Thesis (Ph. D.) -- University of Texas Southwestern Medical Center at Dallas, 2003. / Vita. Bibliography: 180-208.

L'arthrose est une maladie multifactorielle complexe. Parmi les facteurs impliqués dans sa pathogénie, les certains prostaglandines exercent un rôle inflammatoire et d’autres un rôle protecteur. La prostaglandine D2 (PGD2) est bien connue comme une PG anti-inflammatoire, qui est régulée par l’enzyme «Lipocalin prostaglandine D-synthase». Avec l’inflammation de l'arthrose, les chondrocytes essaient de protéger le cartilage en activant certaines voies de récupération dont l'induction du gène L-PGDS. Dans cette étude, nous étudions la voie de signalisation impliquée dans la régulation de l'expression du (L-PGDS) sur les chondrocytes traités avec différents médiateurs inflammatoires. Le but de projet: Nous souhaitons étudier la régulation de la L-PGDS dans le but de concevoir des approches thérapeutiques qui peuvent activer la voie intrinsèque anti-inflammatoire. Méthode et conclusions: In vivo, l'arthrose a été suivie en fonction de l’âge chez la souris ou chirurgicalement suivant une intervention au niveau des genoux de souris. Nous avons confirmé les niveaux d’expression de L-PGDS histologiquement et par immunohistochimie. In vitro, dans les chondrocytes humains qui ont été traités avec différents médiateurs de l'inflammation, nous avons observé une augmentation de l’expression de la L-PGDS dose et temps dépendante. Nous avons montré, in vivo et in vitro que l’inflammation induit une sécrétion chondrocytaire de la L-PGDS dans le milieu extracellulaire. Enfin, nous avons observé la production de différentes isoformes de la L-PGDS en réponse à l'inflammation. / Osteoarthritis is a complex multifactorial disease; many factors are involved in its pathogenesis, among those factors prostaglandins. Some prostaglandins have inflammatory role and some have anti-inflammatory role. Prostaglandin D2 (PGD2) is a well-established anti-inflammatory PG which is synthesized by the Lipocalin prostaglandin-D synthase (L-PGDS) enzyme. Upon the initiation of the inflammatory process in osteoarthritis, chondrocytes try to save themselves by activating salvage pathways, among these pathways is the induction of L-PGDS gene. In this study we are addressing the signaling pathways involved in the regulation of (L-PGDS) gene expression, in chondrocytes treated with different inflammatory mediator. Rationale: understanding the regulation of L-PGDS will allow us to design therapeutic targets that can switch on the intrinsic anti-inflammatory pathway. Method and findings: In vivo, osteoarthritis was induced in mice knees surgically or naturally following aging, the development of osteoarthritis was then confirmed histologically. The expression levels of L-PGDS were detected by Immunohistochemistry. In vitro, human chondrocytes were treated with different inflammatory mediators (interleukin-1 beta, interleukin-17, hydrogen peroxide and tert-Butyl hydroperoxide). Interestingly, in most cases chondrocytes increased expression of L-PGDS in a dose and time dependent manner. We discovered that chondrocytes release L-PGDS into the extracellular space in response to inflammation, in vivo and in vitro. Lastly we observed different isoforms of L-PGDS generated in response to inflammation.

The dioxygenase-cytochrome P450 fusion proteins (DOX-CYP) comprise a heme-containing enzyme family that shares structural and catalytic properties with mammalian prostaglandin H (PGH) synthases. 7,8-Linoleate diol synthase (7,8-LDS) of Gaeumannomyces graminis was first characterized, and DOX-CYP enzymes are of mechanistic and biological interest. The growing number of fungal genome sequences has revealed DOX-CYP homologues in medically and economically important species. The aim of this thesis was to identify novel members of the DOX-CYP fusion protein family. The devastating rice pathogen Magnaporthe oryzae contains two DOX-CYP genes. The fungus synthesizes 7S,8S-dihydroxyoctadecadienoic acid (7,8-DiHODE) by dioxygenation of linoleic acid to 8R-hydroperoxyoctadecadienoic acid (8R-HPODE), and subsequent isomerisation to the diol. 7,8-LDS of M. oryzae was identified by gene deletion, but the infection and reproduction processes of the Δ7,8-LDS strain were not altered. A mutant with constitutive protein kinase A activity profoundly changed the oxygenation profile, possibly due to post-translational modification. The human pathogens Aspergillus fumigatus and A. clavatus contain three DOX-CYP, designated psi producing oxygenase A (ppoA), ppoB, and ppoC, and form three oxylipins: 5S,8R-DiHODE, 8R,11S-DiHODE, and 10R-hydroxyoctadecadienoic acid.  PpoA was identified as 5,8-LDS, and ppoC as 10R-DOX. The 8,11-linoleate hydroperoxide isomerase activity was reduced by two imidazole-containing P450 inhibitors, miconazole and 1-benzylimidazole. PpoB could not be linked to the biosynthesis of 8,11-DiHODE for the following reasons: First, the 8,11-hydroperoxide isomerase activity was retained in A. fumigatus ΔppoB strains. Second, the P450 domain of the deduced ppoB of A. clavatus lacks a heme-thiolate cysteine ligand, presumably essential for hydroperoxide isomerase activity. Linoleate 9R-DOX activities of Aspergillus terreus and Lasiodiplodia theobromae were discovered. 9R-HPODE was further converted into unstable allene oxides, as judged by the accumulation of their hydrolysis products, α- and γ-ketols. These allene oxide synthase activities were specific for 9R-hydroperoxides. The 9R-DOX and AOS were found to have unique characteristics. In conclusion, novel DOX-CYP enzymes were identified in human and plant pathogenic fungi. These enzymes might be involved in biological processes, and show interesting catalytic similarities to human PGH synthase and thromboxane synthase (CYP5A).

L’arthrose (OA) est une maladie dégénérative et multifactorielle caractérisée par une destruction de cartilage, une formation d’ostéophytes et une inflammation au niveau de la membrane synoviale. Le 4-hydroxynonénal (HNE), un produit final de la peroxydation lipidique, a été identifié récemment comme un facteur catabolique et un médiateur inflammatoire dans le cartilage arthrosique humain. Notre projet vise à étudier l’effet du HNE sur la régulation de la prostaglandine E2 synthase-1 microsomale (mPGES-1) et de la protéine activante 5-lipoxygénase (FLAP)/5-lipoxygénase (5-LOX) dans les chondrocytes arthrosiques humains. Lorsque les cellules sont traitées une seule fois avec 10 µM HNE, les résultats de Western blot et de PCR en temps réel montrent que l’expression de la cyclooxygénase-2 (COX-2) et de la mPGES-1 augmente de manière significative et atteint respectivement le maximum après 8 et 16 heures d’incubation puis diminue graduellement. Cependant, lorsque les cellules sont traitées plusieurs fois avec 10 µM HNE à 2 heures d’intervalle, l’expression de la COX-2 et de la mPGES-1 augmente en fonction du temps sans subir une baisse après 24 heures d’incubation. Le HNE induit l’activité du promoteur de la mPGES-1 via l’activation du facteur de transcription Egr-1. L’investigation de la 2ème voie du métabolisme de l’acide arachidonique, à savoir 5-LOX/FLAP, montre que le HNE induit l’expression de FLAP après 24 heures de stimulation et celle de 5-LOX seulement après 48 heures. Ceci semble survenir à l’étape de transcription au cours de laquelle HNE induit l’expression de l’ARNm et l’activité du promoteur du gène 5-LOX. Nous avons démontré aussi que le niveau de leukotriène B4 (LTB4) augmente et suit le même profil que celui de la 5-LOX. L’étude des mécanismes moléculaires susceptibles d’être impliqués dans la régulation de la 5-LOX/FLAP par le HNE montre que ce dernier stimule leur expression via l’action de prostaglandine E2 (PGE2) et du facteur de croissance transformant-beta 1 (TGF-β1). En conclusion, notre étude démontre que le HNE induit à court-terme d’incubation la voie de COX-2/mPGES-1 puis par la suite stimule celle de FLAP/5-LOX à long-terme d’incubation dans les chondrocytes arthrosiques humains. Ces résultats suggèrent que la mPGES-1 et 5-LOX/FLAP sont des potentielles cibles thérapeutiques intéressantes pour contrôler la production de PGE2 et LTB4 dans OA. / 4-hydroxynonenal (HNE), a lipid peroxidation end-product, is produced abundantly in osteoarthritic (OA) articular tissues. Recently, we reported that HNE-induced cyclooxygenase-2 (COX-2) decreased gradually in human OA chondrocytes after 8 h of incubation. This study aimed to investigate whether COX-2 down-regulation is attributed to HNE depletion and is responsible for the switch from COX-2 to 5-lipoxygenase-activating protein (FLAP)/5-lipoxygenase (5-LOX). Treatment of chondrocytes with 10 µM HNE induced prostaglandin E2 (PGE2) release as well as COX-2 and microsomal prostaglandin E2 synthase-1 (mPGES-1) expression at the protein and mRNA levels, with a plateau reached at 8-16 h of incubation, followed by a subsequent decline. However, 8 repeated treatments with 10 µM HNE prevented the reduction of COX-2 and mPGES-1 expression. We demonstrated that HNE induced mPGES-1 promoter activity mainly through transcription factor Egr-1 activation. On the other hand, when COX-2 expression decreased, leukotriene B4 (LTB4) level rose after a long period of stimulation (48 and 72 h). At the mRNA level, HNE induced FLAP and 5-LOX expression after 24 and 48 h of stimulation, respectively. The addition of a nonspecific COX-2 inhibitor (naproxen) to cultured chondrocytes revealed that FLAP and 5-LOX regulation by HNE required PGE2 production. Furthermore, our data showed that 10 µM HNE significantly induced transforming growth factor-beta 1 (TGF-β1) production. The addition of anti-TGF-β antibody to culture medium reduced HNE-induced 5-LOX/FLAP expression by 40%, indicating the involvement of a TGF-β1-dependent mechanism. Our data demonstrate that the shunt to the FLAP/5-LOX pathway in HNE-induced human OA chondrocytes is attributed to COX-2 inhibition, probably due to HNE depletion. PGE2 and TGF-β1 are suggested to be involved in this regulation. Further experiments are in progress to determine other molecular mechanisms underlying this switch in OA chondrocytes.

The expanding therapeutic use of thalidomide (TD) remains limited by its species-specific teratogenicity in humans and rabbits, but not rodents. The R and S isomers of TD may be selectively responsible for its respective therapeutic and teratogenic effects, but rapid in vivo racemization makes this impossible to confirm. Fluorothalidomide (FTD), a fluorinated TD analogue with stable, non-racemizing isomers, may serve as a model compound for determining stereoselective effects. In vivo, FTD was undetectable in plasma, suggesting rapid breakdown, as confirmed in vitro, where FTD hydrolyzed up to 22-fold faster than TD. Unlike TD, FTD in pregnant rabbits and mice was highly toxic and lethal to both dams and fetuses. In rabbit embryo culture, FTD initiated optic (eye) vesicle and hindbrain but not classic limb bud embryopathies. Chemical instability, potent general toxicity and absence of limb bud embryopathies make FTD an unsuitable stereoselective model for TD teratogenesis. TD teratogenesis may involve its bioactivation by embryonic prostaglandin H synthases (PHSs) to a free radical intermediate that increases embryopathic reactive oxygen species (ROS) formation. However, the teratogenic potential of rapidly formed TD hydrolysis products and the determinants of species-specific teratogenesis are unclear. For some teratogens, mouse strains that are resistant in vivo are susceptible in embryo culture, suggesting maternal and/or placental determinants of risk. However, TD and two hydrolysis products, 2-phthalimidoglutaramic acid (PGMA) and 2-phthalimidoglutaraic acid (PGA), were non-embryopathic in CD-1 mouse embryo culture. Also, mice deficient in oxoguanine glycosylase 1 (OGG1), which repairs oxidatively damaged DNA, were resistant to TD embryopathies in culture and in vivo. Therefore, murine resistance to TD teratogenesis is dependent on embryonic factors, rather than maternal/placental determinants or increased DNA repair. In contrast, rabbit embryos exposed in culture to TD, PGMA and PGA exhibited head/brain, otic (ear) vesicle and classic limb bud embryopathies, validating the first mammalian embryo culture model for TD teratogenesis and providing the first evidence of a teratogenic role for TD hydrolysis products. Pretreatment with eicosatetraynoic acid (ETYA), a dual PHS/lipoxygenase inhibitor, or phenylbutylnitrone (PBN), a free radical spin trapping agent, completely blocked TD, PGMA and PGA-initiated embryopathies, implicating a PHS-dependent, ROS-mediated embryopathic mechanism.

The expanding therapeutic use of thalidomide (TD) remains limited by its species-specific teratogenicity in humans and rabbits, but not rodents. The R and S isomers of TD may be selectively responsible for its respective therapeutic and teratogenic effects, but rapid in vivo racemization makes this impossible to confirm. Fluorothalidomide (FTD), a fluorinated TD analogue with stable, non-racemizing isomers, may serve as a model compound for determining stereoselective effects. In vivo, FTD was undetectable in plasma, suggesting rapid breakdown, as confirmed in vitro, where FTD hydrolyzed up to 22-fold faster than TD. Unlike TD, FTD in pregnant rabbits and mice was highly toxic and lethal to both dams and fetuses. In rabbit embryo culture, FTD initiated optic (eye) vesicle and hindbrain but not classic limb bud embryopathies. Chemical instability, potent general toxicity and absence of limb bud embryopathies make FTD an unsuitable stereoselective model for TD teratogenesis. TD teratogenesis may involve its bioactivation by embryonic prostaglandin H synthases (PHSs) to a free radical intermediate that increases embryopathic reactive oxygen species (ROS) formation. However, the teratogenic potential of rapidly formed TD hydrolysis products and the determinants of species-specific teratogenesis are unclear. For some teratogens, mouse strains that are resistant in vivo are susceptible in embryo culture, suggesting maternal and/or placental determinants of risk. However, TD and two hydrolysis products, 2-phthalimidoglutaramic acid (PGMA) and 2-phthalimidoglutaraic acid (PGA), were non-embryopathic in CD-1 mouse embryo culture. Also, mice deficient in oxoguanine glycosylase 1 (OGG1), which repairs oxidatively damaged DNA, were resistant to TD embryopathies in culture and in vivo. Therefore, murine resistance to TD teratogenesis is dependent on embryonic factors, rather than maternal/placental determinants or increased DNA repair. In contrast, rabbit embryos exposed in culture to TD, PGMA and PGA exhibited head/brain, otic (ear) vesicle and classic limb bud embryopathies, validating the first mammalian embryo culture model for TD teratogenesis and providing the first evidence of a teratogenic role for TD hydrolysis products. Pretreatment with eicosatetraynoic acid (ETYA), a dual PHS/lipoxygenase inhibitor, or phenylbutylnitrone (PBN), a free radical spin trapping agent, completely blocked TD, PGMA and PGA-initiated embryopathies, implicating a PHS-dependent, ROS-mediated embryopathic mechanism.

L’arthrose (OA) est une maladie dégénérative et multifactorielle caractérisée par une destruction de cartilage, une formation d’ostéophytes et une inflammation au niveau de la membrane synoviale. Le 4-hydroxynonénal (HNE), un produit final de la peroxydation lipidique, a été identifié récemment comme un facteur catabolique et un médiateur inflammatoire dans le cartilage arthrosique humain. Notre projet vise à étudier l’effet du HNE sur la régulation de la prostaglandine E2 synthase-1 microsomale (mPGES-1) et de la protéine activante 5-lipoxygénase (FLAP)/5-lipoxygénase (5-LOX) dans les chondrocytes arthrosiques humains. Lorsque les cellules sont traitées une seule fois avec 10 µM HNE, les résultats de Western blot et de PCR en temps réel montrent que l’expression de la cyclooxygénase-2 (COX-2) et de la mPGES-1 augmente de manière significative et atteint respectivement le maximum après 8 et 16 heures d’incubation puis diminue graduellement. Cependant, lorsque les cellules sont traitées plusieurs fois avec 10 µM HNE à 2 heures d’intervalle, l’expression de la COX-2 et de la mPGES-1 augmente en fonction du temps sans subir une baisse après 24 heures d’incubation. Le HNE induit l’activité du promoteur de la mPGES-1 via l’activation du facteur de transcription Egr-1. L’investigation de la 2ème voie du métabolisme de l’acide arachidonique, à savoir 5-LOX/FLAP, montre que le HNE induit l’expression de FLAP après 24 heures de stimulation et celle de 5-LOX seulement après 48 heures. Ceci semble survenir à l’étape de transcription au cours de laquelle HNE induit l’expression de l’ARNm et l’activité du promoteur du gène 5-LOX. Nous avons démontré aussi que le niveau de leukotriène B4 (LTB4) augmente et suit le même profil que celui de la 5-LOX. L’étude des mécanismes moléculaires susceptibles d’être impliqués dans la régulation de la 5-LOX/FLAP par le HNE montre que ce dernier stimule leur expression via l’action de prostaglandine E2 (PGE2) et du facteur de croissance transformant-beta 1 (TGF-β1). En conclusion, notre étude démontre que le HNE induit à court-terme d’incubation la voie de COX-2/mPGES-1 puis par la suite stimule celle de FLAP/5-LOX à long-terme d’incubation dans les chondrocytes arthrosiques humains. Ces résultats suggèrent que la mPGES-1 et 5-LOX/FLAP sont des potentielles cibles thérapeutiques intéressantes pour contrôler la production de PGE2 et LTB4 dans OA. / 4-hydroxynonenal (HNE), a lipid peroxidation end-product, is produced abundantly in osteoarthritic (OA) articular tissues. Recently, we reported that HNE-induced cyclooxygenase-2 (COX-2) decreased gradually in human OA chondrocytes after 8 h of incubation. This study aimed to investigate whether COX-2 down-regulation is attributed to HNE depletion and is responsible for the switch from COX-2 to 5-lipoxygenase-activating protein (FLAP)/5-lipoxygenase (5-LOX). Treatment of chondrocytes with 10 µM HNE induced prostaglandin E2 (PGE2) release as well as COX-2 and microsomal prostaglandin E2 synthase-1 (mPGES-1) expression at the protein and mRNA levels, with a plateau reached at 8-16 h of incubation, followed by a subsequent decline. However, 8 repeated treatments with 10 µM HNE prevented the reduction of COX-2 and mPGES-1 expression. We demonstrated that HNE induced mPGES-1 promoter activity mainly through transcription factor Egr-1 activation. On the other hand, when COX-2 expression decreased, leukotriene B4 (LTB4) level rose after a long period of stimulation (48 and 72 h). At the mRNA level, HNE induced FLAP and 5-LOX expression after 24 and 48 h of stimulation, respectively. The addition of a nonspecific COX-2 inhibitor (naproxen) to cultured chondrocytes revealed that FLAP and 5-LOX regulation by HNE required PGE2 production. Furthermore, our data showed that 10 µM HNE significantly induced transforming growth factor-beta 1 (TGF-β1) production. The addition of anti-TGF-β antibody to culture medium reduced HNE-induced 5-LOX/FLAP expression by 40%, indicating the involvement of a TGF-β1-dependent mechanism. Our data demonstrate that the shunt to the FLAP/5-LOX pathway in HNE-induced human OA chondrocytes is attributed to COX-2 inhibition, probably due to HNE depletion. PGE2 and TGF-β1 are suggested to be involved in this regulation. Further experiments are in progress to determine other molecular mechanisms underlying this switch in OA chondrocytes.

Human skin has distinct lipid metabolism and production of bioactive lipid mediators that can be modulated by nutritional supplementation with omega-3 polyunsaturated fatty acids (n-3 PUFA), of which eicosapentaenoic (EPA) and docosahexaenoic (DHA) acids exert anti-inflammatory effects. The aims of this project were to gain better understanding of their individual mechanisms in human epidermis and dermis. HaCaT keratinocytes, 46BR.1N fibroblasts, primary human epidermal keratinocytes and dermal fibroblasts were treated with EPA or DHA for 72h and then sham-irradiated or exposed to 15 mJ/cm2 ultraviolet radiation (UVR). Viability was measured by the MTT assay. The expression of cyclooxygenase-2 (COX-2), microsomal prostaglandin synthase-1 (mPGES-1) and 15-hydroxyprostaglandin dehydrogenase (15-PGDH) proteins was explored by western blotting. Human skin explants (n=4 donors) were cultured for 3 or 6 days and supplemented with EPA, DHA or vehicle. Culture media were collected to evaluate tissue damage and PUFA cytotoxicity (lactate dehydrogenase assay). Epidermal and dermal lipid profiles were assessed by gas chromatography and liquid chromatography coupled to tandem mass spectrometry. Primary keratinocytes were treated with fatty acids and various lipid mediators for 48h. Their effect was determined by the scratch assay and transepithelial electrical resistance. UVR upregulated COX-2 in HaCaT and primary epidermal keratinocytes, but did not affect mPGES-1 and 15-PGDH protein expression. UVR upregulated COX-2 and mPGES-1 in 46BR.1N fibroblasts but had no effect on 15-PGDH expression. The same UVR dose did not alter the expression of COX-2, mPGES-1 and 15-PGDH in primary dermal fibroblasts. Only EPA attenuated COX-2 expression in HaCaT and primary keratinocytes and either EPA or DHA had any effect in 46BR.1N and primary fibroblasts. Skin explants showed initial post-biopsy tissue damage. EPA and DHA supplementation augmented cellular levels of the corresponding fatty acids in both epidermis and dermis to a different extent. Increased uptake of DHA in the dermis was accompanied by reduced arachidonic acid levels. EPA treatment stimulated the production of PGE3 and various HEPE in epidermis, while DHA treatment caused high levels of HDHA species in dermis. N-3 PUFA and their derivatives delayed wound healing, cell migration and epidermal barrier permeability, while n-6 PUFA lipids showed the opposite effect. Overall, these findings suggest that EPA and DHA differently affect skin cells and skin, with EPA preference in epidermis and DHA in the dermis. These results highlight the importance of differential skin responses that could be important in skin health and disease.