Identification of genes involded in the adaptation of "Caenorhabditis elegans" to anoxia

Thesis

Reference

Identification of genes involded in the adaptation of "Caenorhabditis elegans" to anoxia

MENUZ, Vincent

Abstract

L'absence d'oxygène est délétère pour la majorité des organismes. Cependant, le nématode

"C. elegans" est capable de survivre 48h dans un milieu sans oxygène (anoxie). Nous avons montré que le gène "hyl-2" est essentiel, alors que son paralogue, le gène "hyl-1", n'est pas nécessaire à la survie du ver en anoxie. "hyl-1" et "hyl-2" codent pour des céramides synthases. Une analyse des céramides par spectrométrie de masse nous a permis de montrer que HYL-1 synthétise préférentiellement des céramides avec des chaînes d'acide gras de C25/C26 carbones, alors que HYL-2 a une préférence pour les chaînes de C21/C22.

Bien que HYL-1 et HYL-2 soient fonctionnellement homologues à la céramide synthase de levure LAG1, "hyl-1" n'est pas capable de complémenter entièrement une perte de fonction de

"hyl-2". De plus, les cellules du nématode exprimant "hyl-2" sont protégées contre l'anoxie.

Ces résultats suggèrent que les céramides à chaîne d'acides gras courtes (C21/C22), produits préférentiellement par HYL-2, confèrent au ver sa résistance contre l'anoxie.

MENUZ, Vincent. Identification of genes involded in the adaptation of "Caenorhabditis elegans" to anoxia. Thèse de doctorat : Univ. Genève, 2008, no. Sc. 3971

URN : urn:nbn:ch:unige-21679

DOI : 10.13097/archive-ouverte/unige:2167

Available at:

http://archive-ouverte.unige.ch/unige:2167

Disclaimer: layout of this document may differ from the published version.

Identification of genes involved in the

adaptation of Caenorhabditis elegans to anoxia

THESE

Présentée à la Faculté des Sciences de l’Université de Genève pour obtenir le grade de Docteur ès sciences, mention biologie

par

Vincent MENUZ

de Bardonnex (Ge)

Thèse n°3971

GENEVE 2008

TABLE OF CONTENTS

REMERCIEMENTS _________________________________________________ 1

I. RESUME and ABSTRACT ________________________________________ 5 I.1. RESUME _________________________________________________________ 6 I.2. ABSTRACT _______________________________________________________ 8

II. INTRODUCTION _______________________________________________ 11 II.1. USE OF OXYGEN FOR ENERGY PRODUCTION ____________________ 12

II.1.1. Oxygen is essential for efficient ATP production _____________________________ 12 II.1.2. Oxygen homeostasis ___________________________________________________ 13 II.2. METAZOANS CAN ADAPT TO HYPOXIA__________________________ 14 II.2.1. Insuring oxygen delivery for all cells ______________________________________ 14 II.2.2. Acute and chronic responses to hypoxia ____________________________________ 15 II.2.3. HIF-1 is the ringmaster of the chronic response to hypoxia in metazoans __________ 18 II.2.4. HIF-1 mediates the cellular metabolic adaptation to chronic hypoxia _____________ 19 II.2.5. HIF-1 mediates systemic adaptation to chronic hypoxia and tumor progression _____ 21 II.3. SOME METAZOANS CAN ADAPT TO SEVERE HYPOXIA AND ANOXIA23 II.3.1. Severe hypoxia or anoxia: cell necrosis ____________________________________ 23 II.3.2. Ischemic-reperfusion injury: cell apoptosis _________________________________ 24 II.3.3. Animals resistant to severe hypoxia or anoxia _______________________________ 25 II.3.4. Postconditioning and preconditioning______________________________________ 27 II.4. CERAMIDES ____________________________________________________ 31 II.4.1. Synthesis of ceramides _________________________________________________ 31 II.4.2. Subcellular localization _________________________________________________ 34 II.4.3. Different species of ceramide: dealing with the complexity _____________________ 35 II.4.4. Bioactive ceramide: lipid-lipid or lipid-protein interactions _____________________ 37 II.4.5. Biological effect of ceramides____________________________________________ 38 II.5. C.elegans AS A MODEL TO STUDY STRESS INDUCED BY OXYGEN DEPRIVATION _______________________________________________________ 41

II.5.1. C.elegans is a powerful genetic model organism _____________________________ 41 II.5.2. C.elegans shows a good adaptation to environmental oxygen deprivation __________ 42 II.5.3. C.elegans displays anoxia-induced suspended animation _______________________ 44 II.5.4. Implication of the globin protein family of C.elegans during anoxia ______________ 45 II.5.5. The daf-2/daf-16 insulin like pathway is implicated in resistance to anoxia_________ 45 II.6. AIM OF THE PROJECT __________________________________________ 47

III. INITIAL OBSERVATIONS ______________________________________ 49 III.1. Strategy to obtain an anoxic environment ____________________________ 50 III.2. E.coli does not grow in anoxia ______________________________________ 51 III.3. Propidium iodide as rapid death indicator ___________________________ 52 III.4. Post-anoxic reperfusion induced C.elegans death ______________________ 53 III.5. Tolerance to anoxia varies according to the developmental stage and sex __ 53

IV. SENSITIVITY TO ANOXIA ______________________________________ 57

IV.3. Tissue damage in AT10 srf-3(yj10) animals exposed to anoxia ___________ 60 IV.4. The AT10 strain carries another mutation that confers sensitivity to anoxia 61 IV.5. STA animal carries a double point mutation in the hyl-2 gene ___________ 63 IV.6. Characterization of hyl-2(gnv1) animals _____________________________ 66 IV.7. hyl-1 and hyl-2 can both functionally replace ceramide synthase activity in S.

cerevisiae______________________________________________________________ 70 IV.8. hyl-1 mutation confers hyper-resistance to anoxia and HYL-1 is functionally different from HYL-2 in C.elegans ________________________________________ 72 IV.9.Ectopic hyl-2 expression fails to rescue the sensitivity to anoxia of hyl-2(gnv1) 75 IV.10. hyl-1 mutants produce shorter ceramides than hyl-2 mutants ___________ 78

V. EFFECT OF PRECONDITIONING ON RESISTANCE TO ANOXIA ___ 83 V.1. Heat-shock preconditioning increases the resistance of C.elegans to anoxia _ 84 V.2. hyl-2 is necessary for HSP __________________________________________ 85 V.3. 10 day-old worms are not able to precondition _________________________ 88

VI. GENES THAT CONFER RESISTANCE TO ANOXIA________________ 89 VI.1. Forward genetic screen to find mutants hyper-resistant to anoxia ________ 90 VI.2. Characterisation of the rta-1 animal _________________________________ 91 VI.3. rta-1 carries more than one recessive mutation conferring the hyper-resistance to anoxia ______________________________________________________________ 93 VI.4. rta-1 carries a single monogenic mutation that increases its life span ______ 94

VII.DISCUSSION __________________________________________________ 97 VII.1. HYL-2 is required for adaptation of young adult C.elegans to anoxia ____ 98 VII.2. How could C21/C22 Cers and SMs protect C.elegans against anoxia _____ 98 VII.3. Exploring the SL pathway _______________________________________ 101 VII.4. HYL-2 independent survival pathways_____________________________ 102 VII.5. A link between mutations in AT10 srf-3(yj10) animals? _______________ 104 VII.6. Concluding remarks and perspectives _____________________________ 104

VIII.SUPPLEMENTARY MATERIAL _______________________________ 107

IX. MATERIAL AND METHODS ___________________________________ 119 IX.1. Anoxic and heat-shock (36°C) environment _________________________ 120 IX.2. Maintenance, anoxia, heat-shock, life span and hypotonic assays ________ 120

IX.5. Functional complementation of lag1Δlac1Δ cells by yeast LAG1 or C.elegans hyl-1 and hyl-2 ________________________________________________________ 123 IX.6. C. elegans lipid extracts __________________________________________ 123 IX.7. Lipid profiling by Mass Spectrometry ______________________________ 124 IX.8. In Vitro assay of dihydroceramide synthase _________________________ 125 IX.9. ATP measurment _______________________________________________ 125 IX.10. ENU mutagenesis ______________________________________________ 125

X. ABBREVIATIONS _____________________________________________ 127

XI. REFERENCES ________________________________________________ 131

XII.ANNEXE: SUBMITTED ARTICLE ______________________________ 141

MERCI Jean-Claude Martinou pour m’avoir accueilli dans ton laboratoire, pour la grande liberté scientifique et la confiance que tu m’as accordée durant ces nombreuses années. MERCI de m’avoir permis de « monter » le labo du ver. MERCI pour les discussions, tant scientifiques que philosophiques que l’on a pu avoir.

MERCI d’avoir pris tout ce temps pour la correction de cette thèse. MERCI de m’avoir donné l’opportunité de m’occuper de Sébastien pour son master durant un an et demi, de m’avoir laissé la responsabilité des travaux pratiques de troisième année et d’avoir pu participer à l’enseignement du cours de BMC de troisième année. MERCI également pour les différents congrès aux Etats-Unis, en Crête et en France.

MERCI à mes examinateurs, les professeurs Michael Hengartner et Howard Riezman, d’avoir accepté et pris le temps de lire ce travail. MERCI Michael pour l’accueil incroyable que tu m’as fait durant ces trois mois passés dans ton laboratoire à Zürich. MERCI pour tes conseils tout au long de cette thèse. MERCI Howard pour ton investissement incroyable dans le projet, pour le temps passé derrière le spectromètre de masse ainsi que pour toutes les discussions constructives concernant ce travail de thèse.

MERCI Marie pour ton investissement sans faille dans ce projet de thèse. Jamais nous n’aurions osé commencer ce projet sans ton expertise. MERCI de ton amitié et des moments partagés au labo et en dehors.

MERCI Monique pour la qualité de ton travail, ton soutien dans les moments de joie et de difficulté. MERCI pour le yoga, pour les affiches, pour les poissons, pour les centris, pour les synchros, pour les anoxies, pour les maintenances, pour le rangement, pour le nettoyage, et j’en oublie !!

MERCI Sébastien pour avoir été LE disciple parfait durant ton travail de master.

J’espère ne pas avoir été trop tyrannique et que tu continueras à t’intéresser à ce qui t’entoure comme tu le fais.

MERCI à Kate Howell pour le boulot sur les levures. MERCI à Sharon Epstein pour toutes les manips sur les membranes de vers. Un grand MERCI à Isabelle Riezman

pour sa patience devant mon impatience et pour toutes ces mises au point avec le spectromètre de masse.

MERCI aux membres passés et présents du labo Martinou, en vrac: Sylvie, Sandrine, Emilie, Safa, Muriel, Stephanie, Philippe, Dominique, Christophe, Daniel, Romain, Yves, Alexis, Sébastien et Syam.

MERCI à tous les membres du Département de Biologie Cellulaire, en particulier à Jurgi et Guillaume. L’un pour les super discussions de biologie, les conseils et le soutien. L’autre pour les marres, pour les délires et pour le kendo. MERCI à vous deux pour l’Amitié. Jurgi, bonne chance pour ton post-doc et vive ta carrière future.

Guillaume, accroche-toi, c’est long, c’est dur, mais ça fait avancer sa vie. MERCI à Denis et Marie pour les discussions et les précieux conseils durant ces 7 dernières années. MERCI à Lukas et Lilli pour tous les conseils sur la génétique du ver et la transgenèse.

MERCI Ben, Dan, Yann (au fait, t’as bientôt fini ta licence ??), La Gley’s et Alex pour votre soutien et les taches sur mes habits. MERCI aux potes du kendo Yannis, Bruno, Christian, Seb, Alex, Nico, Agata, Boris, Gaël et Crog. MERCI à Anne Cat et ta ménagerie, pour ton écoute et ta présence à tous ces moments curieux de la vie.

MERCI à mes parents sans qui absolument RIEN n’aurait été possible !! MERCI de votre confiance, même lors des mes sorties de route adolescente. MERCI à mon frérot Mathieu et à Audrey. Cool de t’avoir comme frère ! MERCI René, Silva, Nico, Lelia, Ce, Chantal et les Salaming pour votre soutien et votre enthousiasme.

MERCI à Claudia ! MERCI d’avoir accepté les week-ends au labo, les rentrées tardives, les réflexions à toutes les heures du jour et de la nuit, les chaussures pleines de BET dans l’appart, … MERCI de ton soutien et de ton amour. Je t’aime.

I. RESUME and

ABSTRACT

I.1. RESUME

La vie est un processus qui demande de l’énergie et la réduction de l’oxygène (O₂) procure une grande quantité d’énergie. En conséquence, la majorité des organismes a développé la capacité de réduire l’O2 pour la production d’énergie cellulaire, sous la forme d’adénosine triphosphate (ATP). Cette réduction s’effectue dans les mitochondries au cours de la phosphorylation oxydative. Les êtres vivants sont formés de cellules qui ont besoin d’énergie pour fonctionner et le rôle central de l’O2

dans la production de cette énergie le rend extrêmement précieux. C’est pourquoi tous les métazoaires ont développé des mécanismes cellulaires communs afin de faire face aux rapides changements de la concentration d’O2. L’apparition du facteur inductible en hypoxie (HIF-1, hypoxia inducible factor-1) est un événement majeur de l’évolution des organismes métazoaires aérobes. En effet, lorsque la quantité d’O2

devient limitée (hypoxie moyenne, 1-2% O2), HIF-1 active la transcription de gènes qui permettent une adaptation cellulaire et systémique au manque d’O2. Cependant, l’exposition des organismes aérobes à une hypoxie sévère (<1% O2) ou à une anoxie (0% O2) conduit, la plupart du temps, à une issue fatale. De ce fait, il n’est pas surprenant de constater que, dans les sociétés occidentales, la cause principale de décès et de handicap chez l’adulte soit liée à des problèmes d’oxygénation des tissus.

Toutefois, un grand nombre de métazoaires, tels que certains vers, poissons, tortues d’eau et cétacés, ont la capacité de gérer de longues périodes d’hypoxie sévère ou d’anoxie, sans effets délétères. Ces organismes ont développé différents mécanismes qui leurs permettent de gérer les manques sévères d’O₂, tel que la capacité de vivre en état d’hypo-métabolisme, de fermer certains canaux ioniques, ainsi que de redistribuer très spécifiquement l’utilisation de l’ATP. C’est pourquoi l’étude de l’homéostasie de l’O2 chez ces organismes peut à la fois mettre en lumière de nouveaux concepts concernant certains fondements physiologiques de la biologie, mais aussi amener de nouvelles pistes thérapeutiques concernant les problèmes de santé liés aux problèmes d’oxygénation des tissus (infarctus du myocarde, ischmémie cérébrale).

Le nématode Caenorhabditis elegans a la faculté de résister à l’anoxie, ce qui en fait un modèle extrêmement intéressant pour l’étude de l’homéostasie de l’oxygène. Nous avons observé que les larves L1 soumis à une carence alimentaire, les jeunes adultes mâles (72 h post-L1) ainsi que les animaux âgés de 10 jours sont plus résistants à l’anoxie que des vers hermaphrodites au stade 72 h post-L1. Puis,

nous avons montré qu’une courte incubation à 30°C, préalable à l’anoxie, protégeait les animaux contre les effets délétères du manque d’oxygène.

Nous avons ensuite recherché des gènes essentiels à l’adaptation du ver à l’anoxie. Nous avons montré que le gène hyl-2 est essentiel, alors que son paralogue, le gène hyl-1, n’est pas nécessaire à la survie du ver en anoxie. Les gènes hyl-1 et hyl- 2 (homolog of yeast longevity) font partie d’une large famille de gènes eucaryotiques connus sous le nom de gènes Lass (Longevity Assurance) qui sont tous en relation avec le gène de levure LAG1 qui code pour une céramide synthase. Nous avons montré, in vitro, que HYL-1 synthétise préférentiellement des céramides avec des chaînes d’acide gras de C25/C26 carbones, alors que HYL-2 a une préférence pour les chaînes de C21/C22. En accord avec ces résultats, nous avons trouvé que les mutants hyl-2 contiennent moins de céramides C21/C22 et plus de céramides C25/C26. Les mutants hyl-1 ont à l’inverse plus de C25/C26. Bien que HYL-1 et HYL-2 soient fonctionnellement homologues à LAG1, hyl-1 n’est pas capable de complémenter entièrement une perte de fonction de hyl-2. De plus, les cellules qui expriment hyl-2 de façon ectopique sont protégées contre l’anoxie. Ainsi, nos résultats suggèrent que les céramides C21/C22 protègent C.elegans contre l’anoxie.

Enfin, nous avons recherché des gènes qui confèrent une hyper résistance à C.elegans à l’anoxie. Nous avons mutagénisé chimiquement le ver pour induire des mutations susceptibles de le rendre hyper résistant à l’anoxie. Nous avons ainsi isolé un animal, rta-1 (resistant to anoxia-1), qui montrait une très grande résistance à l’anoxie et aux chocs thermiques. En outre, la durée de vie de cet animal s’est avérée être augmentée de 100% par rapport à la normale. Le phénotype de résistance à l’anoxie est dû à des mutations multigéniques, ce qui rend la cartographie des gènes mutés extrêmement difficile. Par contre, le phénotype de longévité de rta-1 est dû à une mutation monogénique récessive, lls-1 (long life-span-1), située sur le chromosome III.

Ce travail met en évidence l’implication de certains types de céramides dans la résistance à l’anoxie chez C.elegans. La compréhension des mécanismes par lesquels les céramides protègent le ver contre l’anoxie pourrait permettre la découverte de nouvelles voies de régulation de l’adaptation des cellules au manque d’oxygène, non seulement chez le nématode, mais aussi chez le mammifère.

I.2. ABSTRACT

Life is a process that uses energy and oxygen (O₂) provides a large free energy release per electron transfer. As a result, the vast majority of the organisms evolved the capacity to reduce O2 for the production of cellular energy, supplied in the form of adenosine tri-phosphate (ATP) by the oxidative phosphorylation in the mitochondria.

Being at the center of life, aerobe organisms have developed a vast number of strategies to deal with a lack of O2 availability. First of all, organisms are made of individual cells that need energy to function. Hence, complex organisms such as vertebrates have evolved complicated systems to deliver O2 to each cell, as well as conserved mechanisms to sense rapid changes in O2 availability. A critical event in metazoan evolution was the emergence of the hypoxia inducible factor-1 (HIF-1), which acts as a universal transcription factor when O2 availability decreases (hypoxia). As a result, in case of mild-hypoxia (1-2% O2), HIF-1 drives cellular and systemic modifications that permit the organism to cope with the lack of O2. Nevertheless, in case of severe hypoxia (<1% O2) or anoxia (0% O2), the majority of aerobe organisms rapidly dies. However, some metazoans like nematodes, fishes, water turtles and cetaceans have developed the ability to live long periods of severe hypoxia or anoxia. Through a regulated hypometabolism, channel arrest and redistribution of ATP. Importantly, all these strategies seem to be independent of the HIF-1 pathway. Investigating the molecular mechanism contributing to the tolerance of these animals to anoxia could allow to unravel novel metabolic regulators. The nematode Caenorhabditis elegans (C.elegans) is a powerful genetic model organism that shows an excellent adaptation to anoxia, making it extremely attractive as a model for oxygen deprivation studies.

Initially, we observed that the resistance to anoxia depended on the stage of development, the gender and the age of the worm. Indeed, unfed L1, young adult males (72 h post-L1) and 10 day-old hermaphrodites were more resistant to anoxia compared to 72 h post-L1 hermaphrodite. Moreover, we observed that incubating young adult nematodes at 30°C two h immediately before anoxia significantly increased their resistance to oxygen deprivation.

We then searched for mutations that would render C.elegans hypersensitive to anoxia. We found that the gene hyl-2 is essential for adaptation of young adults to anoxia. In contrast, hyl-1, a hyl-2 paralog, is dispensable for adaptation to anoxia.

Both hyl-1 and hyl-2 genes (homolog of yeast longevity) belong to a large eukaryotic gene family known as longevity assurance genes (Lass genes) that are all related to the yeast LAG1 gene, which encodes a ceramide synthase (CerS). We demonstrated, in vitro that HYL-1 synthesizes preferentially ceramides with a C25/C26 fatty-acyl chain length, whereas HYL-2 prefers C21/C22 fatty-acyl chains. Accordingly, we found that hyl-2 mutants contain less C21/C22 ceramides and more C25/C26 ceramides. hyl-1 mutants showed the opposite. Although both HYL-1 and HYL-2 are functionally homologous to LAG1 in yeast, hyl-1 rescues only partially hyl-2 loss-of- function. Moreover, cells expressing hyl-2 ectopically are protected against anoxic injury. Altogether, these data strongly suggest that C21/C22 ceramides protect C.elegans against anoxia.

Finally, we searched for animals hyper-resistant to anoxia. We performed a forward genetic screen on wild type animals, using the powerful mutagen N-nitroso- N-ethylurea (ENU), in order to find mutants hyper-resistant to anoxia. We identified one, rta-1 (resistant to anoxia-1) that displays a significant increased resistance to anoxia compared to wild type animals. Interestingly this mutant was also found to have an increased lifespan (47 days) compared to wild type N2 (29 days). Through genetic studies, we were able to determine that the resistance to anoxia is due to multigenic mutations whereas the increased longevity is due to a monogenic mutation, probably residing on chromosome III. Further studies should allow to characterize these mutations.

This work has brought new clues on the implication of specific ceramides species in the adaptation of C.elegans to anoxia. Understanding the molecular mechanisms underlying the role of ceramides in anoxia could allow to define novel signalling pathways that may be needed not only in C.elegans but also in mammalians cells to fight against oxygen deprivation.

II. INTRODUCTION

II.1. USE OF OXYGEN FOR ENERGY PRODUCTION II.1.1. Oxygen is essential for efficient ATP production

Microbial population has largely determined atmospheric evolution of Earth (1).

About 2.3 billion years ago, by a complex biochemical process called photosynthesis that converts CO₂ + H₂O into CH₂O + O₂ in the presence of light, microorganisms such as cyanobacteria have changed the Earth atmosphere from virtually anoxic to about 21% oxygen (O₂, normoxic environment) (2). Oxygen provides a large free energy release per electron transfer. As a result, most multicellular organisms evolved the capacity to reduce oxygen for the production of cellular energy, supplied in the form of adenosine tri-phosphate (ATP), by the oxidative phosphorylation in the mitochondria (3). Briefly, in eukaryotic cells, the glycolytic pathway converts one molecule of glucose into two pyruvate molecules. This pathway produces 2 ATP and 2 molecules of the electron carrier nicotinamide adenine dinucleotide (NADH). The pyruvate is then translocated into mitochondria and converted into CO2 and a 2- carbon acetyl group, which is combined with coenzymeA (CoA) to form acetyl-CoA that enters into the citric acid cycle (Krebs cycle). This biochemical cycle generates 3 molecules of NADH, 1 molecule of flavine adenine dinucleotide (FADH2), 1 molecule of GTP and 2 molecules of CO2. The two electron carrier molecules FADH2

and NADH donate their electron to the process of oxidative phosphorylation that occurs in the inner membrane of the mitochondrion and involves the activity of four protein complexes (Complex I-IV). From complex I to IV, electrons are passed through a series of cytochromes (electron transport chain, ETC) where the energy released by their transfer is used to generate a gradient of protons across the inner membrane of the mitochondrion. This proton gradient serves as the driving force for the production of ATP by activating the mitochondrial ATP synthase (F₁F₀-ATPase).

Molecular oxygen (O2) acts as the final electron acceptor of the ETC at complex IV, a reaction catalysed by the cytochrome c oxydase. At the end of the series of electron transfer through the complexes, cytochrome c passes the low energy electrons to molecules of O₂, which are combined with 2H⁺ to form the transient reactive oxygen species (ROS) H₂O₂ which is then transformed into H₂O (summarized in figure 1) (4, 5). About 90% of the available oxygen in a cell is consumed by mitochondria during the oxidative phosphorylation (6).

Figure 1: Schematic representation of the oxidative phosphorylation process. The electron are transmitted by the electron carrier NADH to the complex I and then are transmitted to each complex by specific cytochromes (cyt) leading to the pumping of protons (H⁺) through the mitochondria inner membrane. The proton gradient is used to phosphorylate ADP into ATP by the ATP synthase (F₁F₀- ATPase). O₂ is the final acceptor of the electron on complex IV and forms the transient ROS H₂O₂, rapidly transformed into H2O (modified from 5).

Together, the glycolysis, the citric acid cycle and the oxidative phosphorylation are extremely efficient and produce 38 ATP for each molecule of glucose consumed. The ATP produced in this process is used to fuel the vast majority of cellular process (7). Consequently, most multicellular organisms depend on oxygen for the production of their metabolic energy (8).

II.1.2. Oxygen homeostasis

Importantly, when in excess and in the presence of the appropriate activators, oxygen gives rise to unstable ROS that have the potential to oxidize cellular proteins, lipids and nucleic acids and, by doing so, may cause cell dysfunction or death (9).

Moreover, the process of oxidative phosphorylation is not 100% efficient and about 20% of H⁺ undergo regulated proton leak, which gives rise to the generation of mitochondrial ROS (10). Hence, complex systems have evolved to deal with the delicate balance that exists between the sufficient supply of oxygen to a cell for energy production and its highly reactive and potentially toxic oxidizing chemical properties (2) (figure 2).

Figure 2: Oxygen homeostasis. Changes in O2 partial pressure (pO2) result in decreased ATP production and/or increased ROS production. The homeostasis state represents the correct balances between too much or too less O₂. When pO₂ is low, the supply increases, when the pO₂ is high, the demand diminish (from 11).

Because of the extreme necessity to use O2 for efficient ATP production, O2

deprivation leads rapidly to a state of metabolic crisis where insufficient ATP levels are produced, a situation that can possibly lead to irreversible cellular damages. While most metazoan possess little naturally evolved tolerance to acute and chronic hypoxia (1-2% O2), acute severe hypoxia (<1% O2) or anoxia (0% O2) usually lead to death.

Indeed, severe oxygen deprivation are encountered in various human pathologic conditions including myocardial and cerebral ischemia, chronic lung disease and cancer progression (7, 12-14).

II.2. METAZOANS CAN ADAPT TO HYPOXIA

Hypoxia is a state of mild-deficient oxygen availability (1-2% O2). We distinguish between environmental hypoxia, which concerns inadequate atmospheric O₂ partial pressure (tension, pO₂) and stagnant hypoxia that involves inadequate blood local circulation, called ischemia, which leads to hypoxia and hypoglycemia and that could happen in a normoxic environment (15).

II.2.1. Insuring oxygen delivery for all cells

Given the importance of oxygen in energy production, physiological systems have evolved to ensure the optimal oxygenation of all cells in each organism. In simple invertebrates with few cells layers, as in C.elegans, the small body size allows direct diffusion of oxygen to the cells (16). In more complex invertebrates, a series of branching tracheal tubes forms a network that fulfills oxygen delivery to cells.

Basically, air enters the trachea and is transported through the network to facilitate the oxygen supply to all cells (17) (figure 3). In vertebrates, due to the high complexity and the size of such organisms, efficient systems of oxygen delivery involving

combined activity of the respiratory and circulatory systems have evolved to insure the absorption of oxygen in all cells (18) (figure 3 b). Whatever the delivery system adopted for a given organism, there exists an oxygen gradient of pO₂ value from atmosphere to individual cells and from cytoplasm to mitochondria (19, 20) (figure 3 c). Using these gradients, organisms have evolved adequate responses to hypoxia (1- 2% O₂) and have developed systemic and cellular O₂ sensing mechanisms (8, 15, 21).

Figure 3: D. melanogaster and human strategy to bring O₂ to cells. a. Visualization of D.melanogaster embryo tracheal branches by antibody staining against lumenal antigens (adapted from 17). DT: dorsal trunk, DB: dorsal branch, VB: visceral branch, LT: lateral trunk, GB: ganglionic branch. b. Schematic representation of complex oxygen delivery by the respiratory and circulatory systems in human (*). c. Schematic representation of the oxygen gradient from the atmosphere to the mitochondria (modified from 20).

*Ref b: http://www.williamsclass.com/SeventhScienceWork/CellsOrganization.htm

II.2.2. Acute and chronic responses to hypoxia

Whatever the organism considered, there are two responses to hypoxia: the acute response, that occurs within seconds to minutes following acute hypoxia and the

chronic response, which reflect an adaptation to chronic hypoxia and that occurs within hours to days (15, 22).

The acute response is characterized by a non-transcriptional immediate adaptation to hypoxia. In vertebrates, particularly in mammals, the survival in acute environmental hypoxia requires rapid respiratory and cardiovascular adjustments to ensure O₂ delivery to the most critical organs such as the brain or the heart (23). In these animals, any slight pO₂ variation is instantaneously sensed by the carotid and aortic bodies that are tiny organs situated in artery close to the heart and that sense any pO2 variation in the blood (24, 25). The neuroepithelial bodies, situated in the lungs, assume a similar function and detect pO2 changes of the inspired air (26, 27).

In these organs, the neurosecretory glomus are the primarily cells responsible for O2

sensing. Briefly, excitation of these cells by hypoxia depends on the presence of membrane K⁺ channels whose activity is inhibited by low pO2. The closure of the K⁺ channels by low pO2 leads to membrane depolarization and to Ca²⁺ influx, followed by a release of transmitter to the extracellular milieu and, in the innervated organs, activation of afferent sensory fibers (23, 28, 29). The information is then carried by chemoreceptor afferent to the brain, which in turn initiates respiratory and cardiovascular reflexes to ensure proper oxygenation in the vital organs (30) (figure 4). The potential identity of the O2 sensor itself remains to be determined (28) and candidate sensors include mitochondria, the AMP-activated protein kinase (AMPK) and the haemoxygenase-2 (7).

Figure 4: Schematic representation of systemic acute hypoxia response in mammals. When O2

availability becomes limited (hypoxia, 1), the K⁺ channels are closed, leading to an influx of Ca²⁺ (2) that leads to the release of transmitter (3), which then activates afferent sensory membranes that drive the information to the CNS (4) and then adjust respiration and vascularization (5). [Ca²⁺]: calcium concentration. (Adapted from 13).

Originally, oxygen sensing was attributed solely to these specialized chemoreceptor cells. In fact, all nucleated cells have the capacity to sense pO₂ and to respond to hypoxia (12). The cellular acute response to hypoxia is mainly regulated by the AMP-activated protein kinase (AMPK), which is found in all eukaryotic cells.

AMPK acts as a cellular energy sensor that is activated by increased AMP:ATP ratio caused by metabolic stresses that interfere with ATP production (e.g., hypoxia) or accelerate ATP consumption (e.g., muscle contraction). Thus, it could be considered as a cellular energy gauge. Basically, when cellular concentration of ATP levels fall because of acute hypoxia, activated AMPK promotes catabolic processes and inhibits anabolic processes by phosphorylating specific substrates. Note that hormones and cytokines can interact with the system, suggesting a role in maintaining energy balance at the whole organism level (31).

While cellular responses to acute hypoxia modify preexisting proteins, chronic exposure to hypoxia leads to chronic responses, which activate gene transcription that in turn lead to important remodeling of systemic and cellular metabolism. A critical event in metazoan evolution was the emergence of a transcription factor, the hypoxia inducible factor-1 (HIF-1), which functions to regulate gene expression in the hypoxic chronic response (32, 33). HIF-1 consists of a basic helix-loop-helix-PAS domain transcription factor, which, in chronic hypoxia, upregulates genes necessary for cellular metabolism, as well as genes that trigger systemic microenvironment remodeling (e.g., angiogenesis). Figure 5 summarizes the acute and chronic response to hypoxia in metazoans.

Figure 5: Schematic diagram of the time course of metabolic responses following acute and chronic hypoxia. Acute responses imply rapid systemic responses, followed by tissue/cell responses.

Simple metazoan only show tissue/cell responses. Chronic response starts after few h and leads, via HIF-1, to important remodeling of systemic and cellular metabolism (adapted from 34).

II.2.3. HIF-1 is the ringmaster of the chronic response to hypoxia in metazoans

HIF-1 is a heterodimer composed of an O2-regulated HIF-1α (or HIF-2α) subunit and a constitutively expressed HIF-1ß subunit (35). Under normoxic condition, HIF-1α is continuously synthetized and degraded (figure 6). The degradation implies the hydroxylation of two specific prolines by the prolyl hydroxylase domain protein 2 (PHD2), which is a dioxygenase that utilizes both O2

and α-ketoglutarate as substrates while generating CO2 and succinate as by-products (36-41). Another putative regulatory protein, ARD1, has been shown to acetylate a specific lysine of HIF-1α (42). Both proline hydroxylation and lysine acetylation of HIF-1α lead to its binding to the von Hippel-Lindau tumor-suppressor protein (VHL), which interacts with the protein Elongin C, thereby recruiting an E3 ubiquitin-protein ligase complex that ubiquitnates HIF-1α and targets it for degradation by the 26S proteasome (figure 6) (43-45). Under hypoxic conditions, PHD2 activity is reduced both by O₂ limitation and inhibition of its catalytic center (39, 46, 47). Importantly, it has been shown that hypoxia induces mitochondrial ROS production (48). PHD2 catalytic center contains Fe(II), which is oxidized and thus inhibited by ROS generated at complex III of the mitochondrial respiratory chain during the hypoxia, hence supporting the HIF-1α stabilization (46, 48, 49). Inhibition of PHD2 leads to the dimerization of HIF-1α with HIF-1ß. This heterodimer binds to specific cis-acting

hypoxia response element (HRE, 5’-RCGTG-3’) in target genes and recruits the trans- acting co-activators p300 and CBP [CREB (camp-response-element-binding protein)- binding protein] (11), which leads to the transcription of more than hundreds different genes (figure 6) (50, 51). This simple model of oxygen activated PHD2 places PHD2 as the direct sensor of O₂ levels (52). HIF-2α is a protein with less than 50% amino acid identity to HIF-1α. It is regulated similarly to HIF-1α (53). However, the genes regulated by HIF-1 and HIF-2 are more or less distinct and are dependent on the cell type (54). The Factor Inhibiting HIF (FIH-1), similarly to PHD2, is another α- ketoglutarate-dependent dioxygenase that hydroxylates a specific asparagin of HIF- 1α. This hydroxylation blocks the interaction of HIF-1α with the coactivators p300 and CBP (figure 6) (55).

A growing number of proteins, such as OS-9, SSAT2, HSP90, RACK1, Siah2 and VDU2 have been identified to interfere with HIF-1 activity, implying different levels of complexity in the regulation of HIF-1 (11).

Figure 6: Schematic representation of HIF-1 regulation. In normoxia, specific prolines of HIF-1α are hydroxylated by PHD2, leading to the binding of VHL and the subsequent degradation of HIF-1α by the proteasome (left panel). Alernatively, FIH hydroxylates a specific asparagin that blocks p300 and CBP co-activators recruitment (left panel). During hypoxia, both absence of O₂ and ROS lead to the inhibition of PHD2 (right panel). HIF-1α is no more degraded and dimerize with HIF-1ß. The HIF- 1 heterodimer triggers transcription of specific genes (right panel) (from 52).

II.2.4. HIF-1 mediates the cellular metabolic adaptation to chronic hypoxia

In all metazoans, HIF-1 plays a central role in determining the cellular metabolic strategy under condition of hypoxia. Hence, the regulation of each of the three phases of aerobic respiration (glycolysis, Krebs cycle and oxidative

phosphorylation) is under regulatory control determined by transcriptional events, which are governed by HIF-1.

In hypoxia, eukaryotic cells have the ability to dramatically enhance the efficiency of the glycolytic pathway, hence becoming the primary source of metabolic energy. This phenomenon is known as the Pasteur effect (56) and is regulated by HIF- 1 (57). In addition, the genes coding for certain glucose transporters are upregulated by HIF-1.

HIF-1 induces the expression of the pyruvate dehydrogenase kinase 1 (PDK1) and of the lactate dehydrogenase A (LDHA). PDK1 phosphorylates and inactivates the pyruvate dehydrogenase, which is responsible for converting pyruvate into acetyl- CoA, causing a decreasing entry of the pyruvate into the Krebs cycle and thereby a decrease in the delivery of reducing equivalents to the electron-transport chain (figure 7 a). Note that decreasing the oxidative phosphorylation leads to a decrease of excessive mitochondrial reactive ROS. Because hypoxic cells already exhibit increased ROS that are necessary for HIF-1α accumulation (48, 49, 58), the induction of PDK1 prevents the persistence of potentially harmful ROS levels (59). Then, LDHA increases the lactic fermentation by promoting the conversion of pyruvate into lactate (figure 7 a) (11, 60).

Interestingly, a recent study has shown that, under hypoxia, HIF-1 induces transcriptional mechanisms that alters the cytochrome c oxidase subunits (complex IV of the electron transport chain), leading to an optimized energy production. Hence, HIF-1 promotes the efficient use of available O2 while also reducing the generation of harmful byproducts of respiration such as H2O2 (11, 61). Briefly, some of the 13 polypeptides (COX subunit) that compose complex IV have multiple isoforms and are subject to regulation. COX4 has 2 isoforms: COX4-1 and COX4-2. During hypoxia, HIF-1 upregulates the expression of COX4-2 and of LON mitochondrial protease, which degrades COX4-1. This facilitates the swapping of subunit COX4-1 for the more efficient subunit COX4-2, hence enhancing mitochondrial respiration (figure 7 b) (62). Interestingly, despite the fact that there is no homolog of HIF-1 in yeast, a similar mechanism of swapping mediated by oxygen concentration exists, suggesting a convergent evolution between yeast and metazoan on hypoxic response (11, 62).

Figure 7: Hypoxia-induced cellular metabolic changes by HIF-1. a. HIF-1 stimulates the expression of glucose transporters and glycolytic enzymes, and thus promotes glycolysis to generate pyruvate.

Moreover, HIF-1 promotes pyruvate reduction to lactate by activating LDH and induces PDK1, which inhibit pyruvate dehdrogenase and blocks conversion of pyruvate to acetyl-CoA, resulting in decreased flux through the Krebs cycle. Decreased Krebs cycle activity results in attenuation of oxidative phosphorylation and excessive mitochondrial ROS production (from (59). b. Schematic representation of the COX4-1-COX4-2 swapping in mammals. HIF-1 induces the mitochondrial LON protease that degrades COX4-1. Simultaneously, HIF-1 induces COX4-2 that replaces the degraded COX4-1 for more efficient aerobic respiration (from 61).

II.2.5. HIF-1 mediates systemic adaptation to chronic hypoxia and tumor progression

Under chronic hypoxia, in addition to cellular adaptation, HIF-1 mediates developmental and physiological pathways by the induction of genes necessary for the reorganization of the microenvironment. For instance, HIF-1 has been shown to modulate the expression of erythropoietin (EPO), which increases the number of red blood cells, thereby increasing the oxygen carrying capacity of the blood circulation (63). Moreover, HIF-1 initiates angiogenesis and vascular remodeling by upregulating the vascular endothelial growth factor (VEGF) (64) and the VEGF receptor-1 (Flt-1) (65, 66). Finally, HIF may also influence different types of cell death including apoptosis, necrosis and autophagy (67). Targets of HIF-1 and HIF-2 are only partly overlapping and the battery of target genes varies considerably from one cell type to another, suggesting that the complete HIF transcriptome is likely to reach thousands of genes (33).

Paradoxically, while HIF evolved to protect metazoans against chronic hypoxia, it is also associated with tumor malignity. Solid tumors are known to contain poorly vascularized regions characterized by hypoxia (14) and HIF-1 has been shown to be a positive regulator of tumorigenesis in a variety of human cancer (68) (figure 8 a). In hypoxic tumor cells, HIF-1 turns on genes such as insulin-like growth factor (ILGF), which induces cellular survival and proliferation (69). Proteins regulating motility (Lysyl oxidase), metastasis (CXCR4, E-cadherin) (14), as well as those that promote tumor invasion and migration, such as matrix metalloproteinase-2 (70) are also upregulated. In a dramatic example, c-Met, a receptor for hepatocyte growth factor (HGF, produces by the stroma), is upregulated by HIF-1. Binding of HGF to c-Met increases cell motility, invasion and metastasis (71, 72) (figure 8 b).

Figure 8: HIF activation in hypoxic solid tumor. a. Red cells represent well-oxygenated tumor cells, while blue cells represent hypoxic tumor cells in which HIF-1 and HIF-2 are stabilized. The induction of HIF implicates the survival of the tumor (from (14). b. Blue cells are hypoxic tumoric cells in which HIF-1 activates the transcription of both VEGF and c-Met. The binding of HGF to c-Met triggers cell motility in the new blood vessels attracted by VEGF, hence leading to metastasis invasion (from 72).

Although HIF is viewed as a universal global and ubiquitous regulator that links O2 availability and gene transcription, it is not the only pathway that links pO2

and transcription activity. Indeed, depending on the severity of the oxygen depletion, the site and the combinatorial nature of the stimuli, hypoxic gene expression can alternatively be mediated, or modified, by different oxygen- and redox-sensitive transcription factors such as the nuclear factor-κB (NF-κB), p53, the C-Fos, c-Jun monomer subunits in activating protein-1 (AP-1), the members of Sp-factor family (i.e. Sp-1, Sp-3), the early growth response protein-1 (Egr-1), the cyclic AMP- response-element binding protein (CREB), the metal-transcription factor-1 (MTF-1) and the nuclear factor for interleukin 6 (NF-IL6 aka C/EBPβ) (73). However, little is known about the activating mechanisms of these factors or their potential cross-talk with the HIF pathway (74).

II.3. SOME METAZOANS CAN ADAPT TO SEVERE HYPOXIA AND ANOXIA

While metazoans have developed strategies to cope with acute and/or chronic hypoxia (1-2%), severe oxygen deprivation such as acute severe hypoxia (<1%) or anoxia (0%) lead, most of the time, to the death of these organisms within minutes.

The majority of mammals, for instance, suffocate if breathing is stopped for more than 3-4 minutes (15).

II.3.1. Severe hypoxia or anoxia: cell necrosis

During acute severe ischemia, metazoans’ cells die by necrosis as a consequence of a loss of ionic integrity of the cell membranes (75). Basically, ion leakage across cell membranes occurs as a result of both intracellular and extracellular ions drifting towards their thermodynamic equilibrium. Maintenance of a homeostatic intracellular environment therefore requires the redistribution of these ions through the use of ATP-dependent pumping systems such as the Na+/K+- ATPase, which consume ATP. Cell death occurs when ATP production fails to meet the energetic maintenance demand of ionic and osmotic equilibrium (6, 75) (figure 9 a). Depending on the cell type, acute severe ischemia does not have the same consequences. Neurons are electrically excitable cell of the nervous system that process and transmit information using ionic channels. Hence, in such cells, the maintenance of the ionic integrity of the membranes is extremely important and 80%

of the ATP produced is dedicated to this purpose (76), rendering them extremely sensitive to acute severe ischemia. Indeed, 15-20 seconds of brain anoxia leads to the loss of electroencephalographic (EEG) activities (15). In contrast to neurons, skeletal muscles use only 20% of ATP to maintain their membrane ionic integrity and show a high tolerance to acute severe ischemia (6, 77).

In addition to the loss of ionic integrity of the membranes, severe ischemia transforms mitochondria from ATP producers to ATP consumer. In an attempt to maintain the mitochondrial membrane potential, the F₁F₀-ATPase runs backwards and uses ATP to actively pump protons from the matrix to the intermembrane space (figure 9 b).

Figure 9: Sequence of events leading to acute severe ischemic cell death. a. Depending of the cell type, such death can occur from seconds to minutes. b. During normoxia, proton are transferred by complex I, III and IV from the matrix to the inner mitochondrial membrane, generating a proton motive force that is used by the ATPase to produce ATP. In severe ischemia, the ATPase runs backwards and pumps actively protons to the inner mitochondrial membrane in an attempt to maintain the mitochondrial membrane potential (adapted from 75).

II.3.2. Ischemic-reperfusion injury: cell apoptosis

Reperfusion is the re-oxygenation of a tissue/organism after an acute severe hypoxia or anoxia (severe ischemia). Although reperfusion is obviously essential for metazoans or organs to recover from a period of severe ischemia, it is known that reperfusion itself can expedite cell death. This is called ischemic/reperfusion (IR) injury (78). IR injury has been shown to occur in a large variety of metazoans (9, 79) and is thought to be responsible, in large part, for the deleterious effects of oxygen deprivation (78, 80), with apoptosis playing a major role (81-83).

Mitochondria play a prominent role in reperfusion. First of all, during severe ischemia, electrons accumulate in the mitochondrial electron chain transport. When molecular oxygen becomes again available, the accumulated electrons suddenly combine with O2, leading to a huge burst of ROS (84, 85). This results in lipid peroxidation, protein oxidation and DNA damage (9, 86-88). Depending of the gravity of the damages, cells undergo necrotic- or apoptotic-induced death (89, 90).

Second, reperfusion induces the opening of the so-called mitochondrial permeability transition pore (MPTP) (91). MPTP are formed by multiprotein complexes that form large nonselective pores in the inner mitochondrial membrane (92) and are opened by Ca²⁺, ROS, ATP depletion and mitochondrial depolarization, all existing during reperfusion (93). The opening of the MPTP during the reperfusion leads to matrix swelling and subsequent rupture of the outer membrane leading to release of proapoptotic proteins and finally to apoptosis (93).

Proteosomal degradation of Mcl-1, a protein belonging to the anti-apoptotic member of the Bcl-2 family, during severe oxygen deprivation (anoxia) may also contribute to apoptosis (94), a phenomenon that does not induce apoptosis per se (95).

In addition, it has been shown that post-ischemic tissues produce tumor necrosis factor-α (TNF-α), which is a crucial inducer of apoptosis in mouse liver (96), in part via ceramide production (97). Consisting with this finding, a recent study has shown accumulation of ceramide in cerebral IR injury (98).

Note that apoptosis, in contrast to necrosis, occurs mainly during reperfusion and almost not during ischemia (81, 82), because apoptosis is an energy-requiring form of cell death and reperfusion is essential to generate the necessary amount of ATP molecules (99).

II.3.3. Animals resistant to severe hypoxia or anoxia

Severe hypoxia and anoxia result in cell death for the majority of the vertebrates.

However, several metazoans are adapted to live with such stresses. For example the harbor seal (P. vitulina, 100), the crucian carp (C. carassius, 101), the Californian blind goby (T. californiensis, 101), the turtle (T. scripta and C. picta, 101) the brine shrimp embryos (A. franciscana, 102), the zebrafish embryos (D. rerio, 103), the fruit fly (D. melanogaster, 104) and the nematode (C.elegans, 105) have the ability to live from h to years in the complete absence of O2. These animals have evolved strategies to cope with the absence of O2.

First of all, animals resistant to severe hypoxia or anoxia have the ability to reallocate cellular energy between essential and non-essential ATP demand processes as energy becomes limited. Hence, protein synthesis is largely inhibited in response to anoxia (106, 107) and the energy spared by the reduction in protein synthesis is reallocated to more critical cell functions involved in osmotic and ionic homeostasis with Na⁺/K⁺ pumping taking the greatest priority (75, 108). Moreover, the absolute ATP demand of the Na⁺/K⁺ pump is reduced during oxygen deprivation, operating at only 25% of its absolute ATP demand in normoxia. The net effect is that, during severe hypoxia or anoxia, cellular energy is unequally balanced between the ion motive ATPases, which use 80% of the total ATP produced, and the protein synthesis (figure 10 a) (109, 110). Animals resistant to severe anoxia or hypoxia exhibit large- scale reduction in absolute ion motive ATPases activity, and then ATP consumption, during severe hypoxia or anoxia. That is certainly the main difference with metazoans sensitive to anoxia. The regulated-hypometabolism of anoxia-tolerant animals allows them to survive prolonged periods of severe hypoxia or anoxia, while the anoxia- sensitive animals show a forced-hypometabolism that rapidly leads to the death (figure 10 b) (15, 34).

Figure 10: ATP turnover of anoxia-tolerant cells exposed to anoxia. a. Histograms show the relative contribution that protein synthesis and ion-motive ATPase make to total ATP turnover in normoxia (left) and after oxygen deprivation (right). The ion motive ATPase functions at only 25% of its absolute ATP demands in normoxia. b. Anoxia-tolerant animals have a regulated hypometabolism that allows them to diminish their ATP consumption, allowing them to live longer in anoxia. Anoxia- sensitive animals show a forced hypometabolism: a rapid use of their total ATP does not allow them to live prolonged periods in anoxia (adapted from 34).

Moreover, the reduction in ionic pump activity in hypoxia- or anoxia-tolerant animals cells is part of a coordinated process of energy conservation wherein severe hypoxia or anoxia initiates a generalized suppression of ion-channel densities and/or channel leak activities. The net effect is that cell membrane permeability is reduced, lowering the energetic costs of maintaining transmembrane ion gradients. This

phenomenon is known as “channel arrest” (figure 11) (111) and is regulated, at least in part, by adenosine (112). In addition, in hypoxia- or anoxia-tolerant animal, the inhibitory F₁-ATPase subunit (IF₁) inhibits the proton pumping of the F₀F₁-ATPase during anoxia (113).

Hence, the key of hypoxia- or anoxia-tolerant animals to survive oxygen deprivation relies on their ability to restrict the use of energy to maintain cell membrane integrity (34).

Figure 11: Schematic representation of the “channel arrest” and inhibition of the F1-ATPase during anoxia in anoxi-tolerant metazoans. Anoxia induces decreases in Na⁺ and K⁺ channel densities, leading to a net reduction in Na⁺/K⁺ ATPase activity, lowering the ATP demand for the membrane ionic integrity maintenance. IF1 inhibits the backward running of the F0F1-ATPase during anoxia (modified from 75).

Finally, to cope with the increased ROS production during the reperfusion period following an oxygen deprivation, anoxia-tolerant animals significantly increase synthesis of specific antioxidant enzymes during the severe hypoxia or anoxia incubation (9, 21).

II.3.4. Postconditioning and preconditioning

Postconditioning is a relatively new concept that consists of a series of mechanical interruptions of reperfusion after ischemia. Such a reperfusion prevents IR injury (114, 115).

Ischemic preconditioning (IP) is defined as an increased tolerance to ischemia and reperfusion induced by a previous sublethal period of ischemia (figure 12 a).

Such a protection has been shown to occur in every species tested so far (116, 117).

Basically, IP induces tolerance both to ischemia itself and to reperfusion. Indeed, IP induces channel arrest (118), maintenance of intra-cellular ion homeostasis and acid-

base balance during prolonged ischemia (119). Moreover, IP gives the cell the ability to function with less ATP during ischemia (120), possibly through the AMPK (121).

Together, these observations suggest that preconditioning adopts strategies that can be compared to those used by metazoans to resist anoxia. Moreover, during reperfusion, preconditioned cells show greater resistance to ROS (122) and an inhibition of MPTP opening (123), resulting in less mitochondrial apoptosis-inducing factors been released compared to non-preconditioned cells (124).

IP offers two windows of protection: an initial strong protective stimulus (early or classical preconditioning) that is brief and a later less powerful but longer lasting protection (late or delayed preconditioning), which may require de novo protein synthesis. There is a transient loss of protection between the two protection windows (figure 12 a and b) (116, 121, 125).

Figure 12: Schematic representation of IP. a. Episodes of non-lethal ischemia (isch) immediately preceding ischemia (early IP) or separated by a 24 h delay before ischemia (late IP). b. Biphasic protection induced by IP. Note that the ischemic protection conferred by IP is lost between the early and late preconditioning (modified from 121).

Importantly, it has been shown that many stressors induce protection against IR (116). For instance, brief period of acute volume loading resulting in myocardial stretch (126) and transient hyperthermia preceding a sustained period of total oxygen deprivation (127-129) have both been shown to limit IR injury. It is thought that similar signaling pathways are shared between classical ischemic preconditioning and

“stress” preconditioning (130). Briefly, adenosine binds to the G-protein-coupled receptor A₁ and A₃, resulting in the phospholipase C (PLC) activation. The hydrolysis of phosphatidyl-4,5-bisphosphate (PIP₂) produces inositol-1,4,5-triphsphate (IP₃) and diacylglycerol (DAG) that activate the protein kinase C (PKC). After its activation, PKC may effect protection by activation of protein tyrosine kinase (PTK), which activates map kinases (MAPK). MAPK consist of three major families (p38MAPK, p46 and p54 c-Jun N-terminal kinases (JNK) and extracellular signal-regulated

kinases (ERK)). Activation of p38MAPK leads to the activation of heat-shock protein 27 (HSP27) that promotes cytoskeleton stability, and thus may protect against reperfusion injury (125). PKC also activates both K_ATP channels that inhibit apoptosis induced by ROS and the ecto-5’-nucleotidase that transforms adenosine mono- phosphate into adenosine, which increases the preconditioning signal (figure 13) (116, 125).

Figure 13: Schematic representation of the signal transduction cascade induced by IP. See main text (modified from 125).

Interestingly, the bioactive ceramide (Cer) has been shown to play a crucial role in IP of rat heart (131) and adding Cer to fibroblasts mimics the effect of IP (132). Importantly, ceramidase (CDase) transforms Cer to the bioactive sphingosine, which is the precursor of another essential mediator of IP: the bioactive lipid sphingosine-1-phosphate (S1P). S1P is known to be a survival factor for a variety of cell types (133-135) and is formed by the phosphorylation of sphingosine catalyzed by the sphingosine kinase (SK) (136). Local S1P protect the heart against IR injury and mediates preconditioning (137). Importantly, it has been shown that IP is able to protect remote cells and organs, which have not been preconditioned by themselves, a phenomenon called remote preconditioning (138, 139). Increasing evidence suggest that high-density lipoproteins (HDL) are a direct agent of the protection conferred by

the remote preconditioning and that S1P is responsible for the beneficial effect of HDL (137). Mechanistically, PKCε is recruited by IP and induces the activation of SK, which phosphorylates the sphingosine to form S1P (140). S1P is then secreted and binds to one of the five S1P receptors (S1PR1 to 5), which is a G protein-coupled receptor (141). Different receptors are differentially expressed in different cells and tissues and coupled to different G proteins, leading to distinct signaling pathways and cellular responses (141, 142) (figure 14).

Figure 14: Schematic representation of the S1P formation following IP. Following IP, the PKCε is activated and phosphorylates SK. Such phosphorylation activates SK, which phosphorylates sphingosine into S1P. S1P is exported outside the cell and binds to one of the five G-coupled receptor S1PR1 to 5 that in turn activates different intracellular pathways depending on the receptor (modified from 142).

II.4. CERAMIDES

Eukaryotic cell membranes possess three main kinds of lipid families: the glycerolipids, the sterols and the sphingolipids (SL) (143). In 1884, J. Thudichum originally named the latter for their enigmatic nature in reference to the sphinx of the mythology. In addition to their structural role in membrane formation, molecules belonging to the SL family have been shown to function as bioactive signaling molecules. Indeed, they are known to have essential roles in membrane microdomains (lipid raft) and to act as both first and second messengers in different signaling pathways controlling, for instance, regulation of cell growth, stress response, apoptosis, differentiation and senescence (144). Ceramide (Cer) is the simplest SL that occupies a central position in sphingolipid biosynthesis/catabolism. It serves as the precursor for all major sphingolipid species in eukaryotes (142). Basically, Cer consists of a sphingoid base to which a fatty acid is attached at C-2 via N-acetylation (figure 15). It serves as a backbone for all complex SL, which are formed by attachment of different head groups at C-1 (145).

Figure 15: “Generic” structural representation of a ceramide. R represents the attachment of any head group at C-1 (from Wikipedia, searching for “ceramide”).

II.4.1. Synthesis of ceramides

Multiple metabolic pathways converge on Cer and so far, we distinguish between the salvage pathway, the recycling of exogenous short-ceramide pathway, the sphyngomyelinase pathway (hydrolysis of the sphyngomyelin) and the main Cer producer, the de novo pathway (figure 16) (144).