Cancer is a leading cause of death that affects millions of people across the globe each year. Photodynamic therapy (PDT) is a relatively new treatment approach for cancer in which anticancer drugs are activated by light at an appropriate wavelength to generate highly cytotoxic reactive oxygen species (ROS) and achieve tumor destruction. Compared with conventional chemo- and radiotherapy, PDT can be performed with minimal invasiveness, local targeting and reduced side effects. However, most of the currently available PDT drugs mainly absorb in the visible part of the spectrum, where light penetration depth into human tissues is very limited. Therefore, increasing the treatment depth of PDT has been considered to be an important approach to improve the effectiveness of PDT for treating larger and thicker tumor masses. In this thesis, we present our investigation into the potential of two-photon activated PDT (2-γ PDT), combination therapy of PDT and chemotherapy, and bioluminescence-activated PDT as a means to increase the treatment depth of this modality. In 2-γ PDT, the photosensitizing agents are activated through simultaneous absorption of two photons. This approach allows the use of near-infrared (NIR) light that can penetrate deeper into tissues and thus, has the potential of treating deep-seated tumors and reducing side effects, while the non-linear nature of two-photon excitation (TPE) may improve tumor targeting. We have evaluated the PDT efficacy of a second-generation photosensitizer derived from chlorophyll a, pyropheophorbide a methyl ester (MPPa), through both one- and two-photon activation. We observed that MPPa had high one-photon (1-γ PDT efficacy against both cisplatin-sensitive human cervical (HeLa) and cisplatin-resistant human lung (A549) and ovarian (NIH:OVCAR-3) cancer cells when activated by femtosecond (fs) laser pulses at 674 nm. At a low light dose of 0.06 J cm-2, the MPPa concentration required to produce a 50% cell killing effect (IC50) was determined to be 5.3 ± 0.3, 3.4 ± 0.3 and 3.6 ± 0.4 μM in HeLa, A549 and NIH:OVCAR-3 cells, respectively. More significantly, we also found that MPPa could be effectively activated at the optimal tissue-penetrating wavelength of 800 nm through TPE. At a light dose of 886 J cm-2, where no measurable photodamage was observed in the absence of MPPa, the IC50 values were measured to be 4.1 ± 0.3, 9.6 ± 1.0 and 1.6 ± 0.3 μM in HeLa, A549 and NIH:OVCAR-3 cells, respectively. We obtained corresponding LD50 (the light dose required to produce a 50% killing effect) values of 576 ± 13, 478 ± 18 and 360 ± 16 J cm-2 for 10 μM MPPa, which were approximately 3-5 times lower than the published 2-γ LD50 of Visudyne® and 20-30 times lower than that of Photofrin®. These results indicate that MPPa may serve as a photosensitizer for both 1- and 2-γ activated PDT treatment of difficult-to-treat tumors by conventional therapies. Indocyanine green (ICG), a dye having an absorption maximum near 800 nm, has been considered to be a potential NIR PDT agent. However, the PDT efficacy of ICG has been found to be very limited probably due to the low yield of cytotoxic ROS. In the present work, we have evaluated the combination effects of ICG-mediated PDT with conventional chemotherapy mediated by two types of chemotherapeutic drugs, namely the type II topoisomerase (TOPII) poisons etoposide (VP-16)/teniposide (VM-26) and the platinum-based drugs cisplatin (CDDP)/oxaliplatin (OXP). Synergistic enhancement of cytotoxicity and increased yields of DNA double strand breaks (DSBs) were observed in HeLa, A549 and NIH:OVCAR-3 cancer cells treated with the combination of ICG-PDT and VP-16. The presence of VP-16 during the laser irradiation process was found to be critical for producing a synergistic effect. An electron-transfer-based mechanism, in which ICG could increase the yield of highly cytotoxic VP-16 metabolites, was proposed for the observed synergistic effects, although direct spectroscopic detection of the reaction products was found to be very challenging. Moreover, we observed a much lower degree of synergy in the human normal fibroblast GM05757 cells than that in the three cancer cell lines investigated. Synergistic effects were also observed in A549 cells treated with the combination of ICG-PDT and VM-26 (i.e. an analog of VP-16). Furthermore, the combination of low-dose CDDP/OXP and ICG-PDT was demonstrated to produce an additive or synergistic effect in selected cancer cell lines. These preliminary results suggest that the combination of ICG-PDT with VP-16/VM-26 or CDDP/OXP chemotherapy may offer the advantages of enhancing the therapeutic effectiveness of ICG-PDT and lowering the side effects associated with the chemotherapeutic drugs. Bioluminescence, the generation of light in living organisms through chemical reactions, has been explored as an internal light source for PDT in recent years. This approach, in principle, does not suffer from the limited tissue penetration depth of light. In the present project, we have evaluated the effectiveness of luminol bioluminescence in activating the porphyrin photosensitizers meso-tetra(4-sulfonatophenyl)porphine dihydrochloride (TPPS4) and Fe(III) meso-tetra(4-sulfonatophenyl)porphine chloride (FeTPPS). The combination treatment induced significant killing of HeLa cells, while additive effects were observed in two normal human fibroblast cell lines (GM05757 and MRC-5). Our observations indicate that bioluminescence of luminol may generate sufficient light for intracellular activation of PDT sensitizers. Furthermore, the combination treatment may have intrinsic selectivity towards cancerous tissues. In summary, we have demonstrated effective killing of cancer cells by MPPa-mediated 1- and 2-γ PDT, combination of ICG-PDT and VP-16/VM-26 or CDDP/OXP chemotherapy, and bioluminescence of luminol activated PDT mediated by TPPS4/FeTPPS. These positive preliminary results indicate that all these three approaches have the potential of increasing the treatment depth of PDT and facilitating the development of more effective PDT treatment strategies.

L’objectif de cette thèse était de mettre au point une analyse critique de vecteurs polymères utilisés pour la thérapie photodynamique (PDT) et de faire le lien avec l’efficacité thérapeutique observée. Pour cela, une analyse complète des vecteurs a été réalisée par des techniques classiques comme la diffusion dynamique de la lumière ou la microscopie électronique, mais aussi grâce au fractionnement flux-force, technique peu utilisée jusqu’à présent dans le domaine des auto-assemblages polymères. Dans un deuxième temps, les auto-assemblages ont été utilisés comme vecteurs d’un photosensibilisateur, le Phéophorbide a, et l’efficacité thérapeutique évaluée en travaillant sur culture cellulaire 2D et 3D de lignées HCT116 (cancer du colon) ou FaDu (cancer tête et cou). Différents vecteurs polymères simples ont tout d’abord été examinés, à savoir des micelles ou des polymersomes à base de copolymères diblocs amphiphiles comme le poly(oxyde d’éthylène-b--caprolactone), le poly(oxyde d’éthylène-b-lactide) ou le poly(oxyde d’éthylène-b-styrène). Ceci a permis d’obtenir des vecteurs présentant des tailles et des morphologies variables. Les résultats en PDT ont montré des comportements différents et une meilleure efficacité en 3D pour les systèmes à base de PEO-PDLLA. La technique de fractionnement flux-force asymétrique (AsFlFFF) a particulièrement été utilisée pour ces vecteurs afin de démontrer la pureté des auto-assemblages. Les connaissances acquises dans cette première partie ont permis de caractériser des vecteurs faits à base de mélanges d’auto-assemblages micelles/vésicules. Ceux-ci ont révélé des phénomènes d’antagonisme ou de synergie dans l’efficacité en PDT, démontrant l’existence de processus complexes au niveau de la réponse cellulaire.Des auto-assemblages figés par réticulation ont aussi été développés, caractérisés et examinés en PDT. Ils se sont avérés extrêmement intéressants pour la PDT sur les cultures cellulaires en 3D, démontrant une efficacité accrue comparée aux systèmes simples. La comparaison de ces résultats avec ceux obtenus en culture 2D pour les mêmes objets a de plus permis de mettre en évidence la différence entre ces deux modèles biologiques. Enfin, des auto-assemblages à base de complexes poly-ioniques ont aussi été formés et caractérisés. Le fractionnement flux-force s’est là encore avéré efficace, mais a nécessité l’utilisation d’une injection spéciale par Frit-inlet. Leur efficacité en PDT s’est avérée faible. / The objective of this study was to critically analyze different polymer self-assemblies used for photodynamic therapy (PDT) and to link this analysis to their therapeutic efficiency. To do that, a thorough characterization of the vectors has been performed by classical techniques such as Dynamic Light Scattering or electron Microscopy, but also using flow fractionation, which has been seldomly used so far for polymeric self-assemblies. In a second step, these have been used as vectors of a photosensitizer, namely Phéophorbide a, and the therapeutic efficiency assessed on both 2D and 3D cell cultures of HCT 116 (colon cancer) and FaDu (head and neck cancer) cells. Different simple polymer vectors have first been evaluated, namely micelles and polymersomes based on diblock amphiphilic copolymers such as poly(ethylene-oxide-b--caprolactone), poly(ethylene-oxide-b-lactide) or poly(ethylene-oxide-b-styrene). This enabled obtaining vectors exhibiting various sizes and morphologies. Results in PDT showed different behaviours and a better efficiency in 3D for PEO-PDLLA. The Asymmetric Flow Field Flow Fractionation was particularly used for these systems to demonstrate their purity. The acquired expertise on this part enabled us to also characterize vectors made of known mixtures of micelles and polymersomes. These revealed antagonism and synergy effects in PDT, demonstrating the presence of complex processes for the cell response. Other self-assemblies consisting of crosslinked systems have also been developed and characterized. These were observed to be particularly efficient for PDT on 3D cell cultures. The comparison of these results with those for the 2D cell culture enabled to highlight the difference between those two biological systems. Finally, self-assemblies based on Polyion Complexes were also formed and characterized. Field Flow Fractionation was once again used as a powerful technique for this, although this implied the use of a special injection device called Frit Inlet. Their PDT efficiency however proved to be low.

La thérapie photodynamique (PDT) est une technique utilisée cliniquement pour traiter certaines maladies de la peau, la dégénérescence maculaire liée à l’âge et certains types de cancer. Elle fait intervenir trois composants : une molécule photosensible ou photosensibilisateur (PS), la lumière et l’oxygène. Après administration du PS, celui-ci va se localiser plus ou moins sélectivement dans les zones tumorales où il est alors activé par irradiation lumineuse à une longueur d’onde et une puissance données. Ceci engendre la formation d’espèces réactives de l’oxygène (ROS) très réactives, dont l'oxygène singulet1 O2, qui entraînent la destruction des tissus tumoraux par nécrose ou apoptose. Afin d’améliorer la sélectivité du traitement, différentes pistes sont actuellement exploitées dont l’élaboration de « photodynamic molecular beacons » (PMB). Dans un PMB, le photosensibilisateur (PS) est associé via un peptide à un inhibiteur 1O2, appelé quencher. Ce quencher inhibe la formation d’1O2 tant que le composé n’a pas atteint sa cible. Une fois la zone cancéreuse atteinte, des enzymes spécifiques clivent le peptide, libérant ainsi le PS qui retrouve alors sa capacité à former de l’1O2. Trouver un couple PS/quencher adéquat reste un challenge en PDT. Les propriétés photophysiques particulières des caroténoïdes et leur aptitude à inhiber la production d’1O2 font de ces derniers des quenchers potentiellement utilisables pour l’élaboration de PMBs. Chez les plantes, les caroténoïdes (carotènes et xanthophylles) sont des pigments associés à la photosynthèse, qui ont deux rôles principaux : un rôle de collecteur de lumière et un rôle photoprotecteur en protégeant le(s) système(s) photosynthétique(s) contre les dommages photooxydatifs liés à une exposition trop intense à la lumière. Ceci s’opère, entre autre, via le cycle des xanthophylles. Cette aptitude à capter de l’énergie présente un intérêt potentiel à ne pas négliger dans la perspective de la conception de PMB utilisables en thérapie photodynamique. Dans le cadre de cette thèse en co-tutelle avec la Faculté des Sciences de Bizerte nous avons ciblé les caroténoïdes présents dans trois fruits produits en Tunisie à savoir les kakis (Diospyros kaki L.), les abricots (Prunus armeniaca L.) et les pêches (Prunus persica L.) connus pour leur richesse globale en ces pigments. Divers procédés d’extractions ont été étudiés : (i) L’extraction de type Soxhlet par solvants organiques à pression atmosphérique, utilisée comme référence, (ii) l'extraction accélérée par solvant organique (ASE : Accelerated solvent Extraction) effectuée sous pression, enfin (iii) l'extraction par CO2 supercritique avec l’éthanol comme cosolvant. Pour ces deux derniers procédés, une approche par plan d’expériences (surfaces de réponses) a été utilisée pour identifier les facteurs clé et les conditions optimales d’extractions de divers caroténoïdes (pression, température, débit, % de cosolvant, temps, nombre de cycles). L'analyse par chromatographie liquide à haute performance couplée à la détection UV-Visible et à la spectrométrie de masse a ensuite permis l'identification et la quantification des caroténoïdes présents dans les extraits obtenus, permettant ainsi de comparer les profils caroténoïdiques propres à chaque fruit et les performances de chaque procédé d’extraction. Cette étude ayant révélé un profil caroténoïdique particulièrement intéressant chez le kaki par rapport aux autres fruits, une extraction et une purification des caroténoïdes de ce fruit par chromatographie liquide haute pression préparative a ensuite été effectuée afin de disposer d’une quantité suffisante de chaque caroténoïde, et parfois de leurs isomères conformationnels, en vue de l’étude de leurs propriétés photophysiques (absorption, émission de fluorescence, inhibition d’1O2) et de l’évaluation de leur intérêt potentiel en tant que quencher d’1O2 dans un édifice de type PMB / Photodynamic therapy (PDT) is a clinically used technique for treating skin diseases, age-relatedmacular degeneration but mainly some types of cancer. PDT involves three components: a photosensitive molecule named photosensitizer (PS), light and oxygen. After administration of the PS, this one will be located more or less selectively in tumoral regions where it is activated by light irradiation at appropriate wavelength and power. This leads to the formation of highly reactive and cytotoxic reactive oxygen species (ROS), especially singlet oxygen, resulting in the destruction of the tumor by necrosis or apoptosis. To improve the treatment selectivity, different strategies are being exploited, one of which is the development of "photodynamic molecular beacons" (PMB). In PMB the photosensitizer is linked via a peptide to an inhibitor of 1O2 (quencher). This quencher inhibits the formation of 1O2 as long as the compound has not reached its target, namely cancer cells. In order to inhibit the toxicity of the PS in non-target cells and restore toxicity only close to the biological target, it is necessary to find an adequate PS/quencher couple. This remains a challenge for PDT. Carotenoids are interesting candidates due to their specific photophysical properties and ability to inhibit 1O2, which makes them potential quenchers for building PMBs. In plants, carotenoids (carotenes and xanthophylls) are pigments involved in the photosynthesis, in which they play two main roles: a light collecting role and a protecting role by preserving the photosynthetic systems against photoxydative damages induced by a too intense light exposure. This protection can for instance occur via the well-known xanthophylls cycle. This capacity to catch energy presents a potential interest that should not be neglected in the framework of the design of PMBs usable in photodynamic therapy. Within the framework as part of this PhD thesis in Cotutelle with the Faculty of Sciences of Bizerte, we focused on carotenoids from three fruits produced in Tunisia: persimmon (Diospyros kaki L.), apricot (Prunus armeniaca L.) and peache (Prunus persica L.), known for their global richness in these natural pigments. Three extraction processes were investigated: (i) the Soxhlet extraction based on the use of organic solvent at atmospheric pressure and used as reference, (ii) the accelerated solvent extraction (ASE) using organic solvent under high pressure, and (iii) the supercritical fluid extraction (SFE) using supercritical CO2 and ethanol as cosolvent. For these two last processes, a design of experiments (Surface Response Design) was used to identify the key factors and optimal extraction conditions of various carotenoids (pressure, temperature, flow, % cosolvent, time, number of cycles). Then, HPLC-PDA coupled with mass spectrometry (MS) enabled the identification and quantification of carotenoids from the extracts. Thus it was possible to compare the profiles in carotenoids content from each fruit as well as the performances of each extraction process. This study showed that the carotenoidic profile in the persimmon was the most interesting as compared to the profiles in the two other fruits. Extraction and purification of the carotenoids from persimmon by preparative high pressure liquid chromatography were then performed in order to have a sufficient amount of each carotenoid and sometimes of their conformational isomers. We finally performed a study of their photophysical properties (absorption, fluorescence emission, 1O2 inhibition) in order to evaluate their potential as 1O2 quencher in molecular construction such as a PMB