cell surface HSPG that is not expressed by RAW264.7 cells. However, additional studies strongly suggested that GRP94.NTD-HSPG cell surface interactions were mediated by the heparan sulfate-containing side chains rather than the presence of a specific cell surface HSPG core protein.</p><p>This dissertation focuses on the critical re-examination of CD91 functions in GRP94 surface binding, uptake, and cross-presentation. Together, these results clarify conflicting data on CD91 function in GRP94 surface binding and endocytosis. This dissertation also describes the identification of cell surface HSPGs as GRP94 binding sites on MEF cells. These studies extend the diversity of surface receptors that recognize of GRP94, and suggest that cell surface HSPG-dependent interactions may contribute to the biology of GRP94-elicited immune responses.</p> / Dissertation

Alternative bone repair strategies are frequently sought after in orthopaedic surgery to address the growing need for improved morbidity and healing rates. This thesis sought to initiate and validate such an alternative, harnessing the flexible nature of a biomaterial substrate and the unique potential of glycosaminoglycan sugars. A novel, biodegradable biomaterial polymer, PHBV, has previously been identified to have the potential to mimic the characteristics of bone necessary for tissue repair and in this study, it was hypothesized that PHBV would be able to support bone formation. When tested in vitro, PHBV was found to support osteoblast cell attachment, proliferation and differentiation, despite its rougher, more hydrophobic surface characteristics compared to tissue culture plastic (TCP). However, unlike the progression of cells on TCP, PHBV caused a developmental delay at each stage of osteogenesis, suggesting a sub-optimal cell-substrate interaction. The expression profiles of genes involved in the maintenance of the extracellular matrix were monitored to investigate this phenomenon further. The results suggested that cells cultured on PHBV appeared to preference 7 against a collagen-based ECM and, instead, trigger an increase in the expression of other factors, such as osteopontin, presumably to modify the biomaterial microenvironment to optimise continued growth and differentiation. This finding led to the next hypothesis that functionalisation of PHBV with suitable compounds could optimise and enhance the osteogenic development at the implant site by facilitating the desired and appropriate cell-substrate interactions. Non-protein factors are often preferred for functionalisation to material scaffolds over proteins, as they are relatively robust and can survive many of the processes used in the manufacture of biomaterials. Glycosaminoglycan (GAG) sugars were appropriate candidates for this purpose, as they are not only abundantly expressed in bone, but more importantly, they are capable of binding and facilitating the activity of growth factors. Furthermore, they are resistant to several environmental influences including changes in pH, heat and desiccation. To identify a GAG that could be integrated with PHBV or any other biomaterial substrate, GAGs were extracted from phenotypically-distinct stages of MG-63 osteosarcoma cells. These GAGs were identified to display gross structural differences, as well as differences in the enzymes synthesising them, between immature and mature osteoblastic cells, with the increased production of a larger GAG species observed as the cells differentiated. Unexpectedly, however, when these GAGs were subsequently dosed back into the media of growing MG-63 cells, their bioactivity did not match the stage at which they had been harvested: all GAG species were able to influence cell survival and growth to varying degrees but were not capable of affecting cell differentiation. However, if these same GAGs were exposed to cells by first being attached to the growth substrate, they induced varying degrees of aggregation in human mesenchymal stem cells (hMSCs), with more mature GAGs producing the most profound effects. Interestingly, a similar phenomenon was not observed when MG-63 cells where cultured in a similar manner. A direct correlation between the GAGs expressed by osteoblasts and the specific cellular processes they functionally influence has yet to be identified. While the experiments presented here demonstrate an effect of GAGs in osteoblastic cell survival, a role for GAGs in the progression of bone formation was not revealed. Loss-of-function studies were therefore necessary to determine the role of GAGs in bone, but this was hampered by the limited availability of procedures that allow the alteration of GAGs and the subsequent detection of these effects. Therefore, a tool to screen the efficacy of a loss of GAG function was developed. TAT-EGFP, a purpose-designed fluorescent GAG-binding peptide, was able to confirm that treatment with sodium chlorate was an effective 8 strategy to hinder GAG expression in MG-63 cells with minimal cytotoxicity to the cells. Following more extensive studies with chlorate treatment, it was found that a recoverable disruption to both proliferation and mineralisation could be induced in MG-63 cells. This suggested a role for GAGs in osteogenesis. A series of experiments then carried out following gene expression microarray analysis indicated that GAG de-sulfation by chlorate gives rise to an S-phase block in the cell cycle and a disruption to the actin cytoskeleton, which appeared to be associated with a change in the activity of cell-surface proteoglycans, most likely syndecan 4. It was also found that cells up-regulated plasma membrane ALP activity and cholesterol synthesis, presumably in an attempt to recover from a chlorate-induced loss in GAG function. Cholesterol is known to be important in establishing connections between membrane elements and the actin cytoskeleton, and its up-regulation here may reflect dysfunctions in these units and a dysfunction in syndecan 4 activity. With further confirmation, this would suggest that syndecan 4 plays a pivotal role in maintaining osteogenesis, in at least MG-63 cells, and that sulfated GAGs function principally to facilitate this role. The effective use of GAGs in bone repair strategies will require further understanding of GAG/syndecan 4/osteogenesis relationship.

L’analyse du mécanisme d’interaction de l’héparane sulfate protéoglycane (HSPG) avec lesprotéines amyloïdes et l’effet de paramètres physiologiques (pH,…) sur ce mécanisme nouspermettent une meilleure compréhension des mécanismes menant à l’amyloïdogenèse.La transthyrétine (TTR), molécule circulant à la fois dans le plasma et dans le liquidecéphalo-rachidien, est une des protéines impliquée dans les amyloses. Elle est responsable desamyloses à la TTR et elle joue un rôle dans la maladie d’Alzheimer en séquestrant la protéinebêta-amyloïde (Aβ). La biochromatographie est un outil très efficace pour analyser lemécanisme entre un ligand et son récepteur dans des conditions modulables et se rapprochantdes conditions biologiques. De plus, les nanotubes de carbones (NTCs) peuvent être utiliséspour détecter ou transporter des molécules qui se lient sur leur surface externe et peuventinteragir avec d’autres composés. Au cours de ce travail, un support particulaire a été utilisé où l’HSPG est immobilisé sur desparticules de silice pré-activées par des résidus amines. Ce support remplissant une colonnechromatographique a permis dans un premier temps d’étudier et de comparer les mécanismesd’association entre l’HSPG et une forme sauvage de la TTR et une forme sénile de la TTRextraite d’un patient décédé des suites d’une amylose sénile à la TTR. Cette étude a montréque pour la TTR sauvage, l’association avec l’HSPG est indépendante du pH et implique desinteractions faibles. Pour la TTR sénile, cette association est dépendante du pH. A pH<6,5, laprotonation d’un résidu histidine est observée. De plus, l’étude des paramètresthermodynamiques et de la compensation enthalpie/ entropie montrent un changement dans lemécanisme de fixation avec l’apparition d’interactions ioniques à pH<6,5. Un pH acide estnécessaire pour dissocier et dénaturer partiellement la TTR. L’affinité de la TTR avec l’HSPGdépend de la structure tétramérique quaternaire de la TTR qui présente alors des résiduscapables de créer des interactions. Dans un deuxième temps, cette colonne a permis d’évaluerl’effet de la TTR et du pH sur la liaison Aβ/HSPG. Comme précédemment, la protonationd’un résidu histidine présent sur la Aβ est observée à pH<6,5. Ce résultat confirme des étudesmenées auparavant sur le précurseur de la protéine Aβ. Les résultats thermodynamiques ontmis en évidence que l’affinité de Aβ avec l’HSPG diminuait avec la concentration croissanteen TTR et l’étude des chromatogrammes associés a montré que la TTR ne séquestrait passeulement Aβ mais la clivait en fragments plus courts qui diminuent son affinité avecl’HSPG. Dans la maladie d’Alzheimer, la TTR exerce une activité protéolytique vis-à-vis deAβ.2. Dans un troisième temps, l’effet de la fonctionnalisation de la TTR par des nanotubes decarbone sur la liaison TTR/HSPG a été étudié. Les résultats obtenus montrent que lesinteractions entre l’HSPG et la TTR-NTC sont de type van der Waals et hydrogène. Lesparamètres thermodynamiques des liaisons TTR/HSPG et TTR-NTC/HSPG sont similairespour des pH>6. A pH<6, il n’existe quasiment pas de différences entre les valeurs obtenues àpH >6 et celles obtenues à pH<6. Les NTCs empêcheraient la formation de liaisons ioniques / The analysis of the heparan sulfate proteoglycan (HSPG) association mechanism withamyloid proteins and the effect of physiological parameters (pH,…) on this mechanism allowa better understanding of the mechanisms leading to amyloidogenesis.Transthyretin (TTR), which circulates in both plasma and cerebrospinal fluid, is one of theproteins involved in amyloidosis. It leads to TTR amyloidosis and plays a role in Alzheimer’sdisease in sequestrating the beta-amyloid (Aβ) protein. Biochromatography is an effectivetool for the analysis of the mechanism involved between a ligand and its receptor inadjustable conditions which could be close to biological conditions. Moreover, carbonnanotubes can be used to detect or to carry molecules which bind on its external surface andcould interact with other molecules. In this work, a particulate support was used where the HSPG was immobilized on the silica particles preactivated by amine residues. This support filling a column was used to study andcompare association mechanisms between HSPG and a wild type TTR form and a senile formof the TTR which was was extracted from a patient who died of cardiac failure with a senilesystemic amyloid. This study showed that the association between wild type TTR and HSPGwas independent of the pH and involved weak interactions.For the senile TTR, this association was dependent on pH. At pH<6,5, a histidine residue wasprotonated. Moreover, the study of both thermodynamical parameters and enthalpy-entropycompensation showed a change in the association mechanism with involvement of more ionicinteractions at pH<6,5. An acidic pH was necessary to dissociate and partially denaturate theTTR. The affinity of the TTR with HSPG depends on the tetrameric quaternary structure ofthe TTR which presents some residues which are able to create interactions.In a second time, this column was used to evaluate the effect of both the TTR and the pH onAβ/HSPG binding. As previously, the protonation of a histidine residue present on Aβ wasobserved at pH<6,5. This result confirmed studies on the Aβ precursor. Thermodynamicalvalues highlighted that the affinity of Aβ for l’HSPG decreased with the increase of TTRconcentration, and the study of the chromatograms associated showed that TTR sequestratedAβ and cut Aβ in smaller fragments which decreased its affinity for HSPG. In Alzheimer’sdisease, TTR has a proteolytic activity on Aβ.In a third time, the effect of the carbon nanotubes (CNTs), immobilized on TTR, on theTTR/HSPG binding was studied. Results showed that interactions between TTR-NTC andHSPG are van der Waals and hydrogen. Thermodynamical data of TTR/HSPG et TTRNTC/HSPG bindings are similar at pH>6. At pH<6, no differences between values obtainedat all pH values for TTR-NTC. NTCs would avoid ionic interactions formations.