Mécanismes D'immunomodulation Par Les

DAMIEN CARIGNAN

MÉCANISMES D’IMMUNOMODULATION PAR LES ALCANOLS ALIPHATIQUES

Thèse présentée à la Faculté des études supérieures et postdoctorales de l’Université Laval dans le cadre du programme de doctorat en biologie cellulaire et moléculaire pour l’obtention du grade de Philosophiae Doctor (Ph.D.)

FACULTÉ DE MÉDECINE UNIVERSITÉ LAVAL QUÉBEC

2013

Résumé

Les alcanols aliphatiques sont des substances ubiquitaires utilisées dans une multitude d’applications domestiques et industrielles. Bien que les effets de l’éthanol sur différentes composantes du système immunitaire soient bien connus, il n’existait presqu’aucune donnée concernant les autres alcanols avant le présent travail. Le premier objectif du projet a été d’étudier les impacts immunologiques d’une exposition aigue aux deux autres alcanols les plus utilisés : l’isopropanol et le méthanol. Nous avons trouvé que l’isopropanol compromet les fonctions effectrices des lymphocytes T, des cellules NK, des monocytes et des macrophages. Par contre, le méthanol agit en synergie avec les stimuli activateurs des lymphocytes T et accroît leur production de cytokines pro-inflammatoires. Les effets biologiques sur ces cellules n’impliquent pas les événements de signalisation précoce en aval des récepteurs activateurs, ils résultent d’une dérégulation sélective de l’activation de facteurs de transcription NFAT avec ou sans la participation d’AP-1. Dans les monocytes activés au LPS, l’isopropanol ne modifie pas les sentiers de signalisation de NF-κB et des MAPK p38 et JNK, mais compromet l’activation de ERK2. Il en résulte une activation défective des sous-unités d’AP-1 c-Fos et JunB.

La deuxième partie du projet a été de vérifier si les n-alcanols (de C1 à C12) avaient des effets immunologiques qui suivent la règle de Meyer-Overton. Cette règle établie une corrélation entre le potentiel anesthésique d’une molécule et son hydrophobicité. Nous avons trouvé que les n-alcanols de C2 à C10 inhibent la sécrétion d’IFN-γ par les lymphocytes T activés d’une manière reliée à leur degré d’hydrophobicité, mais cette corrélation s’interrompt à C11. Les n-alcanols exercent leur effet en aval de la membrane plasmique en altérant progressivement et en fonction de leur taille l’activation du facteur de transcription NFAT; cette tendance s’interrompt aussi à C11. L’activation de la voie NF-κB est altérée par les n-alcanols, mais leur effet s’arrête avant, aux environs de C8. Ces derniers résultats suggèrent l’existence de pochettes d’intéraction de dimensions définies sur des cibles protéiques qui compromettent l’activation de NFAT et NF-κB et altèrent la fonction effectrice des lymphocytes T. L’ensemble de ces travaux contribue à une meilleure compréhension de l’activité biologique des alcanols. ii

Abstract

Aliphatic alkanols are ubiquitous substances used in a variety of household and industrial applications. Although the effects of ethanol on the immune system have been extensively studied, far fewer data is available for the other alkanols. The first objective of the project was to study the immunological impact of acute exposure to the two other most frequently used alkanols, namely isopropanol and methanol. We found that isopropanol is detrimental to the effector functions of T lymphocytes, NK cells, monocytes, and macrophages; they produce less pro-inflammatory cytokines in presence of this alcohol and, in the case of macrophages, phagocytosis is also reduced. Methanol synergizes with activating stimuli to increase their cytokine production. These changes do not involve early signaling events downstream of the cell membrane; they result from the selective dysregulation of the activation of discrete members of the NFAT family of transcription factors with or without the involvement of AP-1. In LPS-activated monocytes, isopropanol does not alter NF-κB and p38/JNK MAPK signaling cascades, but impairs ERK2 activation. The result is a deficient activation of AP-1 subunits c-Fos and JunB. Aliphatic n-alkanols also display anesthetic properties in accordance to their degree of hydrophobicity, following the Meyer- Overton rule. The second part of the project was to determine whether these structurally similar molecules (from C1 to C12) had immunological effects following the same rule. We found that n-alkanols from C2 to C10 inhibit IFN-γ release by activated T lymphocytes in correlation with their hydrophobicity but a cutoff effect was evident at C11. n-Alkanols act downstream of the cell membrane by progressively down-regulating the activation of NFAT in accordance to the size of their aliphatic chain with a clear downward trend that is interrupted at C11. NF-κB signaling is also compromised but the cutoff appears earlier, in the vicinity of C8. Our results suggest the existence of interaction pockets of defined dimensions within intracellular targets that compromise the activation of the NFAT and NF-κB transcription factors and ultimately modulate the effector function of T lymphocytes. Altogether, this work contributes to a better understanding of the biological activity of alkanols.

Avant-Propos

Les chapitres 2 à 5 de cette thèse sont des articles insérés.

Au chapitre 2, on retrouve l’article intitulé : Immunosuppressive effect of isopropanol: down-regulation of cytokine production results from the alteration of discrete transcriptional pathways in activated lymphocytes. Cet article a été publié dans le Journal of Immunology et les auteurs sont: Olivier Désy, Damien Carignan, Manuel Caruso et Pedro O. de Campos-Lima. J’ai contribué aux expérimentations qui ont été effectuées par le premier auteur. J’ai aussi participé à l’élaboration des expériences, à l’analyse des résultats, à la révision et à la correction du manuscript.

Au chapitre 3, l’article: Methanol induces a discrete transcriptional dysregulation that leads to cytokine overproduction in activated lymphocytes a été publié dans le journal Toxicological Sciences. Olivier Désy, Damien Carignan, Manuel Caruso et Pedro O. de Campos-Lima en sont aussi les auteurs. Olivier Désy a réalisé la majeure partie des expériences, mais j’y ai aussi contribué de façon significative en élaborant des protocoles, en réalisant les expériences associées et en analysant les données obtenues. J’ai aussi participé au montage de 2 figures, à la révision et à la correction du manuscript.

Au chapitre 4, The dysregulation of the monocyte/macrophage effector function induced by isopropanol is mediated by the defective activation of distinct members of the AP-1 family of transcription factors, a aussi été publié dans Toxicological Sciences. Les auteurs sont Damien Carignan, Olivier Désy et Pedro O. de Campos-Lima. Je suis co-auteur principal de cet article avec Olivier Désy. Nous avons tous deux contribué pour environ la moitié du travail expérimental contenu dans cet ouvrage. J’ai fait le design de plusieurs expériences, j’ai élaboré de nouveaux protocoles, j’ai analysé les données et monté les figures. J’ai aussi contribué à la rédaction, à la révision et à la correction du manuscript.

Au chapitre 5, le manuscript intitulé The size of the unbranched aliphatic chain determines the immunomodulatory potency of short and long-chain n-alkanols, dont je suis l’auteur principal, n’a pas encore été soumis pour publication. Les co-auteurs sont Olivier Désy, iv

Manuel Caruso et Pedro O. de Campos-Lima. Pour cet article, j’ai fait le design de la plupart des expériences, élaboré la presque totalité des protocoles et fait la majeure partie du travail expérimental. J’ai analysé les données et fait le montage de la plupart des figures. J’ai aussi participé à la rédaction du manuscript et bien sûr à sa révision et à sa correction.

Nous (Olivier Désy, Damien Carignan et Pedro O de Campos-Lima) avons aussi publié un article de revue intitulé Short-term immunological effects of non-ethanolic short-chain alcohols dans le journal Toxicology Letters. Des portions de cet article m’ont servi de point de départ pour la rédaction de segments de mon introduction et de ma discussion. On trouvera une copie du manuscript de cet ouvrage en annexe de cette thèse.

À la science et à l’art qui, ensemble, ont fait de moi ce que je suis devenu. Table des matières

Résumé ...... i Abstract ...... ii Avant-Propos ...... iii Table des matières ...... vi Liste des tableaux ...... ix Liste des figures ...... x Liste des abbréviations ...... xii 1. Introduction ...... 1 Immunotoxicologie de l’Éthanol ...... 4 Effets de l’éthanol sur les cellules de l’immunité innée ...... 5 Effets de l’éthanol sur les cellules de l’immunité adaptative ...... 6 Effets aigus versus effets chroniques ...... 8 Mécanismes moléculaires des effets de l’éthanol sur la fonction immune ...... 9 Mécanismes généraux des effets biologiques des alcools ...... 12 Hypothèses lipidiques ...... 13 Hypothèse protéique ...... 15 Problématique, Hypothèses et Objectifs ...... 22 2. Immunosuppressive effect of isopropanol: down-regulation of cytokine production results from the alteration of discrete transcriptional pathways in activated lymphocytes .. 24 Résumé ...... 25 Abstract ...... 26 Introduction ...... 26 Materials and Methods ...... 27 Results ...... 31 Discussion ...... 36 References ...... 41 Figures and Legends ...... 47 3. Methanol induces a discrete transcriptional dysregulation that leads to cytokine overproduction in activated lymphocytes ...... 56 vii

Résumé ...... 57 Abstract ...... 58 Introduction ...... 58 Materials and Methods ...... 60 Results ...... 64 Discussion ...... 69 References ...... 75 Figures and Legends ...... 82 4. The dysregulation of the monocyte/macrophage effector function induced by isopropanol is mediated by the defective activation of distinct members of the AP-1 family of transcription factors ...... 93 Résumé ...... 94 Abstract ...... 95 Introduction ...... 95 Materials and Methods ...... 97 Results ...... 102 Discussion ...... 107 References ...... 114 Figures and Legends ...... 123 5. The size of the unbranched aliphatic chain determines the immunomodulatory potency of short and long-chain n-alkanols...... 133 Résumé ...... 134 Abstract ...... 135 Introduction ...... 135 Materials and Methods ...... 137 Results ...... 140 Discussion ...... 144 References ...... 149 Figures and Legends ...... 155 6. Discussion ...... 163

viii

Réflexions sur les mécanismes de dérèglement des lymphocytes T par les alcanols ..... 163 Réflexions sur les mécanismes de dérèglement des cellules de l’immunité innée par les alcanols ...... 168 Considérations cliniques et conséquences médicales ...... 172 Conclusion ...... 179 Bibliographie ...... 181 Annexe 1 – Short-term immunological effects of non-ethanolic short-chain alcohols ...... 190

Liste des tableaux

Tableau 1.1 – Production et principaux usages des alcools étudiés ...... 2 Tableau 1.2 – Protéines dont la fonction est altérée par des alcools...... 18 Table 5.1 - Correlation between primary alcohol carbon-chain length and T lymphocyte IFN- release§ ...... 153

Liste des figures

Figure 1.1 – Expositions aiguës aux alcools en Amérique du Nord...... 4 Figure 1.2 - L’éthanol altère l’immunité innée et adaptative...... 9 Figure 1.3 Effets d’une exposition aiguë ou chronique à l’éthanol sur l’activation du TLR4 par le LPS...... 11 Figure 1.4 – Corrélation de Meyer-Overton ...... 13 Figure 1.5 – Concentration de n-alcanols requise pour inhiber de 50% l’activité luciférase...... 16 Figure 1.6 – Caractéristiques de base des pochettes de liaison aux alcools dans les protéines...... 20 Figure 1.7 - Structure moléculaire des composés co-cristallisés en interaction avec des canaux ioniques...... 21 Figure 2.1 - Biological effect of Isopropanol treatment in vitro...... 47 Figure 2.2 - Isopropanol and early TCR signaling ...... 48 Figure 2.3 - Transcriptional effect of isopropanol...... 49 Figure 2.4 - Down-modulation of the effector function of human lymphocytes exposed to isopropanol in vitro...... 51 Figure 2.5 - Immunosuppressive effect of isopropanol in vivo...... 53 Figure 3.1 - Biological effect of methanol treatment on human T lymphocytes in vitro. ... 82 Figure 3.2 - Methanol impact on early TCR signaling and on transcription factor activation...... 84 Figure 3.3 - Differential modulation of NFAT family members by methanol exposure in vitro...... 86 Figure 3.4 - Effect of methanol on proinflammatory cytokine release by human peripheral blood T lymphocytes in vitro...... 89 Figure 3.5 - Immunological impact of methanol in vivo...... 91 Figure 3.6 - Hypothetical model of the methanol transcriptional effect...... 92 Figure 4.1 - Biological effect of isopropanol treatment in vitro on human monocytes. .... 123 Figure 4.2 - Biological effect of isopropanol treatment in vitro on macrophages...... 125 xi

Figure 4.3 - Isopropanol acts downstream of the cell membrane and does not compromise the NF-B signaling pathway...... 127 Figure 4.4 - Isopropanol induces a selective defect in the MAPK signaling cascade and alters the activation of discrete AP-1 family members...... 129 Figure 4.5 - Isopropanol-induced immunosuppression in vivo...... 131 Figure 5.1 - Individual profiles of the immunomodulatory activity of primary alcohols. . 155 Figure 5.2 - Collective profile of the immunomodulatory activity of the homologous series of primary alcohols...... 156 Figure 5.3 - Primary alcohols and TCR early signaling...... 157 Figure 5.4 - Primary alcohols affect NFAT activation...... 159 Figure 5.5 - Impact of primary alcohols on NF-B activation in human primary lymphocytes ...... 161 Figure 6.1 – Modèles d’action hypothétiques des alcanols sur des cellules immunitaires...... 171 Figure 6.2 – L’acide octanoïque inhibe la sécrétion d’IFN-γ par les lymphocytes T...... 177

xii

Liste des abbréviations

AP-1 activator protein-1 ARNm ARN messager CD cluster of differentiation Cellules NK cellules Natural Killer CINC cytokine-induced neutrophil chemoattractant ConA Concanavaline A

EC50 moitié de la concentration maximale effective

ED50 dose effective médiane ERK extracellular signal-regulated kinase

GABAA acide gamme-amino-butyrique de type A GLIC Gloeobacter violaceus pentameric ligand-gated ion channel HGF hepatocyte gowth factor

IC50 moitié de la concentration inhibitrice maximale IFN interferon IL Interleukine IPA Isopropanol IRAK IL-1R-associated kinase IRK1 G-protein-insensitive inwardly rectifying potassium channel JNK c-Jun N-terminal kinase LPS lipopolysaccharide MAPK mitogen activated protein kinase MCP-1 monocyte chemotactic protein-1 MeOH Méthanol MIP-2 macrophage inflammatory protein-2 nAchR récepteur nicotinique de l’acétylcholine NFAT nuclear factor of activated T cells NF-κB nuclear factor-kappa B NMDA N-méthyl-D-aspartate

xiii

PHA phytohémagglutinine PKC protéine kinase C PMA phorbol 12-myristate 13-acétate Poly(I:C) Polyinosinic-polycytidylic acid SEB entérotoxine staphylococcique B TGFβ transforming growth factor β Th T helper TIR Toll/IL-1R TLR Toll-like receptors TNF-α Tumor Necrosis Factor-α

1. Introduction

L’expérience de l’humanité avec des produits contenant de l’alcool peut être retracé au moins jusqu’au 7e millénaire avant Jésus-Christ. Les plus anciennes évidences biochimiques et archéologiques proviennent de l’analyse de fragments de poteries trouvés dans un campement néolithique des plaines centrales de Chine, qui ont révélé des traces d’une concoction fermentée faite de riz, de miel et possiblement de fruits d’aubépine ou de raisins (McGovern et al., 2004). Le développement subséquent du processus de distillation par des alchimistes d’Alexandrie au 1er siècle avant Jésus-Christ, et vraisemblablement longtemps avant par les chinois, a pavé la voie à la caractérisation de l’éthanol et à l’isolation et à la synthèse d’autres alcools (Simmonds, 1919). Avec les années, il est devenu évident que les alcools constituent une large famille de molécules dont les applications s’étendent bien au-delà des premiers usages rituels, récréatifs et médicinaux associés aux boissons contenant de l’éthanol. Ainsi, Robert Boyle a été le premier, au dix-septième siècle, à montrer que la distillation du bois donnait un alcool – le méthanol – qu’il a appellé esprit anonyme, esprit de bois inflammable ou esprit adiaphorétique (adiaphorous spirit) (Hoefer, 1869). De nos jours, le méthanol est surtout produit de manière synthétique et utilisé comme matière première pour la production d’autres produits chimiques; il est aussi aisément disponible pour les consommateurs en tant que constituant de produits familiers tels que des vernis, peintures, liquides lave-glace, antigels et adhésifs (Jammalamadaka and Raissi, 2010). De manière similaire, l’isopropanol - un alcool secondaire - a trouvé sa place dans de nombreux processus industriels depuis sa synthèse au milieu du dix-neuvième siècle par Marcellin Berthelot et Charles Friedel, indépendamment l’un de l’autre. Maintenant la population en général a facilement accès à l’isopropanol sous forme d’alcool à friction et comme ingrédient de gels désinfectants pour les mains et d’autres produits domestiques d’usage courant (Kraut and Kurtz, 2008; Jammalamadaka and Raissi, 2010). Les alcools aliphatiques primaires, comprenant de trois à douze atomes de carbone, ont des usages moins communs que ceux du méthanol, de l’éthanol et de l’isopropanol. Toutefois, plusieurs d’entre eux sont utilisés comme solvants pour des applications industrielles tandis que d’autres sont employés pour la fabrication de parfum et de produits cosmétiques. On 2 retrouve un résumé des ces usages de même que la production annuelle de ces alcools dans le tableau 1.1.

Tableau 1.1 – Production et principaux usages des alcools étudiés Alcanol Production Usages Références Methanol Environ 20 millions de tonnes Production industrielle d’autres (EHC 196, 1997) dans le monde en 1991 composés chimiques, composant de plusieurs produits disponibles dans le commerce. Isopropanol 1 800 000 tonnes dans le Solvant répandu, composant de (EHC 103, 1990) monde en 1995 produits domestiques et d’usage (OECD-SIDS, 2002) personnel, manufacture d’autres produits chimiques. Propanol Plus de 130 000 tonnes dans le Solvant multi-usages industriel et (EHC 102, 1990) monde en 1979. domestique. Intermédiaire dans la manufacture de composés chimiques. Butanol 784 000 tonnes aux États-Unis Intermédiaire industriel, Solvant, (OECD-SIDS, 2001) en 1997. manufacture de produits pharmaceutiques, parfums, agent aromatisant pour aliments et boissons. Pentanol 35 000 tonnes dans le monde Industrie pharmaceutique, (OECD-SIDS, 2006) en 1997 (alcool amylique parfums, arômes, solvants, (Nelson et al., 1989) primaire) fabrication d’explosifs, chimie de synthèse, additif alimentaire. Hexanol 1 580 000 tonnes dans le Solvant, intermédiaire industriel. (OECD-SIDS, 2006) monde § (Nelson et al., 1989) Heptanol Non-disponible Expériences d’électrophysiologie (Takens-Kwak et al., cardiaque, cosmétiques 1992) Octanol 1 580 000 tonnes dans le Parfumerie et cosmétiques, (Opdyke, 1973) monde § traitement expérimental pour le (Bushara et al., 2004) tremblement essentiel, fabrication (OECD-SIDS, 2006) de plastifiants. (Nelson et al., 1990) Nonanol Non-disponible Cosmétiques (Opdyke, 1973) Fabrication de plastifiants. (Nelson et al., 1990) Décanol 1 580 000 tonnes dans le Cosmétiques (Opdyke, 1973) monde § Fabrication de plastifiants. (OECD-SIDS, 2006) (Nelson et al., 1990) Undécanol 1 580 000 tonnes dans le Cosmétiques, additif alimentaire. (OECD-SIDS, 2006) monde § (Opdyke, 1973) Burdock GA (1997) Dodécanol 60 000 tonnes en Europe en intermédiaire pour la synthèse de (Opdyke, 1973) 1993 composés chimiques, ingrédients (OECD-SIDS, 2002) de produits industriels, produits nettoyants, produits cosmétiques. § 1 580 000 tonnes pour un esemble de 15 alcools à longue chaîne comprenant l’hexanol, l’octanol, le décanol et le undécanol. L’utilisation des alcools est désormais très répandue dans les sphères sociale, professionnelle et domestique de la vie moderne. L’accès facile à ces substances et leur

3 présence ubiquitaire dans une multitude de produits créent beaucoup d’opportunités d’empoisonnement. Afin de mieux cerner l’amplitude du problème, il est intéressant de s’attarder au nombre de cas d’expositions aigues rapportés chaque année à des centres antipoison en Amérique du Nord, soit près de 120 000 (Bronstein et al., 2010) (Figure 1.1). Sans surprise, l’éthanol représente la première cause d’empoisonnement avec 85 000 cas d’expositions aigues dont 34 000 non-liés à la consommation de boissons alcoolisées. Il est suivi de l’isopropanol avec un peu moins de 21 000 cas répartis en trois sous-groupes : les alcools à friction, les produits nettoyants et les produits non à friction/non-nettoyant. Les glycols comptent près de 10 000 cas, la plupart liés à des produits d’entretien automobile. En ce qui concerne le méthanol, l’exposition à cette substance à partir de sources alimentaires ou environnementales ne cause généralement pas de problème. Cependant, le méthanol demeure tout de même une importante cause d’intoxication aux alcools et les taux de mortalité qui y sont associé peuvent atteindre 15-36% dans les cas les plus sévères (Liu et al., 1998; Brent et al., 2001). En terme numérique, les centres antipoison reçoivent un peu plus de 2000 rapports de cas d’intoxication aiguë au méthanol par année. L’exposition à d’autres alcools compte environ 1200 cas annuellement.

L’exposition systémique aux alcools produit plusieurs effets biologiques, notamment des dérèglements neurologiques et comportementaux (Kraut and Kurtz, 2008; Jammalamadaka and Raissi, 2010). Il n’est donc pas surprenant que la longue histoire de l’utilisation et de l’abus de l’éthanol dans notre société ait stimulé un intérêt considérable pour sa toxicologie de même que pour son immunotoxicité (OECD-SIDS, 2004). Nous examinerons maintenant les conséquences immunologiques d’une intoxication aiguë à l’éthanol.

Figure 1.1 – Expositions aiguës aux alcools en Amérique du Nord.

Les nombres entre parenthèses représentent le nombre de cas rapportés annuellement à l’association américaine des centres antipoison. Les mélanges de glycols et de méthanol (188 cas) ont été inclus dans la tranche du méthanol.

Immunotoxicologie de l’Éthanol

Une bonne portion de la littérature consacrée aux effets immunologiques de l’éthanol porte sur les effets de cette substance sur la réponse inflammatoire. L’inflammation est une caractéristique de l’immunité innée. Elle est la réponse du corps à une agression physique ou pathogénique, comme un traumatisme ou une blessure, et constitue la première ligne de défense contre les microorganismes pathogènes (Goral et al., 2008). De plus, les processus inflammatoires sont essentiels à l’induction d’une réponse immunitaire adaptative à l’infection. Il est maintenant bien établi qu’une exposition aiguë à l’éthanol chez l’homme et dans des modèles animaux réduit considérablement la chimiotaxie et l’inflammation en diminuant la production de chimiokines et de cytokines pro-inflammatoires. Compte tenu de l’importance de ces mécanismes pour l’immunité innée, ces changements sont donc

5 associés à une plus grande susceptibilité aux infections (Pruett et al., 2004). Les prochaines sections décrivent les effets de l’éthanol sur différentes composantes des branches innée et adaptative du système immunitaire.

Effets de l’éthanol sur les cellules de l’immunité innée

Les cellules de l’immunité innée telles que les leucocytes polynucléaires, les monocytes/macrophages et les cellules natural killer (NK) sont équipées de différents récepteurs capables de reconnaître des motifs moléculaires associées aux pathogènes et de répondre rapidement aux signaux transmis par ceux-ci afin de combattre l’infection. Les récepteurs de ce type les mieux connus sont les Toll-like receptors (TLR) qui reconnaissent une variété de composantes microbiennes et amorcent la signalisation menant à l’activation des fonctions effectrices des cellules de l’immunité innée, comme la phagocytose ou la sécrétion de cytokines et de chimiokines. Les effets de l’éthanol sur l’activation de cellules de l’immunité innée par des ligands de différents TLR ont été largement étudiés et seront brièvement passées en revue dans les paragraphes suivants. De nombreuses études in vivo et in vitro ont montrés qu’un traitement aigu à l’éthanol est capable d’inhiber la production de cytokines pro-inflammatoires en réponse à ces composantes. Par exemple, l’éthanol réduit la sécrétion de tumor necrosis factor-α (TNF-α) et d’interleukine-1β (IL-1β) par des macrophages alvéolaires de souris et des monocytes de sang périphérique humain en réponse au lipopolysaccharide (LPS), un ligand du TLR4 (Nelson et al., 1989), (Verma et al., 1993). Il diminue aussi la production d’IL-6 et d’IL-12 induite par le polyinosinic :polycytidylic acid (poly I:C), un ligand du TLR3, par des macrophages péritonéaux de souris (Pruett et al., 2004). Enfin, des macrophages spléniques murins produisent moins de TNF-α et d’IL-6 lorsque traités à l’éthanol et stimulés au LPS, au peptidoglycane (TLR2) ou à l’ADN microbien (CpG) non-méthylé (TLR9) (Goral and Kovacs, 2005). De plus, l’éthanol réduit aussi la production des chimiokines monocyte chemotactic protein-1 (MCP-1) et IL-8 par les monocytes humains en réponse à une stimulation à

6 l’entérotoxine B staphylococcique (SEB), au phytohémagglutinine (PHA) ou à l’interféron (IFN)-γ (Szabo et al., 1999). Dans un modèle d’infection pulmonaire, l’éthanol inhibe également les chimiokines macrophage inflammatory protein-2 (MIP-2) et cytokine- induced neutrophil chemoattractant (CINC), ce qui, par conséquent, réduit l’influx de neutrophiles (Boe et al., 2003). Les neutrophiles eux-mêmes sont d’ailleurs affectés par une exposition aiguë à l’éthanol : ils sécrètent moins des cytokines IL-8 et TNF-α et la dégranulation de hepatocyte growth factor (HGF) et d’élastase, des molécules responsables du remodelage des tissus environnants, est réduite (Taieb et al., 2002).

Les cellules endothéliales, qui tapissent l’intérieur des vaisseaux sanguins, sont aussi très importantes lors d’une réponse immunitaire. Une fois activées par des médiateurs pro- inflammatoires, elles permettent l’adhésion et le passage de cellules immunitaires vers le site inflammatoire ou infectieux. Elles sécrètent aussi d’avantage de cytokines et de chimiokines. Une intoxication aiguë à l’éthanol diminue le recrutement de leucocytes, la production de chimiokines et l’expression de molécules d’adhésion par les cellules endothéliales in vivo et in vitro (Saeed et al., 2004).

Les effets d’une exposition aiguë à l’éthanol sur les cellules NK sont aussi délétères à leur activité. Dans un modèle de propagation de métastases sensible aux cellules NK chez le rat, une exposition aiguë à l’éthanol inhibe l’activité de ces cellules et accroît le nombre de métastases (Ben-Eliyahu et al., 1996). De plus, dans un modèle murin de consommation aiguë d’alcool, l’activité cytolytique des cellules NK stimulées par le poly (I :C) est réduite (Hebert and Pruett, 2002).

Effets de l’éthanol sur les cellules de l’immunité adaptative

Tel que mentionnée précédemment, l’inflammation contribue au déclenchement de la réponse immunitaire adaptative contre un pathogène donné. Vu les effets délétères

7 susmentionnés de l’éthanol sur la réponse inflammatoire, il est donc cohérent que des composantes importantes de l’immunité acquise soient aussi affectés par une exposition à cet alcool.

Ainsi, il a été montré que l’éthanol réduit la prolifération de lymphocytes T spécifiques à un antigène en diminuant la capacité de présentation de cet antigène par les monocytes. Cet effet est causé par la diminution de production d’IL-1β et l’augmentation de transforming growth factor β (TGFβ) et d’IL-10 par les monocytes, les précurseurs des cellules dendritiques (Szabo et al., 1993; Mandrekar et al., 1996).

Les cellules dendritiques myéloïdes comptent parmi les plus puissantes présentatrices d’antigènes. Elles expriment une variété de molécules d’adhésion et de co-stimulation et produisent des cytokines régulatrices ce qui leur permet d’activer efficacement les lymphocytes T naïfs. Or, des cellules dendritiques générées en présence d’éthanol montrent un phénotype inhibiteur. Elles sont déficientes dans les fonctions associées à la présentation de l’antigène (capture, apprêtement et présentation) aux lymphocytes T. Elles ont aussi une expression réduite de molécules de co-stimulation en plus de produire moins d’IL-12 et d’avantage d’IL-10. Ces cellules dendritiques induisent l’anergie des lymphocytes T et inhibent la réponse T helper (Th)1 (Mandrekar et al., 2004).

Les effets observés d’une exposition aiguë à l’éthanol sur la fonction des lymphocytes T eux-mêmes sont contradictoires. Dans un modèle murin d’intoxication aigue, la prolifération de splénocytes en réponse à la concanavaline A (ConA) demeure à peu près inchangée (Kawakami et al., 1991). Une autre étude montre que l’éthanol augmente la prolifération de cellules de sang périphérique mononuclées (PBMC) stimulées in vitro par le phorbol 12-myristate 13-acétate (PMA) ou la ConA (Szabo et al., 1993). D’autres ont plutôt observé une réduction de la production d’IL-2 et de la prolifération de PBMC traitées à la ConA en présence d’éthanol (Chiappelli et al., 1995). Plus récemment, un autre groupe a montré une réduction de la sécrétion et de la quantité d’ARN messager de l’IL-2 dans des

8 cellules Jurkat et des lymphocytes T CD4 primaires stimulés au PHA ou avec des anticorps contre CD3 et CD28 en présence d’éthanol (Ghare et al., 2011).

Effets aigus versus effets chroniques

Jusqu’ici, les effets décrits de l’éthanol sur la fonction immune ont été ceux causé par une exposition aiguë à cette substance. Or, l’exposition chronique à l’éthanol est fréquente chez les individus alcooliques. Il est bien connu maintenant que ces deux types d’exposition à l’éthanol ont des effets opposés sur la fonction immune : une exposition aiguë est inhibitrice alors que l’exposition chronique augmente la réponse inflammatoire (Goral et al., 2008; Szabo and Mandrekar, 2009). Par exemple, une intoxication chronique à l’éthanol accroît la production de TNF-α par les cellules de Kupffer – des macrophages du foie – en réponse à une stimulation au LPS (Nagy, 2003). La figure 1.2 présentée plus loin résume les effets modulateurs de l’éthanol sur la fonction de différentes cellules immunitaires et leur habileté à répondre à un pathogène. La production de cytokines pro-inflammatoires (comme l’IL-1 et le TNF-α) est inhibé par l’éthanol aigu alors que l’alcool chronique l’augmente. La fonction de présentation de l’antigène des monocytes et des cellules dendritiques est altérée par les deux formes d’exposition à l’éthanol ce qui contribue à l’activation déficiente des lymphocytes T. La production d’IL-12 est aussi réduite. Il en résulte des changements dans la balance entre les cytokines de type Th1 (IFNγ) et Th2 (IL-10). Collectivement, ces changements induits par l’éthanol nuisent à l’élimination des pathogènes et à l’induction d’une réponse adaptative à l’infection chez l’individu exposé à l’alcool (Szabo and Mandrekar, 2009).

Figure 1.2 - L’éthanol altère l’immunité innée et adaptative.

Les expositions chroniques et aiguës à l’éthanol ont de larges effets immunorégulateurs. Adapté de (Szabo and Mandrekar, 2009, Alcoholism Clinical & Experimental Research)

Mécanismes moléculaires des effets de l’éthanol sur la fonction immune

La plupart des mécanismes moléculaires connus expliquant les effets immunomodulateurs de l’éthanol sont liés à l’activation des cellules immunitaires par les TLR (Pruett et al., 2004; Goral et al., 2008). Ces TLR possèdent un court domaine intracellulaire Toll/IL-1R (TIR) qui, suite à la liaison d’un ligand, recrute des molécules adaptatrices et induit les événements de signalisation en aval. Comme plusieurs autres TLR, TLR4 interagit avec la molécule adaptatrice MyD88 qui à son tour recrute IL-1R-associated kinase (IRAK) 4. IRAK4 phosphoryle et active IRAK1 qui va s’associer au TNF receptor-associated factor 6 (TRAF6). L’association entre IRAK1 et TRAF6 est régulée négativement par la protéine IRAK-monocyte (IRAK-M) (Kobayashi et al., 2002). Les cascades de signalisation intracellulaire induite ensuite par le complexe IRAK1/TRAF6 comprennent l’activation du

10 facteur de transcription activator protein-1 (AP-1) via les mitogen-activated protein kinases (MAPK) et l’activation du facteur de transcription nuclear factor-κB (NF-κB) (Takeda and Akira, 2005). Ces deux voies de signalisation sont altérées par une exposition aiguë à l’éthanol. De fait, cet alcool réduit la phosphorylation de p38 et l’activation de NF-κB induites par le LPS dans des leucocytes humains (Arbabi et al., 1999; Mandrekar et al., 1999). L’activation de p38, de NF-κB et de la sous-unité c-Jun du facteur de transcription AP-1 sont inhibés par l’alcool dans des macrophages péritonéaux de souris activés via TLR3 par le poly (I:C) (Pruett et al., 2004). De plus, l’éthanol réduit la phosphorylation de p38 et de l’extracellular-signal regulated kinase (ERK)-1/2 dans des macrophages murins stimulés par des ligands pour les TLR2, 4 et 9 (Goral and Kovacs, 2005). Une étape importante de la signalisation précoce induite par les TLR est l’activation de la protéine IRAK1 et son hyperphosphorylation qui, ultimement, mène à sa dégradation par le protéosome. Il a été montré que l’activité kinase d’IRAK1 est réduite en présence d’éthanol dans des monocytes humains stimulés via TLR4 par le LPS. Par contre, cet effet est absent lorsque les cellules sont stimulées via TLR2 (Oak et al., 2006). Dans une lignée de macrophages murins, la dégradation d’IRAK1 suite à l’activation des cellules par le LPS est diminuée en présence d’éthanol (Dai et al., 2005). Une analyse de l’expression des gènes modifiés par une exposition aiguë à l’éthanol par micropuce à ADN a permis d’identifier, chez des macrophages péritonéaux de souris stimulés au poly(I :C), la suppression d’une boucle d’amplification lié à l’interféron (Pruett et al., 2004). Non seulement l’ARNm de l’IFN et la protéine IFNα sont diminués, mais on note aussi que l’expression de molécules responsables de la signalisation produite par l’IFN et de molécules directement induites par l’IFN est réduite. Tel que mentionné précédemment, une exposition chronique à l’éthanol accroît la production de cytokines pro-inflammatoires. Il a été montré que l’augmentation de la production de TNF-α par les cellules de Kupffer en réponse au LPS était associée à un accroissement de l’activité de p38 et ERK1/2 (Kishore et al., 2001; Kishore et al., 2002). L’ensemble de ces mécanismes est résumé dans la figure 1.3. Il a aussi été proposé que l’éthanol puisse agir au niveau le plus précoce des événements de signalisation en altérant la structure et/ou la fonction des radeaux lipidiques. De fait, suite à

11 la stimulation avec des ligands spécifiques, des TLR membranaires, comme TLR2 et TLR4, sont recrutés dans des radeaux lipidiques. CD14, un composant du complexe du récepteur TLR4, est aussi redistribué dans les microdomaines membranaires suite à la liaison du LPS (Szabo et al., 2007). Des études indiquent qu’une exposition aigüe à l’éthanol altère la redistribution induite par le LPS aux radeaux lipidiques de TLR4 et CD14. Ces changements sont associés à une diminution de la sécrétion de TNF-α et à une réduction de l’activation du facteur de transcription NF-κB. De plus, le traitement de cellules de type macrophages avec des drogues qui perturbent les radeaux lipidiques a des effets similaires (Dai et al., 2005; Dolganiuc et al., 2006).

Éthanol aigu Éthanol chronique

Cytoplasme

Noyau Facteurs de transcription

Figure 1.3 Effets d’une exposition aiguë ou chronique à l’éthanol sur l’activation du TLR4 par le LPS.

Adapté de (Goral et al., 2008, Alcohol)

Mécanismes généraux des effets biologiques des alcools

Les événements moléculaires qui sous-tendent les effets biologiques de l'alcool sont complexes, mais une image plus complète commence à émerger, du moins en partie, en empruntant certains concepts clés d'études réalisées sur des composés anesthésiques

(Eckenhoff, 2001; Streiff et al., 2006; Vedula et al., 2009). Notre compréhension de la façon dont fonctionnent les alcools (et d’autres anesthésiques plus efficaces et/ou moins toxiques) a été porté principalement sur la membrane cellulaire au cours de la majeure partie du XXe siècle (Seeman et al., 1971; Roth and Seeman, 1972; Pringle et al., 1981). Le premier aperçu du mécanisme d'action présumé des alcools est venu de l'observation qu'ils suivent la règle de Meyer-Overton, tout comme d'autres anesthésiques généraux. Cette règle établie une corrélation entre le pouvoir anesthésiant d’un composé donné et son coefficient de partition huile : eau; ainsi, les molécules dotées de chaînes carbonées plus longues ont tendance à être plus puissantes comparativement à celles qui sont moins hydrophobes (Figure 1.4). L'opinion prédominante était que la dissolution des molécules lipophiles dans la bicouche lipidique allait changer ses propriétés physiques compromettant ainsi indirectement la fonction des protéines intégrées (Seeman, 1972; Eckenhoff, 2001). Il existe cependant des exceptions à cette corrélation, la plus notable étant la coupure de l’effet anesthésique (cut-off) après l’atteinte d’une taille moléculaire précise pour une série de molécules homologues. Par exemple, pour la série homologue des n-alcools aliphatiques saturés, le pouvoir anesthésiant augmente progressivement avec la longueur de la chaîne carbonée et atteint un maximum à 12 carbones; après ce point les alcools successifs sont dépourvus d’activité anesthésique malgré qu’ils soient plus hydrophobes que les précédents (Alifimoff et al., 1989).

Figure 1.4 – Corrélation de Meyer-Overton

Relation entre la puissance de différents anesthésiques (montré comme la réciproque de leur EC50 molaire pour l’anesthésie) et leur coefficient de partition huile :eau. Tiré de (Franks, 2006, British Journal of Pharmacology)

Hypothèses lipidiques

Plusieurs hypothèses ont été avancées pour expliquer les mécanismes responsables des perturbations de la bicouche lipidique qui pourraient induire l’anesthésie dont l’expansion du volume membranaire, le désordonnement de la membrane cellulaire et le changement dans les transitions de phase lipidique (Peoples et al., 1996). Selon le modèle de l’expansion du volume membranaire, les alcools ou autres anesthétiques produisent leurs effets comportementaux en occupant un espace précis dans la membrane des neurones, ce qui aurait pour effet d’accroître le volume membranaire. Cette expansion de volume serait attribuable aux effets des molécules d’alcools ou d’anesthétiques sur les lipides et possiblement aussi les protéines membranaires. En conséquence, il résulterait une augmentation des forces latérales dans la membrane qui altérerait le fonctionnement des

14 protéines membranaires intégrales (Seeman, 1972). Toutefois, les changements de volume membranaire induits par des concentrations pertinentes d’alcools (qui induisent l’anesthésie) sont extrêmement petits et peuvent être reproduits par des élévations de température de seulement quelques degrés Celsius. Or, de tels changements de température n’induisent pas d’anesthésie ou de signes d’effets comportementaux d’une intoxication à l’alcool (Franks and Lieb, 1982). Le modèle des transitions de phases lipidiques s’attarde essentiellement au changement de point de fusion des lipides membranaires. Il a été proposé que les alcools et les anesthésiques pourrait altérer l’équilibre entre les phases ordonné et désordonné des lipides membranaires (Lee, 1976; Trudell, 1977). Ainsi, les lipides entourant une protéine telle qu’un canal ionique pourrait devoir être dans un état moins ordonné et plus compressible permettant au canal de s’ouvrir. Donc selon cette hypothèse, les alcools et les anesthésiques affecteraient la fonction des canaux ioniques en réduisant la température à laquelle les lipides entrent dans la phase ordonnée. Cependant, des concentrations cliniques d’anesthésiques diminuent cette température de seulement 1-2°C : une perturbation analogue à celle provoquée par une augmentation de la température corporelle de même magnitude (Eckenhoff, 2001). Selon le modèle du désordre ou de la fluidité membranaire, l’hypothèse fut émise que le désordonnement de l’essentiel de la phase lipidique de la membrane cellulaire par les alcools pourrait perturber la fonction normale de protéines intégrées, telles que les canaux ioniques (Goldstein, 1984). De fait, il a été démontré que la dissolution de n-alcools anesthétiques (octanol, décanol, dodécanol) dans des membranes synaptiques accroît leur désordre tandis que la dissolution de n-alcools plus longs et non-anesthétiques n’a pas d’effet (tetradécanol) ou a même l’effet inverse (hexadécanol, octadécanol). Cela permettrait donc d’expliquer la coupure (cut-off) observée de l’effet anesthésique de la série homologue des n-alcools après le dodécanol (Miller et al., 1989). Cependant, plusieurs réserves existent quant à la validité de cette hypothèse. D’abord, les effets des alcools sur le désordonnement de la membrane cellulaire ne sont généralement mesurables qu’à des concentrations nettement plus élevées que celles de la gamme pharmacologique. Il serait donc difficile d’envisager que des concentrations pertinentes d’un point de vue

15 pharmacologique puissent avoir un effet significatif sur ce désordonnement. D’ailleurs, à des concentrations associées à l’anesthésie, on estime qu’il n’y aurait qu’une seule molécule d’alcool pour 200 molécules de lipides dans la membrane plasmique (Peoples et al., 1996). Ensuite, comme pour l’expansion de la membrane et les transitions de phase lipidique, les changements de désordre membranaire induits par les alcools ou les anesthésiques peuvent être imités par une légère augmentation de température (Pang et al., 1980). Puisqu’une fièvre légère n’induit pas la perte de sensibilité, il s’avère improbable que les hypothèses lipidiques fournissent une explication unitaire de l’anesthésie.

Hypothèse protéique

Une autre interprétation n'impliquant pas la modification de la bicouche lipidique, mais qui envisage plutôt des protéines en tant que cibles directes des alcools, reçu un soutien considérable après qu’il fut montré que des alcools primaires lient la luciférase in vitro en absence d’un contexte lipidique (Franks and Lieb, 1984; Franks and Lieb, 1985). Les alcools compétitionnent avec le substrat luciférine et inhibent l’activité enzymatique en fonction de la taille de leur chaîne carbonée et à des concentrations très similaires à celles induisant l’anesthésie dans des modèles animaux. Puisque les alcools plus longs tel que le n-décanol sont des inhibiteurs de la luciférase plus puissants de plusieurs ordres de grandeurs que les plus petits et moins hydrophobes n-propanol ou éthanol, la règle de Meyer-Overton peut être extrapolée au-delà de la dissolution dans la bicouche lipidique au niveau de sous-domaines protéiques hydrophobes. D’autant plus que les n-alcools lient la luciférase de mieux en mieux à mesure de la croissance de leur chaîne carbonée jusqu’au n- hexanol et au n-heptanol, qui ont une affinité très similaire. La corrélation entre la longueur de la chaîne de carbone et la liaison reprend avec le n-octanol et se poursuit jusqu’au n- dodécanol, après quoi elle disparaît à nouveau (figure 1.5). Ces découvertes suggèrent l’existence d’une pochette de liaison hydrophobe qui aurait la capacité d’accommoder deux molécules de n-hexanol ou une seule de n-dodécanol tandis que les alcools plus longs resteraient partiellement exposés au solvant (Franks and Lieb, 2004). Ces observations fournissent une alternative valable aux hypothèses lipidiques en plus d’une explication

16 simple au respect de la règle de Meyer-Overton. D’ailleurs des sites de liaisons aux anesthésiques qui suivent aussi cette corrélation ont été découvert sur d’autres enzymes : la cytochrome c oxydase (Hasinoff and Davey, 1989) et la protéine kinase C (Slater et al., 1993). Le cut-off de l’effet anesthésique des n-alcools peut aussi être expliqué simplement par la liaison de ces molécules à des pochettes ou des crevasses de dimensions définis sur des cibles protéiques.

Figure 1.5 – Concentration de n-alcanols requise pour inhiber de 50% l’activité luciférase.

ED50 pour l’inhibition de la luciférase en fonction du nombre d’atomes de carbones de la molécule. La ligne csat représente la solubilité maximale des différents alcanols en solution aqueuse. Tiré de (Franks and Lieb, 1985, Nature)

Des travaux ultérieurs impliquant la mutagenèse d’acides aminés spécifiques ont contribué à définir les sites putatifs de liaison aux alcools dans diverses protéines (tableau 1.2). Par exemple, pour l’enzyme adénylate cyclase, dont l’activité est augmentée par l’éthanol, il a été montré que deux régions définies de la protéine sont responsables de cet effet : les 28 premiers acides aminés de la région N-terminale du domaine C1a et les 140 acides aminés

17 de la région C-terminale de la molécule (Yoshimura et al., 2006). En ce qui concerne le récepteur nicotinique de l’acétylcholine (nAchR), des mutations de résidus spécifiques dans le domaine M2 qui forme le pore du récepteur change la sensibilité de la protéine aux anesthésiques (isoflurane, n-hexanol et n-octanol). Plus les acides aminés introduits sont hydrophobes, plus le nAchR est inhibé par ces substances (Forman et al., 1995). Dans le même ordre d’idée, la sensibilité à l’éthanol des récepteurs au glutamate N-méthyl-D- aspartate (NMDA), dont la fonction est inhibée par l’alcool, dépend de résidus précis situés dans les domaines tranmembranaires (TM) 3 et TM4 de la sous-unité NR1. La mutation de ces acides aminés altère la sensibilité à l’éthanol des récepteurs NMDA (Smothers and Woodward, 2006). La plupart des protéines présentées dans le tableau 1.2 sont impliquées dans des processus neurobiologiques et n’ont pas un impact immunologique direct; cependant, il est intéressant de constater qu’il existe peut-être une interrelation entre les événements moléculaires causés par l’alcool qui produisent des effets neurobiologiques et ceux qui mènent à des effets immunologiques. À cet égard, il a été rapporté récemment que l’expression du TLR4 est régulée par le récepteur à l’acide γ-aminobutyrique de type A (GABAA) dans le noyau central de l’amygdale et que ce procédé est associé à la consommation excessive d’alcool (taux d'alcoolémie ≥ 0,08 g% dans une période de 2 h) (Liu et al., 2011). Le récepteur

GABAA possède une cavité de liaison à l'alcool formée par des acides aminés provenant de quatre domaines transmembranaires et a longtemps été considérée comme une cible pour les effets de l'alcool dans le système nerveux central (Harris et al., 2008).

Tableau 1.2 – Protéines dont la fonction est altérée par des alcools.

Cibles définies par des analyses biochimiques Molécules co-cristallisées Référence et de mutagénèse Récepteurs canaux pentamériques (pLGIC)

Récepteur GABAA (Jung et al., 2005) Récepteur glycine (Lobo et al., 2008) Récepteur nicotinique de l’acétylcholine (Forman et al., 1995) Récepteur au glutamate N-méthyl-D-aspartate (Smothers and Woodward, 2006) Homologue bactérien de pLGIC (GLIC) Propofol, desflurane (Nury et al., 2011) Canaux potassiques Shaw2 (Shahidullah et al., 2003) G protein inwardly rectifying potassium channel 2 (Aryal et al., 2009) IRK1 2-méthyl-2,4-pentanediol (Pegan et al., 2006) Molécules d’adhésion L1 (Dou et al., 2011) Protéine de liaison d’odorants LUSH Éthanol, 1-propanol, (Kruse et al., 2003) 1-butanol Enzymes Luciférase Bromoforme (Franks et al., 1998) Alcool déshydrogénase Alcool pentafluorobenzylique, (Ramaswamy et al., trifluoroéthanol 1994) Adénylate cyclase (Yoshimura et al., 2006)

Plus important encore, les données de mutagenèse ont consolidé les résultats obtenus avec le modèle luciférase à établir le principe que les alcools peuvent interagir directement avec des protéines qui appartiennent à des groupes fonctionnels très différents à condition qu'elles affichent des pochettes de liaison appropriées. L'anatomie de ces poches de liaison peut être déduite des caractéristiques les plus fondamentales de la molécule d'alcool soit l'existence d'un groupe hydroxyle lié à un atome

19 de carbone. Le groupe hydroxyle permet aux alcools de se comporter comme des acides faibles et des donneurs de liaisons hydrogène, tandis que la chaîne aliphatique est responsable de leurs propriétés hydrophobes (Ballinger, 1960; Dwyer and Bradley, 2000). Il a été suggéré que les alcools lieraient un site miroir dans les protéines cibles (Dwyer and Bradley, 2000). Selon ce point de vue, on pourrait s'attendre à trouver dans un voisinage proche: (i) un site accepteur de liaison hydrogène, (ii) une charge nette positive localisée pour interagir avec l'atome électronégatif du groupe hydroxyle, et (iii) un sillon hydrophobe qui pourraient résulter de l’empilement d’hélices-α. Ainsi, les alcools déplaceraient les molécules d’eau de ces pochettes et établiraient des interactions atomiques avec les protéines via des liaisons hydrogènes qui seraient stabilisées par des forces de van der Waals au sein du sillon hydrophobe (Klemm, 1998). La déformation locale causée par cette interaction moléculaire pourrait mener à une altération de la fonction des protéines (figure 1.6). Par exemple, le canal potassique neuronal de la drosophile Shaw 2 est inhibé par des concentrations physiologiques de n-alcanols. Ceux-ci se lient probablement à une région formée de 13 acides aminés situés dans la sous-unité formant le pore de la protéine et stabilisent l’état fermé du canal ionique. Des interactions hydrophobes et des forces polaires faibles déterminent la puissance de liaison des alcools à ce site circonscrit. D’ailleurs, la structure α-hélicoïdale de cette région semble critique pour la liaison des alcools et/ou pour l’inhibition allostérique de l’ouverture du canal par les alcanols (Shahidullah et al., 2003). Les caractéristiques structurelles prédites ci-dessus ont été largement confirmées par des données cristallographiques à propos de complexes protéines-alcools (tableau 1.2). La protéine de liaison d’odorants de la drosophile LUSH sert de modèle pour les sites de liaisons aux alcools. LUSH est une protéine non-enzymatique requise pour les réponses comportementales et électrophysiologiques des neurones olfactifs de la drosophile aux n- alcools. La résolution de la structure cristalline de LUSH en présence de 30-50 mM de différents n-alcools (éthanol, propanol et butanol) a révélé la présence d’un seul site de liaison aux alcools situé dans une cavité hydrophobe d’environ 390 Å3. Cette cavité est remplie de molécules d’eau et se situe entre un ensemble d’hélices-α (Kruse et al., 2003). En plus de plusieurs résidus hydrophobes, ce site de liaison présente un jeu de résidus

20 polaires qui forment un réseau de ponts hydrogènes avec les alcools et l’eau. Ainsi, le groupe hydroxyle des alcools établis un lien hydrogène fort avec la thréonine 57 (T57) et un plus faible avec la sérine 52 (S52). L’arrangement optimal aurait T57 donnant un lien hydrogène à l’alcool, qui, à son tour, en donne un à S52, tandis que S52 en donne un au groupe carbonyle de la chaîne principale (Thode et al., 2008).

Figure 1.6 – Caractéristiques de base des pochettes de liaison aux alcools dans les protéines. Tiré de (Désy et al., 2012, Toxicology Letters)

Le schéma décrit l’interaction du n-propanol avec un site protéique hypothétique par une liaison hydrogène et des forces de van der Waals. Les résidus hydrophobes (cercles noirs) aux pourtours de la cavité stabilisent la molécule d’alcool par des forces de van der Waals (ombres bleutées). Le site accepteur de liaison hydrogène est représenté par un croissant orange et le lien hydrogène lui-même par une ligne pointillée orange. Le déplacement des molécules d’eau en dehors de la pochette induit par la liaison de l’alcool est indiqué par des flèches pointillées bleues. Des cavités similaires ont été découvertes par la résolution de la structure cristalline de la G protein-insensitive inwardly rectifying potassium channel (IRK1) et du Gloeobacter

21 violaceus pentameric ligand-gated ion channel (GLIC) en intéraction avec le 2-méthyl-2,4- pentanediol (MPD) et le propofol/desflurane respectivement (Pegan et al., 2006), (Nury et al., 2011).

Figure 1.7 - Structure moléculaire des composés co-cristallisés en interaction avec des canaux ioniques.

(A) 2-méthyl-2,4-pentanediol; (B) Desflurane; (C) Propofol.

La pochette de liaison aux alcools d’IRK1 est formée par les chaînes latérales d’acides aminés hydrophobes (F47, L232, L245, L330 et Y337) et aussi de résidus permettant la formation d’un réseau de liens hydrogènes entre les groupes hydroxyles du MPD et des groupes carbonyles ou hydroxyles de P244, Y242 et Y337 (Aryal et al., 2009). En ce qui concerne GLIC, il a été montré que le desflurane (un anesthésique général volatil) se trouve enfoui profondément à l’intérieur de la cavité et établi des interactions hydrophobiques avec plusieurs résidus (I201, I202, I258, T255, V242); de plus, son atome d’oxygène se trouve à une distance de pont hydrogène de T255. Le propofol (un anesthésique administré par voie intraveineuse) se situe plus près de l’entrée de la cavité et interagi principalement avec T255 et Y254 par des contacts de van der Waals. Le groupe hydroxyle du propofol pourrait former un lien hydrogène avec Y254 (Nury et al., 2011).

Enfin, le site de liaison aux alcanols de l’alcool déshydrogénase est bien connu : la cavité est de nature hydrophobe comme celle des protéines décrites précédemment, mais elle est la seule qui utilise un atome de zinc pour lier le groupe hydroxyle (Ramaswamy et al., 1994). Cette particularité pourrait expliquer l’affinité de liaison aux alcools substantiellement plus élevée pour cette enzyme, soit environ 1 mM.

Problématique, Hypothèses et Objectifs

Il est désormais bien établi que la consommation d’éthanol a des conséquences sur le système immunitaire. Les expositions aigues et chroniques à cet alcool modifient la réponse immune normale aux pathogènes bactériens et viraux. Cependant, les propriétés immunomodulatrices d’autres alcools à courte chaîne d’usage courant, ou même des alcanols en général, demeurent inconnues. En se basant sur des motifs structuraux, nous avons émis l’hypothèse que l’isopropanol et le méthanol, d’autres alcools à courte chaîne dont l’utilisation est répandue, partageraient les effets immunomodulateurs de l’éthanol lors d’une exposition aigue. Nous nous sommes donc fixés comme objectifs d’examiner les effets immunomodulateurs de l’isopropanol et du méthanol et ensuite d’identifier les mécanismes moléculaires responsables des effets observés.

Les alcanols aliphatiques possèdent des propriétés anesthésiantes qui suivent la règle de Meyer-Overton; c'est-à-dire que leur potentiel anesthésique est corrélé à leur degré d’hydrophobicité. Dans cet ordre d’idée, des résultats préliminaires obtenus dans notre laboratoire ont indiqué que l’isopropanol est plus efficace que l’éthanol pour altérer les fonctions effectrices des lymphocytes T. Par ailleurs, la caractérisation des événements moléculaires responsables des effets neurobiologiques des alcools, tel que l’anesthésie, est beaucoup plus avancée que ce qui est connu concernant leurs effets immunologiques. Ainsi, il a été montré par des études biochimiques et structurelles que les alcools pouvaient lier et moduler l’activité de cibles protéiques comme des canaux ioniques ou des molécules d’adhésion, ce qui explique de façon plausible leurs effets neurobiologiques. Par contre, l’identification de cibles protéiques directes dans les cellules immunitaires ne fait que

23 commencer; et les possibles impacts des alcools sur les lipides membranaires ne sont pas écartés non plus (Szabo et al., 2007). La majorité de l’information disponible à ce sujet ne concerne d’ailleurs que l’éthanol.

Ces observations nous ont amené à émettre comme seconde hypothèse que, à l’instar de leurs effets anesthésiques, les n-alcanols possèderaient aussi des effets immunomodulateurs corrélés à la longueur de leur chaîne carbonée, respectant ainsi la règle de Meyer-Overton. Les objectifs associés à cette hypothèse étaient, en premier lieu, d’examiner les effets sur la fonction effectrice des lymphocytes T de la série des n-alcanols comprenant de un à douze carbones; et ensuite, de dégager les mécanismes moléculaires à la base de ces effets. Nous avions aussi comme but d’identifier une ou des cibles protéiques d’intérêt immunologique auxquelles les alcanols pourraient se lier.

2. Immunosuppressive effect of isopropanol: down- regulation of cytokine production results from the alteration of discrete transcriptional pathways in activated lymphocytes

Olivier Désy, Damien Carignan, Manuel Caruso, and Pedro O. de Campos-Lima2

Laval University Cancer Research Center, Quebec City, Quebec, G1R 2J6, Canada

Journal of Immunology, 2008 Aug 15;181(4):2348-55.

Running title: Immunosuppressive properties of isopropanol. Manuscript information: 63,873 characters, 5 figures, 48 references. Keywords: immunosuppression, alcohol, T lymphocyte, NK cell, isopropanol.

Référence complète :

Résumé

L’isopropanol employé couramment pour des usages domestiques et représente la plus importante cause d’empoisonnement aigu aux alcools après l’éthanol. Bien que les effets de l’éthanol sur le système immunitaire aient été abondamment étudiés, il n’existe presqu’aucune donnée au sujet de l’isopropanol. Vu la structure similaire des deux molécules, nous avons émis l’hypothèse que l’isopropanol possédait aussi des propriétés immunomodulatrices. De fait, une exposition aiguë à l’isopropanol in vitro est néfaste pour l’activité des lymphocytes T et des cellules NK à des concentrations aussi faibles que 0,08- 0,16% (13-26 mM). Le traitement à l’IPA ne perturbe pas la signalisation précoce en aval du récepteur, mais a un effet reproductible et dose-dépendant sur la translocation nucléaire des facteurs de transcription NFAT et AP-1. De plus, dans un modèle d’intoxication aiguë à l’isopropanol, les animaux traités à l’alcool subissent une immunosuppression mesurée par la réduction de la présence d’IL-2 et d’IFN- γ dans le sérum en réponse à l’entérotoxine B staphylococcique. L’isopropanol a aussi réduit assez fortement la production de TNF-α de façon à faire survivre des souris à un choc toxique autrement létal induit par l’entérotoxine. Ces résultats suggèrent que l’isopropanol est potentiellement immunosuppresseur pour les systèmes inné et adaptatif et ils ont une portée significative étant donné l’exposition fréquente de la population à ce produit chimique.

Abstract

Isopropanol (IPA) is widely used in household applications and constitutes a leading cause of acute alcohol intoxication second only to ethanol. Although the effects of ethanol on the immune system have been extensively studied, much less data is available on IPA. Given the structural similarity between the two molecules, we hypothesized that IPA could as well have immune modulatory properties. We report here that acute IPA exposure is detrimental to human T lymphocyte and NK cell activity in vitro in concentrations as low as 0.08- 0.16% (13-26 mM). IPA treatment did not affect receptor-mediated early signaling but had a reproducible and dose-dependent effect on the nuclear translocation of the nuclear factor of activated T cells (NFAT) and activator protein-1 (AP-1). Furthermore, we show in a model of acute IPA intoxication that animals became immunosuppressed as judged by their reduced ability to release IL-2 and IFN-γ in the serum in response to the staphylococcal enterotoxin B. This effect was also associated to the down-regulation of TNF-α production and was sufficiently strong to rescue susceptible animals from enterotoxin-induced toxic shock. Our results suggest that IPA is potentially immunosuppressive to the adaptive and innate immune system and have broad significance given the exposure of the general population to this ubiquitous chemical.

Introduction

Short-chain alcohols have a multitude of biological effects, including cardiac and central nervous system depression. In addition, a considerable body of evidence indicates that ethanol is capable of modulating the immune function mediated by T cells, monocytes, macrophages and neutrophils (1-4). Ethanol also inhibits the leukocyte/endothelial cell interaction thereby limiting the inflammatory response (4). Although the in vitro and in vivo effects of ethanol have been well characterized, much less data is available on other alcohols. Isopropanol (IPA) 3 exposure is the second most common cause of acute alcohol intoxication in North America with about 20,000 cases reported each year to poison centers (5). IPA is readily available to most consumers as rubbing alcohol and as an ingredient of hand-sanitizing gels and other commonly used household solutions. In addition, IPA is

27 widely utilized in hospitals as an antiseptic for surgical scrub and for patient care. Occupational exposure may also occur in numerous industrial applications. Previous studies addressed the impact of IPA exposure on the central nervous system, general hematologic parameters, carcinogenesis, vascular permeability, urinary system, reproduction and development (6-9). However, no detailed analysis of the potential impact of IPA on the immune system is available. Given the structural similarity between IPA and ethanol, we hypothesized that IPA could also have immune modulatory properties. We report here that IPA is detrimental to human T lymphocyte and NK cell activity in vitro in concentrations as low as 0.08-0.16% w/v (or 13-26 mM). These results were further substantiated in a mouse model of acute IPA intoxication in which animals were immunosupressed as judged by their reduced capacity to produce inflammatory cytokines. This immunosupression was sufficiently strong to protect susceptible animals from superantigen-induced lethal shock. Our results have broad significance taking into account the potential exposure of the general population to this ubiquitous chemical.

Materials and Methods

Cell Isolation, Culture, Activation and Proliferation Analysis Mononuclear cells were prepared from peripheral blood from healthy volunteers by density gradient centrifugation using Ficoll-Hypaque (GE Healthcare, Piscataway, NJ). Written informed consent was obtained from all donors. More than 95% pure populations of human NK cells (CD56+) and T cells (CD8+/CD4+) were obtained by using antibody-based EasySep® separation kits with magnetic nanoparticles according to the manufacturer’s instructions (StemCell Technologies, Vancouver, Canada). Cells were kept in complete medium: RPMI 1640 (Invitrogen Canada, Burlington, Canada) supplemented with 10% heat-inactivated FBS (BioCell Inc., Drummondville, Canada). Isopropanol was purchased from BDH (Toronto, Canada) In most experiments, T cells were activated for 5 h at 37°C with anti-CD3/CD28 antibody- coated magnetic beads (Invitrogen). When indicated, alternative T cell activation protocols

28 were used: a) pre-treatment for 20 min on ice with 1 µg/ml mouse anti-human CD3 monoclonal antibody (CD3-2, Mabtech, Nacka Strand, Sweden) followed by washing, and incubation for 3 min at 37°C with 10 µg/ml goat anti-mouse IgG (Invitrogen); b) treatment with 10 ng/ml phorbol 12-myristate 13-acetate (PMA, Sigma, St Louis, MO) and 200 ng/ml ionomycin (Sigma) for 5 h. Human NK cells were activated for 24 h by treatment with 50 µg/ml polyinosinic:polycytidylic acid (poly (I:C)) (Sigma) in presence of 10 U/ml IL-2 and 0.01 ng/ml IL-12 (Feldan Bio Inc., St-Laurent, Canada). CFSE Staining: Freshly purified T cells were labeled with 10 µM CFSE (Invitrogen) in PBS/1% FBS for 10 min at 37°C and further incubated in RPMI/10% FBS for 5 min on ice; then, cells were washed and activated with anti-CD3/CD28 antibody-coated beads for 5 h with or without 0.6% (w/v) IPA. The activating beads were magnetically removed; the cells were washed, and incubated for 72 h in 96-well plates (106 cells/ml) in complete medium without exogenous IL-2. FACS analysis of cell divisions and surface marker expression was performed on a XL flow cytometer (Beckman Coulter Inc., Miami, FL). Western Blot and Luciferase Assay Western blot: Purified T cells were activated for 3 minutes at 37°C by anti-CD3/anti-IgG antibodies as described above with or without 0.6% (w/v) IPA. The cells were lysed in sodium dodecyl sulfate (SDS) sample buffer (2% w/v SDS, 0.25 M β-mercaptoethanol, 10% v/v glycerol, 0.05 M Tris-HCl, pH 6.8, 0.004% w/v bromophenol blue); lysates were separated in 12% polyacrylamide gels and blotted onto nitrocellulose filters (Hybond-C, GE Healthcare, Piscataway, NJ). The membranes were first probed with ZAP-70-specific antibodies: rabbit anti-human ZAP-70 (99F2, 1/1000, Cell Signaling Technology, Danvers, MA) and mouse anti-human ZAP-70 (pY319)/Syk (pY352) (17a, 1/5000, BD Biosciences, Mississauga, Canada); then, they were washed, and incubated with 1/15000 dilutions of the antibodies IRDye 800CW goat anti-rabbit IgG and IRDye 680 goat anti-mouse IgG (Li-Cor Biosciences, Lincoln, NE). Detection and quantification was performed with the Odyssey Infrared Imaging System (Li-Cor Biosystems). Luciferase assay: Jurkat-luc cells were stimulated with PMA/ionomycin with or without IPA treatment as indicated in the text. Lysates for luciferase assays were prepared with the passive lysis buffer (E1941, Promega, Madison, WI), mixed with reaction solution (25 mM

glycylglycine, pH7.8, 10 mM MgSO4, 5 mM ATP), and analysed in a Berthold Lumat 9501 luminometer (Berthold, Nashua, NH) after the addition of D-luciferin (Fisher Scientific, Pittsburgh, PA) to a final concentration of 0.1 mM. Relative luciferase units were calculated in relation to the unstimulated negative control after normalization to total protein content measured by the Bradford assay (Bio-Rad, Hercules, CA). Generation of Jurkat-luc cells: Stable Jurkat-luc cells were generated by lentiviral transduction. Vector generation and transduction conditions: The synthetic promoter used in our studies contains three copies of the human distal IL-2 NFAT binding site placed upstream of the -77 to +45 region of the human IL-2 promoter. The firefly luciferase gene driven by this synthetic promoter was cloned in the Nhe I and Xho I sites of a modified version of pRRL-5pme (10) to generate pLV-iluc. This parental version of pRRL-5pme also carries the zeocin resistance gene driven by the PGK promoter. Lentiviral supernates were generated in 293T cells by transient transfection (11). Three plasmids: pMDLg/RRE, pRSV-rev and pMD.VSV-G were cotransfected with pLV-iluc. The supernates containing lentivirus were harvested 48 and 72 hours after transfection, filtered through a 0.45 m filter, and frozen at -80oC until use. Jurkat cells were transduced overnight in 24-well plates at 2.5 x 105 cells/ml with 0.5 ml viral supernate plus 0.5 ml fresh medium and 8 g/ml polybrene (Sigma). Stable Jurkat-luc cells were generated after two rounds of 4-day zeocin selection (Invitrogen).

Cytokine analysis Measurements of human or murine IL-2 and IFN-γ in cell culture supernates and murine IL-2 and IFN-γ in serum samples were performed with specific cytokine ELISA kits according to the manufacturer’s instructions (Mabtech). Briefly, 96-well plates were coated with the relevant capture antibody (hIL-2: IL-2 I, mIL-2: 1A12, hIFN-γ: 1-D1K, mIFN-γ: AN18) and incubated with serially diluted standards or unknown samples; then, they were washed and incubated with a biotinylated detection antibody (hIL-2: IL-2 II- biotin, mIL-2: 5H4-biotin, hIFN-γ: 7-B6-1-biotin, mIFN-γ: R4-6A2-biotin) followed by streptavidin-horseradish peroxidase. The plates were read at 450 nm (or 620 nm) after treatment with 3,3',5,5'- Tetramethylbenzidine (TMB) substrate solution.

Murine TNF-α levels were measured with the mouse TNF-α Enzyme Immunometric Assay Kit (Assay Designs, Ann Arbor, MI) according to the manufacturer’s instructions. Inhibitor compounds (Sigma) were used as follows: cyclosporin A: 1 µg/ml; BAY 11-7082: 5 µM; SP600125: 25 µM; and PD98059: 50 µM. ELISA-based Transcription Factor Activation Assay Nuclear proteins were extracted using the Active Motif Nuclear Extract kit (Active Motif, Carlsbad, CA) according to the manufacturer's instructions; the total protein concentration of the lysates was determined by the Bradford assay (Bio-Rad). NFAT activation was measured with the TransAM NFATc1 kit; c-Fos and c-Jun activation was measured with the TransAm AP-1 kit; p50 and p65 activation was measured with the TransAm NF-κB kit. ELISA-based TransAm kits were used according to the manufacturer’s instructions (Active Motif). Briefly, nuclear extracts were incubated with plate-bound transcription factor- specific oligonucleotides; the plates were washed, and further incubated with transcription factor-specific antibodies. Addition of a horseradish-conjugated secondary antibody and the TMB substrate produced a colorimetric reaction measurable in a spectrophotometer. Cytotoxicity assays The cytotoxic activity was analysed in standard 4 h 51Cr-release assays as reported 51 o elsewhere (11). The targets were labeled with Na CrO4 for 2 h at 37 C. Tests were performed in presence or absence of 0.6% (w/v) IPA in triplicate. Freshly purified NK cells and peripheral T lymphocytes were used as effectors. Target cells were the NK-sensitive K562 cell line and a control autologous lymphoblastoid cell line in the NK cells assays; Targets for T cell assays were the OKT3 hybridoma and a control hybridoma specific to MAGE-A9 (kindly provided by Dr. Alain Bergeron, Laval University). In vivo studies 7-13-week-old female BALB/c mice were bought from The Jackson Laboratory (Bar Harbor, ME). All tests respected the ethical guidelines set by the Institutional Animal Protection Committee (CPA-CHUQ). Animals received subcutaneously 5 µg staphylococcal enterotoxin B (SEB) (Toxin Technology Inc., Sarasota, FL) for cytokine induction and were sacrificed by CO2 asphyxiation 2 h or 4 h after administration for IL- 2/TNF-α or IFN-γ serum analysis, respectively.

For the analysis of murine T cell subsets, mice received 20 µg SEB intravenously with (or without) an intraperitoneal injection of 2 g/kg isopropanol. The animals were sacrificed 160 min later for spleen isolation; CD4+ and CD8+ T cell populations were purified by using antibody-based cell separation kits with magnetic nanoparticles (StemCell Technologies) according to the manufacturer’s instructions. Purified cells (97-98% pure) were cultured in vitro without isopropanol for 18 h and the supernates were checked for the presence of IFN-γ by ELISA. Toxic shock was induced with a subcutaneous injection of 10 µg SEB after presensitization with 20 mg D-galactosamine (Sigma) as reported elsewhere (12). IPA was injected intraperitoneally (2 g/kg). Mice were checked hourly for 72 h. Animals that survived the 72 h experiment were followed for 5 days. The blood alcohol concentration (BAC) was determined by gas chromatography with a 3900 GC unit (Varian, Palo Alto, CA). Statistical Analysis One-way ANOVA followed by Dunnett’s multiple-comparison posttest was performed with GraphPad Prism (GraphPad Software Inc., San Diego, CA) on data presented in all figures, except when indicated otherwise. The Student’s t test was used in figures 2.2B, 2.4C and 2.4D. Survival curves were determined by the Kaplan-Meier method (fig. 5E). p values < 0.05 were considered significant.

Results

Isopropanol interferes with the production of IL-2 and the proliferative capacity of antigen-receptor-activated peripheral T lymphocytes The IL-2 gene is transcribed following antigen-specific activation of the T cell receptor (TCR). In this study, we have investigated first whether IPA exposure in vitro would have any impact on the ability of human peripheral lymphocytes to produce IL-2 once activated by antibody cross-linking of the TCR. These cells produced less IL-2 when treated with IPA as measured in the culture supernates by ELISA (fig. 2.1A, black bars). The reduction in cytokine production was observed at IPA doses as low as 0.16%. The observed alcohol effect was not the consequence of nonspecific cytotoxicity as the cell viability of IPA-

32 treated samples in the concentration range that produced 36 to 86% IL-2 inhibition was similar to that of untreated control cells (fig. 2.1A, gray bars). Given the importance of the IL-2 autocrine loop for the expansion of antigen-specific cells in vivo, we asked if the reduced production of IL-2 translates into a lower proliferative capacity of the activated lymphocytes. Purified T cells were labeled with CFSE, activated, and analysed 72 h after by flow cytometry. Although the 5-h TCR cross-linking led to 1-4 divisions in about 40% of the cells in absence of exogenous IL-2 (mean: 39.69 ± 2.06 SEM, n: 4), cell proliferation was much less pronounced in presence of IPA treatment (mean: 18.82 ± 4.53 SEM, n: 4). The means of IPA-treated and untreated cells differ with a p value of 0.0058 (Student’s t test). One representative experiment is shown in figure 2.1B. Early signaling following TCR activation is preserved in isopropanol-treated lymphocytes IPA could conceivably interact with the T cell receptor directly thereby severing antigen- dependent signal transduction in lymphocytes. To address this possibility we have examined the phosphorylation status of the key downstream adaptor molecule ZAP-70. Figure 2.2A shows that signaling through the T cell receptor itself was not affected by IPA, as the ZAP-70 activation proceeded as efficiently as in untreated cells following anti-CD3 antibody cross-linking. Figure 2.2B presents the densitometric analysis of three Western blots, one of which is depicted in panel A. Next, we have examined the cytosolic Calcium levels induced by T cell activation in presence of IPA. Intracellular Calcium oscilations regulate a variety of biological processes, including antigen-dependent, TCR-mediated T cell activation (13). IPA did not affect the cytosolic Calcium increase observed after TCR triggering or treatment with the Calcium ionophore ionomycin (data not shown). Isopropanol blocks IL-2 production via transcriptional inhibition The lack of an obvious impact of the IPA treatment on early TCR signaling led us to examine the possibility of a negative effect on IL-2 transcription. For this purpose, we have generated a stable Jurkat subline carrying the firefly luciferase gene driven by a synthetic IL-2 minimal promoter shown previously to be responsive to PMA/ionomycin (Jurkat-luc) (14-15). IPA was capable of inhibiting in a dose-dependent manner the luciferase activity triggered by PMA/ionomycin in these cells (fig. 2.3A). IPA concentrations as low as 0.3%

33 had a significant dampening effect on IL-2 transcription as indicated by a 24% reduction in luciferase activity. Similar results were obtained by anti-CD3 antibody cross-linking in Jurkat-luc cells (data not shown). Nuclear translocation of transcription factors is affected by isopropanol in activated T cells The promoter used in the experiments shown in figure 2.3A contains three copies of the binding site for the nuclear factor of activated T cells (NFAT) placed upstream of the IL-2 core promoter and is highly responsive to the Ca++/NFAT signaling pathway (14-15). However, the regulation of the IL-2 gene following TCR triggering is more complex and involves the participation of transcription factors activated by two additional major signal transduction pathways (16). In order to dissect further the relative impact of IPA on these molecules, we have measured the nuclear translocation of NFAT (NFATc1), nuclear factor- κB (NF-κB: p50/p65), and activator protein-1 (AP-1: c-Jun/c-Fos) in TCR-stimulated purified human T cells exposed to different concentrations of IPA. Figure 2.3B shows that lymphocyte activation by anti-CD3/CD28 antibodies led to a 2-fold increase in the amount of NFAT in the nucleus. The same stimulation in presence of 0.6% IPA led only to a 1-fold increase in nuclear NFAT (or 54% of the maximal NFAT nuclear content above the unstimulated cell baseline). This effect was dose-dependent with the highest inhibition observed for the highest IPA concentrations. The calcineurin inhibitor cyclosporine A was used as a control in the same stimulatory conditions with little variation in nuclear NFAT content (17% less than the nuclear content baseline of unstimulated cells). Activation of AP-1 in presence of IPA followed the same pattern observed for NFAT with 55% of the maximal c-Jun nuclear content above the unstimulated cell baseline achieved at the highest IPA concentration (fig. 2.3E). Activation by anti-CD3/CD28 antibodies in presence of the c-Jun N-terminal kinase inhibitor SP600125 produced 42% of the maximal c-Jun nuclear content above the unstimulated cell baseline. The same T cell stimulatory conditions produced 45.4% of the maximal c-Fos nuclear content in presence of the highest IPA concentration and 88.3% in presence of the MEK1 inhibitor PD98059 (fig. 2.3F). In contrast to the results obtained with NFAT and AP-1, activation of NF-κB remained unaffected by IPA treatment at all tested concentrations (fig. 2.3C/D). The compound BAY

11-7082, an inhibitor of IκBα phosphorylation, reduced the nuclear content of p50/p65 in the same experiments to levels lower than those of the unstimulated control.

Production of the inflammatory cytokine interferon-γ by human peripheral T lymphocytes and NK cells is inhibited by isopropanol The NFAT and AP-1 families of transcription factors play a major role in the expression of several cytokines, including IFN-γ (17-22). The finding that the inhibition of IL-2 production in IPA-treated T cells is associated with reduced nuclear translocation of c-Jun, c-Fos and NFAT led us to speculate that IFN-γ expression would also be affected in these cells. First, we have examined the IFN-γ release in peripheral T lymphocytes following activation with anti-CD3/CD28 antibodies in presence of different IPA concentrations (fig. 2.4A). Treatment with 0.16% IPA led to a 35% reduction in IFN-γ release. The higher IPA concentrations tested, 0.3%, 0.6% and 1.2% produced inhibitory effects of 55%, 84% and 98%, respectively. Cyclosporin A in the same experimental conditions led to a virtually complete IFN-γ inhibition (data not shown). The above results have encouraged us to extend our analysis to another immune cell capable of producing large amounts of IFN-γ. Purified human NK cells have been stimulated in vitro with poly (I:C) in presence of IL-2 and IL-12 and exposed to different quantities of IPA. Figure 4B shows that concentrations as low as 0.08% were active. Treatment with 0.08%, 0.16% and 0.3% IPA reduced the IFN-γ release by stimulated NK cells by 31%, 40% and 87%, respectively. The two highest IPA concentrations produced an almost complete inhibitory effect with only background levels of IFN-γ being released. NK cells were > 95% viable at all IPA concentrations tested (data not shown). Isopropanol reduces the cytotoxic activity of T lymphocytes and NK cells in vitro The identification of the negative impact of isopropanol on the production of IFN-γ by T and NK cells has prompted us to examine if this effect was extended to other effector functions. Panel C in figure 4 shows that the cytotoxic activity of purified human peripheral T lymphocytes against OKT3 hybridoma cells was inhibited by about 20% in presence of IPA. OKT3 cells display the activating anti-CD3 antibody and work as a T cell target (23).

Similarly, the cytotoxicity of purified human NK cells against K562 target cells was inhibited by 30-40% in presence of IPA (Figure 2.4D).

Isopropanol inhibits the production of IL-2 and IFN-γ in vivo After having demonstrated the negative impact of IPA treatment on IL-2 and IFN-γ production by lymphocytes in vitro, we have examined the relevance of these findings in a mouse model of acute IPA intoxication. Mice were administered IPA intraperitoneally, 2 g/kg, to generate a mean blood alcohol concentration of 200 mg/dl after 30 min (198 ± 3.704 SEM, n: 10). Induction of IL-2 and IFN-γ production in vivo was achieved by subcutaneous injection of the superantigen staphylococcal enterotoxin B. T lymphocytes with the relevant TCR Vβ chains undergo transient activation, cell proliferation, and begin massive cytokine production (12). As anticipated, injected SEB induced IL-2 levels of > 12 ng/ml after 2 h (fig. 2.5A) and > 1.6 ng/ml IFN-γ after 4 h (fig. 2.5B). IPA dampened the cytokine production substantially: 49.2% or 6.3 ng/ml IL-2 detected at 2 h, and 86.6% or 0.2 ng/ml IFN-γ detected at 4h. The differences in cytokine serum levels between animals treated and untreated with IPA were statistically significant as indicated in figures 2.5A and 2.5B.

Isopropanol treatment impacts both CD4+ and CD8+ T lymphocytes in vivo Both CD4+ and CD8+ T lymphocytes may respond to SEB to produce cytokines. We have examined next how these T cell subsets were affected by isopropanol in our in vivo experimental model. BALB/c mice were injected with SEB intravenously to activate CD4+ and CD8+ T cells carrying the responsive TCRs; isopropanol was provided intraperitoneally. It has been shown that a minimal T cell/APC conjugate time of 2 h is required for commitment to cytokine production and cell proliferation (24). Basing on this premise, we have followed the animals for 160 minutes, after which they were sacrificed; their splenocytes were harvested, and separated into CD4+ and CD8+ T cell populations. Purified T cell subsets were incubated for 18 h at 37oC and culture supernates were analysed by ELISA for IFN-γ production. Figure 2.5C shows that isopropanol exposure in

36 vivo reduced the IFN-γ release by CD4+ T cells in vitro by 74%. Similarly, the IFN-γ production in vitro by CD8+ T cells decreased about 70% when animals from which they derived were exposed to IPA for 160 min (fig. 2.5D).

The immunosuppressive effect of IPA confers protection to animals from SEB- induced toxic shock The ability of IPA to prevent or substantially reduce the production of cytokines in response to SEB in vivo led us to wonder whether this immunosuppressive effect could delay the development of toxic shock in susceptible animals. In order to investigate this possibility, we have sensitized BALB/c mice with 20 mg D-galactosamine by intraperitoneal injection. All sensitized animals succumbed to toxic shock within 14 h after receiving SEB subcutaneously (12/12) (fig. 2.5E). In contrast, a single injection of 2 g/kg IPA had a dramatic effect since it delayed or completely aborted the development of SEB- induced toxic shock in all animals. Seven out of twelve animals survived the 72-h experiment; they were followed for up to 5 days after SEB injection and were indistinguishable from sensitized control groups receiving PBS or IPA as regards activity, grooming, eating and drinking behavior. No animal injected with IPA (12/12) or PBS (12/12) in absence of SEB died. In addition to IL-2 and IFN-γ, the inflammatory cytokine TNF-α is copiously produced and plays a central role in the pathophysiology of superantigen-induced lethal shock (12). TNF- α gene transcription was shown by other investigators to be regulated by NFAT (25). Our results revealed IPA as a negative regulator of NFAT nuclear translocation in vitro and as an immunosuppressive agent capable of protecting mice from toxic shock. Therefore, it was a logical assumption that IPA treatment would block TNF-α production in vivo. Indeed, figure 2.5F shows that mice treated with SEB produced significant amounts of TNF-α in the first 2 h after injection as opposed to animals that received IPA + SEB.

Discussion

Alcohols have the ability to partition into cell membranes and to denature proteins by promoting the formation of α-helices and/or by disrupting tertiary structures; these effects

37 are largely nonspecific and are typically observed at high concentrations (> 2% w/v or > 500 mM) (26). At more physiologically relevant concentrations, alcohols have been shown to induce loss of function of specific proteins, such as: ion channels, neurotransmitter receptors, enzymes, and adhesion molecules (27-29). Structural and biophysical data suggest that binding to the target proteins occurs at discrete sites that are constituted by hydrophobic pockets lined by nonpolar amino acids (26-29). As suggested for other short- chain alcohols (26, 30), IPA could displace water molecules from such pockets and establish contact with the proteins via hydrogen bonds that would be stabilized by van der Waals forces in the hydrophobic region. These interactions would ultimately produce a local distortion and alteration in protein function. Another interpretation for the effects of IPA would result from the possible interference with the capacity of membrane micro-domains to recruit and/or retain molecules involved in signaling, thus compromising the formation of the immunological synapse. Given the central role played by lipid rafts in amplifying receptor-mediated signals in immune cells (3), it is conceivable that IPA could affect surface molecules, such as the TCR, directly by inducing unfavorable conformational changes or, indirectly, by disrupting lipid-protein interactions. A similar model has recently been evoked to explain the ethanol inhibition of LPS-mediated Toll-like receptor 4 (TLR4) signaling in macrophages (3). Upon engagement of the relevant ligand, the TCR triggers a phosphorylation cascade that is followed by a biphasic increase in intracellular Ca++. The initial wave derives from the intracellular stores and is rapidly trailed by the extra-cellular influx regulated by Ca++ release-activated Ca++ (CRAC) channels (13). We initially examined if IPA would exert its immunosuppressive effect by interference with the CRAC-regulated Ca++ influx and the subsequent calcineurin-dependent activation of NFAT. There was some support for this possibility given the reported association of other alcohols with ion channels, often altering their function (26-27, 29). We failed to show any IPA-induced change in the pattern of intracellular Calcium increase that follows TCR triggering or ionomycin treatment (data not shown). The fact that TCR-mediated early signaling as measured by ZAP-70 phosphorylation and Calcium release is preserved indicates that the effect of IPA is downstream of the cell membrane. We cannot discard, however, that higher IPA

38 concentrations could also affect lipid rafts in a way reminiscent of the model suggested for ethanol on the TLR4 receptor (3) but this remains to be experimentally tested. The inhibition reported here was observed in vitro at IPA concentrations as low as 0.08% (13 mM) as measured by IFN-γ release in NK cells and 0.16% (26 mM) as measured by IL- 2 and IFN-γ release in T cells. These concentrations are equal to or lower than those of ethanol used in previous studies that reported a statistically significant impact on immune cells (1-4). Many of the biological effects of ethanol on the immune system have been associated to a reduced nuclear translocation of NF-κB, a transcription factor capable of binding the promoter regions of multiple cytokines (3-4). In contrast, we found that IPA does not affect NF-κB but has a reproducible and dose-dependent effect on the nuclear translocation of AP-1 and NFAT. This finding supports the view that IPA exerts its impact on immune cells through the interaction with selective pathways rather than a membrane- based nonspecific down-modulation of the immune cell activation. NFAT and AP-1 have been shown to modulate synthesis of the three cytokines examined in this paper (IL-2, IFN- γ and TNF-α) (14-15, 17-22, 25). As expected, NFAT nuclear translocation and cytokine release was blocked by cyclosporine A in activated lymphocytes. Nevertheless, IPA differed from cyclosporine A in that it did not target the calcineurin phosphatase activity (data not shown). Thus it must interact directly with NFAT or with downstream molecules involved in its nuclear translocation, such as importin β1, or molecules involved in its phosphorylation in the nucleus, such as glycogen synthase kinase 3. Similarly, IPA may also interact directly with c-Jun and/or c-Fos compromising the formation and/or function of the AP-1 dimer as the phosphorylation pattern of upstream molecules such as p38 remains unchanged (data not shown). In order to address the immunosuppressive effects of IPA in an in vivo setting, we have used a model of acute intoxication. Deliberate or accidental ingestion of IPA ranks second as a cause of alcohol poisoning according to the 2005 annual report of the American Association of Poison Control Centers (5). Acute intoxication usually occurs in alcoholic patients, children, and suicidal individuals (31). The blood IPA concentration can reach levels as high as 560 mg/dl (0.56% or 93 mM); many of the reported measurements have been made hours after ingestion and may underestimate the serum amounts present in the

39 early phase of the intoxication (32-37). Nevertheless, a concentration above 400 mg/dl is usually considered life-threatening and demands a more aggressive intervention such as dialysis (38-39). In our model, we have injected mice intraperitoneally with 2 g/kg IPA to generate a blood alcohol concentration of 200 mg/dl (0.2% or 33 mM) after 30 min; this level is under the reported average sublethal blood concentration in intoxicated humans (310 mg/dl) (40) and is well within the concentration range that we have shown to be biologically active in vitro (starting at 0.08-0.16% or 13-26 mM). Our results indicate that during this state of acute intoxication the animals are immunosuppressed as judged by their reduced ability to release IL-2 or IFN-γ in the serum in response to SEB. The magnitude of this immunosupression was further assessed by monitoring the survival of animals injected with SEB after presensitization with D-galactosamine. We were initially uncertain about the outcome of this particular experiment as we thought that it might be too stringent. In fact, all presensitized animals developed a fulminating toxic shock syndrome with a median survival of 9 h after SEB injection. Nevertheless, in contrast to the untreated animals, the syndrome did not occur or had its development delayed in all mice treated with IPA and the majority survived. This is in line with the massive suppressive effect of IPA on the production of IL-2, IFN-γ and TNF-α. It is believed that the production of copious amounts of IFN-γ, and especially of TNF-α, plays a major role in the development of the syndrome (12, 25). An obvious consequence of our findings is the assumption that any individual acutely intoxicated by IPA may also be acutely immunosuppressed. This assumption could be easily tested in a clinical setting and may provide the basis for the establishment of precautionary measures in the emergency room to deal with this predicament. The issue would be particularly relevant in circumstances in which an underlying infection or trauma complicate the clinical picture. The implications of our results could be also extrapolated to areas other than acute systemic intoxication. IPA is used in many industrial applications and occupational exposure by inhalation or other routes may occur (8). One of the major weaknesses of the literature on IPA is the virtual absence of solid information about chronic toxicity in humans. Our experiments were not designed to address this issue. Nevertheless the results reported here

40 for acute exposure indicate that immunological parameters may serve as a sensitive endpoint for IPA toxicity that could be included in future studies on the long-term effects of this chemical. Intact adult skin is not an efficient route of IPA absorption; the skin permeation coefficient -4 (kp) is estimated to be in the order of 4-15 x 10 cm/h (41-43). Yet, dermal absorption does occur (41, 44-45) and a few cases of systemic IPA intoxication after topical exposure have been reported in the literature (46-48). IPA is present, often in a high concentration (60- 95%), in hand sanitizer gels/solutions and many household products readily available over the counter. Taking into consideration that our results show a significant biological effect in vitro with IPA concentrations as low as 0.08-0.16% (13-26 mM), it is reasonable to question if the application to the skin of a product that is 500-1000 times more concentrated would have similar consequences even in the context of poor dermal absorption. As regards intact normal adult skin, it is likely that the immunosuppressive effect, if any, would be transitory and only relevant to the immune cells present in the treated skin itself. Nevertheless, future studies to address this issue are warranted given the widespread and poorly regulated use of this chemical. The in vitro data presented here suggests that acute IPA exposure reduces the ability of lymphocytes to produce proinflammatory cytokines, and thus may compromise the innate and adaptive immune system; in addition, acute intoxication led to acute immunosupression in vivo, an effect that was sufficiently strong to rescue susceptible animals from enterotoxin-induced toxic shock. These results are directly relevant in the context of acute IPA intoxication and constitute a rational for the inclusion of immunological endpoints into the design of future studies to address chronic and topical IPA exposure.

Acknowledgements The authors wish to thank: G. Crabtree for the IL-2 minimal promoter; J. Gosselin and M. Dufour for Calcium experiments; L. Nadeau for gas chromatography; D. Richard for access to the infrared imaging system; P. Salmon and D. Trono for lentivectors; D. L’Héreault and H. Dombrowski for blood collection; J. Charron, F. Lyon, J. Filley and M. Vauzelle for helpful discussions; Anne Julien and Richard Boutet for valuable technical assistance.

References

1. Goral, J., and E. J. Kovacs. 2005. In vivo ethanol exposure down-regulates TLR2, TLR4, and TLR9-mediated macrophage inflammatory response by limiting p38 and ERK1/2 activation. J. Immunol. 174: 456-463. 2. Taieb, J., C. Delarche, F. Ethuin, S. Selloum, T. Poynard, M. A. Gougerot-Pocidalo, and S. Chollet-Martin. 2002. Ethanol-induced inhibition of cytokine release and protein degranulation in human neutrophils. J. Leukoc. Biol. 72: 1142-1147. 3. Szabo, G., A. Dolganiuc, Q. Dai, and S. B. Pruett. 2007. TLR4, ethanol, and lipid rafts: a new mechanism of ethanol action with implications for other receptor-mediated effects. J. Immunol. 178: 1243-1249. 4. Saeed, R. W., S. Varma, T. Peng, K. J. Tracey, B. Sherry, and C. N. Metz. 2004. Ethanol blocks leukocyte recruitment and endothelial cell activation in vivo and in vitro. J. Immunol. 173: 6376-6383. 5. Lai, M. W., W. Klein-Schwartz, G. C. Rodgers, J. Y. Abrams, D. A. Haber, A. C. Bronstein, and K. M. Wruk. 2006. 2005 Annual Report of the American Association of Poison Control Centers' national poisoning and exposure database. Clin. Toxicol. 44: 803- 932. 6. Burleigh-Flayer, H. D., M. W. Gill, D. E. Strother, L. W. Masten, R. H. McKee, T. R. Tyler, and T. Gardiner. 1994. Isopropanol 13-week vapor inhalation study in rats and mice with neurotoxicity evaluation in rats. Fundam. Appl. Toxicol. 23: 421-428. 7. Burleigh-Flayer, H., R. Garman, D. Neptun, C. Bevan, T. Gardiner, R. Kapp, T. Tyler, and G. Wright. 1997. Isopropanol vapor inhalation oncogenicity study in Fischer 344 rats and CD-1 mice. Fundam. Appl. Toxicol. 36: 95-111. 8. Kapp, R. W., Jr., C. Bevan, T. H. Gardiner, M. I. Banton, T. R. Tyler, and G. A. Wright. 1996. Isopropanol: summary of TSCA test rule studies and relevance to hazard identification. Regul. Toxicol. Pharmacol. 23: 183-192. 9. Kasuga, F., S. Inoue, T. Asano, and S. Kumagai. 1992. Effects of isopropanol on the development of inflammatory reactions in rats. Food Chem. Toxicol. 30: 631-634.

10. Dull, T., R. Zufferey, M. Kelly, R. J. Mandel, M. Nguyen, D. Trono, and L. Naldini. 1998. A third-generation lentivirus vector with a conditional packaging system. J. Virol. 72: 8463-8471. 11. Lalonde, A., J. Avila-Cariño, M. Caruso, and P. O. de Campos-Lima. 2007. Rescue of the immunotherapeutic potential of a novel T cell epitope in the Epstein-Barr virus latent membrane protein 2. Virology 361: 253-262. 12. Miethke, T., C. Wahl, K. Heeg, B. Echtenacher, P. H. Krammer, and H. Wagner. 1992. T cell-mediated lethal shock triggered in mice by the superantigen staphylococcal enterotoxin B: critical role of tumor necrosis factor. J. Exp. Med. 175: 91-98. 13. Gallo, E. M., K. Cante-Barrett, and G. R. Crabtree. 2006. Lymphocyte calcium signaling from membrane to nucleus. Nat. Immunol. 7: 25-32. 14. Durand, D. B., J. P. Shaw, M. R. Bush, R. E. Replogle, R. Belagaje, and G. R. Crabtree. 1988. Characterization of antigen receptor response elements within the interleukin-2 enhancer. Mol. Cell. Biol. 8: 1715-1724. 15. Shaw, J. P., P. J. Utz, D. B. Durand, J. J. Toole, E. A. Emmel, and G. R. Crabtree. 1988. Identification of a putative regulator of early T cell activation genes. Science 241: 202-205. 16. Attema, J. L., R. Reeves, V. Murray, I. Levichkin, M. D. Temple, D. J. Tremethick, and M. F. Shannon. 2002. The human IL-2 gene promoter can assemble a positioned nucleosome that becomes remodeled upon T cell activation. J. Immunol. 169: 2466-2476. 17. Barbulescu, K., C. Becker, J. F. Schlaak, E. Schmitt, K. H. Meyer zum Buschenfelde, and M. F. Neurath. 1998. IL-12 and IL-18 differentially regulate the transcriptional activity of the human IFN-gamma promoter in primary CD4+ T lymphocytes. J. Immunol. 160: 3642-3647. 18. Jones, B., and J. Chen. 2006. Inhibition of IFN-gamma transcription by site-specific methylation during T helper cell development. EMBO J. 25: 2443-2452. 19. Macian, F. 2005. NFAT proteins: key regulators of T-cell development and function. Nat. Rev. Immunol. 5: 472-484.

20. Porter, C. M., and N. A. Clipstone. 2002. Sustained NFAT signaling promotes a Th1-like pattern of gene expression in primary murine CD4+ T cells. J. Immunol. 168: 4936-4945. 21. Teixeira, L. K., B. P. Fonseca, A. Vieira-de-Abreu, B. A. Barboza, B. K. Robbs, P. T. Bozza, and J. P. Viola. 2005. IFN-gamma production by CD8+ T cells depends on NFAT1 transcription factor and regulates Th differentiation. J. Immunol. 175: 5931-5939. Yang, D. D., D. Conze, A. J. Whitmarsh, T. Barrett, R. J. Davis, M. Rincon, and R. A. Flavell. 1998. Differentiation of CD4+ T cells to Th1 cells requires MAP kinase JNK2. Immunity 9: 575-585. Geisberg, M., and B. Dupont. 1992. Cytolytic effector function is present in resting peripheral T lymphocytes. Int. immunol. 4: 1273-1281. Lee, K. H., A. D. Holdorf, M. L. Dustin, A. C. Chan, P. M. Allen, and A. S. Shaw. 2002. T cell receptor signaling precedes immunological synapse formation. Science 295:1539-1542. 25. Tsytsykova, A. V., and A. E. Goldfeld. 2000. Nuclear factor of activated T cells transcription factor NFATp controls superantigen-induced lethal shock. J. Exp. Med. 192: 581-586. 26. Dwyer, D. S., and R. J. Bradley. 2000. Chemical properties of alcohols and their protein binding sites. Cell. Mol. Life Sci. 57: 265-275. 27. Jung, S., M. H. Akabas, and R. A. Harris. 2005. Functional and structural analysis of the GABAA receptor alpha 1 subunit during channel gating and alcohol modulation. J. Biol. Chem. 280: 308-316. 28. Ren, H., Y. Honse, and R. W. Peoples. 2003. A site of alcohol action in the fourth membrane-associated domain of the N-methyl-D-aspartate receptor. J. Biol. Chem. 278: 48815-48820. 29. Shahidullah, M., T. Harris, M. W. Germann, and M. Covarrubias. 2003. Molecular features of an alcohol binding site in a neuronal potassium channel. Biochemistry 42: 11243-11252. 30. Klemm, W. R. 1998. Biological water and its role in the effects of alcohol. Alcohol 15: 249-267.

31. Zaman, F., A. Pervez, and K. Abreo. 2002. Isopropyl alcohol intoxication: a diagnostic challenge. Am. J. Kidney Dis. 40: E12. King, L. H. Jr, K. P. Bradley, and D. L. Shires Jr. 1970. Hemodialysis for isopropyl alcohol poisoning. JAMA. 211:1855. Daniel, D. R., B. H. McAnalley, and J. C. Garriott. 1981. Isopropyl alcohol metabolism after acute intoxication in humans. J. Anal. Toxicol. 5:110-112. Rosansky, S. J. 1982. Isopropyl alcohol poisoning treated with hemodialysis: kinetics of isopropyl alcohol and acetone removal. J. Toxicol. Clin. Toxicol. 19:265-271. Gaudet, M.P., and G. L. Fraser. 1989. Isopropanol ingestion: case report with pharmacokinetic analysis. Am. J. Emerg. Med. 7:297-299. Rich, J., R. T. Scheife, N. Katz, and L. R. Caplan. 1990. Isopropyl alcohol intoxication. Arch. Neurol. 47:322-324. 37. Mueller-Kronast, N., A. A. Rabinstein, L. Voung, and A. M. Forteza. 2003. Isopropanol intoxication mimicking basilar artery thrombosis. Neurology 61:1456-1457. 38. Emadi, A., and L. Coberly. 2007. Intoxication of a hospitalized patient with an isopropanol-based hand sanitizer. N. Engl. J. Med. 356: 530-531. 39. Lacouture, P. G., S. Wason, A. Abrams, and F. H. Lovejoy Jr. 1983. Acute isopropyl alcohol intoxication. Diagnosis and management. Am. J. Med. 75: 680-686. 40. Ekwall, B., and C. Clemedson. 1997. Time-related lethal blood concentrations from acute human poisoning of chemicals. Part 2: The monographs. No. 10 Isopropyl alcohol. CTLU, Stanga, Sweden. 1-7. 41. Boatman, R. J., L. G. Perry, L. A. Fiorica, J. C. English, R. W. Kapp Jr., C. Bevan, T. R. Tyler, M. I. Banton, and G. A. Wright. 1998. Dermal absorption and pharmacokinetics of isopropanol in the male and female F-344 rat. Drug Metab. Dispos. 26: 197-202. 42. Clewell, H. J., 3rd, P. R. Gentry, J. M. Gearhart, T. R. Covington, M. I. Banton, and M. E. Andersen. 2001. Development of a physiologically based pharmacokinetic model of isopropanol and its metabolite acetone. Toxicol. Sci. 63: 160-172. 43. Frasch, H. F. 2002. A random walk model of skin permeation. Risk Anal. 22: 265- 276.

44. Cronin, M. T., J. C. Dearden, G. P. Moss, and G. Murray-Dickson. 1999. Investigation of the mechanism of flux across human skin in vitro by quantitative structure- permeability relationships. Eur. J. Pharm. Sci. 7: 325-330. 45. Turner, P., B. Saeed, and M. C. Kelsey. 2004. Dermal absorption of isopropyl alcohol from a commercial hand rub: implications for its use in hand decontamination. J. Hosp. Infect. 56: 287-290. 46. Arditi, M., and M. S. Killner. 1987. Coma following use of rubbing alcohol for fever control. Am. J. Dis. Child. 141: 237-238. 47. Dyer, S., M. B. Mycyk, W. R. Ahrens, and M. Zell-Kanter. 2002. Hemorrhagic gastritis from topical isopropanol exposure. Ann. Pharmacother. 36: 1733-1735. 48. Leeper, S. C., A. L. Almatari, J. D. Ingram, and K. E. Ferslew. 2000. Topical absorption of isopropyl alcohol induced cardiac and neurologic deficits in an adult female with intact skin. Vet. Hum. Toxicol. 42: 15-17.

Footnotes 1 This work was supported by grants from the National Sciences and Engineering Research Council of Canada and the Canada Foundation for Innovation. 2 Address correspondence and reprint requests to Dr. Pedro O. de Campos-Lima, Laval University Cancer Research Center, McMahon St 9, Quebec City, QC, G1R 2J6, Canada. 3 Abbreviations used in this paper: IPA, isopropanol, isopropyl alcohol, 2-propanol; SEB, staphylococcal enterotoxin B.

Figures and Legends

Figure 2.1 - Biological effect of Isopropanol treatment in vitro.

(A) Isopropanol inhibits IL-2 production in peripheral blood T lymphocytes: Purified T cells were stimulated with anti-CD3/CD28 antibody-coated beads for 5 h in presence of 0.16%, 0.3%, 0.6% and 1.2% (w/v) isopropanol. The IL-2 concentration in the supernates was measured by ELISA and is depicted as means ± SEM in the black bar histogram. The cell viability is shown as means ± SEM in the gray bar histogram (** p < 0.01 relative to the st control group, n: 4). (B) Flow cytometric analysis of CFSE-labeled cells: one representative experiment of four is presented. Figure symbols: st indicates cell activation in absence of isopropanol. (-) represents the unstimulated control in absence of isopropanol. IPA indicates presence of 0.6% (w/v) isopropanol.

Figure 2.2 - Isopropanol and early TCR signaling

(A) Isopropanol treatment does not affect ZAP-70 phosphorylation following TCR activation: T cells were stimulated for 3 min with anti-CD3/anti-IgG in presence or absence of 0.6% (w/v) IPA and processed for SDS/PAGE. Western blots were probed with the monoclonal antibodies 17a (anti-human phospho-ZAP-70) and 99F2 (anti-human total ZAP-70). One representative blot of three is shown. (B) Relative quantitation in relation to total ZAP-70 from 3 Western blots is shown as mean densitometric units ± SEM. Figure symbols are as in fig. 1. ns: p > 0.05.

Figure 2.3 - Transcriptional effect of isopropanol.

(A) Jurkat-luc cells were stimulated with PMA/ionomycin in presence of 0.16%, 0.3%, 0.6%, and 1.2% (w/v) isopropanol for 5 h. Samples were lysed and assayed for luciferase activity. Results are presented as mean relative luciferase units/µg of protein ± SEM (n: 3). (B) Isopropanol inhibits NFAT nuclear translocation: T cells were stimulated with anti- CD3/CD28 antibody-coated beads for 5 h in presence of 0.6%, 1.2% and 1.6% (w/v) isopropanol. Nuclear extracts were incubated with immobilized NFAT-binding oligonucleotides in 96-well plates; the amount of retained transcription factor was assessed

50 with a NFAT-specific antibody by ELISA. Data is presented as mean optical density (OD) units ± SEM (n: 6). (C/D) NF-κB nuclear translocation is not affected by isopropanol: T cells were stimulated as described in panel B in presence of the indicated amounts of isopropanol. Nuclear extracts were incubated with p50- or p65-binding oligonucleotides in 96-well plates; the amount of bound transcription factor was assessed with a p50- or a p65-specific antibody by ELISA. Data is presented as mean OD units ± SEM (n: 3 for C; n: 3 for D). (E/F) Isopropanol inhibits AP-1 nuclear translocation: T cells were stimulated as described in panel B in presence of the indicated amounts of isopropanol. Nuclear extracts were incubated with c-Jun- or c-Fos-binding oligonucleotides in 96-well plates; the amount of bound transcription factor was assessed with a c-Jun- or a c-Fos-specific antibody by ELISA. Data is presented as mean OD units ± SEM (n: 3 for E; n: 3 for F). The dashed line in 3B, 3C/D and 3E/F represents the baseline of the relevant transcription factor nuclear content in unstimulated cells. ns: p > 0.05, * p < 0.05, ** p < 0.01 relative to the st control group. Stimulation in presence of inhibitor compounds are indicated as follows: cyclosporine A: CsA; BAY 11-7082: BAY; SP600125: inhib. (panel E); PD98059: inhib. (panel F); other symbols are as in fig. 1.

Figure 2.4 - Down-modulation of the effector function of human lymphocytes exposed to isopropanol in vitro.

(A) Isopropanol inhibits IFN-γ production in human peripheral blood T lymphocytes: Purified T cells were stimulated with anti-CD3/CD28 antibody-coated beads for 5 h in presence of 0.08%, 0.16%, 0.3%, 0.6% and 1.2% (w/v) isopropanol. The IFN- concentration in the supernates was measured by ELISA. Panel symbols: st indicates cell activation in absence of isopropanol. (-), represents the unstimulated control in absence of

isopropanol. Results are presented as means ± SEM (ns: p > 0.05, ** p < 0.01 relative to the st control group; n: 3). (B) Isopropanol inhibits IFN-γ production in human NK cells: NK cells were stimulated by poly (I:C) in presence of IL-2/IL-12 for 24 h. Isopropanol was used in different concentrations, 0.08%, 0.16%, 0.3%, 0.6% and 1.2% (w/v), as shown. The IFN-γ concentration in the supernates was measured by ELISA. Panel symbols: st indicates cell activation in absence of isopropanol. (-) represents the unstimulated control in absence of isopropanol (complete medium plus IL-2/IL-12 without poly (I:C)). Results are presented as means ± SEM (* p < 0.05, ** p < 0.01 relative to the st control group; n: 3). (C) Isopropanol decreases the cytotoxic activity of human T cells: The effector function of peripheral T cells against OKT3 hybridoma cells was analysed by Chromium-release assays in presence or absence of 0.6% (w/v) IPA as described in materials and methods. The effector/target ratios are indicated. Means ± SEM of untreated (grey columns) and IPA-treated (black columns) cells are shown in the bar histogram. The two means (treated versus untreated cells) were compared with the t test for each ratio: ns: p >0.05, ** p <

0.01, *** p < 0.001; n: 4. (D) Isopropanol reduces the cytotoxic activity of human NK cells. The cytolytic function of peripheral NK cells against K562 cells was measured by Chromium-release assays in presence or absence of 0.6% (w/v) IPA as described in materials and methods. Means ± SEM of untreated (grey columns) and IPA-treated (black columns) cells are shown. The effector/target ratios are indicated. The two means (treated versus untreated cells) were compared with the t test for each ratio: ns: p > 0.05, ** p < 0.01, *** p < 0.001; n: 4.

Figure 2.5 - Immunosuppressive effect of isopropanol in vivo.

(A) IL-2 production is compromised in mice acutely exposed to isopropanol: BALB/c mice received 5 µg SEB subcutaneously plus the intraperitoneal injection of either 2 g/kg isopropanol (SEB/IPA group) or saline (SEB/pbs group). Control groups received either

54 saline subcutaneously plus isopropanol intraperitoneally (pbs/IPA) or saline only (pbs/pbs). Animals were sacrificed 2 h after injections, and serum IL-2 levels were quantified by

ELISA. Results are presented as means ± SEM (* p < 0.05, ** p < 0.01 relative to the SEB/pbs group, n: 6/group). (B) IFN-γ production is compromised in mice acutely exposed to isopropanol: BALB/c mice were injected as above and the animals were sacrificed after 4 h. Serum IFN-γ was measured by ELISA. Means ± SEM are shown (** p < 0.01 relative to the SEB/pbs group, n: 6/group). (C) Isopropanol treatment in vivo inhibits the production of IFN-γ in vitro by CD4+ T cells: Purified CD4+ T cells were prepared from the spleens of animals that were treated (or not) with 2 g/kg isopropanol for 160 minutes before the sacrifice as described in materials and methods. The plotted results reflect the IFN-γ release after an 18-h culture period in vitro. SEB indicates that cells derived from animals that received only SEB (20 µg). SEB-IPA indicates that lymphocytes derived from mice that were injected with both SEB (iv) and isopropanol (ip). (-) indicates that cells came from animals that received neither SEB nor isopropanol. Results are presented as means ± SEM (** p < 0.01 relative to the SEB group, n: 5/group). (D) Isopropanol exposure in vivo inhibits IFN-γ release in vitro by CD8+ T lymphocytes: Purified CD8+ T cells were prepared from the spleens of animals that were treated (or not) with isopropanol as described in panel 5C. The IFN-γ release was measured in supernates by ELISA after an 18-h culture period in vitro. The symbols SEB, SEB-IPA, and (-) are as in panel 5C. Results are presented as means ± SEM (** p < 0.01 relative to the SEB group, n: 4/group). (E) Isopropanol protects mice from SEB-induced toxic shock: BALB/c mice were presensitized with 20 mg D-galactosamine; then, they were injected with 10 µg SEB subcutaneously plus 2 g/kg isopropanol intraperitoneally (SEB+IPA group). Alternatively, the presensitized animals were injected with 10 µg SEB subcutaneously plus saline intraperitoneally (SEB+pbs group). The Kaplan-Meier survival curve is presented (p < 0.0001, n: 12/group).

(F) TNF-α production is compromised in mice acutely exposed to isopropanol: Experimental groups were treated and labeled as in panel A. Animals were sacrificed 2 h after injections and serum TNF-α levels were quantified by ELISA. Results are presented as means ± SEM (** p < 0.01 relative to the SEB/pbs group, n: 6/group).

3. Methanol induces a discrete transcriptional dysregulation that leads to cytokine overproduction in activated lymphocytes

Olivier Désy,* Damien Carignan,* Manuel Caruso,* and Pedro O. de Campos-Lima*1 * Laval University Cancer Research Center, Quebec City, Quebec, G1R 2J6, Canada Toxicological Sciences, 2010 Oct;117(2):303-13

1 Correspondence: Dr. Pedro O. de Campos-Lima, Laval University Cancer Research Center, McMahon St 9, Quebec City, QC, G1R 2J6, Canada. Phone: 1 418 525 4444, fax: 1 418 691 5439, e-mail: [email protected]

Short title: Immunomodulation by methanol

Résumé

Le méthanol représente une cause importante d’intoxication aiguë aux alcools; c’est une substance ubiquitaire qui se retrouve autant à la maison que sur les lieux de travail. Il existe une littérature étoffée concernant l’impact immunologique de l’éthanol et, dans une moindre mesure, de l’isopropanol. Toutefois, on en connaît très peu à ce sujet sur le méthanol. En se basant sur la structure de ces molécules, nous avons émis l’hypothèse que le méthanol devait partager les propriétés immunodépressives des deux autres alcools à courte chaîne. Nous rapportons que le méthanol accroît la capacité proliférative des lymphocytes T et agit en synergie avec les stimuli activateurs pour augmenter la production de cytokines. Cet accroissement de la production de cytokines (IL-2, IFN-γ et TNF-α) par les cellules T a été observé in vitro à des concentrations de méthanol commençant aussi bas que 0,08% (25mM). Le méthanol n’a pas affecté la signalisation précoce en aval du récepteur à l’antigène, mais a promue l’activation sélective et différentielle d’un membre de la famille de facteurs de transcription NFAT, soit NFATc2. Ces résultats ont été corroborés dans un modèle murin d’intoxication aiguë au méthanol pour lequel on observe une augmentation de la libération de cytokines pro- inflammatoires dans le sérum. Ces données suggèrent donc que le méthanol a un effet immunologique qui lui est propre et d’une importance significative étant donné la présence répandue de ce solvant multi-usage parmi la population.

Abstract

Methanol is an important cause of acute alcohol intoxication; it is ubiquitously present at home and in the workplace. Although the existing literature provides a reasonable insight into the immunological impact of ethanol and to a much lesser extent of isopropanol, much less data is available on methanol. We hypothesized on structural grounds that methanol would share the immunosuppressive properties of the two other short-chain alcohols. We report here that methanol increases the proliferative capacity of human T lymphocytes and synergizes with the activating stimuli to augment cytokine production. The cytokine up- regulation was observed in vitro at methanol concentrations as low as 0.08% (25 mM) as measured by IL-2, IFN-, and TNF- release in T cells. Methanol did not affect the antigen receptor-mediated early signaling but promoted a selective and differential activation of the NFAT family of transcription factors. These results were further substantiated in a mouse model of acute methanol intoxication in which there was an augmented release of proinflammatory cytokines in the serum in response to the staphylococcal enterotoxin B. Our results suggest that methanol has a discrete immunological footprint of broad significance given the exposure of the general population to this multi-purpose solvent.

Keywords: immunomodulation, leukocyte, T lymphocyte, cytokine, methanol, immunotoxicology

Introduction

The use of short-chain alcohols is deeply ingrained into a variety of aspects of our everyday lives. These substances have a multitude of biological effects and a considerable body of evidence indicates that ethanol is capable of modulating the immune function mediated by T cells, monocytes, macrophages and neutrophils (Goral and Kovacs, 2005; Saeed et al., 2004; Szabo et al., 2007; Taieb et al., 2002). In addition, we have recently reported that isopropanol is detrimental to human T lymphocyte and NK cell activity and acute intoxication may lead to acute immunosuppression (Désy et al., 2008). Although the

59 existing literature on short-chain alcohols provides a reasonable insight into the immunological impact of ethanol and to a much lesser extent of isopropanol, much less data is available on other alcohols. Exposure of the general population to dietary and environmental sources of methanol, the simplest alcohol, usually does not amount to dangerous levels (Shelby et al., 2004). Yet, methanol exposure is an important cause of acute alcohol intoxication with more than 2,000 cases reported each year to poison centers in North America (Bronstein et al., 2008). Methanol is readily available to consumers as a component of several household solutions such as varnishes, paints, windshield washer fluids, antifreeze, and adhesives (Lanigan, 2001). The ingestion of as little as 6 mL of methanol can cause toxicity; on the other hand there are individuals who survived the intake of more than 500 mL of this alcohol (Brahmi et al., 2007; Hantson et al., 2000a; Martens et al., 1982). Regardless of the type of treatment provided, hemodialysis, alcohol dehydrogenase inhibition or substrate competition, some 15-36% of the intoxicated patients die (Brent et al.; 2001; Hunderi et al., 2004; Liu et al., 1998). In surviving patients, the blood alcohol concentration (BAC) recorded at hospital admission ranges broadly from 20- 1,290 mg/dL (0.02-1.29% or 6.24-402 mmol/L) (Brahmi et al., 2007; Brent et al.; 2001; Cowen et al., 1964; Gonda et al., 1978; Hantson et al., 2000a, 2005; Hovda et al., 2005; Hunderi et al., 2004; Kostic and Dart, 2003; Liu et al., 1998; Lushine et al., 2003; Martens et al., 1982; Verhelst et al., 2004; Wu et al., 1995). The highest methanol concentrations are often associated with permanent visual impairment. However, there are documented cases of treated patients who recover without visual sequelae despite BAC levels higher than 600 mg/dL (0.6% or 187 mmol/L) (Brent et al.; 2001; Lushine et al., 2003; Martens et al., 1982; Wu et al., 1995). Previous studies addressed the effect of methanol on the central nervous system, the sensory system, the gastrointestinal tract, renal function, metabolism, genetic stability, carcinogenesis, reproduction and development (Barceloux et al., 2002; Hantson and Mahieu, 2000b; Shelby et al., 2004). In addition, existing reports suggest a potential immunosuppressive activity of methanol in rats (Parthasarathy et al., 2007; Zabrodskii et al., 2008). However, the latter studies involved repeated exposure of the rodents to large doses of the alcohol over several days or weeks, a situation that is not likely to be met in

60 the clinical setting of acute intoxication. To our knowledge, no detailed analysis of the impact of methanol on the human immune system and of the underlying mechanisms of such potential effect is yet available. Given the structural similarity between methanol, ethanol and isopropanol, we hypothesized that methanol would share the immune modulatory properties of the two other alcohols and that these biological effects would be mediated through similar molecular mechanisms. We report here that methanol works in a different way by inducing the up-regulation of the proliferative capacity and effector function of human T lymphocytes in vitro in concentrations as low as 0.08% w/v (or 25 mM). We also show that a specific transcriptional dysregulation underlies the immunological impact of methanol. These results are further substantiated in a mouse model of acute methanol intoxication in which the production of proinflammatory cytokines is dysregulated. All concentrations tested in vitro and in vivo were within the clinically relevant range observed in the first hours of acute methanol intoxication. The full understanding of the pathophysiology of methanol poisoning, including the putative immune dysregulation, may be instrumental in rescuing patients that do not respond to therapy. Our results have broad significance taking into account the potential exposure of the general population to this ubiquitous chemical.

Materials and Methods

Cell isolation, culture, activation, and proliferation analysis. This study was approved by the Institutional Clinical Research Ethics Committee (L’Hôtel-Dieu de Québec/Centre hospitalier universitaire de Québec - L’HDQ-CHUQ). Mononuclear cells were prepared from the peripheral blood from healthy volunteers by density gradient centrifugation using Ficoll-Hypaque (GE Healthcare, Piscataway, NJ, USA). Written informed consent was obtained from all donors. More than 95% pure populations of human T cells (CD8+/CD4+) were obtained with EasySep® separation kits (StemCell Technologies, Vancouver, Canada). Cells were kept in complete medium: RPMI 1640 (Invitrogen Canada, Burlington, Canada) supplemented with 10% heat-inactivated FBS (BioCell Inc.,

Drummondville, Canada). Methanol (99.9% pure) was purchased from Fisher Scientific, Pittsburgh, PA, USA. In most experiments, T cells were activated with anti-CD3/CD28 antibody-coated magnetic beads (Invitrogen) at 37°C for 1 h or 5 h as specified in the text. When indicated, three alternative T cell activation protocols were used: a) pre-treatment for 20 min on ice with 1 g/mL mouse anti-human CD3 monoclonal antibody (CD3-2, Mabtech, Nacka Strand, Sweden), followed by washing and incubation at 37°C for 3 min with 10 g/mL goat anti- mouse IgG (Invitrogen); b) pre-treatment for 20 min on ice with 1 g/mL anti-human CD3 (CD3-2) and 5 g/mL anti-human CD28 (CD28.2, BioLegend, San Diego, CA, USA) mouse monoclonal antibodies, followed by washing and incubation at 37°C for 15 min with 10 g/mL anti-IgG; and c) treatment with 10 ng/mL phorbol 12-myristate 13-acetate (PMA) and 200 ng/mL ionomycin (Sigma, St Louis, MO, USA) for 5 h. Carboxyfluorescein diacetate succinimidyl ester (CFSE) staining: Freshly purified T cells were labeled with 10 M CFSE (Invitrogen) in PBS/1% FBS for 10 min at 37°C and further incubated in RPMI/10% FBS for 5 min on ice; then, cells were washed and activated with anti-CD3/CD28 antibody-coated beads for 5 h with or without 0.6% (w/v) methanol. The activating beads were magnetically removed; the cells were washed, and incubated for 72 h in 96-well plates (106 cells/mL) in complete medium without exogenous IL-2. FACS analysis of cell divisions and surface marker expression was performed on a XL flow cytometer (Beckman Coulter Inc., Miami, FL, USA).

Cytokine analysis. Measurements of human IL-2, IFN-γ, and TNF-α in cell culture supernatants and murine IL-2 and IFN-γ in serum samples were performed with specific cytokine enzyme-linked immunosorbent assay (ELISA) kits according to the manufacturer’s instructions (Mabtech). Murine TNF-α levels were measured with the mouse TNF-α ELISA MAX Standard Kit (BioLegend) following the protocol provided by the manufacturer.

ELISA-based transcription factor activation assay. Nuclear proteins were extracted using the Active Motif Nuclear Extract kit (Active Motif, Carlsbad, CA, USA) according to the

62 manufacturer's instructions; the total protein concentration of the lysates was determined by the Bradford assay (Bio-Rad, Hercules, CA, USA). The nuclear translocation of the nuclear factor of activated T cells (NFAT), activator protein-1 (AP-1), and nuclear factor-B (NF- B) was analysed as follows: NFATc1 activation was measured with the TransAM NFATc1 kit; NFATc2 activation was measured with a modified NFATc1 TransAM kit (the anti-NFATc2 4G6-G5 antibody was used to replace the original anti-NFATc1 primary antibody in the kit); c-Fos and c-Jun activation was measured with the TransAm AP-1 kit; and p50 and p65 activation was measured with the TransAm NF-B kit. ELISA-based TransAm kits were used according to the manufacturer’s instructions (Active Motif). Briefly, nuclear extracts were incubated with plate-bound transcription factor-specific oligonucleotides; the plates were washed, and further incubated with transcription factor- specific antibodies. Addition of a horseradish-conjugated secondary antibody and the 3,3',5,5'-tetramethylbenzidine substrate produced a colorimetric reaction measurable in a spectrophotometer. Inhibitor compounds (Sigma) were used as follows: cyclosporin A: 1 g/mL; BAY 11- 7082: 5 M; SP600125: 25 M; SB202190: 10 M; and PD98059: 50 M.

Electrophoretic mobility shift assay (EMSA). Purified T cells were stimulated with anti- CD3/CD28 antibody-coated beads for 1 h in presence of the indicated amounts of methanol. Nuclear extracts were prepared with the Active Motif Extraction kit and used for DNA-binding analysis with a double-stranded oligonucleotide carrying the NFAT recognition sequence from the distal ARRE-2 in the human IL-2 promoter (5’- GGAGGAAAAACTGTTTCATAGAAGGCGT-3’). The EMSA probe was end-labeled with [-32P] ATP by T4 polynucleotide kinase treatment. Binding reactions were performed with 5 g of nuclear protein in 10 mM Tris-HCl (pH 7.5), 1 mM MgCl2, 0.5 mM EDTA, 0.5 mM DTT, 50 mM NaCl, 6% glycerol, 1 g BSA and 0.2 g poly(dI)-poly(dC) at room temperature for 20 min. Protein-DNA complexes were separated in a 6% nondenaturing polyacrylamide gel. Supershift assays were performed by preincubating the samples on ice with antibodies against NFATc1 (7A6) or c2 (G1-D10) (Santa Cruz Biotechnology, Santa Cruz, CA, USA) before the addition of the labeled probe.

Western blot and luciferase assay. Western blot: Purified T lymphocytes or Jurkat cells were activated as mentioned above for 3 minutes (ZAP-70) or 15 minutes (NFAT) in presence or absence of 0.6% (w/v) methanol. Samples for ZAP-70 analysis were lysed in SDS buffer (Désy et al., 2008) and those for NFAT analysis were prepared with the Nuclear Extract kit (Active Motif). Proteins were separated by SDS-PAGE and blotted as described (Désy et al., 2008). Primary antibodies for ZAP-70 were the rabbit anti-human total ZAP-70 (99F2, 1/1000, Cell Signaling Technology, Danvers, MA, USA) and the mouse anti-human phospho-ZAP-70 (17a, 1/5000, BD Biosciences, Mississauga, Canada); In the case of NFAT analysis, the primary antibodies were the mouse anti-human NFATc2 (4G6-G5, 1/200, Santa Cruz Biotechnology) and the control rabbit anti-HDAC-1 (H-51, 1/200, Santa Cruz Biotechnology). Secondary antibodies were the IRDye 800CW goat anti- rabbit IgG and/or IRDye 680 goat anti-mouse IgG (1/20000, Li-Cor Biosciences, Lincoln, NE, USA). Detection and quantification was performed with the Odyssey Infrared Imaging System (Li-Cor Biosciences). Luciferase assay: The generation of Jurkat cells carrying the firefly luciferase gene driven by the NFAT synthetic promoter is described elsewhere (Désy et al., 2008). Jurkat- luciferase cells were stimulated with PMA/ionomycin with or without methanol treatment as indicated in the text. Lysates for luciferase assays were prepared with the passive lysis buffer (E1941, Promega, Madison, WI, USA) and analysed in a Lumat 9501 luminometer (Berthold, Nashua, NH, USA). Relative luciferase units were calculated in relation to the unstimulated negative control after normalization to total protein content measured by the Bradford assay (Bio-Rad).

In vivo studies. 7-13-week-old female BALB/c mice were bought from The Jackson Laboratory (Bar Harbor, ME, USA). All tests respected the ethical guidelines set by the Institutional Animal Protection Committee (CPA-CHUQ). Food and water were provided ad libitum. Animals received subcutaneously 5 g of staphylococcal enterotoxin B (SEB, Toxin Technology Inc., Sarasota, FL, USA) for cytokine induction and were sacrificed by

CO2 asphyxiation 4 h after administration for IL-2, IFN-, and TNF- serum analysis.

Methanol was injected intraperitoneally (2 g/kg). The blood alcohol concentration (BAC) was determined by gas chromatography with a 3900 GC unit (Varian, Palo Alto, CA, USA).

Statistical analysis. One-way ANOVA followed by Dunnett’s multiple-comparison posttest was performed with GraphPad Prism (GraphPad Software Inc., San Diego, CA, USA) on data presented in all figures. p values < 0.05 were considered significant.

Results

The immune modulatory effect of methanol is distinct from that of other short-chain alcohols We have previously reported that isopropanol down-regulates the production of IL-2 in

vitro and in vivo (Désy et al., 2008) and our expectation was that the closely related methanol would exhibit a similar dampening effect on lymphocytes. In the present study, we have investigated first whether methanol exposure in vitro would have any impact on the ability of human peripheral T lymphocytes to produce IL-2 once activated by antibody cross-linking of the T cell receptor (TCR). Although this cytokine peaked at 24-48h post- stimulation it was readily measurable within 5 h (mean: 1.15 ng/mL ± 0.07 SEM, n: 6). Unlike our earlier results with isopropanol, activated T cells produced more IL-2 when treated with methanol as measured in the culture supernatants by ELISA (fig. 3.1A, black bars). The increase in cytokine production was observed at methanol doses as low as 0.08%. The observed effect was decoupled from any nonspecific cytotoxicity as the cell viability of methanol-treated samples in the concentration range that up-regulated IL-2 by 27-70% was similar to that of untreated control cells (fig. 3.1A, gray bars). Methanol in absence of stimulation did not have any effect. Given the importance of the IL-2 autocrine loop for the expansion of antigen-specific cells in vivo, we have asked if the increased release of IL-2 translated into a higher proliferative capacity of the activated lymphocytes. Purified peripheral T cells were labeled with CFSE and activated for 5 h; then, they were washed and further incubated for 72 h before analysis

65 by flow cytometry. Figure 1B shows that although the 5-h TCR cross-linking led to 1-4 divisions in about 20% of the cells in absence of exogenous IL-2 (mean: 21.35% ± 0.71 SEM, n: 3), cell proliferation was more pronounced if methanol was present during the 5-h stimulation (mean: 32.82% ± 1.31 SEM, n: 3). Significant differences between treated and untreated samples were also found when cells were analysed according to the number of divisions completed by the end of the experiment (fig. 3.1B).

Methanol does not compromise early signaling following TCR activation of lymphocytes Methanol could act on the cell membrane in a way reminiscent of the effect of ethanol on ion channels and neurotransmitter receptors (Arevalo et al., 2008; Aryal et al., 2009). It could conceivably interact with the T cell receptor directly, thereby amplifying antigen- dependent signal transduction in lymphocytes. To address this possibility, we have examined the phosphorylation status of ZAP-70, a key downstream molecule that is activated soon after TCR engagement. Figure 2A shows that signaling through the T cell receptor itself was not affected by methanol because the ZAP-70 activation proceeded as efficiently as in untreated cells following anti-CD3 antibody cross-linking.

IL-2 up-regulation does not involve the transcriptional pathways affected by other short-chain alcohols Since methanol did not present an obvious impact on early TCR signaling, the possibility of a positive regulation of IL-2 transcription by the alcohol was examined. A precedent in favor of this scenario was set by earlier studies which revealed that ethanol, and more recently isopropanol, mediate their impact on the immune system through the reduction of the nuclear translocation of transcription factors capable of binding the promoter regions of key cytokine genes (Désy et al., 2008; Mandrekar et al., 2007; Saeed et al., 2004; Szabo et al., 2007). In order to dissect the relative impact of methanol on transcription factors relevant to IL-2 production, we have measured the nuclear translocation of NF-B (p50/p65) and AP-1 (c-Jun/c-Fos) in TCR-stimulated purified human T cells exposed to different concentrations of methanol. Panels B/C in fig. 3.2 show that lymphocyte

66 activation by anti-CD3/CD28 antibodies led to 1.7 and 2.7-fold nuclear increase in the amount of NF-B, p50 and p65, respectively. However, the activation of NF-B remained unaffected by methanol treatment at all tested concentrations. The compound BAY 11- 7082, an inhibitor of IB phosphorylation, reduced the nuclear content of p50/p65 in the same experiments to levels lower than those of the unstimulated control. AP-1 followed the same pattern observed for NF-B with increase in the amount of c-Jun (10-fold) and c-Fos (20-fold) in the nucleus following activation by anti-CD3/CD28 antibodies (fig. 3.2D/E). None of the examined concentrations of methanol had an effect on AP-1 c-Jun/c-Fos nuclear content. Instead, lymphocyte stimulation in presence of the c-Jun N-terminal kinase inhibitor SP600125 or the combination of inhibitor drugs PD98059 (MEK1)/SB202190 (p38) produced about 50% of the maximal c-Jun/c-Fos nuclear content (st).

Methanol increases IL-2 production in activated T cells via differential modulation of NFAT family members Given the apparent lack of effect of methanol on the pathways leading to the activation of NF-B and AP-1, we have turned our attention to the NFAT family of transcription factors. NFAT synergizes with NF-B and AP-1 to promote the maximal activation of the IL-2 gene (Attema et al., 2002). First, we have tested the impact of different concentrations of methanol on a stable Jurkat subline carrying a firefly luciferase gene that is highly responsive to the Ca++/NFAT signaling pathway. This reporter gene is driven by a synthetic minimal IL-2 promoter containing three copies of the human distal IL-2 NFAT binding site (Désy et al., 2008; Durand et al., 1988; Shaw et al., 1988). Methanol was capable of augmenting the luciferase activity triggered by PMA/ionomycin in these cells in a dose- dependent manner (fig. 3.3A). Methanol concentrations as low as 0.3% had a significant enhancement effect on IL-2 transcription as indicated by a 52% increase in luciferase activity in relation to the untreated positive control (st). The up-regulation reached 81% at the highest dose. Similar results were obtained by anti-CD3 antibody cross-linking in Jurkat-luciferase cells (data not shown). The interference of methanol with NFAT activation that was suggested by the results of the reporter assays was corroborated by data obtained from electrophoretic mobility shift

67 assays. Nuclear extracts were prepared from human peripheral blood T lymphocytes that had been activated or not in presence of methanol. Figure 3.3B shows one representative experiment in which there was more binding of the probe to NFAT in the nuclear extracts from methanol-treated samples in a dose dependent fashion. The percentage increase in binding in relation to the untreated positive control (st) was measured by densitometry in three independent experiments and plotted in the associated histogram. A supershift analysis with antibodies against NFATc1 and NFATc2 is also presented in figure 3.3B. The figure shows that in our experimental conditions, one hour-stimulation with anti- CD3/CD28 antibodies, both NFAT proteins are available in the nucleus to form complexes on the target promoter regions. NFATc1 and c2 are highly homologous in their DNA-binding domains and are capable of recognizing the same nucleotide sequences. The similarity between these transcription factors is also reflected in the distribution of calcineurin docking sites and phosphorylated serine-rich motifs (Macian, 2005; Srinivasan and Frauwirth, 2007). Thus, we predicted that the up-regulatory transcriptional effect of methanol would extend to both NFAT family members given their overall structural similarity. The results of the transcription factor activation assays presented in figures 3.3C and 3.3D disprove that hypothesis. Although T cell activation increased the nuclear content of NFATc1 1.6-fold, there was no additional change when methanol was added even at the highest concentration (fig. 3.3C). T cell activation also led to a 1.7-fold increase in NFATc2 nuclear content but activation in presence of methanol raised the NFATc2 nuclear content 2.2-fold (29% more than activation alone) at the lowest concentration and 2.5-fold (47% more than activation alone) at the highest concentration (fig. 3.3D). The calcineurin inhibitor cyclosporine A was used as a control in the same stimulatory conditions and reduced NFATc1 and c2 to about 50% of the nuclear content of resting T cells. The methanol-induced effect on NFATc2 was corroborated by western blot analysis of the nuclear fraction of activated T cells which showed a 27% increase in the dephosphorylated form of the transcription factor in alcohol- treated samples (fig. 3.3E).

Release of proinflammatory cytokines by human peripheral T lymphocytes is up- regulated by methanol The DNA consensus motif 5´-T/AGGAAA-3´ recognized by NFAT is found in the 5’ region of several genes encoding proteins of immunological relevance (Macian, 2005; Porter and Clipstone, 2002). Thus, our finding that methanol increases IL-2 production through the dysregulation of NFATc2 could herald a wider impact of this alcohol on the immune system. The potentially extended immune modulatory effect of methanol was examined by checking two other cytokines that are central to the immune response and that are known to be inducible by the Ca++/NFAT signaling pathway: TNF- and IFN- (Macian, 2005; Porter and Clipstone, 2002; Tsytsykova and Goldfeld, 2000). Peripheral T lymphocytes were activated with anti-CD3/CD28 antibodies in presence of different methanol dilutions. Similarly to IL-2, TNF- and IFN- peaked at later times but were already measurable 5 h after stimulation (mean: 0.65 ng/mL ± 0.04 SEM, n: 3, for TNF- and 3.93 ng/mL ± 0.41 SEM, n: 4, for IFN-. The cytokine release increased in a dose- dependent fashion in methanol-exposed cells, with an average augmentation ranging from 33% to 133% for TNF- (fig. 4A) and 32% to 86% for IFN- (fig. 3.4B). Next, we examined whether the immune modulatory impact of methanol was equally extended to different T cell subsets by looking at their IFN- production. Peripheral T lymphocytes were separated in CD8+ (fig. 3.4C) and CD4+ (fig. 3.4D) subpopulations and activated with anti-CD3/CD28 antibodies in presence or absence of methanol. Figure 3.4C shows that methanol concentrations as low as 0.08% possessed an immune modulatory effect. Treatment with 0.08%, 0.16% and 0.3% methanol up-regulated the IFN- release by stimulated CD8+ cells relative to the untreated positive control (st) by 41%, 51% and 67%, respectively. The highest methanol concentration almost doubled the production of the cytokine in stimulated cells. A similar trend was observed with CD4+ lymphocytes: methanol concentrations of 0.16%, 0.3% and 0.6% raised the IFN- release above the level achieved by activation alone by 45%, 59% and 92%, respectively (Fig. 3.4D). Lymphocytes were > 97% viable at all tested methanol concentrations (data not shown and fig. 3.1A).

Methanol up-regulates the production of IL-2 and proinflammatory cytokines in vivo After having demonstrated the up-regulatory impact of methanol treatment on IL-2, TNF- and IFN- production by lymphocytes in vitro, we examined the relevance of these findings in a mouse model of acute methanol intoxication. Mice were administered methanol intraperitoneally, 2 g/kg, to generate a mean blood alcohol concentration of 238 mg/dL after 30 min (238.4 ± 13.7 SEM, n: 3). Cytokine production was triggered in vivo by subcutaneous injection of the superantigen staphylococcal enterotoxin B. In these conditions, T lymphocytes with the relevant TCR V chains undergo transient activation, cell proliferation, and begin massive cytokine production (Tsytsykova and Goldfeld, 2000). As anticipated, SEB alone induced IL-2 (18.8 ng/mL, fig. 3.5A), IFN- (1.1 ng/mL, fig. 3.5B), and TNF- (43.3 pg/mL, fig. 3.5C). Methanol treatment enhanced the SEB- stimulated cytokine production substantially: 33% for IL-2/IFN- and 47% for TNF-. Serum levels of 25.1 ng/mL IL-2, 1.5 ng/mL IFN-, and 64.1 pg/mL TNF- were detected 4 h after SEB/methanol administration. The differences in cytokine serum levels between animals treated and untreated with methanol were statistically significant as indicated in figure 3.5.

Discussion

Short-chain alcohols have the ability to partition into cell membranes and to denature proteins by promoting the formation of -helices and/or by disrupting tertiary structures; these effects are largely nonspecific and are typically observed at high concentrations (> 500 mM) (Dwyer and Bradley, 2000). At more physiologically relevant concentrations, alcohols have been shown to induce loss of function of specific proteins such as: ion channels, neurotransmitter receptors, enzymes, and adhesion molecules (Aryal et al., 2009; Jung et al., 2005; Ren et al., 2003; Shahidullah et al., 2003). Structural and biophysical data suggest that binding to the target proteins occurs at discrete sites that are constituted by hydrophobic pockets lined by nonpolar amino acids (Aryal et al., 2009; Dwyer and

Bradley, 2000; Jung et al., 2005; Ren et al., 2003; Shahidullah et al., 2003). These interactions would ultimately produce a local distortion and alteration in protein function.

Many of the biological effects of ethanol on the immune system have been attributed to reduced nuclear translocation of NF-B (Saeed et al., 2004; Szabo et al., 2007). In addition to ethanol, isopropanol is the only other alcohol that has been studied with regard to the transcriptional regulation of its immune modulatory properties (Désy et al., 2008). Isopropanol differs from ethanol in that it does not compromise NF-B but reduces the nuclear content of the transcription factors AP-1 and NFAT. We have assumed on the grounds of structural similarity to the above alcohols that methanol might possess an acute immunosuppressive effect possibly involving an alteration of similar activation pathways. Instead, our results revealed that methanol synergizes with the activating stimuli and augments the T cell cytokine production. Methanol per se did not induce any effect in our experiments indicating that its target(s) must be pre-activated (figs. 3.1, 3.3, 3.4, 3.5 and data not shown). The source of T cell activation in healthy humans may derive from the exposure to environmental non-self antigens or from the reactivation of persistent infection by common agents such as Epstein-Barr virus that often lead to clonal T cell expansion (de Campos-Lima et al., 1997; Lalonde et al., 2007). Antigen-independent activation of T cells also occurs in physiological conditions and is important to maintain the T cell repertoire in the steady state (Goldrath et al., 2002). It is noteworthy that the inebriation that follows the excessive consumption of any alcohol, including methanol, is frequently associated to trauma and, thus, may increase the risk of infection and pathogen-driven T cell activation (Fitzgerald et al., 2007; Saxena et al., 1987). The cytokine release up-regulation reported here was observed in vitro at methanol concentrations as low as 0.08% (25 mM) as measured by IL-2, TNF- and IFN- release in T cells. These concentrations are equal to or lower than those of short-chain alcohols used in previous studies that reported a statistically significant impact on immune cells (Goral and Kovacs, 2005; Saeed et al., 2004; Szabo et al., 2007; Taieb et al., 2002). Another aspect that distinguishes the methanol effect is the lack of interference with the pathways leading to the activation of NF-B and AP-1. Methanol does affect the NFAT nuclear content but very differently from isopropanol.

Although the NFATc1 activation remained unaffected even at the highest doses tested, methanol induced an enhancement of the NFATc2 nuclear content following TCR engagement. Conceivably, alcohols could modulate the immune cell activation by interfering with several steps along the cascade that relay the signal generated by the recognition of an antigenic ligand on the cell membrane to the nucleus. Alcohols could partition into the cell membrane altering the capacity of micro-domains to recruit and/or retain molecules involved in signaling, thus compromising the formation of the immunological synapse. The inhibitory effect of ethanol on the LPS-induced Toll-like receptor 4 (TLR4) signaling in macrophages has been suggested to unfold along these lines (Szabo et al., 2007). Methanol could affect surface molecules, such as the TCR, directly by inducing conformational changes or, indirectly, by disrupting lipid-protein interactions. We tested this possibility by measuring ZAP-70 phosphorylation in methanol-treated samples. This tyrosine-kinase has a critical role in signaling downstream of the TCR; it binds to the CD3-zeta and phosphorylates the key adaptor protein LAT (Smith-Garvin et al., 2009). Our results indicate that methanol acts downstream of the cell membrane as ZAP-70 activation proceeds normally following TCR triggering. It is still possible however, that methanol could affect lipid rafts at higher concentrations not examined in this study. The phosphorylation cascade downstream of ZAP-70 leads to an increase in intracellular Ca++ (Smith-Garvin et al., 2009); the phosphatase calcineurin senses the elevated Ca++ levels and, in presence of calmodulin, activates the cytoplasmic moieties of the NFAT transcription factors thereby allowing their migration into the nucleus (Smith- Garvin et al., 2009). Ethanol and longer alcohols have been shown to bind calmodulin at discrete sites enhancing its affinity for Ca++ and changing its impact on calcineurin function (Ohashi et al., 2004). Methanol could conceivably change the phosphatase activity of calcineurin perhaps indirectly via calmodulin binding. We have tested this hypothesis by measuring the calmodulin/calcineurin activity in activated cells and found it to be preserved in presence of methanol (data not shown). In addition, we have observed that the calcineurin inhibitor cyclosporine A overrides the alcohol-induced NFATc2 up-regulation and the consequent increase in cytokine release (data not shown).

Methanol could also influence the shuttling of NFATc2 into the nucleus by interfering with the function of importin 1. However, one would expect the closely related NFATc1 to be affected; a more general impact extended to other transcription factors that bind importin 1, such as c-Fos, would be anticipated as well (Malnou et al., 2007). The similarity of the post-stimulation nuclear content of c-Fos and NFATc1 in methanol-treated and untreated samples does not support this contention. By the same token, methanol could interfere with the activity of maintenance kinases that keep the NFAT in the cytoplasm in resting T cells, such as casein kinase 1 (CK1). However, the differential modulation of NFATc2 is unlikely to result from the dysregulation of the activity of CK1 because this enzyme recognizes the serine-rich motif SRR1 in both NFATc1 and c2 (Macian, 2005; Okamura et al., 2004; Srinivasan and Frauwirth, 2007). Another possibility is that the alcohol inhibits the nuclear export of the NFAT proteins. Constitutive kinases, such as the glycogen synthase kinase 3 (GSK3) and CK1, as well as inducible kinases, such as p38 and the c-Jun N terminal kinase (JNK), have been implicated in the rephosphorylation of the NFAT proteins that enables their export from the nucleus (Macian, 2005; Okamura et al., 2004; Srinivasan and Frauwirth, 2007). GSK3 and CK1 act on both NFAT c1/c2 nuclear factors making these enzymes unlikely alcohol targets (Beals et al., 1997; Macian, 2005; Okamura et al., 2004; Srinivasan and Frauwirth, 2007). The same reasoning applies to the recently identified dual-specificity tyrosine phosphorylation regulated kinase (DYRK) 1A, a priming kinase that allows additional phosphorylation by GSK3 and CK1 of both NFATc1/c2 nuclear factors (Arron et al., 2006; Gwack et al., 2006). As for the inducible kinases, JNK specifically targets the nuclear factor that is unchanged by methanol treatment, NFATc1 (Chow et al., 2000). On the other hand, p38 specifically phosphorylates NFATc2 at the first serine of the SRR1 motif (Gómez del Arco et al., 2000); methanol could conceivably induce an augmented retention of this nuclear factor by inhibiting p38. Nevertheless p38 post-activation expression, phosphorylation pattern, and kinase activity were not different in alcohol-treated and untreated samples (data not shown). We favor the possibility that methanol interacts directly with NFATc2 perhaps inducing a conformational change that masks the nuclear-export signal or reduces the accessibility of phosphorylation sites (Fig. 3.6). This matter will be addressed in future studies.

Most of the acute toxicity of methanol has been attributed to its liver-produced formate metabolite (Barceloux et al., 2002). In this work, we have focused on the direct immune toxicity of the parent compound in vitro. In addition, we have used an in vivo model that largely excludes the metabolite components of the acute intoxication. Rodents break down formate very efficiently and even alcohol doses substantially higher than the one we have used in the in vivo experiments are not sufficient to induce formate accumulation (Lanigan, 2001; Shelby et al., 2004). Thus the experimental dose and timeframe adopted in our animal model facilitate the analysis of the immune impact attributable to methanol during the first hours of acute intoxication. To put in perspective the relevance of the alcohol concentrations chosen for our studies, one should consider a few aspects of the methanol pharmacokinetics in the context of acute intoxication. Methanol is readily absorbed by the human gastrointestinal tract with peak blood levels achieved at 30-60 min after ingestion (Barceloux et al., 2002). As methanol per se does not have a very conspicuous clinical footprint and causes only a transient mild inebriation, the patients usually seek help much later after ingestion when the formate concentration builds up causing metabolic acidosis, visual disturbances, severe central nervous system depression and gastrointestinal symptoms (Barceloux et al., 2002; Liu et al., 1998). The blood methanol concentration upon presentation at the hospital varies; the mean value was 179 mg/dL (0.179% or 56 mmol/L) in a large retrospective study conducted in the metropolitan Toronto area and 170 mg/dL (0.17% or 53 mmol/L) in the patients who entered the prospective trial of fomepizole for the treatment of methanol poisoning (Brent et al., 2001; Liu et al., 1998). These concentrations measured at the hospital are much lower than the peak values in the first hour post-ingestion because it may take up to two days for the patients to look for professional help. In saturating conditions of acute intoxication, methanol liver clearance follows zero-order kinetics with an estimated rate of 8.5 mg/dL/h (Jacobsen et al., 1988). Thus, a patient who presents at 48 h post- ingestion a methanol blood level just barely above the accepted treatment threshold of 20 mg/dL (0.02% or 6.24 mmol/L) would have had an earlier peak concentration of 420 mg/dL (0.42% or 131 mmol/L). This analysis is in line with a recent retrospective study of 173 poisoned patients for whom time-related blood alcohol data were available (Kostic and

Dart, 2003); the individuals who presented themselves 24-48 h after ingestion had a median blood concentration of 110 mg/dL (0.11% or 34 mmol/L) as opposed to those few who sought medical attention within 6 h after ingestion and had a median methanol concentration of 300 mg/dL (0.3% or 94 mmol/L). Therefore, all methanol concentrations in our in vitro experiments, including the highest concentration used, are clinically relevant as they can either be measured upon hospital admission in surviving patients or can be inferred to have occurred during the first hours post-ingestion (Brahmi et al., 2007; Brent et al., 2001; Gonda et al., 1978; Hantson et al., 2000a; Kostic and Dart, 2003; Lushine et al., 2003; Martens et al., 1982; Verhelst et al., 2004; Wu et al., 1995). In our in vivo model, we have injected mice intraperitoneally with 2 g/kg methanol to generate a blood alcohol concentration of 238 mg/dL (0.238% or 74 mmol/L) after 30 min; in clinical terms, this time-related concentration compares to the peak methanol level in humans who present at 24 h post-ingestion a blood concentration of 42.5 mg/dL (0.04% or 13 mmol/L) (Jacobsen et al., 1988; Kostic and Dart, 2003). In addition, the dose we have chosen is well within the concentration range that we have shown to be biologically active in vitro (starting at 0.08% or 25 mM). Our results indicate that during this state of acute intoxication the animals had an enhanced immune reaction as judged by their ability to release IL-2, IFN- and TNF- in the serum in response to SEB. One very peculiar aspect of methanol poisoning is the somewhat selective toxic injury inflicted on the basal ganglia, particularly the putamina, which may lead to Parkinsonism

(Barceloux et al., 2002; Reddy et al., 2007). There is evidence implicating the proinflammatory cytokine tumor necrosis factor in the pathogenesis of the most common cause of Parkinsonism, Parkinson’s disease (McCoy et al., 2008; Mogi et al., 1994). Our data show that NFATc2 is up-regulated in methanol-treated cells; this transcription factor is capable of binding the promoter regions of multiple cytokine genes (Porter and Clipstone, 2002), including TNF-. Indeed, we have found that methanol up-regulates the TNF- production both in vitro and in vivo (fig. 3.4 and 3.5). These results feed the speculation that the methanol toxicity on the basal ganglia could be initiated by a conducive proinflammatory cytokine environment generated in the early phase of acute intoxication. By the same token, the up-regulation of proinflammatory cytokines could play a role in the

75 pathophysiology of methanol-induced acute pancreatitis, a condition that has been overlooked so far but that may be identified in half of the patients with severe acute intoxication (Hantson and Mahieu, 2000b). These are tantalizing possibilities well worth investigating in future studies. If confirmed, they may provide the basis for the establishment of anti-inflammatory measures to counteract the proinflammatory properties of methanol in acutely poisoned patients. To our knowledge this is the first in depth characterization of the immune toxicity of methanol in human lymphocytes. Our findings are directly relevant in the context of acute methanol intoxication and constitute a rationale for the inclusion of immunological endpoints into the design of future studies to address methanol exposure.

Funding This work was supported by grants from the National Sciences and Engineering Research Council of Canada (327062-07) and the Canada Foundation for Innovation (4087) to P.O.d.C.-L.

Acknowledgements The authors wish to thank: G. Crabtree for the NFAT promoter; L. Nadeau for gas chromatography; D. Richard for access to the infrared imaging system; D. L’Héreault and H. Dombrowski for blood collection; J. Charron and M. Vauzelle for helpful discussions; A. Julien and R. Boutet for valuable technical assistance. The authors declare no financial or commercial conflict of interest.

References

Arevalo, E., Shanmugasundararaj, S., Wilkemeyer, M. F., Dou, X., Chen, S., Charness, M. E., and Miller, K. W. (2008). An alcohol binding site on the neural cell adhesion molecule L1. Proc. Natl. Acad. Sci. U. S. A. 105, 371-375.

Arron, J. R., Winslow, M. M., Polleri, A., Chang, C. P., Wu, H., Gao, X., Neilson, J. R., Chen, L., Heit, J. J., Kim, S. K., Yamasaki, N., Miyakawa, T., Francke, U., Graef, I. A., and Crabtree, G. R. (2006). NFAT dysregulation by increased dosage of DSCR1 and DYRK1A on chromosome 21. Nature 441, 595-600. Aryal, P., Dvir, H., Choe, S., and Slesinger, P. A. (2009). A discrete alcohol pocket involved in GIRK channel activation. Nat. Neurosci. 12, 988-995. Attema, J. L., Reeves, R., Murray, V., Levichkin, I., Temple, M. D., Tremethick, D. J., and Shannon, M. F. (2002). The human IL-2 gene promoter can assemble a positioned nucleosome that becomes remodeled upon T cell activation. J. Immunol. 169, 2466-2476. Barceloux, D.G., Bond, G. R., Krenzelok, E. P., Cooper, H., and Vale, J. A. (2002). American Academy of Clinical Toxicology practice guidelines on the treatment of methanol poisoning. J. Toxicol. Clin. Toxicol. 40, 415-446. Beals, C. R., Sheridan, C. M., Turck, C. W., Gardner, P., and Crabtree, G. R. (1997). Nuclear export of NF-ATc enhanced by glycogen synthase kinase-3. Science 275, 1930- 1934. Brahmi, N., Blel, Y., Abidi, N., Kouraichi, N., Thabet, H., Hedhili, A., and Amamou, M. (2007). Methanol poisoning in Tunisia: report of 16 cases. Clin. Toxicol. 45, 717-720. Brent, J., McMartin, K., Phillips, S., Aaron, C., and Kulig, K. (2001). Fomepizole for the treatment of methanol poisoning. N. Engl. J. Med. 344, 424-429. Bronstein, A. C., Spyker, D. A., Cantilena Jr, L. R., Green, J. L., Rumack, B. H., and Heard, S. E. (2008). 2007 Annual Report of the American Association of Poison Control Centers' National Poison Data System (NPDS): 25th Annual Report. Clin. Toxicol. 46, 927-1057. Chow, C. W., Dong, C., Flavell, R. A., and Davis, R. J. (2000). c-Jun NH(2)-terminal kinase inhibits targeting of the protein phosphatase calcineurin to NFATc1. Mol. Cell. Biol. 20, 5227-5234. Cowen, D. L. (1964). Extracorporeal dialysis in methanol poisoning. Ann. Intern. Med. 61, 134-135.

77 de Campos-Lima, P. O., Levitsky, V., Imreh, M., Gavioli, R., and Masucci, M. G. (1997). Epitope-dependent selection of highly restricted or diverse T-cell repertoires in response to persistent infection by Epstein-Barr virus. J. Exp. Med. 186, 83-89. Désy, O., Carignan, D., Caruso, M., and de Campos-Lima, P. O. (2008). Immunosuppressive effect of isopropanol: down-regulation of cytokine production results from the alteration of discrete transcriptional pathways in activated lymphocytes. J. Immunol. 181, 2348-2355. Durand, D. B., Shaw, J. P., Bush, M. R., Replogle, R. E., Belagaje, R., and Crabtree, G. R. (1988). Characterization of antigen receptor response elements within the interleukin-2 enhancer. Mol. Cell. Biol. 8, 1715-1724. Dwyer, D. S., and Bradley, R. J. (2000). Chemical properties of alcohols and their protein binding sites. Cell. Mol. Life Sci. 57, 265-275. Fitzgerald, D. J., Radek, K. A., Chaar, M., Faunce, D. E., DiPietro, L. A., and Kovacs E. J. (2007). Effects of acute ethanol exposure on the early inflammatory response after excisional injury. Alcohol Clin. Exp. Res. 31, 317-323. Goldrath, A. W., Sivakumar, P. V., Glaccum, M., Kennedy, M. K., Bevan, M. J., Benoist, C., Mathis, D., and Butz, E. A. (2002). Cytokine requirements for acute and Basal homeostatic proliferation of naive and memory CD8+ T cells. J. Exp. Med. 195, 1515- 1522. Gómez del Arco, P., Martínez-Martínez, S., Maldonado, J. L., Ortega-Pérez, I., and Redondo, J. M. (2000). A role for the p38 MAP kinase pathway in the nuclear shuttling of NFATp. J. Biol. Chem. 275, 13872-13878. Gonda, A., Gault, H., Churchill, D., and Hollomby, D. (1978). Hemodialysis for methanol intoxication. Am. J. Med. 64, 749-758. Goral, J., and Kovacs, E. J. (2005). In vivo ethanol exposure down-regulates TLR2, TLR4, and TLR9-mediated macrophage inflammatory response by limiting p38 and ERK1/2 activation. J. Immunol. 174, 456-463. Gwack, Y., Sharma, S., Nardone, J., Tanasa, B., Iuga, A., Srikanth, S., Okamura, H., Bolton, D., Feske, S., Hogan, P. G., and Rao, A. (2006). A genome-wide Drosophila RNAi screen identifies DYRK-family kinases as regulators of NFAT. Nature 441, 646-650.

Hantson, P., Haufroid, V., and Mahieu, P. (2000a). Survival with extremely high blood methanol concentration. Eur. J. Emerg. Med. 7, 237-240. Hantson, P., and Mahieu, P. (2000b). Pancreatic injury following acute methanol poisoning. J. Toxicol. Clin. Toxicol. 38, 297-303. Hantson, P., Haufroid, V., and Wallemacq, P. (2005). Formate kinetics in methanol poisoning. Hum. Exp. Toxicol. 24, 55-59. Hovda, K. E., Hunderi, O. H., Tafjord, A. B., Dunlop, O., Rudberg, N., and Jacobsen, D. (2005). Methanol outbreak in Norway 2002-2004: epidemiology, clinical features and prognostic signs. J. Intern. Med. 258, 181-190. Hunderi, O. H., Hovda, K. E., Lie, B., Listerud, S., Hartmann, A., and Jacobsen, D. (2004). Metanolforgiftningene I 2002. Tidsskr. Nor. Laegeforen. 124, 3199-3202. Jacobsen, D., Webb, R., Collins, T. D., and McMartin, K. E. (1988). Methanol and formate kinetics in late diagnosed methanol intoxication. Med. Toxicol. Adverse Drug Exp. 3, 418- 423. Jung, S., Akabas, M. H., and Harris, R. A. (2005). Functional and structural analysis of the GABAA receptor alpha 1 subunit during channel gating and alcohol modulation. J. Biol. Chem. 280, 308-316. Kostic, M.A., and Dart, R. C. (2003). Rethinking the toxic methanol level. J. Toxicol. Clin. Toxicol. 41, 793-800. Lalonde, A., Avila-Carino, J., Caruso, M., de Campos-Lima, P. O. (2007). Rescue of the immunotherapeutic potential of a novel T cell epitope in the Epstein-Barr virus latent membrane protein 2. Virology. 361, 253-62. Lanigan, S. (2001). Final report on the safety assessment of Methyl Alcohol. Int. J. Toxicol. 20, 57-85. Liu, J.J., Daya, M. R., Carrasquillo, O., and Kales, S. N. (1998). Prognostic factors in patients with methanol poisoning. J. Toxicol. Clin. Toxicol. 36, 175-181. Lushine, K. A., Harris, C. R., and Holger, J. S. (2003). Methanol ingestion: prevention of toxic sequelae after massive ingestion. J. Emerg. Med. 24, 433-436. Macian, F. (2005). NFAT proteins: key regulators of T-cell development and function. Nat. Rev. Immunol. 5, 472-484.

Malnou, C.E., Salem, T., Brockly, F., Wodrich, H., Piechaczyk, M., and Jariel-Encontre, I. (2007). Heterodimerization with Jun family members regulates c-Fos nucleocytoplasmic traffic. J. Biol. Chem. 282, 31046-31059. Mandrekar, P., Jeliazkova, V., Catalano, D., and Szabo, G. (2007). Acute alcohol exposure exerts anti-inflammatory effects by inhibiting IkappaB kinase activity and p65 phosphorylation in human monocytes. J. Immunol. 178, 7686-7693. Martens, J., Westhovens, R., Verberckmoes, R., Delooz, H., and Daenens, P. (1982). Recovery without sequelae from severe methanol intoxication. Postgrad. Med. J. 58, 454- 456. McCoy, M. K., Ruhn, K. A., Martinez, T. N., McAlpine, F. E., Blesch, A., and Tansey, M. G. (2008). Intranigral Lentiviral Delivery of Dominant-negative TNF Attenuates Neurodegeneration and Behavioral Deficits in Hemiparkinsonian rats. Mol. Ther. 16, 1572–1579. Mogi, M., Harada, M., Riederer, P., Narabayashi, H., Fujita, K., and Nagatsu, T. (1994). Tumor necrosis factor-alpha (TNF-alpha) increases both in the brain and in the cerebrospinal fluid from parkinsonian patients. Neurosci. Lett. 165, 208–210. Ohashi, I., Pohoreki, R., Morita, K., and Stemmer, P. M. (2004). Alcohols increase calmodulin affinity for Ca2+ and decrease target affinity for calmodulin. Biochim. Biophys. Acta. 1691, 161-167. Okamura, H., Garcia-Rodriguez, C., Martinson, H., Qin, J., Virshup, D. M., and Rao, A. (2004). A conserved docking motif for CK1 binding controls the nuclear localization of NFAT1. Mol. Cell. Biol. 24, 4184-4195. Parthasarathy, N. J., Srikumar, R., Manikandan, S., Narayanan, G. S., and Devi, R. S. (2007). Effect of methanol intoxication on specific immune functions of albino rats. Cell. Biol. Toxicol. 23, 177-187. Porter, C. M., and Clipstone, N. A. (2002). Sustained NFAT signaling promotes a Th1-like pattern of gene expression in primary murine CD4+ T cells. J. Immunol. 168, 4936-4945. Reddy, N. J., Lewis, L. D., Gardner, T. B., Osterling, W., Eskey, C. J., and Nierenberg, D. W. (2007). Two cases of rapid onset Parkinson's syndrome following toxic ingestion of ethylene glycol and methanol. Clin. Pharmacol. Ther. 81, 114-121.

Ren, H., Honse, Y., and Peoples, R. W. (2003). A site of alcohol action in the fourth membrane-associated domain of the N-methyl-D-aspartate receptor. J. Biol. Chem. 278, 48815-48820. Saxena, K., Lerud, K., and Cicero, J. J. (1987). Methanol intoxication in multiple trauma. Am. J. Emerg. Med. 5, 60-63. Saeed, R. W., Varma, S., Peng, T., Tracey, K. J., Sherry, B., and Metz, C. N. (2004). Ethanol blocks leukocyte recruitment and endothelial cell activation in vivo and in vitro. J. Immunol. 173, 6376-6383. Shahidullah, M., Harris, T., Germann, M. W., and Covarrubias, M. (2003). Molecular features of an alcohol binding site in a neuronal potassium channel. Biochemistry 42, 11243-11252. Shaw, J. P., Utz, P. J., Durand, D. B., Toole, J. J., Emmel, E. A., and Crabtree, G. R. (1988). Identification of a putative regulator of early T cell activation genes. Science 241, 202-205. Shelby, M., Portier, C., Goldman, L., Moore, J., Iannucci, A., Jahnke, G., and Donkin, S. (2004). NTP-CERHR expert panel report on the reproductive and developmental toxicity of methanol. Reprod. Toxicol. 18, 303-390. Smith-Garvin, J. E., Koretzky, G. A., and Jordan, M. S. (2009). T cell activation. Annu. Rev. Immunol. 27, 591-619. Srinivasan, M., and Frauwirth, K. A. (2007). Reciprocal NFAT1 and NFAT2 nuclear localization in CD8+ anergic T cells is regulated by suboptimal calcium signaling. J. Immunol. 179, 3734-3741. Szabo, G., Dolganiuc, A., Dai, Q., and Pruett, S. B. (2007). TLR4, ethanol, and lipid rafts: a new mechanism of ethanol action with implications for other receptor-mediated effects. J. Immunol. 178, 1243-1249. Taieb, J., Delarche, C., Ethuin, F., Selloum, S., Poynard, T., Gougerot-Pocidalo, M. A., and Chollet-Martin, S. (2002). Ethanol-induced inhibition of cytokine release and protein degranulation in human neutrophils. J. Leukoc. Biol. 72, 1142-1147.

Teixeira, L. K., Fonseca, B. P., Vieira-de-Abreu, A., Barboza, B. A., Robbs, B. K., Bozza, P. T., and Viola, J. P. (2005). IFN-gamma production by CD8+ T cells depends on NFAT1 transcription factor and regulates Th differentiation. J. Immunol. 175, 5931-5939. Tsytsykova, A. V., and Goldfeld, A. E. (2000). Nuclear factor of activated T cells transcription factor NFATp controls superantigen-induced lethal shock. J. Exp. Med. 192, 581-586. Verhelst, D., Moulin, P., Haufroid, V., Wittebole, X., Jadoul, M., and Hantson, P. (2004). Acute renal injury following methanol poisoning: analysis of a case series. Int. J. Toxicol. 23, 267-273. Wu, A. H., Kelly, T., McKay, C., Ostheimer, D., Forte, E., and Hill, D. (1995). Definitive identification of an exceptionally high methanol concentration in an intoxication of a surviving infant: methanol metabolism by first-order elimination kinetics. J. Forensic Sci. 40, 315-320. Zabrodskii, P. F., Germanchuk, V. G., and Mandych, V. G. (2008). Inhibition of function of T cell subpopulations and decrease in cytokine production during subacute poisoning with various toxicants. Bull. Exp. Biol. Med. 146, 234-236.

Figures and Legends

Figure 3.1 - Biological effect of methanol treatment on human T lymphocytes in vitro.

(A) Methanol enhances IL-2 production in peripheral blood T lymphocytes: Purified T cells were stimulated with anti-CD3/CD28 antibody-coated beads for 5 h in presence of 0.08%,

0.16%, 0.3%, and 0.6% (w/v) methanol. The IL-2 concentration in the supernatants was measured by ELISA and is depicted as means ± SEM in the black bar histogram. The cell viability is shown as means ± SEM in the gray bar histogram (** p < 0.01 relative to the st positive control group, n: 6). Panel symbols: st, stimulated cells in absence of methanol; (-), unstimulated cells; (-)M, unstimulated cells in presence of 0.6% (w/v) methanol; MeOH, methanol. (B) Proliferation of peripheral blood T cells in presence of methanol: Freshly purified T lymphocytes were labeled with CFSE and stimulated with anti-CD3/CD28 antibody-coated beads for 5 h with or without methanol. After removal of the activating beads and the alcohol, the cells were further incubated for 72 h in absence of exogenous IL-2. The number of cell divisions was counted by flow cytometric analysis and the mean percentage of dividing cells ± SEM was plotted as a histogram. White bars: unstimulated cells; black bars: stimulated cells; grey bars: cells stimulated in presence of 0.6% (w/v) methanol. For each cell division group, the means of the negative control (white) and of the alcohol- treated stimulated sample (grey) were compared to the mean of the positive control (black); ** p < 0.01, n: 3.

Figure 3.2 - Methanol impact on early TCR signaling and on transcription factor activation.

(A) Methanol treatment does not affect ZAP-70 phosphorylation following TCR activation: human peripheral blood T cells were stimulated for 3 min with anti-CD3/anti- IgG in presence (st-MeOH) or absence (st) of 0.6% (w/v) methanol and processed for SDS/PAGE. One representative blot of three is shown (Mean densitometric units ± SEM: 29.8 ± 6.9 (st); 28.5 ± 7.9 (st-MeOH), p > 0.05).

(B/C) NF-B nuclear translocation does not change with isopropanol treatment: human peripheral blood T cells were stimulated with anti-CD3/CD28 antibody-coated beads for 5 h in presence of 0.16%, 0.3% and 0.6% (w/v) methanol. Nuclear extracts were incubated with immobilized p50- or p65-binding oligonucleotides in 96-well plates; the amount of bound transcription factor was assessed with a p50- or a p65-specific antibody by ELISA. The nuclear transcription factor ratio was calculated by dividing the sample value by the value of the unstimulated control in absence of methanol. Data is presented as means ±

SEM (ns: p > 0.05, ** p < 0.01 relative to the st control group, n: 3 for B and n: 3 for C). (D/E) Methanol does not affect AP-1 nuclear translocation: human peripheral blood T cells were stimulated as described in panels B/C in presence of the indicated amounts of methanol. Nuclear extracts were incubated with immobilized c-Jun- or c-Fos-binding oligonucleotides in 96-well plates; the amount of bound transcription factor was assessed with a c-Jun- or a c-Fos-specific antibody by ELISA. The nuclear transcription factor ratio was calculated by dividing the sample value by the value of the unstimulated control in absence of methanol. Data is presented as means ± SEM (ns: p > 0.05, * p < 0.05 relative to the st control group, n: 3 for D and n: 3 for E). Stimulation in presence of inhibitor compounds is indicated as follows: BAY: BAY 11- 7082 (panels 2B and 2C); inhib.: SP600125 (panel D); inhib.: PD98059/SB202190 (panel E); other symbols are as in fig. 1A.

Figure 3.3 - Differential modulation of NFAT family members by methanol exposure in vitro.

(A) Methanol increases the activation of a synthetic promoter containing NFAT binding sites: Jurkat-luciferase cells were stimulated with PMA/ionomycin in presence of 0.16%, 0.3%, and 0.6% (w/v) methanol for 5 h. Samples were lysed and assayed for luciferase activity. Results are presented as mean relative luciferase units/g of protein ± SEM (ns: p

> 0.05, * p < 0.05, ** p < 0.01 relative to the st control group, n: 4). (B) Effect of methanol on the DNA-binding capacity of NFAT: Nuclear extracts were prepared from purified human peripheral blood T cells that have been stimulated with anti- CD3/CD28 antibody-coated beads for 1 h in presence of the indicated methanol concentrations and were analysed by EMSA. One representative autoradiograph is depicted. The histogram under the autoradiograph presents the percentage NFAT binding above the level of the stimulated positive control (st). The densitometric analysis was prepared from 3 independent experiments. The detached right-hand side of the autoradiograph shows the supershift analysis with the anti-NFATc1 (7A6) and anti- NFATc2 (G1-D10) antibodies. The NFAT complex is indicated on the left and the supershifted band is indicated by the arrow on the right. NS: nonspecific band. (C) Methanol does not change NFATc1 nuclear content: Purified human peripheral blood T lymphocytes were stimulated with anti-CD3/CD28 antibody-coated beads for 1 h in presence of 0.16%, 0.3% and 0.6% (w/v) methanol. Nuclear extracts were incubated with immobilized NFAT-binding oligonucleotides in 96-well plates; the amount of retained transcription factor was assessed with an NFATc1-specific antibody by ELISA. The nuclear NFATc1 ratio was calculated by dividing the sample value by the value of the unstimulated control in absence of methanol. Data is presented as means ± SEM (ns: p >

0.05, ** p < 0.01 relative to the st control group, n: 3). (D) Methanol up-regulates NFATc2 nuclear content: Purified human peripheral blood T cells were stimulated as described in panel C in presence of the indicated amounts of methanol. Nuclear extracts were incubated with immobilized NFAT-binding oligonucleotides in 96-well plates; the amount of retained transcription factor was assessed with an NFATc2-specific antibody by ELISA. The nuclear NFATc2 ratio was calculated by dividing the sample value by the value of the unstimulated control in absence of methanol.

Data is presented as means ± SEM (* p < 0.05, ** p < 0.01 relative to the st control group, n: 3). (E) Measurement of NFATc2 in the nucleus by Western blot. Jurkat cells were stimulated for 15 min with anti-CD3/CD28/anti-IgG in presence (st-MeOH) or absence (st) of 0.6% (w/v) methanol and the nuclear fractions were processed by SDS/PAGE. Histone deacetylase (HDAC-1) was used as a loading control. One representative blot of four is shown (Mean densitometric units ± SEM: 26.2 ± 2.6 (st); 35.9 ± 3.1 (st-MeOH), p < 0.05, n: 4). Figure symbols: CsA indicates stimulation in presence of cyclosporine A; other symbols are as in fig. 1A. The dashed line in 3A indicates the promoter activation level of the stimulated positive control (st). The dashed lines in 3C and 3D indicate the ratio of the relevant NFAT in the stimulated positive control (st).

Figure 3.4 - Effect of methanol on proinflammatory cytokine release by human peripheral blood T lymphocytes in vitro.

(A) Methanol increases TNF- production in peripheral blood T lymphocytes: T cells were stimulated with anti-CD3/CD28 antibody-coated beads for 5 h in presence of 0.08%, 0.16%, 0.3%, and 0.6% (w/v) methanol. The TNF- concentration in the supernatants was measured by ELISA. Results are presented as means ± SEM (* p < 0.05, ** p < 0.01 relative to the st control group; n: 3). (B) Methanol increases IFN- production in peripheral blood T lymphocytes: T cells were stimulated and treated as in panel A. The amount of IFN- released in the supernatants was

measured by ELISA. Results are presented as means ± SEM (ns: p > 0.05, ** p < 0.01 relative to the st control group; n: 4). (C) Methanol increases IFN- production in CD8+ T lymphocytes: Purified CD8+ T cells were stimulated and treated as in panel A. The IFN- concentration in the supernatants was measured by ELISA. Results are presented as means ± SEM (** p < 0.01 relative to the st control group; n: 7). (D) Methanol increases IFN- production in CD4+ T lymphocytes: Purified CD4+ T cells were stimulated and treated as in panel A. ELISA results are presented as means ± SEM

(ns: p > 0.05, * p < 0.05, ** p < 0.01 relative to the st control group; n: 4). Figure symbols are as in fig. 1A.

Figure 3.5 - Immunological impact of methanol in vivo.

(A) IL-2 production is up-regulated in mice acutely exposed to methanol: BALB/c mice received 5 g SEB subcutaneously plus the intraperitoneal injection of either 2 g/kg methanol (SEB-MeOH group) or saline (SEB/pbs group). Additional control groups received either saline subcutaneously plus methanol intraperitoneally (pbs-MeOH) or saline only (pbs/pbs). Animals were sacrificed 4 h after injections, and serum IL-2 levels were quantified by ELISA. Results are presented as means ± SEM (** p < 0.01 relative to the SEB/pbs group, n: 6/group). (B) IFN- production increases in mice acutely exposed to methanol: BALB/c mice were injected and sacrificed as above. Serum IFN- was measured by ELISA. Means ± SEM are shown (** p < 0.01 relative to the SEB/pbs group, n: 5/group). (C) Acute exposure to methanol augments TNF- production: BALB/c mice were injected and sacrificed as above. The mean TNF- serum concentration ± SEM was determined by ELISA (** p < 0.01 relative to the SEB/pbs group, n: 7/group).

Figure 3.6 - Hypothetical model of the methanol transcriptional effect.

The diagram illustrates a simplified model of NFATc2 activation in absence (A) or presence (B) of methanol. T cell receptor triggering increases cytoplasmic Ca++ thereby activating calmodulin and unleashing the phosphatase activity of calcineurin. Dephosphorylation exposes the nuclear-localization signal (nls) and allows the translocation of NFATc2 into the nucleus. Multiple kinases (GSK3, CK1, DYRK1A, and p38) rephosphorylate the NFAT proteins leading to their export from the nucleus; only p38 is indicated. Methanol may interact directly with NFATc2 producing a conformational change that masks the nuclear-export signal (nes) or reduces the accessibility of phosphorylation sites.

4. The dysregulation of the monocyte/macrophage effector function induced by isopropanol is mediated by the defective activation of distinct members of the AP-1 family of transcription factors

Damien Carignan,*1 Olivier Désy,*1 and Pedro O. de Campos-Lima*2 * Laval University Cancer Research Center, Quebec City, Quebec, G1R 2J6, Canada Toxicological Sciences, 2012 Jan; 125(1):144-56.

1 D.C. and O.D. contributed equally to this work. 2 Correspondence: Dr. Pedro O. de Campos-Lima, Laval University Cancer Research Center, McMahon St 9, Quebec City, QC, G1R 2J6, Canada. Phone: 1 418 525 4444, fax: 1 418 691 5439, e-mail: [email protected]

Short title: Immunomodulation by isopropanol.

Résumé

L’isopropanol est la deuxième cause la plus commune d’intoxication aiguë aux alcools à courte chaîne. Les effets immunomodulateurs des alcools à chaîne courte autre que l’éthanol s’exercent par une interférence avec l’activation du nuclear factor of activated T cells (NFAT) avec ou sans la participation additionnelle de l’activator protein-1 (AP-1). Dans ce chapitre, nous examinons l’effet immunomodulateur de l’isopropanol dans des conditions non-dépendantes de NFAT : soit, la sécrétion de cytokines inflammatoires par des monocytes stimulés au lipopolysaccharide (LPS). Notre hypothèse était que, dans ce contexte, l’isopropanol n’aurait qu’un effet ténu ou encore aucun effet. À notre étonnement, le défaut d’activation d’AP-1 s’est avéré suffisant pour causer une dérégulation sévère et dose-dépendante de l’activité des monocytes in vitro, et ce à des concentrations aussi basses que 0,16% (26 mM). Trois effets ont été observés : l’interleukine (IL)-1β et l’IL-8 demeurent inchangés; l’IL-6 est accru; tandis que le TNF-α et CCL2 sont diminués. Les fonctions effectrices des macrophages dérivés des monocytes sont aussi compromises. Nos résultats montrent que, dans des monocytes activés au LPS, la signalisation précoce en aval du Toll-like receptor 4 (TLR4) est préservée puisque l’isopropanol ne change pas l’activité kinase de l’IL-1 receptor associated kinase 1 (IRAK1). Les sentiers de signalisation de NF-ΚB et des MAPK p38 et JNK sont demeurés insensibles à l’alcool. À l’opposé, l’activation de l’extracellular signal-regulated protein kinase (ERK) et ultimement de c-Fos et JunB est réduite. La dérégulation des cytokines induite par l’alcool a été confirmée dans un modèle murin d’intoxication à l’isopropanol dont la production de TNF-α en réponse au LPS a pratiquement été abolie. L’ampleur de cet effet de l’isopropanol est telle que des souris ont pu être sauvées d’un choc toxique induit par le LPS. Ces données s’ajoutent au peu de connaissances disponibles sur l’immunotoxicologie de l’isopropanol, un produit chimique ubiquitaire auquel la population est grandement exposée.

Abstract

Isopropanol is the second most common cause of short-chain alcohol acute intoxication. Non-ethanolic short-chain alcohols mediate their immunomodulatory effect by interfering with NFAT activation with or without additional AP-1 involvement. In the present study, we examined the immunomodulation induced by isopropanol in conditions that are not reliant on NFAT: the inflammatory cytokine response of lipopolysaccharide-stimulated monocytes. Our hypothesis was that isopropanol acute exposure would have an attenuated effect or no consequence in this setting. To our surprise, the impairment of AP-1 activation was sufficient to mediate a severe and dose-dependent phenotype in human monocytes in vitro at alcohol concentrations as low as 0.16% (or 26 mM). There were three outcomes: IL-1/IL-8 were unaltered; IL-6 was up-regulated; and TNF-/CCL2 were down-regulated. The effector function of human monocyte-derived macrophages was also compromised. Our results showed that TLR4 early signaling was preserved, as isopropanol did not change the IRAK1 kinase activity in LPS-stimulated cells. The NF-B signaling cascade and the p38/JNK modules of the MAPK pathway were alcohol-insensitive. Conversely, the activation of ERK and, ultimately, of c-Fos and JunB were impaired. The alcohol-induced cytokine dysregulation was confirmed in a mouse model of isopropanol intoxication in which the production of TNF- in response to LPS challenge was virtually abolished. The magnitude of this alcohol effect was sufficiently high to rescue animals from LPS-induced toxic shock. Our data contribute to the dismal body of information on the immunotoxicology of isopropanol, one of the most ubiquitous chemicals to which the general population is significantly exposed. Keywords: immunomodulation, immunosuppression, isopropanol, monocyte, macrophage, cytokine, immunotoxicology

Introduction

Deliberate or accidental exposure to isopropanol is the second most common cause of acute alcohol intoxication in North America with over 20,000 cases recorded each year

(Bronstein et al., 2009). Alcoholic patients, children, and suicidal individuals are the ones most likely to be poisoned (Zaman et al., 2002). Isopropanol is used in multiple industrial applications but is also easily available to most consumers as rubbing alcohol and as an ingredient of several household products such as hand-sanitizing gels. Although there is a considerable body of information on the general biological effects of isopropanol in animal models of acute, subchronic and chronic exposure (Burleigh-Flayer et al., 1994, 1997; Kapp et al., 1996; Kasuga et al., 1992), surprisingly little is known about its immunotoxicology. Conversely, it is well documented that ethanol modulates the immune system directly or indirectly by interfering with the function of a variety of cells such as T lymphocytes, monocytes, macrophages, dendridic cells, neutrophils, and endothelial cells (Goral and Kovacs, 2005; Oak et al., 2006; Saeed et al., 2004; Szabo et al., 2007; Taieb et al., 2002; Zhao et al., 2003). In our previous work, we have demonstrated that the molecular events that underlie the immunomodulation induced by short-chain alcohols have different flavors that are specific to each alcohol despite their considerable structural similarity (Désy et al., 2008, 2010). Although many of the biological effects of ethanol on the immune system have been attributed to a dysregulation of the nuclear factor-B (NF- B) signaling pathway, other short-chain alcohols seem to keep this signaling cascade unaltered (Désy et al., 2008, 2010; Oak et al., 2006; Saeed et al., 2004; Szabo et al., 2007). Instead, they mediate their immunomodulation by interfering with the activation of the nuclear factor of activated T cells (NFAT) family of transcription factors with or without additional involvement of the activator protein-1 (AP-1). Thus, while methanol up- regulates NFATc2 nuclear translocation in lymphocytes (Désy et al., 2010), isopropanol down-regulates the activation of NFATc1 and AP-1 in T lymphocytes and NK cells (Désy et al., 2008). In the present study, we sought to examine the immunomodulation induced by isopropanol in a stimulation model that is less reliant on the NFAT family of transcription factors. We chose to study the monocyte inflammatory cytokine response to lipopolysaccharide (LPS) because it involves a well-defined signal transduction pathway that leads to NF-B and AP-1 activation and does not require NFAT (Kawai and Akira, 2010). Our hypothesis was that isopropanol acute exposure would have an attenuated effect or no consequence in this setting. Our results revealed that the impairment of AP-1

97 activation was sufficient to cause a severe and dose-dependent phenotype in human monocytes in vitro at alcohol concentrations as low as 0.16% (or 26 mM). Similarly to what was reported for T lymphocytes and NK cells (Désy et al., 2008), isopropanol did not change the NF-B signaling cascade in activated monocytes; nevertheless, it produced an immune dysregulation that was mediated by ERK and, ultimately, by the c-Fos and JunB members of the AP-1 family of transcription factors. The immunosuppressive potential of this alcohol was validated in vivo and had sufficiently high magnitude to rescue mice from LPS-induced toxic shock.

Materials and Methods

Cell isolation, culture, and stimulation. This study was approved by the Institutional Clinical Research Ethics Committee (L’Hôtel-Dieu de Québec/Centre hospitalier universitaire de Québec - L’HDQ-CHUQ). Mononuclear cells were prepared from the peripheral blood from healthy volunteers by density gradient centrifugation using Ficoll- Hypaque (GE Healthcare, Piscataway, NJ). Written informed consent was obtained from all donors. Monocytes were isolated from mononuclear cells by plastic adherence (Fuss et al., 2009; Szabo and Mandrekar, 2008) and maintained in RPMI 1640 (Invitrogen Canada, Burlington, Canada) supplemented with 10% heat-inactivated FBS (BioCell Inc., Drummondville, Canada). Monocyte-derived macrophages were generated by cultivating monocytes in RPMI 1640 supplemented with 18% heat-inactivated autologous serum for 8 days as described (Szabo and Mandrekar, 2008). Human primary cells were used in most in vitro experiments to strengthen the quality of the data. Established cell lines were used only in a few instances to limit the volume of blood drawn from the donors. The human monocytic line Mono Mac 6 (Ziegler-Heitbrock et al., 1988) and the murine macrophage cell line P388D1 (Koren et al., 1975) were cultivated in RPMI 1640 containing 10% FBS. Murine monocytes were purified from bone marrow or from spleens (Swirski et al., 2009) by using an antibody-based negative selection kit with magnetic nanoparticles (StemCell Technologies, Vancouver, Canada) according to the manufacturer’s instructions.

Monocytes and macrophages were activated for cytokine production for 24 h at 37°C with 1 g/ml ultra pure lipopolysaccharide from Escherichia coli O111:B4 (List Biological Laboratories, Campbell, CA). When indicated, alternative cell activation protocols were used: protocol a, 15 min stimulation with 1 g/ml LPS at 37oC (Western blot and in vitro kinase assay); protocol b, 1 h stimulation with 1 g/ml LPS at 37oC (Transcription factor activation assays); protocol c, 1 h stimulation with labeled E. coli at 37oC (Flow cytometry- based phagocytosis analysis); and protocol d, 2 h stimulation with labeled E. coli at 37oC (Microscopy-based phagocytosis analysis). The data on primary cells in the various assays utilized in this work was obtained from multiple donors. The means ± SEM from independent experiments are indicated in each figure. In addition, two technical replicates per sample were used within each independent ELISA experiment. Isopropanol was purchased from BDH (Toronto, Canada) and was 99.5% pure.

Cytokine analysis. Measurements of human TNF-, IL-1, IL-6, and IL-8 or murine IL-6 in cell culture supernatants and murine TNF-, IL-6, and CCL2 in serum samples were performed with specific cytokine ELISA kits according to the manufacturer’s instructions (BioLegend, San Diego, CA). Human CCL2 levels were measured with the human CCL2 (MCP-1) ELISA Ready-SET-Go! kit (Ebioscience, San Diego, CA) as recommended by the manufacturer. Supernatants from human cells were used undiluted for IL-1 analysis or diluted 1:20, 1:10, 1:100, and 1:4 for TNF-, IL-6, IL-8, and CCL2 measurements, respectively. Supernatants from murine cells were diluted 1:4 for IL-6 analysis. Murine sera were diluted 1:25, 1:50, and 1:5 for TNF-, IL-6, and CCL2 measurements, respectively. Briefly, 96-well plates were coated with the relevant capture antibody and incubated with serially diluted standards or unknown samples; then, they were washed and incubated with the biotinylated detection antibody followed by streptavidin-horseradish peroxidase. The plates were read at 450 nm after sequential treatment with 3,3',5,5'- tetramethylbenzidine (TMB) substrate solution and phosphoric acid.

It is worth mentioning that short chain alcohols may interfere with the sensitivity of certain TNF- ELISA kits at high concentrations (von Maltzan and Pruett, 2011). In our experiments, TNF- ELISA assays were conducted with supernatants diluted 20 times and harvested after a 24 h-long incubation period with LPS at 37oC. Isopropanol has an evaporation rate of 2.83 as compared to 1.4 for ethanol and 0.3 for water (n-BuAc = 1.0); thus, only negligible isopropanol amounts were still present at the time of processing of diluted supernatants. As remarked for the ELISA assays for TNF- measurement in cell supernatants, the serum samples were also diluted (25 times) before analysis. Only trace amounts of isopropanol were left in the diluted serum, < 0.01% or < 1.5 mM, without accounting for metabolic clearance, which is likely to reduce these levels even further. Inhibitor compounds (Sigma, St Louis, MO) were used as follows: BAY 11-7082: 5 M; SB202190: 10 M; SP600125: 25 M; and PD98059: 50 M.

Western blot. Purified cells were activated for 15 min at 37°C with LPS as described above with or without 0.6% (w/v) isopropanol. The cells were washed in alcohol-free buffer and lysed in sodium dodecyl sulfate (SDS) sample buffer (2% w/v SDS, 0.25 M - mercaptoethanol, 10% v/v glycerol, 0.05 M Tris-HCl, pH 6.8, 0.004% w/v bromophenol blue); lysates were separated in 12% polyacrylamide gels and blotted onto nitrocellulose filters (Hybond-C, GE Healthcare, Piscataway, NJ). All antibodies were purchased from Cell Signaling Technology (Danvers, MA) unless indicated otherwise. The membranes were first probed with the following antibodies: IKK/ detection, anti-phospho-IKK/ rabbit monoclonal antibody (16A6, 1:1000) and anti--tubulin mouse monoclonal antibody for protein loading control (TUB 2.1, 1:4000, Sigma); p38 detection, anti-phospho-p38 MAPK mouse monoclonal antibody (28B10, 1:2000) and anti-total p38 rabbit polyclonal antibody (In-house, 1:10000); JNK detection, anti-phospho-SAPK/JNK rabbit polyclonal antibody (Thr183/Tyr185, 1:1000) and anti--tubulin mouse monoclonal antibody for protein loading control (TUB 2.1, 1:4000); and ERK detection, anti-phospho-p44/42 MAPK mouse monoclonal antibody (E10, 1:2000) and anti-total p42 MAPK rabbit polyclonal antibody (In-house, 1:5000, Huot et al., 1995). The membranes were subsequently washed and

100 incubated with 1/15000 dilutions of the antibodies IRDye 800CW goat anti-rabbit IgG and/or IRDye 680 goat anti-mouse IgG (Li-Cor Biosciences, Lincoln, NE). Detection and quantification were performed with the Odyssey Infrared Imaging System (Li-Cor Biosystems).

Interleukin-1 receptor-associated kinase 1 (IRAK1) immunoprecipitation and in vitro kinase assay. Mono Mac 6 cells were stimulated with LPS for 15 min at 37oC in presence or absence of 0.6% (w/v) isopropanol. Cells were washed in alcohol-free PBS and submitted to lysis and IRAK1 immunoprecipitation with reagents from the Roche protein G immunoprecipitation kit (Roche Diagnostics, Laval, Canada) according to the manufacturer’s instructions. The anti-IRAK1 rabbit polyclonal antibody (Millipore, Billerica, MA) was used to generate immunocomplexes. Sample input was equalized for protein content after quantification with the DC Protein Assay (Bio-Rad, Hercules, CA). Washed immunocomplexes were resuspended in kinase buffer (25 mM Tris HCl, pH 7.2,

10 mM MgCl2, 5 mM -glycerophosphate, 0.1 mM Na3VO4, and 2 mM DTT) and added to the myelin basic protein substrate in presence of 32P--ATP and cold ATP for 30 min at 30oC. The kinase reaction was stopped by adding SDS sample buffer and an aliquot was separated in 12% SDS-PAGE. Gels were dried and subsequently analysed and quantified in a phosphoimager.

Phagocytosis assays. P388D1 cells were incubated at 37oC with E. coli bioparticles labeled with either pHrodo or Alexa Fluor 488 (Invitrogen Canada) in presence or absence of isopropanol as indicated. Phagocytosis was analysed with a TE300 microscope (Nikon, Melville, NY) and quantified with the MetaVUE software (Molecular Devices, Sunnyvale, CA). Confocal imaging was performed with a FluoView FV1000 microscope and the FluoView application software (Olympus Canada, Markham, Canada). Alternatively, samples were analysed in a XL flow cytometer (Beckman Coulter Inc., Miami, FL). Quenching of surface-bound Alexa Fluor 488 bacteria was achieved by incubating the cells with 0.2% (w/v) trypan blue in PBS.

101

ELISA-based transcription factor activation assay. Cells were washed in alcohol-free buffer before the generation of nuclear lysates. Nuclear proteins were extracted using the Active Motif Nuclear Extract kit (Active Motif, Carlsbad, CA) according to the manufacturer's instructions; the total protein concentration of the lysates was determined by the Bradford assay (Bio-Rad). c-Fos, Fra1, FosB, c-Jun, JunD and JunB activation was measured with the TransAm AP-1 kit; p65 activation was measured with the TransAm NF- B kit. ELISA-based TransAm kits were used according to the manufacturer’s instructions (Active Motif). Briefly, nuclear extracts were incubated with plate-bound transcription factor-specific oligonucleotides; the plates were washed, and further incubated with transcription factor-specific antibodies. Addition of a horseradish-conjugated secondary antibody and the TMB substrate produced a colorimetric reaction measurable in a spectrophotometer.

In vivo studies. 7-13-week-old female BALB/c mice were bought from The Jackson Laboratory (Bar Harbor, ME). All tests respected the ethical guidelines set by the Institutional Animal Protection Committee (CPA-CHUQ). Food and water were provided ad libitum. Animals received 5 g LPS subcutaneously for cytokine induction and were sacrificed by CO2 asphyxiation 90 min or 180 min after administration for TNF- or IL- 6/CCL2 serum analysis, respectively. Toxic shock was induced with a subcutaneous injection of 0.2 g LPS after presensitization with 20 mg D-galactosamine (Sigma) as reported elsewhere (Tsytsykova and Goldfeld, 2000). Animals were checked hourly in the beginning of the protocol and were followed for 5 days. Isopropanol was injected intraperitoneally (2 g/kg) in the experiments for cytokine induction analysis and for toxic shock protection. Statistical analysis. One-way ANOVA followed by Dunnett’s multiple-comparison posttest was performed with GraphPad Prism (GraphPad Software Inc., San Diego, CA) on data presented in all figures, except when indicated otherwise. The Student’s t test was used in figures 4.3A, 4.3B and 4.4A. Survival curves were determined by the Kaplan-Meier method (fig. 4.5E). p values < 0.05 were considered significant.

102

Results

Isopropanol changes the cytokine/chemokine release profile of activated human monocytes Monocytes are capable of releasing several cytokines in response to Gram-negative bacteria (da Silva Correia et al., 2001; Miyake, 2006). The acquisition of effector function is a highly choreographed process that may begin with the recognition of pathogen-derived lipoglycans, as illustrated by the engagement of lipopolysaccharides by the MD2/Toll-like receptor 4 (TLR4) complex, and culminates with the transcriptional activation of several genes encoding immunologically relevant proteins (Miyake, 2006). In the present study, we have investigated whether isopropanol exposure in vitro would have any impact on the ability of human monocytes to produce cytokines once activated by lipopolysaccharides. Three distinct outcomes were observed as illustrated in figure 4.1. The releases of the proinflammatory cytokine IL-1β and of the chemokine IL-8 were unaltered even at the highest isopropanol concentrations tested. The production of IL-6 was up-regulated by 53% at the lowest alcohol dilution and more than doubled at the highest isopropanol concentration as measured in the culture supernatants by ELISA. In contrast, monocytes treated with isopropanol produced lower amounts of the proinflammatory cytokine TNF- and of the chemokine CCL2. This down-regulation was observed at isopropanol doses as low as 0.16%. The observed effect was not the consequence of nonspecific cytotoxicity as the cell viability of alcohol-treated samples in the concentration range that produced inhibition levels of 30-41% for TNF- and of 25-64% for CCL2 in figure 4.1 was similar to that of untreated control cells (data not shown and Désy et al., 2008). Moreover, in agreement with our previous data on the effect of isopropanol and the closely related methanol on lymphocytes (Désy et al., 1998, 2010), alcohol treatment alone had no impact on the release of any tested cytokine by monocytes. This was evident even for IL-6, whose negligible production in unstimulated cells was not augmented by isopropanol (Mean release ± SEM: 105 ± 5.8 pg/ml in unstimulated cells and 97 ± 11.9 pg/ml in alcohol- treated unstimulated cells, n: 3).

103

The immunomodulatory effect of isopropanol extends to differentiated macrophages Monocytes are attracted to injured sites where they differentiate into macrophages (Swirski et al., 2009), which play a pivotal role in the regulation of the inflammatory response. The finding that cytokine release by human monocytes is affected by isopropanol led us to predict a similar impact on the function of derived macrophages. Panel A in figure 4.2 shows that exposure to this alcohol in vitro creates an identical profile of cytokine dysregulation in monocytes and macrophages with down-regulation of TNF-/CCL2 and up-regulation of IL-6 production. Besides their secretory function, tissue macrophages are also active phagocytes involved in pathogen clearance and removal of necrotic material (Park et al., 2011). Thus, it was conceivable that the alcohol-induced dysfunction could extend to their ability to ingest large particles. We have tested this assumption by examining the internalization of fluorescent bacteria in a murine macrophage cell line in presence and absence of isopropanol. Figure 4.2 shows representative micrographs (panel B) and a quantitative assessment of the cellular fluorescence intensity (panel C); phagocytosis was reduced by 34% at 0.3% isopropanol exposure and by more than 50% at the highest concentration tested. These results were further substantiated by flow cytometric analysis, which revealed similar levels of phagocytosis impairment, 37% and 67%, in cells exposed to 0.3% and to the highest alcohol concentration, respectively (Figure 4.2, panel D).

Early signaling events triggered by Toll-like receptor 4 activation are not compromised in human monocytes acutely exposed to isopropanol Isopropanol could exert its modulatory effect via direct interaction with molecules that initiate or facilitate signaling in response to lipoglycans. LBP lipid transferase initially promotes the incorporation of LPS aggregates into the plasma membrane and their subsequent transfer to CD14 (Miyake, 2006). The latter ultimately presents LPS to the MD2/TLR4 complex, inducing receptor clustering, recruitment of key adaptor molecules to the TLR4 cytoplasmic domain, and downstream signaling (da Silva Correia et al., 2001; Miyake, 2006). In order to address the possibility that isopropanol would interfere with the

104 membrane-based early events that characterize lipoglycan recognition, we have examined the kinase activity of IRAK1. This enzyme is recruited by the MyD88 adaptor protein early on following LPS stimulation and is central to the activation of NF-B and AP-1 transcription factors and the resulting production of proinflammatory cytokines (Kawai and Akira, 2010). Panel A in figure 4.3 shows that IRAK1 is not affected by alcohol exposure as measured by its ability to phosphorylate a model substrate.

The biological effect of isopropanol is not mediated by an alteration of the NF-B signaling pathway in LPS-activated human monocytes Once we have established that isopropanol does not change early TLR4 signaling, we checked whether NF-B activation was compromised by examining the phosphorylation status of the IB kinase (IKK). This enzyme consists of regulatory () and catalytic ( and ) subunits, whose activity is found downstream of IRAK1/TRAF6/TAK1 in the TLR4 signal transduction pathway (Hayden and Ghosh, 2008; Israël, 2010); upon LPS stimulation, the IKK complex undergoes transautophosphorylation or is directly phosphorylated by TAK1 on serine residues in the T loop domain of the catalytic subunits (Hayden and Ghosh, 2008). Western blot analysis with an antibody that recognizes phosphorylated IKKs (Ser176/180 in IKK and Ser177/181 in IKK revealed that IKK activation is similar in alcohol-treated and untreated samples following LPS exposure (Figure 4.3, panel B). These findings were supported by the measurement of the nuclear translocation of the NF-B p65 subunit in TLR4-stimulated human monocytes treated or not with isopropanol. Figure 4.3C shows that monocytes activated by LPS remained unaffected by alcohol treatment. The compound BAY 11-7082, an inhibitor of IB phosphorylation, reduced the nuclear content of p65 in the same experiments to levels lower than those of the unstimulated control.

Isopropanol selectively modulates the activation of the extracellular signal-regulated protein kinase (ERK) module of the mitogen-activated protein kinase (MAPK)

105 superfamily and reduces the nuclear translocation of discrete components of the AP-1 transcription factor in LPS-treated human monocytes As our results have revealed that the NF-B pathway is insensitive to isopropanol exposure, we have decided next to examine the three modules of the MAPK signaling cascade, p38, JNK, and ERK, in LPS-stimulated monocytes. Panel A in figure 4.4 shows that there is no clear alcohol effect on the phosphorylation of p38 and JNK. Conversely, the phosphorylation of the activation loop residues Thr185 and Tyr187 in ERK2 was 37% less efficient in presence of isopropanol. One of the best-studied effects of ERK is the initiation of transcription of the immediate early gene c-fos through the activation of Elk-1 (Yoon and Seger, 2006). The c-Fos transcription factor is detectable as early as 20 min after stimulation and is readily phosphorylated at Ser374 by ERK and on additional sites by ERK- activated kinases or by an extended ERK activity (Lallemand et al., 1997; Murphy et al., 2002; Yoon and Seger, 2006). In order to dissect the impact of isopropanol on ERK substrates, we have measured the nuclear translocation of phosphorylated c-Fos in TLR4- stimulated monocytes exposed to different concentrations of the alcohol. Figure 4.4B, left panel, shows that monocyte activation by LPS led to a 3-fold increase in the amount of c- Fos in the nucleus. The same stimulation in presence of 0.16% isopropanol led to a 2.4-fold increase in nuclear c-Fos (or 68% of the maximal c-Fos nuclear content above the unstimulated cell baseline). The alcohol effect was dose-dependent and the increase of this transcription factor in the nucleus was down to 1.7 fold (or 33% of the maximal content) at the highest isopropanol concentration. The same stimulatory conditions in presence of a combination of inhibitor drugs of the MAPK signaling cascade, PD98059 (MEK1)/SB202190 (p38)/SP600125 (JNK), produced levels of nuclear c-Fos lower than those of the unstimulated cells. We have examined next the fate of other members of the AP-1 family of transcription factors in the same experimental conditions as above. Our results have revealed that isopropanol exposure does not have any impact on the nuclear translocation of c-Jun, JunD, Fra1 and FosB (data not shown). We have found, however, a measurable alcohol effect on JunB activation. Similar to c-Fos, the JunB protein appears rapidly after cell activation and is encoded by an immediate early gene whose transcription is induced by two important

106 types of ERK substrates, the kinases RSK2 and MSK1/2 (Cargnello and Roux, 2011; Lallemand et al., 1997). The right panel in figure 4.4B shows that LPS activation produced a 2.2-fold augmentation in nuclear JunB. Monocytes stimulated in presence of 0.16% isopropanol experienced a 1.8-fold increase in nuclear JunB (or 68% of the maximal JunB nuclear content above the unstimulated cell baseline). The alcohol effect was also dose- dependent and the LPS-induced increase in nuclear JunB was 1.4 fold (or 36% of the maximal content) at the highest alcohol concentration. The JunB nuclear content in TLR4- activated monocytes was similar to that of control unstimulated cells when the PD98059/SB202190/SP600125 kinase inhibitors were used.

Isopropanol inhibits the production of TNF- and CCL2 in vivo and confers protection from LPS-induced toxic shock The biological effect of isopropanol on the immune response to LPS was subsequently tested in a mouse model of acute alcohol intoxication. Mice were administered isopropanol intraperitoneally, 2 g/kg, to generate a mean blood alcohol concentration of 200 mg/dl after 30 min. The production of TNF-, IL-6, and CCL2 was induced in vivo by subcutaneous injection of LPS. As anticipated, LPS triggered TNF- levels of > 1700 pg/ml after 90 min (Fig. 4.5A) and > 13000 pg/ml CCL2 after 180 min (Fig. 4.5D). Isopropanol exposure led to a statistically significant down-regulation of the serum release of these cytokines: 86% or 240 pg/ml TNF- detected at 90 min, and 76% or 3195 pg/ml CCL2 detected at 180 min. The differences in cytokine serum levels between animals treated and untreated with the alcohol were statistically significant as indicated in figures 4.5A and 4.5D. Similarly, we expected that isopropanol would increase IL-6 production in this mouse model in a way that would resemble its impact on LPS-stimulated human monocytes in vitro. To our surprise, LPS challenge in vivo was not associated with higher serum levels of IL-6 in the context of isopropanol intoxication (Fig. 4.5B). This finding made us consider the possibility that murine and human primary monocytes would respond differently to LPS as regards the alcohol modulation of IL-6 release. Indeed, upon LPS-stimulation, purified murine monocytes were not responsive to isopropanol in vitro as illustrated in figure 4.5C.

107

Our results revealed that isopropanol is a powerful negative regulator of the inflammatory cytokine TNF- both in vitro and in vivo. TNF- is copiously produced and plays a central role in the pathophysiology of LPS-induced lethal shock (Galanos and Freudenberg, 1993; Tsytsykova and Goldfeld, 2000). Therefore, it was our assumption that isopropanol treatment could protect mice from toxic shock. In order to address this issue, we have sensitized BALB/c mice with 20 mg D-galactosamine by intraperitoneal injection. All sensitized animals developed toxic shock within 16 h following subcutaneous LPS challenge (10/10) (Fig. 4.5E). A very different picture was produced by alcohol administration, a single injection of 2 g/kg had a dramatic effect since it completely prevented LPS-induced toxic shock, and all animals survived. These mice were followed for 5 days after LPS challenge and were indistinguishable from sensitized control groups receiving PBS or isopropanol. No death occurred in groups injected with isopropanol (10/10) or PBS (10/10) in absence of LPS.

Discussion

Short-chain alcohols mediate a variety of biological effects through nonspecific mechanisms at high concentrations, usually in the 500 mM range (Dwyer and Bradley, 2000). Nevertheless, they may interact with specific targets and induce loss of function of ion channels, neurotransmitter receptors, enzymes, and adhesion molecules at more physiologically relevant levels (Jung et al., 2005; Ren et al., 2003; Shahidullah et al., 2003). The inhibitory effect of ethanol on LPS-induced TLR4 triggering in macrophages has been suggested to result from its partition into cell membranes (Dai et al., 2005; Szabo et al., 2007); in this setting, the alcohol would reduce the capacity of micro-domains to recruit and/or retain relevant molecules and compromise the cascade that relays the signal generated by receptor ligation on the cell membrane to the nucleus. In a similar fashion, it was conceivable that isopropanol could change LPS binding to CD14 and the stability and/or conformation of the MD2/TLR4 complex thereby dampening or virtually aborting downstream signaling. This scenario would be compatible with our findings that mostly revealed loss of function of LPS-stimulated monocytes after isopropanol treatment, IL-6

108 up-regulation being the notable exception. Our results, however, showed that TLR4 early signaling is preserved during isopropanol acute exposure (Figure 4.3A). We found that isopropanol acts downstream of the cell membrane as it does not change the kinase activity of IRAK1 in LPS stimulated cells. This is in sharp contrast with ethanol, which affects IRAK1 activity in line with the model described above. Further down in the activation cascade, the kinase TAK1 is capable of triggering the two major pathways leading to proinflammatory cytokine production during LPS stimulation, MAPKs and NF-B signaling (Cargnello and Roux, 2011; Israël, 2010; Kawai and Akira, 2010). Our data suggest that isopropanol initiates its effect downstream of TAK1, as the function of this enzyme remains unchanged since the activation of the p38, JNK, and NF- B signaling cascades proceeds normally (Figs. 4.3B, 4.3C and 4.4A). We believe that isopropanol may interact directly with ERK2 because ERK1 phosphorylation is not affected, indicating that the upstream MEK kinases are functional (Fig. 4.4A). Although MEK1/2 are not completely interchangeable, MEK1 can phosphorylate both ERK1/2 in vitro (Xu et al., 2001) and in vivo in MEK2-deficient mice (Bélanger et al., 2003). Thus, it is unlikely that MEK2 loss of function could account for the lower ERK2 phosphorylation observed in isopropanol-treated samples because the MEK2 kinase activity could be operationally replaced by that of MEK1. Unphosphorylated ERK2 is virtually idle but undergoes a dramatic conformational change upon cell activation resulting from the posttranslational modification of the Thr183 and Tyr185 residues (Yoon and Seger, 2006). This change is accompanied by the acquisition of a catalytic activity five orders of magnitude higher than basal levels and the ability to translocate into the nucleus (Chuderland and Seger, 2005). There are multiple ERK substrates and, among these, c-Fos has been studied extensively (Yoon and Seger, 2006). The nuclear content of c-Fos is determined by the level of direct ERK-mediated phosphorylation, which stabilizes the c-Fos protein, and by the very fast transcriptional activation of the c-fos gene via Elk-1, which is also an ERK-dependent process (Cargnello and Roux, 2011; Murphy et al., 2002; Yoon and Seger, 2006). In agreement with the results presented in figure 4.4, lower ERK activity is expected to be associated with a diminished c-Fos phosphorylation and nuclear presence (Murphy et al., 2002). c-Fos dimerizes with

109

Jun proteins to form the AP-1 transcription factor (Lallemand et al., 1997). It is noteworthy that the promoter region of the gene that encodes the chemokine CCL2 displays two AP-1 binding sites (Kok et al., 2009; Martin et al., 1997; Shyy et al., 1995; Sutcliffe et al., 2009). Thus, the lower activation and nuclear translocation of c-Fos that are associated to isopropanol exposure could account for the down-regulation of CCL2 production by stimulated monocytes as illustrated in figure 1. Nevertheless, although there are AP-1 sites also in the proximal TNF- promoter, monocytes stimulated with whole bacteria or LPS assemble rather unique TNF- enhanceosomes that include AP-1 heterodimers lacking c- Fos (Barthel et al., 2003; Tsai et al., 2000). It is conceivable that the other transcription factor that we found to be altered in our experiments, JunB, mediates the TNF- down- regulation induced by acute alcohol exposure. The transcription of the immediate early gene junB is coordinated by the ERK-activated substrate kinases RSK2 and MSK1/2, which activate CREB by phosphorylation at Ser133 thereby allowing the recruitment of CBP and p300 to the junB promoter (Cargnello and Roux, 2011). JunB appears as a protein 20- 40 min after the initiation of MAPK signaling and is expected to be down-regulated when the catalytic activity of ERK is diminished (Lallemand et al., 1997). Figure 4.4B, right panel, shows that the nuclear content of transcriptionally active JunB is indeed compromised by isopropanol. In support to this scenario, a recent report described the participation of JunB in the induction of TNF- in LPS-stimulated myeloid cells and showed JunB binding to the TNF- promoter by chromatin immunoprecipitation (ChIP) (Gomard et al., 2010). At this point in time, we cannot exclude that additional factors encoded by immediate early genes downstream of ERK may contribute to the alcohol effect. The Elk-1-dependent Egr-1 transcription factor, for instance, was shown to integrate the TNF- enhanceosome and to participate in the transcriptional activation of this gene in LPS-stimulated monocytic cells (Barthel et al., 2003; Shi et al., 2002; Tsai et al., 2000). This in line with the observation that the MEK1 (ERK) inhibitor PD98059 blocks Elk-1, Egr-1, and TNF- expression in LPS-stimulated cells (Guha et al., 2001; Shi et al., 2002). The finding that isopropanol acute exposure was associated to JunB down-regulation can also help us to understand the paradoxical up-regulation of IL-6 presented in figure 4.1. It

110 has recently been shown by ChIP and luciferase reporter assays that JunB-containing AP-1 complexes act as powerful repressors of transcriptional activity in the context of the IL-6 promoter (Pflegerl et al., 2009). Most importantly, JunB-deficient cells produce larger amounts of IL-6 than their wild-type counterparts (Meixner et al., 2008; Pflegerl et al., 2009). Therefore, the selective and simultaneous reduction of c-Fos and JunB in the nucleus could account for the alcohol-induced down-regulation of TNF-/CCL2 and up- regulation of IL-6 in LPS-stimulated cells in vitro. It is not surprising that isopropanol induces a similar pattern of cytokine modulation in monocytes and macrophages (Figs. 4.1 and 4.2A) as these cells represent partially overlapping differentiation stages of the same lineage (Valledor et al., 1998). This modulation is likely to be mediated by the dysregulation of the c-Fos and JunB transcription factors as discussed above. Moreover, we have found that macrophages were less efficient in internalizing bacteria in presence of isopropanol (Fig. 4.2B/C/D), a finding that is reminiscent of the impact of ethanol on phagocytosis (Boé et al., 2010; Goral et al., 2008; Karavitis et al., 2008). ERK has been reported to participate in the Fc receptor- and complement-mediated phagocytosis in neutrophils and macrophages (García-García et al., 2002; García-García and Rosales, 2002; Jehle et al., 2006; Mansfield et al., 2000). MAPK signaling also contributes to the uptake of non-opsonized particles by macrophages, although the p38 pathway is believed to play the predominant role (Blander and Medzhitov, 2004). The reduced phagocytosis observed in our experiments with isopropanol may result from impaired ERK signaling. With over 150 direct phosphorylation targets (Yoon and Seger, 2006), ERK kinases are truly pleiotropic and have several possible paths to impact phagocytosis. Thus, the alcohol effect could be indirectly mediated by one of the first identified targets of ERK, cytosolic phospholipase A2, which has been suggested to be involved in phagocytosis (García-García et al., 2002; García-García and Rosales, 2002; Lin et al., 1993); alternatively, such an effect could be a consequence of reduced phosphorylation of cytoskeletal elements by ERK (Mansfield et al., 2000; Yoon and Seger, 2006). This matter will be addressed in future investigations. The immunomodulation by isopropanol of the monocyte/macrophage effector function was demonstrated in vitro at concentrations as low as 0.16% (26 mM), which are comparable to

111 the concentrations of ethanol associated to a biological effect on immune cells in other studies (Goral and Kovacs, 2005; Oak et al., 2006; Saeed et al., 2004; Szabo et al., 2007; Taieb et al., 2002; Zhao et al., 2003). The potential of isopropanol to modulate the TLR4- mediated inflammatory cytokine response to LPS was confirmed in vivo in a mouse model of acute alcohol intoxication. To put in perspective the relevance of the alcohol doses used in our in vivo experiments, we should consider the blood alcohol concentrations in acutely poisoned patients. Although there are thousands of cases of isopropanol intoxication recorded each year (Bronstein et al., 2010), clinical reports with detailed information on time and volume of ingestion are relatively scarce in the medical literature. Patients may survive blood isopropanol concentrations as high as 560 mg/dl (0.56% or 93 mM), while others may succumb to much lower concentrations (Lacouture et al., 1983). This disparity comes from the fact that clinical measurements have often been made hours after ingestion and underestimate the serum alcohol levels present in the early phase of the intoxication (Daniel et al., 1981; Gaudet and Fraser, 1989; King et al., 1970; Mueller-Kronast et al., 2003; Rich et al., 1990; Rosansky, 1982). Concentrations above 400 mg/dl are considered life-threatening and generally require dialysis (Emadi and Coberly, 2007; Lacouture et al., 1983). We have injected mice with 2 g/kg isopropanol to generate a blood alcohol concentration of 200 mg/dl (0.2% or 33 mM) after 30 min; this level is lower than the reported average sublethal isopropanol blood concentration in severely intoxicated humans (310 mg/dl after 7 h) (Ekwall and Clemedson, 1997) and is comparable to the concentration range that is active in vitro (starting at 0.16% or 26 mM). As suggested by the results of our in vitro experiments, intoxicated mice had a severe impairment in their response to LPS as measured by a substantial drop in serum TNF- and CCL2 (7.4- and 4- fold, respectively). There was, however, no significant effect of isopropanol on IL-6 production in animals challenged with LPS. The latter result suggested that mouse and human cells are differently susceptible to alcohol modulation of IL-6 production. This assumption was confirmed by exposing purified murine monocytes in vitro to LPS in presence or absence of isopropanol. We found that murine cells responded to LPS but were indeed insensitive to alcohol modulation of the IL-6 release. The regulation of the IL-6 promoter is rather intricate and involves the concerted action of several transcription factors (Dendorfer et al., 1994;

112

Vanden Berghe et al., 1999). Although the proximal region of the mouse and human IL-6 promoters share consensus sequences for AP-1, NF-B, NFAT, C/EBP, and CREB, there is divergence further upstream from the transcription start site (Allen et al., 2010; Samuel et al., 2008). It is conceivable that the differential alcohol modulation of the IL-6 response to LPS reflects a distinct set of interactions of JunB-containing AP-1 dimers with other factors that are recruited in accordance with the contextual profile of each promoter; this issue, however, goes beyond the scope of this paper. Overall, the immunological impact of isopropanol in vivo is suppressive and corroborates our previous data obtained with lymphocytes. The magnitude of the TNF- down-regulation led us to speculate that isopropanol could rescue mice from LPS-induced toxic shock syndrome. We found that all animals injected with LPS after presensitization with D-galactosamine developed a fulminant toxic shock with a median survival of 10.5 h. In stark contrast, all mice treated with isopropanol survived without any signs of the syndrome. Our results have clinical implications given the possibility that patients acutely intoxicated with isopropanol may also be acutely immunosuppressed. This scenario should be considered in cases of severe poisoning, especially if underlying infection and/or trauma is present. Moreover, the general consumer has easy access to a wide range of products that contain high concentrations of isopropanol such as hand sanitizers and rubbing alcohol. Isopropanol is readily absorbed by the gastrointestinal tract but its absorption through intact adult skin is very poor (Brown et al., 2007; Kirschner et al., 2009; Kraut and Kurtz 2008). With that in mind, limited transdermal absorption may occur with an estimated skin -4 permeation coefficient (kp) of 4-15 x 10 cm/h (Boatman et al., 1998; Clewell et al., 2001; Cronin et al., 1999; Frasch, 2002; Turner et al., 2004). Although topical application was documented in a few cases of isopropanol poisoning (Arditi and Killner, 1987; Dyer et al., 2002; Leeper et al., 2000), most conventional topical uses are likely to produce only negligible systemic alcohol levels (Brown et al., 2007; Kirschner et al., 2009). In addition, the direct immunomodulatory impact of isopropanol on healthy skin, if any, is likely to be transitory and confined to resident/infiltrating immune cells. From another standpoint, however, one should perhaps be cautious when applying isopropanol-containing gels or solutions to diseased skin. It should be noticed that a typical 70% alcohol solution is over

113

400 times more concentrated than the minimal biologically active amount in vitro (0.16% or 26 mM). In this regard, a recent report has identified the down-regulation of JunB and the associated up-regulation of IL-6 production in keratinocytes in systemic lupus erythematosus (Pflegerl et al., 2009). As the JunB/IL-6 dysregulation could potentially be exacerbated by isopropanol, one should be prudent in using topical isopropanol in these patients until further investigation is conducted to clarify the risks, if any. The current work extends our previous findings in that it reveals yet another mechanism of isopropanol-induced immunosuppression, which is based on the disablement of downstream events in the TLR4 signaling cascade. Acute exposure to this alcohol dysregulates the effector function of monocyte/macrophages in vitro and compromises the cytokine response to LPS challenge in vivo to an extent that protects animals from otherwise lethal toxic shock. Our data contribute to the existing dismal body of information on the immunotoxicology of isopropanol, one of the most ubiquitous chemicals in the world.

Funding This work was supported by grants from the National Sciences and Engineering Research Council of Canada (327062-07) and the Canada Foundation for Innovation (4087) to P.O.d.C.-L.

Acknowledgements The authors wish to thank: M. Caruso for helpful discussions and critical reading of the manuscript; D. Richard for access to the infrared imaging system; J. P. McNamee for providing the Mono Mac 6 cell line; Y. Fradet and A. Bergeron for providing the P388D1 cell line; D. L’Héreault, S. Comeau and H. Dombrowski for blood collection; C. St-Pierre for microscopy technical support; J. Charron, L. Caron, H. Lambert and E. Chouinard for helpful discussions. The authors declare no financial or commercial conflict of interest.

114

References

Allen, D. L., Uyenishi, J. J., Cleary, A. S., Mehan, R. S., Lindsay, S. F., and Reed, J. M. (2010). Calcineurin activates interleukin-6 transcription in mouse skeletal muscle in vivo and in C2C12 myotubes in vitro. Am. J. Physiol. Regul. Integr. Comp. Physiol. 298, R198- 210. Arditi, M., and Killner, M. S. (1987). Coma following use of rubbing alcohol for fever control. Am. J. Dis. Child. 141, 237-238. Barthel, R., Tsytsykova, A. V., Barczak, A. K., Tsai, E. Y., Dascher, C. C., Brenner, M. B., and Goldfeld, A. E. (2003). Regulation of tumor necrosis factor alpha gene expression by mycobacteria involves the assembly of a unique enhanceosome dependent on the coactivator proteins CBP/p300. Mol. Cell. Biol. 23, 526-533. Bélanger, L. F., Roy, S., Tremblay, M., Brott, B., Steff, A. M., Mourad, W., Hugo, P., Erikson, R., and Charron, J. (2003). Mek2 is dispensable for mouse growth and development. Mol. Cell. Biol. 23, 4778-4787. Blander, J. M., and Medzhitov, R. (2004). Regulation of phagosome maturation by signals from toll-like receptors. Science 304, 1014-1018. Boatman, R. J., Perry, L. G., Fiorica, L. A., English, J. C., Kapp, R. W., Jr., Bevan, C., Tyler, T. R., Banton, M. I., and Wright, G. A. (1998). Dermal absorption and pharmacokinetics of isopropanol in the male and female F-344 rat. Drug Metab. Dispos. 26, 197-202. Boé, D. M., Richens, T. R., Horstmann, S. A., Burnham, E. L., Janssen, W. J., Henson, P. M., Moss, M., and Vandivier, R. W. (2010). Acute and chronic alcohol exposure impair the phagocytosis of apoptotic cells and enhance the pulmonary inflammatory response. Alcohol Clin. Exp. Res. 34, 1723-1732. Bronstein, A. C., Spyker, D. A., Cantilena, L. R., Jr., Green, J. L., Rumack, B. H., and Giffin, S. L. (2010). 2009 Annual Report of the American Association of Poison Control Centers' National Poison Data System (NPDS): 27th Annual Report. Clin. Toxicol. 48, 979- 1178.

115

Brown, T. L., Gamon, S., Tester, P., Martin, R., Hosking, K., Bowkett, G. C., Gerostamoulos, D., and Grayson, M. L. (2007). Can alcohol-based hand-rub solutions cause you to lose your driver's license? Comparative cutaneous absorption of various alcohols. Antimicrob. Agents Chemother. 51, 1107-1108. Burleigh-Flayer, H. D., Gill, M. W., Strother, D. E., Masten, L. W., McKee, R. H., Tyler, T. R., and Gardiner, T. (1994). Isopropanol 13-week vapor inhalation study in rats and mice with neurotoxicity evaluation in rats. Fundam. Appl. Toxicol. 23, 421-428. Burleigh-Flayer, H., Garman, R., Neptun, D., Bevan, C., Gardiner, T., Kapp, R., Tyler, T., and Wright, G. (1997). Isopropanol vapor inhalation oncogenicity study in Fischer 344 rats and CD-1 mice. Fundam. Appl. Toxicol. 36, 95-111. Cargnello, M., and Roux, P. P. (2011). Activation and function of the MAPKs and their substrates, the MAPK-activated protein kinases. Microbiol. Mol. Biol. Rev. 75, 50-83. Chuderland, D., and Seger, R. (2005). Protein-protein interactions in the regulation of the extracellular signal-regulated kinase. Mol. Biotechnol. 29, 57-74. Clewell, H. J., 3rd, Gentry, P. R., Gearhart, J. M., Covington, T. R., Banton, M. I., and Andersen, M. E. (2001). Development of a physiologically based pharmacokinetic model of isopropanol and its metabolite acetone. Toxicol. Sci. 63, 160-172. Cronin, M. T., Dearden, J. C., Moss, G. P., and Murray-Dickson, G. (1999). Investigation of the mechanism of flux across human skin in vitro by quantitative structure-permeability relationships. Eur. J. Pharm. Sci. 7, 325-330. Dai, Q., Zhang, J., Pruett, S. B. (2005). Ethanol alters cellular activation and CD14 partitioning in lipid rafts. Biochem. Biophys. Res. Commun. 332, 37-42. Daniel, D. R., McAnalley, B. H., and Garriott, J. C. (1981). Isopropyl alcohol metabolism after acute intoxication in humans. J. Anal. Toxicol. 5, 110-112. da Silva Correia, J., Soldau, K., Christen, U., Tobias, P. S., and Ulevitch, R. J. (2001). Lipopolysaccharide is in close proximity to each of the proteins in its membrane receptor complex. transfer from CD14 to TLR4 and MD-2. J. Biol. Chem. 276, 21129-21135. Dendorfer, U., Oettgen, P., and Libermann, T. A. (1994). Multiple regulatory elements in the interleukin-6 gene mediate induction by prostaglandins, cyclic AMP, and lipopolysaccharide. Mol. Cell. Biol. 14, 4443-4454.

116

Désy, O., Carignan, D., Caruso, M., and de Campos-Lima, P. O. (2008). Immunosuppressive effect of isopropanol: down-regulation of cytokine production results from the alteration of discrete transcriptional pathways in activated lymphocytes. J. Immunol. 181, 2348-2355. Désy, O., Carignan, D., Caruso, M., and de Campos-Lima, P. O. (2010). Methanol induces a discrete transcriptional dysregulation that leads to cytokine overproduction in activated lymphocytes. Toxicol. Sci. 17, 303-313. Dwyer, D. S., and Bradley, R. J. (2000). Chemical properties of alcohols and their protein binding sites. Cell. Mol. Life Sci. 57, 265-275. Dyer, S., Mycyk, M. B., Ahrens, W. R., and Zell-Kanter, M. (2002). Hemorrhagic gastritis from topical isopropanol exposure. Ann. Pharmacother. 36, 1733-1735. Ekwall, B., and Clemedson, C. (1997). Time-related lethal blood concentrations from acute human poisoning of chemicals. Part 2: The monographs. No. 10 Isopropyl alcohol. pp. 1-7. CTLU, Stanga, Sweden. Emadi, A., and Coberly, L. (2007). Intoxication of a hospitalized patient with an isopropanol-based hand sanitizer. N. Engl. J. Med. 356, 530-531. Frasch, H. F. (2002). A random walk model of skin permeation. Risk Anal. 22, 265-276. Fuss, I. J., Kanof, M. E., Smith, P. D., and Zola, H. (2009). Isolation of whole mononuclear cells from peripheral blood and cord blood. Curr. Protoc. Immunol. 7, 7.1. Galanos, C., and Freudenberg, M. A. (1993). Mechanisms of endotoxin shock and endotoxin hypersensitivity. Immunobiology 187, 346-356. García-García, E., Rosales, R., and Rosales, C. (2002). Phosphatidylinositol 3-kinase and extracellular signal-regulated kinase are recruited for Fc receptor-mediated phagocytosis during monocyte-to-macrophage differentiation. J. Leukoc. Biol. 72, 107-114. García-García, E., and Rosales, C. (2002). Signal transduction during Fc receptor-mediated phagocytosis. J. Leukoc. Biol. 72, 1092-1108. Gaudet, M.P., and Fraser, G. L. (1989). Isopropanol ingestion: case report with pharmacokinetic analysis. Am. J. Emerg. Med. 7, 297-299.

117

Gomard, T., Michaud, H. A., Tempé, D., Thiolon, K., Pelegrin, M., and Piechaczyk, M. (2010). An NF-kappaB-dependent role for JunB in the induction of proinflammatory cytokines in LPS-activated bone marrow-derived dendritic cells. PLoS One 5, e9585. Goral, J., Karavitis, J., and Kovacs, E. J. (2008). Exposure-dependent effects of ethanol on the innate immune system. Alcohol 42, 237-247. Goral, J., and Kovacs, E. J. (2005). In vivo ethanol exposure down-regulates TLR2, TLR4, and TLR9-mediated macrophage inflammatory response by limiting p38 and ERK1/2 activation. J. Immunol. 174, 456-463. Guha, M., O'Connell, M. A., Pawlinski, R., Hollis, A., McGovern, P., Yan, S. F., Stern, D., and Mackman, N. (2001). Lipopolysaccharide activation of the MEK-ERK1/2 pathway in human monocytic cells mediates tissue factor and tumor necrosis factor alpha expression by inducing Elk-1 phosphorylation and Egr-1 expression. Blood 98, 1429-1439. Hayden, M. S., and Ghosh, S. (2008). Shared principles in NF-kappaB signaling. Cell 132, 344-362. Huot, J., Lambert, H., Lavoie, J. N., Guimond, A., Houle, F., and Landry, J. (1995). Characterization of 45-kDa/54-kDa HSP27 kinase, a stress-sensitive kinase which may activate the phosphorylation-dependent protective function of mammalian 27-kDa heat- shock protein HSP27. Eur. J. Biochem. 227, 416-427. Israël, A. (2010). The IKK complex, a central regulator of NF-kappaB activation. Cold Spring Harb. Perspect. Biol. 2, a000158. Jehle, A. W., Gardai, S. J., Li, S., Linsel-Nitschke, P., Morimoto, K., Janssen, W. J., Vandivier, R. W., Wang, N., Greenberg, S., Dale, B. M., Qin, C., Henson, P. M., and Tall, A. R. (2006). ATP-binding cassette transporter A7 enhances phagocytosis of apoptotic cells and associated ERK signaling in macrophages. J. Cell. Biol. 174, 547-556. Jung, S., Akabas, M. H., and Harris, R. A. (2005). Functional and structural analysis of the GABAA receptor alpha 1 subunit during channel gating and alcohol modulation. J. Biol. Chem. 280, 308-316. Kapp, R. W., Jr., Bevan, C., Gardiner, T. H., Banton, M. I., Tyler, T. R., and Wright, G. A. (1996). Isopropanol: summary of TSCA test rule studies and relevance to hazard identification. Regul. Toxicol. Pharmacol. 23, 183-192.

118

Karavitis, J., Murdoch, E. L., Gomez, C. R., Ramirez, L., and Kovacs, E. J. (2008). Acute ethanol exposure attenuates pattern recognition receptor activated macrophage functions. J. Interferon Cytokine Res. 28, 413-422. Kasuga, F., Inoue, S., Asano, T., and Kumagai, S. (1992). Effects of isopropanol on the development of inflammatory reactions in rats. Food Chem. Toxicol. 30, 631-634. Kawai, T., and Akira, S. (2010). The role of pattern-recognition receptors in innate immunity: update on Toll-like receptors. Nat. Immunol. 11, 373-384. King, L. H., Jr., Bradley, K. P., and Shires, D. L., Jr. (1970). Hemodialysis for isopropyl alcohol poisoning. JAMA 211, 1855. Kirschner, M. H., Lang, R. A., Breuer, B., Breuer, M., Gronover, C. S., Zwingers, T., Böttrich, J. G., Arndt, A., Brauer, U., Hintzpeter, M., Burmeister, M. A., and Fauteck, J. D. (2009). Transdermal resorption of an ethanol- and 2-propanol-containing skin disinfectant. Langenbecks Arch. Surg. 394, 151-157. Kok, S. H., Hong, C. Y., Kuo, M. Y., Wang, C. C., Hou, K. L., Lin, Y. T., Galson, D. L., and Lin, S. K. (2009). Oncostatin M-induced CCL2 transcription in osteoblastic cells is mediated by multiple levels of STAT-1 and STAT-3 signaling: an implication for the pathogenesis of arthritis. Arthritis Rheum. 60, 1451-1462. Koren, H. S., Handwerger, B. S., and Wunderlich, J. R. (1975). Identification of macrophage-like characteristics in a cultured murine tumor line. J. Immunol. 114, 894-897. Kraut, J. A., and Kurtz, I. (2008). Toxic alcohol ingestions: clinical features, diagnosis, and management. Clin. J. Am. Soc. Nephrol. 3, 208-225. Lacouture, P. G., Wason, S., Abrams, A., and Lovejoy, F. H., Jr. (1983). Acute isopropyl alcohol intoxication. Diagnosis and management. Am. J. Med. 75, 680-686. Lallemand, D., Spyrou, G., Yaniv, M., and Pfarr, C. M. (1997). Variations in Jun and Fos protein expression and AP-1 activity in cycling, resting and stimulated fibroblasts. Oncogene 14, 819-830. Leeper, S. C., Almatari, A. L., Ingram, J. D., and Ferslew, K. E. (2000). Topical absorption of isopropyl alcohol induced cardiac and neurologic deficits in an adult female with intact skin. Vet. Hum. Toxicol. 42, 15-17.

119

Lin, L.L., Wartmann, M., Lin, A. Y., Knopf, J. L., Seth, A., and Davis, R. J. (1993). cPLA2 is phosphorylated and activated by MAP kinase. Cell 72, 269-278. Mansfield, P. J., Shayman, J. A., and Boxer, L. A. (2000). Regulation of polymorphonuclear leukocyte phagocytosis by myosin light chain kinase after activation of mitogen-activated protein kinase. Blood 95, 2407-2412. Martin, T., Cardarelli, P. M., Parry, G. C., Felts, K. A., and Cobb, R. R. (1997). Cytokine induction of monocyte chemoattractant protein-1 gene expression in human endothelial cells depends on the cooperative action of NF-kappa B and AP-1. Eur. J. Immunol. 27, 1091-1097. Meixner, A., Zenz, R., Schonthaler, H. B., Kenner, L., Scheuch, H., Penninger, J. M., and Wagner, E. F. (2008). Epidermal JunB represses G-CSF transcription and affects haematopoiesis and bone formation. Nat. Cell. Biol. 10, 1003-1011. Miyake, K. (2006). Roles for accessory molecules in microbial recognition by Toll-like receptors. J. Endotoxin Res. 12, 195-204. Mueller-Kronast, N., Rabinstein, A. A., Voung, L., and Forteza, A. M. (2003). Isopropanol intoxication mimicking basilar artery thrombosis. Neurology 61, 1456-1457. Murphy, L. O., Smith, S., Chen, R. H., Fingar, D. C., and Blenis, J. (2002). Molecular interpretation of ERK signal duration by immediate early gene products. Nat. Cell Biol. 4, 556-564. Oak, S., Mandrekar, P., Catalano, D., Kodys, K., and Szabo, G. (2006). TLR2- and TLR4- mediated signals determine attenuation or augmentation of inflammation by acute alcohol in monocytes. J. Immunol. 176, 7628-7635. Park, H., Ishihara, D., and Cox, D. (2011). Regulation of tyrosine phosphorylation in macrophage phagocytosis and chemotaxis. Arch. Biochem. Biophys. 510, 101-111. Pflegerl, P., Vesely, P., Hantusch, B., Schlederer, M., Zenz, R., Janig, E., Steiner, G., Meixner, A., Petzelbauer, P., Wolf, P., Soleiman, A., Egger, G., Moriggl, R., Kishimoto, T., Wagner, E. F., and Kenner, L. (2009). Epidermal loss of JunB leads to a SLE phenotype due to hyper IL-6 signaling. Proc. Natl. Acad. Sci. USA 106, 20423-20428.

120

Ren, H., Honse, Y., and Peoples, R. W. (2003). A site of alcohol action in the fourth membrane-associated domain of the N-methyl-D-aspartate receptor. J. Biol. Chem. 278, 48815-48820. Rich, J., Scheife, R. T., Katz, N., and Caplan, L. R. (1990). Isopropyl alcohol intoxication. Arch. Neurol. 47, 322-324. Rosansky, S. J. (1982). Isopropyl alcohol poisoning treated with hemodialysis: kinetics of isopropyl alcohol and acetone removal. J. Toxicol. Clin. Toxicol. 19, 265-271. Saeed, R. W., Varma, S., Peng, T., Tracey, K. J., Sherry, B., and Metz, C. N. (2004). Ethanol blocks leukocyte recruitment and endothelial cell activation in vivo and in vitro. J. Immunol. 173, 6376-6383. Samuel, J. M., Kelberman, D., Smith, A. J., Humphries, S. E., and Woo, P. (2008). Identification of a novel regulatory region in the interleukin-6 gene promoter. Cytokine 42, 256-264. Shahidullah, M., Harris, T., Germann, M. W., and Covarrubias, M. (2003). Molecular features of an alcohol binding site in a neuronal potassium channel. Biochemistry 42, 11243-11252. Shi, L., Kishore, R., McMullen, M. R., and Nagy, L. E. (2002). Lipopolysaccharide stimulation of ERK1/2 increases TNF-alpha production via Egr-1. Am. J. Physiol. Cell. Physiol. 282, C1205-1211. Shyy, J. Y., Lin, M. C., Han, J., Lu, Y., Petrime, M., and Chien, S. (1995). The cis-acting phorbol ester "12-O-tetradecanoylphorbol 13-acetate"-responsive element is involved in shear stress-induced monocyte chemotactic protein 1 gene expression. Proc. Natl. Acad. Sci. USA 92, 8069-8073. Sutcliffe, A. M., Clarke, D. L., Bradbury, D. A., Corbett, L. M., Patel, J. A., and Knox, A. J. (2009). Transcriptional regulation of monocyte chemotactic protein-1 release by endothelin-1 in human airway smooth muscle cells involves NF-kappaB and AP-1. Br. J. Pharmacol. 157, 436-450. Swirski, F. K., Nahrendorf, M., Etzrodt, M., Wildgruber, M., Cortez-Retamozo, V., Panizzi, P., Figueiredo, J. L., Kohler, R. H., Chudnovskiy, A., Waterman, P., Aikawa, E.,

121

Mempel, T. R., Libby, P., Weissleder, R., and Pittet, M. J. (2009). Identification of splenic reservoir monocytes and their deployment to inflammatory sites. Science 325, 612-616. Szabo, G., Dolganiuc, A., Dai, Q., and Pruett, S. B. (2007). TLR4, ethanol, and lipid rafts: a new mechanism of ethanol action with implications for other receptor-mediated effects. J. Immunol. 178, 1243-1249. Szabo, G., and Mandrekar, P. (2008). Human monocytes, macrophages and dendritic cells – Alcohol treatment methods. In Alcohol: Methods and Protocols (L. E. Nagy, Ed.), pp. 113-124. Humana Press, Totowa, NJ. Taieb, J., Delarche, C., Ethuin, F., Selloum, S., Poynard, T., Gougerot-Pocidalo, M. A., and Chollet-Martin, S. (2002). Ethanol-induced inhibition of cytokine release and protein degranulation in human neutrophils. J. Leukoc. Biol. 72, 1142-1147. Tsai, E. Y., Falvo, J. V., Tsytsykova, A. V., Barczak, A. K., Reimold, A. M., Glimcher, L. H., Fenton, M. J., Gordon, D. C., Dunn, I. F., and Goldfeld, A. E. (2000). A lipopolysaccharide-specific enhancer complex involving Ets, Elk-1, Sp1, and CREB binding protein and p300 is recruited to the tumor necrosis factor alpha promoter in vivo. Mol. Cell. Biol. 20, 6084-6094. Tsytsykova, A. V., and Goldfeld, A. E. (2000). Nuclear factor of activated T cells transcription factor NFATp controls superantigen-induced lethal shock. J. Exp. Med. 192, 581-586. Turner, P., Saeed, B., and Kelsey, M. C. (2004). Dermal absorption of isopropyl alcohol from a commercial hand rub: implications for its use in hand decontamination. J. Hosp. Infect. 56, 287-290. Valledor, A. F., Borràs, F. E., Cullell-Young, M., and Celada, A. (1998). Transcription factors that regulate monocyte/macrophage differentiation. J. Leukoc. Biol. 63, 405-417. Vanden Berghe, W., De Bosscher, K., Boone, E., Plaisance, S., and Haegeman, G. (1999). The nuclear factor-kappaB engages CBP/p300 and histone acetyltransferase activity for transcriptional activation of the interleukin-6 gene promoter. J. Biol. Chem. 274, 32091- 32098. von Maltzan, K., and Pruett, S. B. (2011). ELISA assays and alcohol: increasing carbon chain length can interfere with detection of cytokines. Alcohol 45, 1-9.

122

Yoon, S., and Seger, R. (2006). The extracellular signal-regulated kinase: multiple substrates regulate diverse cellular functions. Growth Factors 24, 21-44. Xu, B.-e., Stippec, S., Robinson, F. L., and Cobb, M. H. (2001). Hydrophobic as well as charged residues in both MEK1 and ERK2 are important for their proper docking. J. Biol. Chem. 276, 26509-26515. Zaman, F., Pervez, A., and Abreo, K. (2002). Isopropyl alcohol intoxication: a diagnostic challenge. Am. J. Kidney Dis. 40, E12. Zhao, X. J., Marrero, L., Song, K., Oliver, P., Chin, S. Y., Simon, H., Schurr, J. R., Zhang, Z., Thoppil, D., Lee, S., Nelson, S., and Kolls, J. K. (2003). Acute alcohol inhibits TNF- alpha processing in human monocytes by inhibiting TNF/TNF-alpha-converting enzyme interactions in the cell membrane. J. Immunol. 170, 2923-2931. Ziegler-Heitbrock, H. W. L., Thiel, E., Fiitterer, A., Herzog, V., Wirtz, A., and Riethmiiller, G. (1988). Establishment of a human cell line (Mono Mac 6) with characteristics of mature monocytes. Int. J. Cancer 41, 456-461.

123

Figures and Legends

Figure 4.1 - Biological effect of isopropanol treatment in vitro on human monocytes.

Primary monocytes were stimulated with LPS for 24 h in presence of 0.16%, 0.3% and 0.6% (w/v) isopropanol. The concentrations of TNF-, IL-1, IL-6, IL-8 and CCL2 in the supernatants were measured by ELISA and are depicted as means ± SEM (ns: p > 0.05, **

124 p < 0.01 relative to the (+) control group; n: 4 for TNF-, n: 3 for IL-6/IL-8, n: 5 for IL- 1/CCL2). Figure symbols: (+) indicates cell activation in absence of isopropanol; (-) represents the unstimulated control in absence of isopropanol. IPA indicates isopropanol.

125

Figure 4.2 - Biological effect of isopropanol treatment in vitro on macrophages.

(A) Isopropanol treatment affects the cytokine production of human primary macrophages: Primary cultures of human macrophages (M) and monocytes (Mo) were stimulated with LPS in presence or absence of 0.6% (w/v) isopropanol and the supernatants were harvested after 24 h for measurement of TNF-, IL-6, and CCL2 by ELISA. Concentrations are shown as means ± SEM. The positive controls (LPS-stimulation in absence of isopropanol)

126 are shown as black columns; the negative controls (unstimulated cells in absence of isopropanol) are shown as white columns; and the experimental samples (LPS-stimulation in presence of isopropanol) are shown as grey columns (ns: p > 0.05, * p < 0.05, ** p < 0.01 relative to the positive control (black) group in each histogram; n: 4 for  TNF- , n: 3 for  IL-6, n: 5 for Mo CCL2 and n: 3 for M CCL2). The arrows indicate the mean cytokine production in alcohol-treated unstimulated macrophages. (B) Isopropanol interferes with macrophage phagocytosis: P388D1 cells were incubated with pHrodo E. coli for 2 h in presence or absence of 0.6% (w/v) isopropanol. These bioparticles fluoresce in acidic environment upon internalization. The panels show representative micrographs of cells without bioparticles (B-I), of cells incubated with bioparticles on ice in absence (B-II) or presence (B-III) of isopropanol, and of cells incubated with bioparticles at 37oC in absence (B-IV) or presence (B-V) of isopropanol. (C) Phagocytosis quantification by microscopy: P388D1 cells were incubated with pHrodo E. coli at 37oC for 2 h in presence of 0.16%, 0.3% and 0.6% (w/v) isopropanol. The fluorescence intensity values above background measured with the MetaVUE software are shown as means ± SEM (MFI; ns: p > 0.05, ** p < 0.01 relative to the (+) control group; n: 4). Background levels were established by incubating P388D1 cells with bioparticles in absence of isopropanol on ice for 2 h. (+) indicates incubation of P388D1 cells with bioparticles and without isopropanol at 37oC for 2 h. IPA indicates isopropanol. (D) Phagocytosis quantification by flow cytometry: P388D1 cells were incubated with Alexa Fluor 488 E. coli bioparticles at 37oC for 1 h in presence of 0.16%, 0.3% and 0.6% (w/v) isopropanol. Quenching of surface-bound bacteria was performed as described in materials and methods. The mean fluorescence intensity ± SEM is presented (MFI; ns: p > 0.05, ** p < 0.01 relative to the (+) control group; n: 6). (-) indicates incubation of P388D1 cells in absence of bioparticles and without isopropanol at 37oC for 1 h; ICE indicates incubation of P388D1 cells with bioparticles and without isopropanol on ice for 1 h. IPA indicates isopropanol.

127

Figure 4.3 - Isopropanol acts downstream of the cell membrane and does not compromise the NF-B signaling pathway.

(A) Isopropanol does not interfere with the activation of IRAK1: Mono Mac 6 cells were stimulated with LPS for 15 min in presence ((+)IPA) or absence (+) of 0.6% (w/v)

128 isopropanol; cell lysates were equalized for protein content and were used for IRAK1 immunoprecipitation as described in materials and methods. The enzymatic activity in the immunocomplexes was assessed by measuring 32P--ATP incorporation into the myelin basic protein (pMBP) substrate after SDS-PAGE. One representative autoradiograph of four is depicted (Mean densitometric units ± SEM: 16.25 ± 2.0 (+); 15.31 ± 1.8 ((+)IPA), p > 0.05). (-) represents the unstimulated control in absence of isopropanol. (-)IPA represents the unstimulated control in presence of 0.6% (w/v) isopropanol. (B) Isopropanol treatment does not affect IKK/ phosphorylation following LPS stimulation: Human primary monocytes were stimulated for 15 min with LPS in presence ((+)IPA) or absence (+) of 0.6% (w/v) isopropanol and processed for SDS/PAGE. One representative Western blot is shown. The relative quantification of phosphorylated IKK/ (pIKK) in relation to -tubulin expression is depicted in mean densitometric units (DU) ± SEM on the right-hand side histogram (ns: p > 0.05, n: 3). (-) represents the unstimulated control in absence of isopropanol. (-)IPA represents the unstimulated control in presence of 0.6% (w/v) isopropanol. LC indicates loading control (-tubulin). (C) Nuclear translocation of the NF-B p65 subunit is not affected by isopropanol: Primary human monocytes were stimulated with LPS for 1 h in presence or absence of isopropanol. Nuclear extracts were incubated with immobilized NF-B-binding oligonucleotides in 96- well plates; the amount of bound transcription factor was assessed with a p65-specific antibody by ELISA. The nuclear transcription factor ratio was calculated by dividing the sample value by the value of the unstimulated control in absence of isopropanol. Data is presented as means ± SEM (ns: p > 0.05, * p < 0.05, ** p < 0.01 relative to the (+) control group, n: 6). The dashed line represents the baseline of p65 nuclear content in unstimulated cells. (-)IPA represents the unstimulated control in presence of 0.6% (w/v) isopropanol. (+) indicates LPS stimulation in absence of isopropanol. (+)IPA indicates LPS stimulation in presence of 0.6% (w/v) isopropanol. BAY indicates LPS stimulation in presence of the BAY 11-7082 inhibitor compound without isopropanol.

129

Figure 4.4 - Isopropanol induces a selective defect in the MAPK signaling cascade and alters the activation of discrete AP-1 family members.

(A) Isopropanol treatment inhibits the LPS-induced phosphorylation of ERK2 without compromising p38, JNK, and ERK1 activation: Human primary monocytes were stimulated for 15 min with LPS in presence ((+)IPA) or absence (+) of 0.6% (w/v) isopropanol and processed for SDS/PAGE. The upper blots show the expression of the phosphorylated forms (P) of p38, JNK p46/p54 and ERK1/2. The lower blots present the protein loading controls (LC, total p38; -tubulin; and total ERK2). Relative quantifications

130 of the phosphorylated proteins in relation to the relevant loading control are depicted in mean densitometric units (DU) ± SEM under the representative Western blots (ns: p > 0.05, ** p < 0.01; n: 6 for p38, n: 5 for JNK, n: 3 for ERK). (-) represents the unstimulated control in absence of isopropanol. (-)IPA represents the unstimulated control in presence of 0.6% (w/v) isopropanol. (B) Isopropanol inhibits the nuclear translocation of c-Fos and JunB: Primary human monocytes were stimulated with LPS for 1 h in presence of 0.16%, 0.3% and 0.6% (w/v) isopropanol. Nuclear extracts were incubated with immobilized AP-1-binding oligonucleotides in 96-well plates; the amount of bound transcription factor was assessed with a c-Fos- or JunB-specific antibody by ELISA. The nuclear transcription factor ratio was calculated by dividing the sample value by the value of the unstimulated control in absence of isopropanol. Data is presented as means ± SEM (* p < 0.05, ** p < 0.01 relative to the (+) control group; n: 7 for c-Fos, n: 6 for JunB). The dashed line represents the baseline of the relevant transcription factor nuclear content in unstimulated cells. (+) indicates LPS stimulation in absence of isopropanol. (-)IPA represents the unstimulated control in presence of 0.6% (w/v) isopropanol. Inhib. indicates LPS stimulation in presence of PD98059/SB202190/SP600125 kinase inhibitors without isopropanol. IPA indicates isopropanol.

131

Figure 4.5 - Isopropanol-induced immunosuppression in vivo.

(A) TNF- production is virtually stopped in mice acutely exposed to isopropanol: BALB/c mice received 5 g LPS subcutaneously plus the intraperitoneal injection of either 2 g/kg isopropanol (LPS/IPA group) or saline (LPS/pbs group). Control groups received

132 either saline subcutaneously plus isopropanol intraperitoneally (IPA/pbs) or saline only (pbs/pbs). Animals were sacrificed 90 min after injections, and serum TNF-levels were quantified by ELISA. Results are presented as means ± SEM (** p < 0.01 relative to the LPS/pbs group, n: 6/group). (B) Acute exposure to isopropanol does not change the IL-6 production in vivo in response to LPS: BALB/c mice were injected as above and the animals were sacrificed after 180 min. Serum IL-6 was measured by ELISA. Means ± SEM are shown (ns: p > 0.05, ** p < 0.01 relative to the LPS/pbs group, n: 6/group). (C) Isopropanol treatment in vitro does not interfere with the ability of mouse monocytes to produce IL-6 in response to LPS: Purified primary murine monocytes were stimulated with 1 g/ml LPS for 24 h in presence ((+)IPA) or absence (+) of 0.6% (w/v) isopropanol. The levels of IL-6 in the supernatants were measured by ELISA and are shown as means ± SEM (ns: p > 0.05, ** p < 0.01 relative to the (+) control group; n: 3). (-) represents the unstimulated control in absence of isopropanol. (-)IPA represents the unstimulated control in presence of 0.6% (w/v) isopropanol. IPA indicates isopropanol. (D) CCL2 production is compromised in mice acutely exposed to isopropanol: Experimental groups were treated and labeled as in panel A. Animals were sacrificed 180 min after injections, and serum CCL2 levels were quantified by ELISA. Results are presented as means ± SEM (** p < 0.01 relative to the LPS/pbs group, n: 6/group). (E) Isopropanol confers full protection from LPS-induced toxic shock syndrome: BALB/c mice were presensitized with 20 mg D-galactosamine; then, they were injected with 0.2 g LPS subcutaneously plus 2 g/kg isopropanol intraperitoneally (LPS+IPA group). Alternatively, the presensitized animals were injected with 0.2 g LPS subcutaneously plus saline intraperitoneally (LPS+pbs group). The Kaplan-Meier survival curve is presented (p < 0.0001, n: 10/group).

5. The size of the unbranched aliphatic chain determines the immunomodulatory potency of short and long-chain n-alkanols.

Damien Carignan, Olivier Désy, Manuel Caruso and Pedro O. de Campos-Lima* Laval University Cancer Research Center, Quebec City, Quebec, G1R 2J6, Canada

* Correspondence: Dr. Pedro O. de Campos-Lima, Laval University Cancer Research Center, McMahon St 9, Quebec City, QC, G1R 2J6, Canada. Phone: 1 418 525 4444, fax: 1 418 691 5439, e-mail: [email protected]

Short title: Immunomodulation by short and long-chain n-alkanols. Keywords: n-alkanol, alcohol, immunomodulation, T cell, IFN-

134

Résumé

Les alcanols aliphatiques sont des substances ubiquitaires qui affichent des propriétés anesthésiques reliées à leur degré d’hydrophobicité. Une abondante littérature a documentée la capacité de l’éthanol, et d’autres alcools à courte chaîne, à modifier le système immunitaire en interférant avec la fonction de plusieurs types de cellules. Nous avons pensé que, puisque les n-alcanols aliphatiques non-ramifiés sont structurellement très similaires, ils pourraient avoir un impact immunologique qui réflète leur puissance anesthésique. Dans cet article, nous rapportons l’impact de la série homologue des alcools de C1 à C12 sur la capacité des lymphocytes primaires humains à produire de l’IFN-γ une fois activé par leur récepteur à l’antigène. Le méthanol a accru la production d’IFN-γ tandis que les alcools de C2 à C10 ont réduit la sécrétion de cette cytokine. L’activité de la série des n-alcanols a été observée sur une très large gamme de concentrations allant de niveaux de l’ordre des mM pour les alcools à chaîne courte jusqu’à des niveaux de l’ordre des µM pour les alcools C9-C10. Nous avons observé une corrélation claire entre l’activité immunomodulatrice et l’hydrophobicité des composés, mais un effet de coupure était

évident à C11. Les n-alcanols ont agit en aval de la membrane cellulaire puisque la signalisation précoce associée au récepteur à l’antigène a été préservée. L’activation du facteur de transcription NFAT est progressivement altérée en fonction de la taille de la chaîne aliphatique des n-alcanols en suivant une tendance claire vers le bas qui éa été interrompue à C11. L’activation de la voie NF-κB est aussi affectée par les n-alcanols, mais leur effet s’arrête avant, aux environs de C8. Le patron de la dérégulation transcriptionnelle et de l’immunomodulation induite par la série des n-alcanols suggère la présence de poches d’intéraction de dimensions définies sur des cibles protéiques intracellaires qui compromettent l’activation de NFAT et NF-κB et, finalement modulent la fonction effectrice des lymphocytes T. Ce travail contribue à une meilleure compréhension de l’activité biologique des alcanols.

135

Abstract

Aliphatic n-alkanols are ubiquitous substances that display general anesthetic properties in accordance to their degree of hydrophobicity. Extensive literature has documented the capacity of ethanol and other short-chain alcohols to modulate the immune system by interfering with the function of several types of immune cells. We reasoned that because unbranched aliphatic n-alkanols are structurally very similar they might have an immunological impact that mirrors their anesthetic potency. In this article, we report the impact of the homologous C1-C12 alcohol series on the ability of human primary lymphocytes to produce IFN- once activated by the T cell receptor. Methanol enhanced

IFN- production while C2-C10 alcohols reduced the release of this cytokine. The activity of the n-alkanol series was observed within a very wide concentration window ranging from mM levels for short-chain alcohols to M amounts in the case of C9-C10 alcohols. There was a clear correlation between immunomodulatory activity and hydrophobicity of the compounds but a cutoff effect was evident at C11. n-Alkanols were shown to act downstream of the cell membrane because T cell receptor early signaling was preserved. The activation of the nuclear factor of activated T cells (NFAT) was down-regulated progressively in accordance to the size of the n-alkanol aliphatic chains with a clear downward trend that was interrupted at C11. The nuclear factor-B (NF-B) signaling was also compromised but the cutoff appeared earlier in the vicinity of C8. The pattern of immunomodulation and transcriptional dysregulation induced by the n-alkanol series suggested the existence of interaction pockets of defined dimensions within intracellular targets that compromise the activation of NFAT and NF-B transcription factors and ultimately modulate the effector function of the T lymphocyte. This work contributes to a better understanding of the biological activity of n-alkanols.

Introduction

Aliphatic n-alkanols constitute a large family of ubiquitous substances composed of short- chain (C1-C5) and long-chain (C6-C22) alcohols that are used in a variety of domestic and

136 industrial applications (Désy et al., 2012; Veenstra, 2009). Extensive literature has documented the capacity of ethanol to modulate the immune system directly or indirectly by interfering with the function of T lymphocytes, monocytes, macrophages, dendridic cells, neutrophils, and endothelial cells (Goral and Kovacs, 2005; Oak et al., 2006; Saeed et al., 2004; Szabo et al., 2007; Taieb et al., 2002; Zhao et al., 2003). Recent results have provided evidence that two other short-chain alcohols also possess a discernible immunomodulatory footprint. Thus, isopropanol was shown to down-regulate the effector function of T cells, NK cells and monocytes (Désy et al., 2008; Carignan et al., 2012). Conversely, methanol enhances the inflammatory cytokine release from activated T lymphocytes (Désy et al., 2010). A common thread in the mechanisms that underlie the biological effect of the above short-chain alcohols is their obvious impact on transcriptional pathways that are important for immune cell function. Many of the effects of ethanol on the immune system have been associated to the dysfunctional activation of the nuclear factor- B (NF-B) (Oak et al., 2006; Saeed et al., 2004; Szabo et al., 2007), and at least in lipopolysaccharide-activated macrophages, these biological consequences have been suggested to result from a change in the dynamics of protein recruitment into rafts on the cell membrane (Dai et al., 2005; Szabo et al., 2007). Instead, the impact of isopropanol and methanol initiates downstream of the cell membrane and is mediated by the dysregulation of distinct members of the nuclear factor of activated T cells (NFAT) family of transcription factors with or without additional involvement of the activator protein-1 (AP- 1) (Désy et al., 2012). The work described in this article was set off by two observations: first, unbranched aliphatic n-alkanols display general anesthetic properties that correlate with their degree of hydrophobicity (McCreery and Hunt 1978; Pringle et al., 1981; Franks and Lieb 1984, 1985); and second, our own preliminary data shows that isopropanol is substantially more effective than ethanol in down-regulating the effector function of T lymphocytes. We reasoned that because unbranched aliphatic n-alkanols are structurally very similar they might have an immunological impact that mirrors their anesthetic potency observed within the C1-C12 range. We have chosen the production of IFN- by human primary T lymphocytes as the readout to test our hypothesis because this cytokine is essential for the

137 innate and adaptive immune response (Billiau and Matthys 2009). The present article reports that indeed the size of the aliphatic chains determines the immunomodulatory potency of n-alkanols. Furthermore, our results suggest the existence of discrete molecular targets downstream of the cell membrane, which display defined alcohol interaction cutoffs.

Materials and Methods

Cell isolation, culture, and stimulation. This study was approved by the Institutional Clinical Research Ethics Committee (L’Hôtel-Dieu de Québec/Centre hospitalier universitaire de Québec - L’HDQ-CHUQ). Mononuclear cells were prepared from the peripheral blood from healthy volunteers by density gradient centrifugation using Ficoll- Hypaque (GE Healthcare, Piscataway, NJ). Written informed consent was obtained from all donors. Monocyte-depleted populations were more than 95% pure human T cells (CD8+/CD4+) and were kept in complete medium: RPMI 1640 (Invitrogen Canada, Burlington, Canada) supplemented with 10% heat-inactivated FBS (BioCell Inc., Drummondville, Canada). T cells were activated with anti-CD3/CD28 antibody-coated magnetic beads (Invitrogen) at 37°C for 6 h. When indicated, three alternative T cell activation protocols were used: (i) Pre-treatment for 20 min on ice with 1 g/mL mouse anti-human CD3 monoclonal antibody (CD3-2, Mabtech, Nacka Strand, Sweden) and 2.5 g/mL mouse anti-human CD28 (CD28.2, Biolegend, San Diego, CA), followed by incubation at 37°C for 3 min with a 10-fold excess of goat anti-mouse IgG (Sigma); (ii) Treatment was identical to the one described in (i) except for the incubation step at 37°C that was performed for 5 h; and (iii) Treatment with 10 ng/ml phorbol 12-myristate 13-acetate (PMA) and 200 ng/ml ionomycin (Sigma, St Louis, MO) for 5 h.

Alcohols. 99.9% Methanol (C1) was purchased from Fisher Scientific, Pittsburgh, PA,

USA. > 99% pure Ethanol (C2) was purchased from Commercial Alcohols, Brampton,

138

Canada. 99.5% 1-butanol (C4) was purchased from BDH (Toronto, Canada). 99.9% 1- propanol (C3), 99% 1-pentanol (C5), 99% 1-hexanol (C6), 99% 1-heptanol (C7), 99% 1- octanol (C8), 98% 1-nonanol (C9), 99% 1-decanol (C10), 99% 1-undecanol (C11), and 98%

1-dodecanol (C12) were purchased from Alfa-Aesar, Ward Hill, MA, USA. A two- to five- times concentrated stock solution of C1-C7 alcohols was prepared in serum-free RPMI and diluted to the final experimental concentration. For C8-C12 alcohols, a stock solution was first prepared in DMSO and further diluted in assay medium to the desired experimental concentration with extensive vortexing. The final DMSO concentration was never higher than 0.2%.

Cytokine analysis. Measurements of human IFN- in cell culture supernatants were performed with the ELISA MAX™ Standard kit according to the manufacturer’s instructions (BioLegend). Supernatants from human cells were diluted 1:100; thus, only negligible alcohol amounts were still present at the time of processing of diluted supernatants. These trace quantities had no impact on the assay itself. Briefly, 96-well plates were coated with the relevant capture antibody and incubated with serially diluted standards or unknown samples; then, they were washed and incubated with the biotinylated detection antibody followed by streptavidin-horseradish peroxidase. The plates were read at 450 nm after sequential treatment with 3,3',5,5'-tetramethylbenzidine (TMB) substrate solution and phosphoric acid.

ELISA-Based Transcription Factor Activation Assay. Cells were washed in alcohol-free buffer before the generation of nuclear lysates. Nuclear proteins were extracted using the Active Motif Nuclear Extract Kit (Active Motif, Carlsbad, CA) according to the manufacturer’s instructions; the total protein concentration of the lysates was determined by the Bradford assay (Bio-Rad, Hercules, CA). p65 activation was measured with the ELISA-based TransAM NF-B Kit by following to the manufacturer’s instructions (Active Motif). Briefly, nuclear extracts were incubated with plate-bound p65–specific oligonucleotides; the plates were washed and further incubated with p65–specific antibodies. Addition of a horseradish-conjugated secondary antibody and the 3,3’,5,5’-

139 tetramethylbenzidine substrate produced a colorimetric reaction measurable in a spectrophotometer.

Western blot. For ZAP-70 analysis, purified T cells were activated for 3 minutes at 37°C by anti-CD3/anti-CD28/anti-IgG antibodies as described above with or without the relevant alcohol at the indicated molar concentration. The cells were washed in alcohol-free buffer and lysed in sodium dodecyl sulfate (SDS) sample buffer (2% w/v SDS, 0.25 M - mercaptoethanol, 10% v/v glycerol, 0.05 M Tris-HCl, pH 6.8, 0.004% w/v bromophenol blue); lysates were separated in 10% polyacrylamide gels and blotted onto nitrocellulose filters (Hybond-C, GE Healthcare, Piscataway, NJ). The membranes were first probed with ZAP-70-specific antibodies: rabbit anti-human ZAP-70 (99F2, 1/1000, Cell Signaling Technology, Danvers, MA) and mouse anti-human ZAP-70 (pY319)/Syk (pY352) (17a, 1/5000, BD Biosciences, Mississauga, Canada); For NFATc1 analysis, purified T cells were activated for 5 h by PMA/ionomycin as described above with or without the relevant alcohol at the indicated molar concentration. Nuclear extracts were prepared with the Active Motif kit; samples were separated in 7.5% polyacrylamide gels and blotted onto nitrocellulose filters (Hybond-C, GE Healthcare, Piscataway, NJ). The membranes were first probed with mouse anti-human NFATc1 monoclonal antibody (7A6, 1/500, Biolegend) and rabbit anti-human HDAC1 polyclonal antibody (H-51, 1/2000, Santa Cruz Biotechnology, Santa Cruz, CA). Blots were then washed and incubated with 1/15000 dilutions of the antibodies IRDye 680 goat anti-rabbit IgG and IRDye 800CW goat anti-mouse IgG (Li-Cor Biosciences, Lincoln, NE). Detection and quantification was performed with the Odyssey Infrared Imaging System (Li-Cor Biosystems).

Luciferase assay. The generation of Jurkat cells carrying the firefly luciferase gene driven by the NFAT synthetic promoter is described elsewhere (Désy et al., 2008). The stable Jurkat cell line expressing the constitutive luciferase was produced by transduction with an MFG vector that carries the luciferase gene driven by the Moloney murine leukemia virus long terminal repeat (LTR) (Qiao et al. 2002). Jurkat cells carrying the inducible or

140 constitutive luciferase were stimulated with PMA/ionomycin with or without methanol treatment as indicated in the text. Lysates for luciferase assays were prepared with the passive lysis buffer (E1941; Promega, Madison, WI) and analyzed in a Lumat 9501 luminometer (Berthold, Nashua, NH). Relative luciferase units were calculated in relation to the unstimulated negative control after normalization to total protein content measured by the Bradford assay (Bio-Rad).

Statistical analysis. One-way ANOVA followed by Dunnett’s multiple-comparison posttest was performed with GraphPad Prism (GraphPad Software Inc., San Diego, CA) on data presented in table 5.1 and figure 5.3. p values < 0.05 were considered significant.

Results

Primary alcohols modulate the secretion of interferon- by human primary lymphocytes according to the size of their aliphatic chain One major effector function of T cells triggered by the engagement of their antigen receptor is the synthesis of IFN- (Billiau and Matthys 2009). In the present study, we have investigated whether exposure in vitro to a panel of primary aliphatic alcohols spanning the

C1-C12 range would have any impact on the ability of human primary lymphocytes to produce IFN- once activated by the T cell receptor. Preliminary assays were performed to determine the optimal concentration window for analysis of each alcohol, which were then tested in 3-6 independent experiments with cells isolated from different donors. Table 5.1 illustrates the results obtained with five alcohol concentration points starting from no alcohol (group 0) to the highest concentration (group IV). The mean viability of cells exposed to the highest concentration of each alcohol was always > 97% as indicated. Three distinct outcomes were observed. As anticipated, the release of IFN- was up-regulated by methanol exposure with about 70% increase measured in the culture supernatants by

ELISA at the highest alcohol concentration. In contrast, T lymphocytes treated with C2-C10

141 alcohols produced lower amounts of this cytokine, with inhibition levels of 68-96% recorded at the highest concentration tested. C11-C12 alcohols had little or no impact on IFN- secretion and the only significant effect was observed with the highest 1-undecanol concentration, which led to 30% reduction in cytokine release. The biological effect of alcohols is also depicted in figure 5.1. Upon examination of the plotted curves, it becomes apparent that although all C2-C10 alcohols exhibit a clear inhibitory impact on IFN- production, they do so at very different concentrations in accordance with the size of their carbon chain. Thus, short-chain alcohols are effective in low mM levels, while C9-C10 alcohols work at M amounts.

The homologous series of primary alcohols has an immunomodulatory activity profile with a plateau in the mid range and a discrete cutoff effect It is well established that the anesthetic potency of the homologous series of primary alcohols as measured by the loss of the righting reflex in tadpoles exhibits progressive intensity that parallels the increase in size of their carbon chains. This pattern, however, is abruptly interrupted at C12 with substantial loss of activity at C13 and the observation of a virtual disappearance of activity at C14 and longer alcohols (Pringle et al., 1981). In addition, the potency levels off at C6-C7 in a lipid-free in vitro model of anesthetic-protein interaction (Franks and Lieb, 1985, 2004). In order to better characterize the profile of immunomodulatory activity of the alcohol series and to verify how close it resembles the pattern of anesthetic potency, the concentration values associated to 50% of the biological effect of each alcohol were calculated from the corresponding curves in figure 5.1 and then plotted against the maximal achievable concentration in aqueous media in figure 5.2. The

50% effect was generally within each experimental curve; in the case of C11, the last experimentally achievable data point corresponded to 30% inhibition and higher concentrations were beyond its maximal solubility. C12 was totally ineffective. In agreement with the data obtained in the studies on anesthetic potency, examination of the resulting curve for the immunomodulatory potency of primary alcohols revealed a qualitative match with a plateau at C5-C6 and a cutoff at C11.

142

The homologous series of primary alcohols shares the lack of impact on early signaling events triggered by activation of the T cell receptor Our previous work has shown that methanol and isopropanol initiate their biological effect downstream of the cell membrane in lymphocytes (Désy et al., 2008, 2010). It remained possible, however, that the longer C4-C11 alcohols would act on the cell membrane in a way reminiscent of the effect of ethanol on ion channels and neurotransmitter receptors (Jung et al., 2005; Aryal et al., 2009). In this scenario, longer n-alkanols could plausibly interact with the T cell receptor directly and blunt antigen-dependent signal transduction. We have tested this possibility by checking the phosphorylation status of ZAP-70, a major player in early T cell signaling. As shown in figure 5.3, none of the C4-C11 alcohols interferes with signaling through the T cell receptor as demonstrated by their lack of effect on ZAP-70 activation following anti-CD3 antibody cross-linking. A similar picture was obtained by the analysis of a downstream target of ZAP-70, the Linker of Activated T Cells or LAT, whose phosphorylation was not affected by the above n-alkanols (data not shown).

The homologous series of primary alcohols reduce the activation and nuclear content of the nuclear factor of activated T cells in primary human lymphocytes according to the length of their aliphatic chains We have previously shown that the secondary alcohol isopropanol mediates the attenuation of lymphocyte effector functions, including the capacity to secrete IFN-, by down- regulating the activation of NFATc1. As the experiments reported here revealed that primary alcohols with aliphatic chains spanning C2-C11 similarly reduce IFN- production in human primary lymphocytes, the underlying dysregulation of the activation of this transcription factor was a plausible assumption. We have set out to examine this possibility by measuring NFATc1 in nuclear lysates from lymphocytes that have been activated in the presence of each one of the various n-alkanols. Moreover, we have chosen to activate the T cells with a stimulus that bypasses the membrane to further corroborate the view that the studied chemicals work without relying on early TCR signaling. Indeed, the Western blot analysis in figure 5.4, panel A, shows that ionophore-activated lymphocytes display lower amounts of nuclear NFATc1 when exposed to n-alkanols. The concentrations used were

143 those capable of inducing 75% of the biological effect as measured by IFN- secretion and ranged from mM amounts for the short-chain molecules to M levels for the longer moieties as described in the figure legend. NFATc1 levels in C7-C10-treated samples hovered around half the maximal content of activated T cells (55.8% for C7, 50.8% for C8,

54% for C9, 58.8% C10). This inhibitory activity on NFATc1 activation was not significantly found in activated lymphocytes exposed to C11, which exhibited close to 75% of the maximal content of activated T cells. Next, we have confirmed the above results by using a more sensitive luciferase assay with a stable T cell line carrying an NFAT-responsive promoter. This time we have activated the T cells through the antigen and co-stimulatory receptors with anti-CD3/CD28 antibody- coated beads. Figure 5.4, panel B, shows a significant downward trend in promoter activity (ANOVA p < 0.0001) that parallels the increase in size of the alcohol aliphatic chains. The percentage inhibition of NFAT activation measured in this assay was 34.2%, 36.6%,

48.2%, 57.4%, 53.4%, and 56.5% for C6, C7, C8, C9, C10, and C11, respectively. It is noteworthy that alkanols may directly affect the luciferase activity in certain experimental conditions. However, we have conducted the luciferase assays in absence of alkanols with lysates prepared from washed cells. Moreover, to further validate our results, we have tested the effect of each alcohol on a stable T cell line that carries an integrated cassette in which the luciferase gene is driven by a constitutive promoter. Figure 5.4, panel

C, shows that C1-C9 and C12 did not change the constitutive luciferase activity in our experimental conditions. Instead, we noticed that C10 and C11 change the luciferase activity, and C10 did so significantly. Thus, in the specific case of C10 and C11, the luciferase assay is not conclusive as to their impact on NFAT activation. However, the Western blot analysis in figure 5.4, panel A, complements these results by revealing that C10 does indeed reduce NFATc1 in the nucleus. Similarly to the Western blot analysis, the alcohol concentrations used in the luciferase assays were progressively lower so that the longer the alcohol the lower the amount needed to produce the observed effect. Given the higher sensitivity of the luciferase assay, we have chosen the concentrations that are depicted in figure 5.2, which are required to produce

50% of the biological effect as measured by IFN- release in lymphocytes. Thus, while C6

144

reduced the NFAT-responsive promoter activity by one third, the dampening effect of C9 reached about 60%. Nevertheless, 1.4 mM of C6 and only 106 M of C9 were needed to produce this outcome, thereby stressing the higher potency of the long-chain molecules.

Nuclear translocation of the nuclear factor-B is affected by n-alkanols in activated T cells Although the dysregulation of the nuclear factor-B (NF-B) signaling cascade seems to play an important part in the immunological effects of ethanol, we have previously found that this pathway is not altered in the case of two other short-chain alcohols, methanol and isopropanol. In order to dissect further the immunomodulation mechanism of the homologous series of primary alcohols, we have measured the nuclear translocation of p65 in TCR-stimulated purified human T cells exposed to the same alcohol concentrations previously used for Western blot analysis. The p65 protein is the Rel transactivating component of the major and most common NF-B heterodimer. Figure 5.5 shows that lymphocyte activation by anti-CD3/CD28 antibodies led to a 1.6-fold increase in the amount of p65 in the nucleus. Conversely, the same stimulation in the presence of 860 M C7 barely moved the p65 nuclear content, leading to a dismal increase that represented 12.8% of the maximal p65 nuclear content above the unstimulated cell baseline. The effect of the homologous series of primary alcohols was dependent on the size of the aliphatic chain and produced a V shaped curve with the highest inhibition observed in the middle for C6-C8 and little or no effect detected for C1-C3 on the one end and C11-C12 on the other.

Discussion

This article reports three novel findings: (i) the immunomodulatory capacity of the homologous series of n-alkanols displays a clear correlation to hydrophobicity and is reminiscent of their well-established general anesthetic potency; (ii) the pattern of immunomodulatory activity of the alcohol series suggests the existence of an interaction pocket of defined dimensions; and (iii) the immunomodulatory target(s) of n-alkanols is

145

(are) located downstream of the cell membrane. Most of the research on the mechanism of action of alcohols and other general anesthetics had first been focused on the cell membrane and finds its roots in the independent contributions of Hans H. Meyer and Charles E. Overton over a century ago (Meyer 1899; Overton 1901). Their seminal work established the basis for one of the most tested correlations in biomedicine, namely, that between anesthetic potency of a given compound and its oil: water partition coefficient. In face of the strong experimental evidence in support of this correlation, it was natural to assume that anesthetics (and by extension alcohols) would work under a unifying theory in which the dissolution of lipophilic molecules in the lipid bilayer could modify its physical properties and compromise indirectly the function of embedded proteins (Eckenhoff 2001; Seeman 1972). Nevertheless, several incongruities have cast doubts about the validity of the different flavors of the lipid theory to explain the mechanism of action of alcohols and anesthetics (Franks and Lieb 1982). To single out a few, alcohol-induced changes in lipid phase transitions are minute and comparable in magnitude to the effect of mild hyperthermia, which obviously lacks the consequences predicted by the lipid theory (Eckenhoff 2001). Another inconsistence was revealed by the finding of substantial differences in biological activity between stereoisomers that are supposed to have the same impact on lipid phase transition (Dickinson et al., 2000). Also, a cutoff effect was evident when a homologous series of n-alkanols (or n-alkanes) was tested with increasing anesthetic activity being observed up to a discrete alcohol size after which it disappeared even though the larger ineffective molecules were highly lipophilic (Franks and Lieb 1985). Finally, compounds that possess high lipid solubility and fail to display anesthetic activity regardless of their size have been described in obvious discordance with the central tenet of the lipid theory (Koblin et al., 1994). On the one hand, our results indicate that the Meyer-Overton correlation do apply to the alcohol immunomodulatory activity, but on the other hand, they do not lend support to the interpretation that membrane alterations are the major underlying process. First, there is a clear cutoff effect, and second, the inability of these molecules, including the longer more hydrophobic C9-C10 alcohols, to change ZAP-70 phosphorylation strongly suggests that by

146 whatever means they operate they do so downstream of the cell membrane. Previous studies reported putative alcohol-binding sites in several proteins, and there are now crystallographic data for some of them, including ion channels, enzymes, and the odorant binding protein LUSH (Franks et al., 1998; Kruse et al., 2003; Nury et al., 2011; Pegan et al., 2006; Ramaswamy et al., 1994). A feature that was consistently found in these structural analyses of the alcohol-binding site was the identification of hydrogen bond acceptor site(s) and of a hydrophobic groove in close vicinity (Désy et al., 2012). Thus, it is plausible to interpret the Meyer-Overton correlation without resorting to lipid solubility if one applies their concept to the context of hydrophobic protein subdomains. In this scenario, alcohols or similar molecules would dislodge water from hydrophobic pockets and use hydrogen bonds and van der Waals forces to stabilize their binding with potential conformational and functional consequences (Klemm 1998). The hydrophobic nature of the molecules would still dictate the outcome of the interaction and the volume constraints of the relevant cavities would account for the cutoff effect. It is noteworthy that about half of the molecular mass of the cell membrane is in fact protein (Eckenhoff 2001). Nevertheless, at least in principle, all cellular compartments could provide suitable protein targets for alcohols. One of the first and arguably most convincing pieces of evidence that proteins may be the actual mechanistic targets of alcohol action is the demonstrated ability of the homologous series of primary alkanols to inhibit luciferase activity in vitro in absence of a lipid context (Franks and Lieb 1984, 1985). These experiments revealed an almost linear correlation between luciferase inhibition and general anethestic potency. Alcohols were shown to inhibit luciferase activity progressively better from C1 to C6 and then from C8 to C12; there was a clear cutoff effect at C16 and two identifiable activity plateaus between C6-C7 and between C12-C16 in which the binding affinity was nearly the same. These findings implied the existence of a hydrophobic binding pocket in the enzyme with sufficient volume to lodge two 1-hexanol molecules or a single 1-dodecanol molecule (Franks and Lieb 2004). Subsequent X-ray structural studies confirmed the existence of specific binding sites that coincide with the substrate-binding pocket (Franks et al., 1998). There is a striking similarity between the overall profile of the immunomodulatory activity of primary

147 alcohols reported here and that associated to their ability to inhibit luciferase (Franks and Lieb 1985). In both situations, there was a first inflection in the response curve, which leveled off in the mid range at C5-C6 in our case (Fig. 5.2). A second inflection appeared at

C10 in our case and at C12 for luciferase. There was no obvious leveling off of the immunomodulatory activity before or after this point as opposed to the plateau of luciferase activity observed at C12-C16. In both cases, however, the second inflection led to the crossing of the maximum solubility curve at C11 and C16, exactly where the cutoff points for IFN- production and luciferase activity have been experimentally determined to occur, respectively. The analysis of the results reported here and the analogy to the luciferase inhibition data permit us to predict that the alcohol immunomodulatory effect is mediated by the alteration of one (and conceivably more than one) intracellular protein target via interaction to a hydrophobic pocket. This cavity is likely to be sufficiently big to accommodate two molecules of 1-pentanol, as suggested by the leveling off of IFN- release in the mid-range of the curve depicted in figure 5.2, or a single molecule of 1- decanol. Additional methyl groups are likely to contribute to the binding energy by augmenting the van der Waals interactions up to C5. From this point on, two molecules would be too big to fit simultaneously and part of their aliphatic chains would remain exposed to the external polar environment. Efficient binding of single molecules would drive the increase in potency from C7-C10 but steric constraints would resume at C11. It is noteworthy that the amount of C11 required to produce 30% inhibition, which is within the soluble range, is comparable to the concentration of C10 that induces the same biological effect as can be inferred from table 5.1. The complete loss of activity at higher concentrations, however, is not necessarily due to molecular size but to solubility. The immunomodulatory cutoff is lower than that observed for anesthetic potency or for luciferase activity (Franks and Lieb 1985; Pringle et al., 1981), suggesting that the interaction occurs within a somewhat smaller cavity. Given that early signaling seems to be preserved in the presence of all alcohols tested here (fig. 5.3) and that our previous work on short-chain alcohols has shown their exquisite specificity in inducing the dysregulation of transcription factors that play a role in the activation of immunologically relevant genes (Désy et al., 2008, 2010), we favor the

148 hypothesis that the putative intracellular targets are the transcription factors themselves or molecules placed immediately upstream in their signaling cascades. The linear correlation between the size of the aliphatic chain and NFAT inhibition within the C2-C10 range and the loss of activity at C11 are reminiscent of the findings obtained with IFN- release and support the previous assumption of existence of a pocket sufficiently large to accommodate

C10. In the case of NF-B, the results obtained in the p65 activation experiments are compatible with the existence of an additional alcohol-binding site of smaller dimensions within a member of this activation pathway whose maximal capacity would fit C7 or C8.

The reduced NF-B activation would synergize with NFAT inhibition, notably between C6-

C8. The combined end result of n-alkanol action on these two transcriptional pathways would be the progressive disablement of immune cell function between C2-C10 and loss of effect at C11-C12. The analysis of figure 5.1 shows a clear dichotomy between the pattern of response elicited by methanol and that triggered by all the other n-alcohols. Either methanol binds to the same pocket but produces a different conformational change or, perhaps, its smaller size and lower hydrophobicity allow binding to different molecular structures in the same or in a distinct set of targets. In any case, the crystal structure of bromoform bound to luciferase demonstrates that a molecule not much bigger than methanol with a single carbon atom is capable of binding the same cavities that accommodate longer alcohols (Franks et al., 1998). The data reported in this article have confirmed our original hypothesis and revealed that all primary alcohols have an immunological impact that mirrors their anesthetic potency. This effect correlates well with hydrophobicity but the site of their action does not involve the membrane. The discrete alcohol modulation cutoff and the general profile of the IFN- release curve suggest the binding to a hydrophobic pocket in a common intracellular protein target. NFATc1 is a conceivable candidate, whose activation is inhibited by alcohols in a way that closely resembles the reduction in IFN- production and exhibits a similar cutoff. Moreover, our data also reveals an additional target in the NF-B pathway with an earlier cutoff, whose inhibition could synergize with the NFATc1 down-regulation in the nucleus and ultimately reduce IFN- release. The biological effect of alcohols,

149 encompassing the anesthetic and immunomodulatory properties, may result from interactions with many potential molecular targets that display suitable hydrophobic cavities. It is also likely to depend on cell type- and activation state-dependent protein expression profiles. Thus, in principle, our results should be interpreted within the context of the activation of human primary T cells through their antigen receptor. Altogether, our work contributes to a better understanding of the biological activity of the ubiquitous family of n-alkanols and provides additional insight into the mechanism that underlies their impact on the function of T lymphocytes in particular.

Acknowledgements The authors wish to thank S. Comeau and H. Dombrowski for blood collection. The authors declare no financial or commercial conflict of interest. This work was supported by grants from the National Sciences and Engineering Research Council of Canada (327062-07) and the Canada Foundation for Innovation (4087) to P.O.d.C.-L.

References

Aryal, P., Dvir, H., Choe, S., Slesinger, P.A. (2009). A discrete alcohol pocket involved in GIRK channel activation. Nat. Neurosci. 12, 988-995. Bell, G.H. (1973). Solubilities of normal aliphatic acids, alcohols and alkanes in water. Chem. Phys. Lipids 10, 1-10. Billiau, A., Matthys, P. (2009). Interferon-gamma: a historical perspective. Cytokine Growth Factor Rev. 20, 97-113. Carignan, D., Désy, O., de Campos-Lima, P.O. (2012). The dysregulation of the monocyte/macrophage effector function induced by isopropanol is mediated by the defective activation of distinct members of the AP-1 family of transcription factors. Toxicol. Sci. 125:144-156. Dai, Q., Zhang, J., Pruett, S. B. (2005). Ethanol alters cellular activation and CD14 partitioning in lipid rafts. Biochem. Biophys. Res. Commun. 332, 37-42. Désy, O., Carignan, D., Caruso, M., de Campos-Lima, P. O. (2008). Immunosuppressive

150

effect of isopropanol: down-regulation of cytokine production results from the alteration of discrete transcriptional pathways in activated lymphocytes. J. Immunol. 181, 2348- 2355. Désy, O., Carignan, D., Caruso, M., de Campos-Lima, P. O. (2010). Methanol induces a discrete transcriptional dysregulation that leads to cytokine overproduction in activated lymphocytes. Toxicol. Sci. 17, 303-313. Désy, O., Carignan, D., de Campos-Lima, P.O. (2012). Short-term immunological effects of non-ethanolic short-chain alcohols. Toxicol. Lett. 210:44-52. Dickinson, R., White, I., Lieb, W.R., Franks, N.P. (2000). Stereoselective loss of righting reflex in rats by isoflurane. Anesthesiology 93, 837–843. Eckenhoff, R.G. (2001). Promiscuous ligands and attractive cavities: how do the inhaled anesthetics work? Mol. Interv. 1, 258-268. Franks, N.P., Jenkins, A., Conti, E., Lieb, W.R., Brick, P. (1998). Structural basis for the inhibition of firefly luciferase by a general anesthetic. Biophys. J. 75, 2205-2211. Franks, N.P., Lieb, W.R. (1982). Molecular mechanisms of general anaesthesia. Nature 300, 487-493. Franks, N.P., Lieb, W.R. (1984). Do general anaesthetics act by competitive binding to specific receptors? Nature 310, 599-601. Franks, N.P., Lieb, W.R. (1985). Mapping of general anaesthetic target sites provides a molecular basis for cutoff effects. Nature 316, 349-351. Franks, N.P., Lieb, W.R. (2004). Seeing the light: protein theories of general anesthesia. Anesthesiology 101, 235-237. Goral, J., Kovacs, E. J. (2005). In vivo ethanol exposure down-regulates TLR2, TLR4, and TLR9-mediated macrophage inflammatory response by limiting p38 and ERK1/2 activation. J. Immunol. 174, 456-463. Jung, S., Akabas, M. H., Harris, R. A. (2005). Functional and structural analysis of the

GABAA receptor alpha 1 subunit during channel gating and alcohol modulation. J. Biol. Chem. 280, 308-316. Klemm, W.R. (1998). Biological water and its role in the effects of alcohol. Alcohol 15, 249-267.

151

Koblin, D.D., Chortkoff, B.S., Laster, M.J., Eger, E.I. 2nd, Halsey, M.J., Ionescu, P. (1994). Polyhalogenated and perfluorinated compounds that disobey the Meyer-Overton hypothesis. Anesth. Analg. 79, 1043–1048. Kruse, S.W., Zhao, R., Smith, D.P., Jones, D.N. (2003). Structure of a specific alcohol- binding site defined by the odorant binding protein LUSH from Drosophila melanogaster. Nat. Struct. Biol. 10, 694-700. McCreery, M.J., Hunt, W.A., (1978). Physico-chemical correlates of alcohol intoxication. Neuropharmacology 17, 451-461. Meyer, H.H. (1899). Zur Theorie der Alkoholnarkose. Erste Mittheilung. Welche Eigenschaft der Anästhetica bedingt ihre narkotische Wirkung? Arch. Exp. Pathol. Pharmakol. 42, 109-118. Nury, H., Van Renterghem, C., Weng, Y., Tran, A., Baaden, M., Dufresne, V., Changeux, J.P., Sonner, J.M., Delarue, M., Corringer, P.J. (2011). X-ray structures of general anaesthetics bound to a pentameric ligand-gated ion channel. Nature 469, 428-431. Oak, S., Mandrekar, P., Catalano, D., Kodys, K., Szabo, G. (2006). TLR2- and TLR4- mediated signals determine attenuation or augmentation of inflammation by acute alcohol in monocytes. J. Immunol. 176, 7628-7635. Overton, E. (1901). Studien über die Narkose, Zugleich ein Beitrag zur allgemeinen Pharmakologie. Gustav Fischer, Jena, Germany. Pegan, S., Arrabit, C., Slesinger, P.A., Choe, S. (2006). Andersen's syndrome mutation effects on the structure and assembly of the cytoplasmic domains of Kir2.1. Biochemistry 45, 8599-8606. Pringle, M.J., Brown, K.B., Miller, K.W. (1981). Can the lipid theories of anesthesia account for the cutoff in anesthetic potency in homologous series of alcohols? Mol. Pharmacol. 19, 49-55. Qiao, J., Roy, V., Girard, M.H., Caruso, M. (2002). High translation efficiency is mediated by the encephalomyocarditis virus internal ribosomal entry sites if the natural sequence surrounding the eleventh AUG is retained. Hum. Gene Ther. 13, 881-887. Ramaswamy, S., Eklund, H., Plapp, B.V. (1994). Structures of horse liver alcohol dehydrogenase complexed with NAD+ and substituted benzyl alcohols. Biochemistry

152

33, 5230-5237. Saeed, R. W., Varma, S., Peng, T., Tracey, K. J., Sherry, B., Metz, C. N. (2004). Ethanol blocks leukocyte recruitment and endothelial cell activation in vivo and in vitro. J. Immunol. 173, 6376-6383. Seeman, P. (1972). The membrane actions of anesthetics and tranquilizers. Pharmacol. Rev. 24, 583-655. Szabo, G., Dolganiuc, A., Dai, Q., Pruett, S. B. (2007). TLR4, ethanol, and lipid rafts: a new mechanism of ethanol action with implications for other receptor-mediated effects. J. Immunol. 178, 1243-1249. Taieb, J., Delarche, C., Ethuin, F., Selloum, S., Poynard, T., Gougerot-Pocidalo, M. A., Chollet-Martin, S. (2002). Ethanol-induced inhibition of cytokine release and protein degranulation in human neutrophils. J. Leukoc. Biol. 72, 1142-1147. Veenstra, G., Webb, C., Sanderson, H., Belanger, S.E., Fisk, P., Nielsen, A., Kasai, Y., Willing, A., Dyer, S., Penney, D., Certa, H., Stanton, K., Sedlak, R. (2009). Human health risk assessment of long chain alcohols. Ecotoxicol. Environ. Saf. 72, 1016-1030. Zhao, X. J., Marrero, L., Song, K., Oliver, P., Chin, S. Y., Simon, H., Schurr, J. R., Zhang, Z., Thoppil, D., Lee, S., Nelson, S., Kolls, J. K. (2003). Acute alcohol inhibits TNF- alpha processing in human monocytes by inhibiting TNF/TNF-alpha-converting enzyme interactions in the cell membrane. J. Immunol. 170, 2923-2931.

153

Table 5.1 - Correlation between primary alcohol carbon-chain length and T lymphocyte IFN- release§

CONCENTRATION 1-OH 2-OH 3-OH 4-OH 5-OH 6-OH 7-OH 8-OH 9-OH 10-OH 11-OH # GROUP

0 101 ± 2 105 ± 6.7 98.2 ± 4 88.8 ± 2.4 100.3 ± 7 105.2 ± 0.7 100.7 ± 1.9 119.9 ± 8.1 99.7 ± 3.4 109.6 ± 4.1 107.7 ± 7.9

70.5 ± 8.3* 80.8 ± 4.9** 75.1 ± 5** 58.9 ± 7.6** 82.3 ± 3.2* I 122 ± 7.5 ns 110.6 ± 7.6 ns 73.1 ± 8.6 ns 78.8 ± 3.1 ns 96.5 ± 4.8 ns 91.3 ± 6.5 ns

60.9 ± 8.7** 50.4 ± 2.7** 46.9 ± 5.4** 60.2 ± 4.4** 51.1 ± 4** 44.6 ± 7.8** 76.2 ± 3.1** 83.6 ± 2.5** II 134 ± 3 ns 89.2 ± 11.6 ns 90.7 ± 4.9 ns

158.6 ± 14 42.7 ± 3.9 46.9 ± 13.4 25.7 ± 3.7 23.5 ± 4.9 24.1 ± 3.6 24.8 ± 3.9 20.5 ± 2.4 55 ± 4.7 59.9 ± 4 III * ** ** ** ** ** ** ** ** ** 84.7 ± 8 ns

167 ± 19.8 3.4 ± 2.4 13.4 ± 4.1 13.4 ± 4 11.7 ± 3.8 7 ± 2.7 8.2 ± 4.2 3.4 ± 1.4 32.4 ± 7.3 22.9 ± 3 69.8 ± 3.4 IV ** ** ** ** ** ** ** ** ** ** **

∫∫ % VIABILITY 98.9 ± 0.8 99.2 ± 0.2 98.9 ± 0.1 99.2 ± 0.4 97.5 ± 1.6 99.8 ± 0.2 98.9 ± 0.1 99.3 ± 0.3 98.9 ± 1.1 99.4 ± 0.6 99.4 ± 0.6

# T cells were activated by anti-CD3/anti-CD28 antibodies in the absence of alcohols (group 0) or in the absence of alcohols and the presence of 0.2% DMSO (DMSO). Alternatively, T cells were activated by anti-CD3/anti-CD28 antibodies in the presence of the indicated alcohols (groups I-IV). Group I represents the lowest and group IV the highest tested alcohol concentrations with intermediary ascending values in the groups in-between. The experimental molar concentration for each alcohol was: Group I (1-OH, 23.4 mM; 2-OH, 12.5 mM; 3-OH, 6.2 mM; 4-OH, 2.0 mM; 5-OH, 1134 M; 6-OH, 489 M, 7-OH, 215.1 M; 8-OH, 96 M; 9-OH, 21.7 M; 10-OH, 14.6 M; 11-OH, 29 M); Group II (1-OH, 46.8 mM; 2-OH, 25 mM; 3-OH, 6.7 mM; 4-OH, 2.4 mM; 5-OH, 1418 M; 6-OH, 979 M, 7- OH, 430.3 M; 8-OH, 192 M; 9-OH, 43.3 M; 10-OH, 29.2 M; 11-OH, 43.5 M); Group III (1-OH, 93.6 mM; 2-OH, 50 mM; 3-OH, 7.5 mM; 4-OH, 2.7 mM; 5-OH, 1700 M; 6-OH, 1958 M, 7-OH, 860.6 M; 8-OH, 384 M; 9-OH, 86.7M; 10-OH, 58.4 M; 11-OH, 58 M); and Group IV (1-OH, 187 mM; 2-OH, 100 mM; 3-OH, 8.3 mM; 4-OH, 3.0 mM; 5-OH, 1985 M; 6-OH, 3915 M, 7-OH, 1721M; 8-OH, 768M; 9-OH, 173.3M; 10-OH, 116.9M; 11-OH, 75.4M). The results for 12- OH are not shown because there was no effect at all tested concentrations, including that corresponding to its maximal aqueous solubility.

§ Data are shown as percentage of the IFN- release from the TCR-activated DMSO control. Means ± SEM are indicated (ns: p > 0.05, * p < 0.05, ** p < 0.01 relative to group 0; n: 3 for 1-OH, 2-OH, 8- OH, 10-OH and 11-OH; n: 4 for 4-OH, 5-OH and 9-OH; n: 6 for 3-OH, 6-OH and 7-OH).

154

∫∫ Viability was assessed after exposure to the highest alcohol concentration.

Figure symbols: 1-OH (C1), methanol; 2-OH (C2), ethanol; 3-OH (C3), 1-propanol; 4-OH (C4), 1- butanol; 5-OH (C5), 1-pentanol; 6-OH (C6), 1-hexanol; 7-OH (C7), 1-heptanol; 8-OH (C8), 1-octanol; 9-

OH (C9), 1-nonanol; 10-OH (C10), 1-decanol; 11-OH (C11), 1-undecanol; 12-OH (C12), 1-dodecanol.

155

Figures and Legends

Figure 5.1 - Individual profiles of the immunomodulatory activity of primary alcohols.

Primary alcohols alter IFN- production in peripheral blood T lymphocytes: Purified T cells were stimulated with anti-CD3/CD28 antibody-coated beads for 6 h in the presence of C1-

C12 alcohols within the molar concentration windows indicated. The IFN- levels released in the supernatants were measured by ELISA and used to calculate the % alteration in effector function in relation to the positive control activated in absence of alcohols (100%).

Resulting data indicating inhibition of IFN- release for C2-C11 or enhancement of IFN- production for C1 are depicted in the form of curves. The corresponding data points are shown in table 1 and were compiled from independent experiments with samples from at least three different donors for each alcohol as detailed in the table footnote. Panel symbols:

1-OH, C1, methanol; 2-OH, C2, ethanol; 3-OH, C3, 1-propanol; 4-OH, C4, 1-butanol; 5-OH,

C5, 1-pentanol; 6-OH, C6, 1-hexanol; 7-OH, C7, 1-heptanol; 8-OH, C8, 1-octanol; 9-OH, C9,

1-nonanol; 10-OH, C10, 1-decanol; 11-OH, C11, 1-undecanol.

156

Figure 5.2 - Collective profile of the immunomodulatory activity of the homologous series of primary alcohols.

The concentrations required to induce 50% of the immunomodulatory effect were calculated from individual response curves for each alcohol and are depicted as linked empty circles. The maximal aqueous solubility for C5-C12 alcohols was plotted as filled circles in a straight line following the equation described by Bell (Bell 1973). The curve inflections are inside grey boxes. The number of carbons identifies the alcohols.

157

Figure 5.3 - Primary alcohols and TCR early signaling.

Treatment with primary alcohols does not affect ZAP-70 phosphorylation following TCR activation: Peripheral blood T cells were stimulated for 3 min with anti-CD3/CD28/anti- IgG in the absence or the presence of each alcohol at the concentrations that produce 75% of the biological effect as measured by IFN- release (augmentation for C1 and inhibition for C2-C10). The chosen concentrations for C2-C10 were: C1:, C2:, C3:, C4:, C5:, C6:, C7:, C8:,

C9:, C10:, C11:, C12:. The maximal water soluble concentrations were used in the case of

C11-C12,, respectively. Samples were processed for SDS/PAGE and one representative blot of three is shown. Relative quantification in relation to total ZAP-70 is presented underneath as mean densitometric units ± SEM; ns: p > 0.05, ** p < 0.01, n: 3. None of the alcohols had an effect on ZAP-70 phosphorylation in absence of TCR stimulation (not shown).

158

Figure symbols: (-), unstimulated cells in the absence of alcohols; (+), TCR-stimulated cells in the absence of alcohols. TCR-activated samples treated with alcohols are indicated by their number of carbons (1-12). LC: Loading control (Total ZAP-70). OH: Alcohol.

159

Figure 5.4 - Primary alcohols affect NFAT activation.

(A) Measurement of NFATc1 in the nucleus by Western blot. Human peripheral blood T cells were stimulated with PMA/ionomycin for 5 hours in the presence or absence of the relevant alcohol. The chosen concentration for each alcohol were those capable of inducing 75% of the biological effect or the maximal water soluble amount as described in the legend of figure 3. Nuclear extracts were prepared from each sample and processed for SDS/PAGE; one representative blot of four is shown. Relative quantification in relation to histone deacetylase (HDAC-1) is presented underneath as mean densitometric units ± SEM; * p < 0.05, ** p < 0.01 relative to the (+) control group, n: 4. Non significant values, p > 0.05, are not labeled. The relative NFATc1 content in relation to the (+) control group is also presented as % values. The mean of the (+) control group, which represents 100%, and the 50% value are indicated by dashed lines. (B) A) Primary alcohols modulate the activation of a synthetic promoter containing NFAT

160 binding sites. Jurkat-luciferase cells were stimulated with anti-CD3/CD28 antibody-coated beads for 5 h in presence or absence of the relevant alcohol. Grey columns show TCR- activated samples. Black columns represent unstimulated samples. The concentrations used in the experiments were those capable of inducing 50% of the biological effect as measured by IFN- release for C2-C10 or those representing the maximal water solubility for C11-C12 (Figure 2). Samples were lysed and assayed for luciferase activity. Results are presented as mean relative luciferase units per microgram of protein ± SEM; * p < 0.05, ** p < 0.01 relative to the (+) control group, n: 6. Non significant values, p > 0.05, are not labeled. The relative NFATc1 activation in relation to the (+) control group is also presented as % values. The mean of the (+) control group, which represents 100%, and the 50% value are indicated by dashed lines. (C) Impact of primary alcohols on the enzymatic activity of constitutively expressed luciferase. A stable Jurkat subclone carrying the luciferase gene driven by the Moloney LTR was stimulated by anti-CD3/CD28 antibody-coated beads for 5 h in the presence of the same alcohol concentrations used in panel B. Samples were lysed and assayed for luciferase activity. Results are presented as mean relative luciferase units per microgram of protein ± SEM; * p < 0.05 relative to the (+) control group, n: 3. Non significant values, p > 0.05, are not labeled. The relative luciferase activity is also presented as % values in relation to the (+) control group. The mean of the (+) control group, which represents 100%, is indicated by a dashed line.

Figure symbols: (-), unstimulated cells in absence of alcohols; (+), TCR-stimulated cells in absence of alcohols. TCR-activated samples treated with alcohols are indicated by their number of carbons (1-12). D: TCR-stimulated cells in the absence of alcohols and in the presence of 0.2% DMSO. LC: Loading control (HDAC-1). OH: Alcohol.

161

Figure 5.5 - Impact of primary alcohols on NF-B activation in human primary lymphocytes

Treatments with C6-C8 primary alcohols inhibit NF-B activation following TCR activation: Peripheral blood T cells were stimulated for 5 h with anti-CD3/anti-IgG in the absence or the presence of each alcohol at the concentrations that produce 75% of the biological effect as measured by IFN-γ release. The relevant concentrations are listed in the legend of figure 3. Nuclear extracts were incubated with immobilized p65-binding oligonucleotides in 96-well plates; the amount of retained transcription factor was assessed with a p65-specific antibody by ELISA. The nuclear p65 ratio was calculated by dividing the sample value by the value of the unstimulated control in absence of alcohols. Data are presented as means ± SEM; * p < 0.05, ** p < 0.01 relative to the (+) control group, n: 5. Non significant values, p > 0.05, are not labeled. Figure symbols: (+), TCR-stimulated cells in absence of alcohols. (-)OH, unstimulated cells in absence of alcohols and in the presence of C7 (C7 was chosen as a representative alcohol

162 because it is placed in the middle of the alcohol series). TCR-activated samples treated with alcohols are indicated by their number of carbons (1-12). OH: Alcohol.

6. Discussion

Tel que démontré dans les chapitres précédents, l’exposition à l’isopropanol et aux alcools aliphatiques primaires de 2 à 10 carbones est néfaste pour l’activité des lymphocytes in vitro tandis que le méthanol agit en synergie avec les stimuli activateurs pour accroître la production de cytokines par les lymphocytes. L’isopropanol a aussi des effets délétères sur la fonction des monocytes et des macrophages in vitro. Les dérèglements immunologiques induits par le méthanol et l’isopropanol peuvent aussi être observés dans des modèles animaux d’intoxication aiguë comme on peut en juger par la sécrétion anormale de cytokines pro-inflammatoires en réponse à l’entérotoxine B staphylococcique ou au lipopolysaccharide.

Réflexions sur les mécanismes de dérèglement des lymphocytes T par les alcanols

Le méthanol, l’isopropanol et les autres alcools aliphatiques primaires ont tous un impact immunomodulateur marqué sur les lymphocytes T, bien que l’effet du méthanol soit à l’inverse de celui des autres molécules de la famille. D’ailleurs, les résultats du chapitre 5 sont les premiers à montrer un effet délétère de l’éthanol sur une fonction effectrice importante de ces cellules immunitaires : la sécrétion d’IFN-γ. Les lymphocytes T sont un composant important du système immunitaire adaptatif et toute altération de leur fonctionnement normal pourrait engendrer de profondes répercussions sur la santé (Smith- Garvin et al., 2009). Ces cellules utilisent leur récepteur à l’antigène (TCR) pour scruter la surface de cellules présentatrices d’antigènes (APC) à la recherche de peptides du non-soi présenté par les molécules du complexe majeur d’histocompatibilité. Au moment de l’activation, l’interface cellule T-APC subit une réorganisation majeure qui mène à l’agrégation du récepteur à l’antigène, du co-récepteur et de la molécule CD28 d’un côté et de leur ligand respectif de l’autre (Janeway, 2009). Il est concevable que les alcanols puissent interférer avec cette réorganisation massive de molécules membranaires qui précèdent la formation de la synapse immunologique. Cette 164 supposition se base sur la capacité reconnue des alcools à s’immiscer dans les membranes cellulaires (Roth and Seeman, 1972; Nizza and Gawrisch, 2009). De fait, la maturation de la synapse immunologique dépend de la fusion de multiples radeaux lipidiques de tailles nanométriques en domaines membranaires plus imposants de tailles micrométriques. Ces microdomaines membranaires constituent la plateforme opérationnelle pour la signalisation par le TCR (Harder and Sangani, 2009). Selon ce modèle, les différents alcools pourraient perturber les interactions lipides-protéines, désordonner la structure des microdomaines membranaires et interrompre leur coalescence. Le résultat final serait une synapse immunologique déficiente et une activation lymphocytaire anormale. Néanmoins, tel que montré dans les chapitres précédents, la signalisation précoce induite par le TCR et mesurée par la phosphorylation de ZAP70 demeure inchangée par le traitement avec les différents alcanols; ceux-ci initient plutôt leurs effets en aval de la membrane cellulaire. Pour les alcools à courte-chaîne (méthanol, isopropanol, n-propanol, n-butanol), ces observations s’accordent avec le fait que le degré d’insertion d’une molécule d’alcool dans le cœur hydrophobe de la bicouche lipidique dépend directement de la longueur de sa chaîne carbonée. Les alcools plus courts tendent à demeurer à la surface de la membrane, au niveau de l’interface entre les lipides et l’eau (Nizza and Gawrisch, 2009), alors que les alcools à longue chaîne s’intègrent bien dans la bicouche en alignant leur région aliphatique avec les chaînes hydrocarbonées des molécules lipidiques (Westerman et al., 1988). Les courts alcools ne pénétreraient donc pas suffisamment la membrane pour en modifier significativement la fonction. Malgré cela, même les plus longs alcools testés (n-pentanol jusqu’à n-undecanol) ne modifient pas la phosphorylation de ZAP70 ni de LAT. On peut donc en conclure que, bien qu’ils puissent s’insérer dans la membrane, l’effet produit par ces alcools, s’il en est un, n’est pas suffisant pour altérer significativement la formation d’une synapse immunologique pleinement fonctionnelle du côté du lymphocyte T. Les alcanols pourraient aussi exercer leurs effets immunomodulateurs en interagissant directement avec des cibles protéiques. Utilisés à des concentrations élévés (>500mM), ils induisent la formation d’hélices α et peuvent détruire les structures tertiaires menant ainsi à une dénaturation effective (Dwyer and Bradley, 2000). À des concentrations plus physiologiques, ils induisent toutefois des changements structuraux plus subtils qui peuvent

165

être associés à des modifications fonctionnelles de protéines spécifiques telles que celles énumérées en introduction dans le tableau 1.2. Les données expérimentales concernant la phosphorylation de ZAP70 déjà discutées sont incompatibles avec un effet direct des alcanols sur le TCR lui-même ou encore sur une autre protéine participant à l’initiation de la signalisation suite à la reconnaissance de l’antigène. Les cibles potentielles des alcools pourraient donc se retrouver vraisemblablement quelque part dans la chaîne d’événements moléculaires qui transmet le signal d’activation de la membrane plasmique au noyau du lymphocyte. Une fois la kinase ZAP70 phosphorylée, elle phosphoryle à son tour la queue cytoplasmique de la protéine LAT, une protéine adaptatrice transmembranaire, et déclenche ainsi une cascade de phosphorylation suivie par le relargage cytoplasmique de Ca2+ depuis les réserves intracellulaires et par un influx extracellulaire régulé par les canaux calciques activés par le Ca2+ (CRAC) (Gallo et al., 2006; Smith-Garvin et al., 2009). L’augmentation de la concentration intracellulaire de Ca2+ est détectée par la protéine phosphatase calcineurine qui, en compagnie de la calmoduline, active le facteur de transcription NFAT et permet son transport au noyau. Bien qu’il ait été montré que les alcanols altéraient le fonctionnement de certains canaux ioniques (Shahidullah et al., 2003; Aryal et al., 2009), l’isopropanol ne modifie pas l’influx de calcium dépendant des canaux CRAC dans des lymphocytes activés. De plus, l’activité phosphatase de la calcineurine est maintenue intacte en présence de méthanol ou d’isopropanol. Cette observation est importante puisqu’il appert que l’éthanol puisse lier la calmoduline et ainsi accroître son affinité pour le Ca2+ tout en modifiant ses capacités de liaison à la calcineurine (Ohashi et al., 2004). Par contre, ces observations ont été faites à des concentrations d’éthanol de 400 mM et 1M qui sont bien au-delà des valeurs physiologiques généralement acceptées pour cet alcool (25- 100 mM). Dans le même article qui décrit les résultats précédents, les auteurs ont aussi montré une inhibition de l’activité phosphatase de la calcineurine par le n-butanol et le n- pentanol avec des IC50 approximatifs de 70 et 30 mM respectivement. Aux concentrations de ces deux alcanols que nous avons utilisés au chapitre 5 (IC50 de 3,0 et 1,4 mM), l’inhibition de la calcineurine obtenue par ces auteurs est négligeable et non-significative. Il semble donc raisonnable de spéculer que la situation serait la même pour la suite de la série des n-alcanols et, bien que cela reste à démontrer expérimentalement, il est peu probable

166 qu’une réduction de l’activité de la calcineurine soit responsable des effets biologiques provoqués par ces molécules. Tel que mentionné en introduction, plusieurs des impacts immunologiques de l’éthanol ont été associés à un dérèglement dans l’activation du facteur de transcription NF-κB. La situation s’avère toute autre en ce qui concerne les autres alcools à courte chaîne (de 1 à 5 carbones, incluant l’isopropanol), puisqu’aucune de ces molécules n’a d’impact sur la translocation nucléaire de la sous-unité p65 de NF-κB. D’ailleurs le méthanol et l’isopropanol n’affectent pas non plus la translocation des autres sous-unités qui composent NF-κB. Cette disparité entre les résultats décrits pour l’éthanol et ceux présentés ici pourrait s’expliquer d’une part par le fait que l’impact de l’éthanol sur NF-κB n’a été décrit que pour des monocytes et des macrophages et non pour des lymphocytes T. D’autre part, l’effet de l’éthanol sur la signalisation du TLR4 induite par le lipopolysaccharide dans des cellules de l’immunité innée débuterait au niveau de la membrane cellulaire (Dai et al., 2005; Szabo et al., 2007); il est donc cohérent que NF-κB en soit affecté puisque son activation dépend ultimement de ces événements de signalisation précoces. Au contraire de cela, nous avons montré aux chapitres 2, 3 et 5 que dans les lymphocytes T les alcanols exercent plutôt leurs effets en aval de la membrane plasmique en ciblant des étapes ultérieures des sentiers de signalisation. La situation est toutefois différente pour les alcools primaires de C6 à C8, puisque ceux-ci réduisent l’activation de p65 par un mécanisme qui n’implique pas une inhibition des premiers événements de signalisation en aval du TCR. Un événement moléculaire de la cascade subséquente qui mène à l’activation de NF-κB est possiblement altéré par ces alcanols. La protéine kinase C θ (PKCθ) est recrutée à la membrane plasmique du lymphocyte et activée par Lck peu de temps après l’initiation de la signalisation via le TCR. Cette kinase est celle qui initie l’activation de la voie de signalisation qui mène à la translocation nucléaire de NF-κB; cependant, les événements moléculaires qui relient les deux extrémités de ce sentier demeurent mal connus. Immédiatement en amont du noyau, on sait toutefois que la phosphorylation de l’inhibitor of NF-κB (IκB) par le complexe IκB kinase (IKK), mène à son ubiquitination et à sa dégradation ce qui libère NF-κB et permet son import nucléaire (Smith-Garvin et al 2009).

Il a déjà été montré que différents alcanols pouvaient inhiber l’activité des isoformes α, βI/II

167 et γ de la protéine kinase C (Slater et al., 1993; Slater et al., 1997). La situation pourrait être similaire pour la PKCθ, mais celle-ci serait ciblée préférentiellement par les alcools primaires qui possèdent de 6 à 8 atomes de carbones. Une étape plus en aval encore inconnue pourrait aussi être inhibé par ces molécules tout comme la phosphorylation de IκB par IKK.

Les résultats présentés dans cette thèse montrent clairement que les alcanols sont capables d’induire une dérégulation partiellement sélective de membres distincts de la famille de facteurs de transcription NFAT. Cette famille est composée de cinq membres desquels, NFATc1, c2 et c3 ont été bien caractérisé au niveau du système immunitaire (Hogan et al., 2003). NFATc3 est exprimé dans les thymocytes et joue un rôle important pour le développement des lymphocytes T (Oukka et al., 1998). NFATc2 est le membre de la famille le plus abondamment exprimé au niveau du cytoplasme des cellules T au repos (non-activées). NFATc1 est aussi exprimée dans tous les sous-types de lymphocytes T périphériques, mais à un bien plus faible niveau. Cependant, l’isoforme NFATc1αA est fortement inductible suite à l’activation de la cellule T et participe à une boucle d’autorégulation (Macian, 2005). Le méthanol et l’isopropanol sont tous deux capables d’altérer le fonctionnement normal de membres de la famille NFAT, mais ils le font d’une manière assez différente. Tel que montré au chapitre 2, l’isopropanol réduit l’activation de NFATc1, alors que le méthanol (chapitre 3) augmente la quantité de NFATc2 dans le noyau des cellules T activées. Une autre caractéristique qui distingue ces deux alcools l’un de l’autre est l’absence d’interférence du méthanol avec les voies de signalisation menant à l’activation du facteur de transcription AP-1. Dans le cas de l’isopropanol, la réduction de la translocation nucléaire de NFATc1 coopère avec l’activation déficiente d’AP-1 pour réduire les fonctions effectrices des lymphocytes activés. En ce qui concerne la série des n- alcanols aliphatiques primaires, les résultats du chapitre 5 montrent que les molécules de C7

à C10 réduisent significativement l’activation de NFATc1. Il n’est pas exclu que la translocation nucléaire de NFATc2 soit aussi affectée par les alcools de cette série, puisque les résultats des essais luciférase montrent une forte tendance vers l’inhibition qui débute à

C3. D’ailleurs, l’effet de l’isopropanol sur NFATc2 n’est pas connu non plus. La question

168 de l’impact des autres alcools autre que le méthanol sur ce facteur de transcription mériterait sans doute d’être approfondie.

La corrélation linéaire observée entre la taille de la chaîne carbonée des alcanols et l’inhibition de NFAT de C2 à C10 et la perte subséquente de cette activité à C11 s’accordent très bien avec les résultats sur la sécrétion d’IFN-γ. Ensembles ces observations suggèrent l’existence d’une cavité de liaison aux alcanols sur une cible protéique intracellulaire impliquée dans l’activation de NFATc1 qui pourrait accommoder jusqu’à 10 atomes de carbones. L’ensemble de nos données, nous fait considérer NFATc1 lui-même comme étant un candidat probable pour être cette cible. Les résultats sur l’inhibition de NFATc1 par l’isopropanol supportent aussi cette hypothèse. Ainsi, les alcanols aliphatiques pourraient lier directement NFATc1 ce qui entrainerait un changement de conformation favorisant l’export nucléaire ou la rétention cytoplasmique. À l’opposée, le méthanol interagirait avec NFATc2 et le changement de conformation produit masquerait partiellement le signal d’exportation nucléaire ou réduirait l’accessibilité des sites de phosphorylation.

Réflexions sur les mécanismes de dérèglement des cellules de l’immunité innée par les alcanols

Dans les chapitres 2 et 4, nous avons montrés que l’isopropanol avait aussi la capacité d’altérer les fonctions effectrices des cellules NK, des monocytes et des macrophages. Notamment, ces cellules perdent leur habileté à sécréter des cytokines pro-inflammatoires et ce, autant in vitro qu’in vivo. La réduction de la production de TNF-α est d’ailleurs assez importante pour protéger des souris traitées à l’isopropanol d’un choc toxique induit par des doses autrement létales de LPS. Ces dérèglements sont provoqués par des changements biochimiques similaires à ceux qui opèrent dans les lymphocytes T traités aux alcools, mais implique des cibles moléculaires différentes. Ainsi, la détection de patrons moléculaires associés aux pathogènes par leurs TLR spécifiques mène à l’activation de sentiers de signalisation qui aboutissent au déclenchement de la transcription de gènes d’intérêt

169 immunologique. La reconnaissance du LPS entraîne le regroupement de récepteurs, comme CD14 et TLR4, et le recrutement d’importantes protéines de signalisation au niveau de la région cytoplasmique de TLR4 (Miyake, 2006). Parmi celles-ci, la protéine adaptatrice encodée par le gène myeloid differentiation primary response (88) (MyD88) est responsable du recrutement de la interleukin-1 receptor-associated kinase 1 (IRAK1), qui à son tour joue un rôle clé dans l’activation des facteurs de transcription NF-κB et AP-1 (Kawai and Akira, 2010). En présence d’isopropanol, la signalisation précoce associée à TLR4 opère normalement puisque l’activité kinase d’IRAK1 demeure inchangée. Cette observation contraste avec ce qui a été rapporté pour l’éthanol (Oak et al., 2006) qui réduit l’activité kinase de la même protéine. Le même groupe qui a publié ce résultat avait précédemment suggéré que l’effet d’une exposition aigue à l’éthanol sur les monocytes pouvait résulter d’une altération de la distribution de TLR4 dans les radeaux lipidiques (Dolganiuc et al., 2006), ce qui serait compatible avec une réduction de l’activité kinase d’IRAK1. Cependant, dans des travaux plus récents, ce groupe a plutôt attribué l’effet sur IRAK1 à l’augmentation de l’expression de la protéine régulatrice IRAK-M (Oak et al., 2006; Mandrekar et al., 2009). Bien qu’il semble que l’éthanol puisse agir au niveau de la membrane, tout comme l’isopropanol il peut aussi avoir d’autres cibles en aval de celle-ci. Pour l’isopropanol, l’existence d’une cible moléculaire en aval est d’ailleurs confirmée au chapitre 4 par la découverte que celui-ci compromet spécifiquement le fonctionnement normal de la voie de signalisation ERK dans des monocytes primaires. De fait, la phosphorylation d’ERK2 est diminuée en réponse à une stimulation au LPS alors que les autres MAPK demeurent inchangées. Il en résulte une réduction sélective du contenu nucléaire des sous-unités c-Fos et JunB du facteur de transcription AP-1, ce qui entraîne un dérèglement substantiel de la fonction effectrice de ces cellules.

En résumé des deux dernières sections de cette discussion, les propriétés immunomodulatrices des alcanols à courte chaîne sur les cellules des systèmes innés et adaptatifs résultent principalement de l’activation déficiente des facteurs de transcription NFAT et/ou AP-1. La situation est différente en ce qui concerne les alcanols de 6 à 8 carbones, puisque ceux-ci perturbent également l’activation de la sous-unité p65 de NF-κB

170 dans les lymphocytes T. En termes généraux, les alcanols – exception faite du méthanol – limitent la production de cytokines des lymphocytes T par la réduction de la translocation nucléaire de NFAT accompagnée d’une réduction similaire d’AP-1 (Isopropanol) ou de

NF-κB (C6 à C8). Dans des conditions d’activation qui ne dépendent pas de NFAT, comme la réponse des monocytes au LPS, l’activation déficiente d’AP-1 est suffisante pour produire un effet délétère dose-dépendant sur le phénotype activé de ces cellules. Il est important de noter que NFAT, AP-1 et NF-κB participent tous à la régulation transcriptionnelle de toutes les cytokines ou chimiokines dont la production est altérée par les alcanols (IL-2, IFN-γ, TNF-α et CCL-2) (Attema et al., 2002; Kok et al., 2009). Même l’augmentation paradoxale de la sécrétion d’IL-6 par les monocytes traités à l’isopropanol peut être expliquée par la réduction de la translocation nucléaire de JunB; cette sous-unité du facteur de transcription AP-1 agit comme un répresseur du promoteur de l’IL-6 (Pflegerl et al., 2009). La figure 6.1 présenté à la page suivante décrit les mécanismes d’immunomodulation hypothétiques que nous proposons pour les alcanols et les types cellulaires étudiés dans cet ouvrage.

171

Figure 6.1 – Modèles d’action hypothétiques des alcanols sur des cellules immunitaires.

(A) Le panneau de gauche montre l’activation de NFATc1 dans des lymphocytes T stimulés par leur TCR en présence d’isopropanol (IPA). Cet alcool pourrait intéragir directement avec NFATc1 et causer un changement de conformation qui limiterait sa localisation nucléaire. Le panneau de droite montre un modèle simplifié de l’activation d’ERK dans des monocytes activés au LPS en absence ou en présence d’IPA. L’alcool réduit la phosphorylation et l’activation d’ERK2 ce qui aboutit à une diminution de l’activation de c-Fos et JunB.

172

(B) Le diagramme illustre l’activation de NFATc2 en présence de méthanol (MeOH) dans des lymphocytes T stimulés par leur TCR. Le MeOH pourrait intéragir avec NFATc2 et induire un changement conformationnel qui masquerait le signal d’export nucléaire (nes) ou réduirait l’accessibilité des sites de phosphorylation. (C) Ces schémas illustrent les possibles effets des n-alcanols sur NF-κB (gauche), NFATc1 (centre) et

ERK/AP-1 (droite) dans des lymphcytes T activés. Les n-alcanols C6-C8 inhibe une étape de la signalisation menant à l’activation de p65. Ils pourraient (i) agir au niveau d’IκB et réduire l’accessibilité de ses sites de phosphorylation, (ii) diminuer l’activité kinase de IKK ou (iii) lier p65

directement et limiter sa translocation nucléaire (gauche). Les n-alcanols C7-C10 pourraient intéragir directement avec NFATc1 et causer un changement de conformation qui limiterait sa localisation nucléaire (centre). Les n-alcanols pourraient peut-être inhiber l’activation d’ERK ou AP-1, par analogie avec les résulats obtenus avec l’IPA.

Considérations cliniques et conséquences médicales

Les dérèglements transcriptionnels provoqués par les alcanols décrits dans cette thèse ont d’importantes répercussions immunologiques. Les conséquences cliniques les plus immédiates se trouvent sans doute dans le domaine de la médecine d’urgence. Les deux premiers alcools étudiés ici, soit l’isopropanol et le méthanol, sont très facilement accessibles à la population en général et en quantité assez abondante comme constituants d’une multitude de produits domestiques. Les personnes alcooliques, les enfants et les individus suicidaires composent la majorité des patients admis à l’urgence suite à une intoxication aigue aux alcools (Barceloux et al., 2002; Zaman et al., 2002). La plupart de ces cas d’intoxications surviennent après une ingestion orale délibérée ou accidentelle de l’une de ces substances (Kraut and Kurtz, 2008). Le méthanol et l’isopropanol sont tous deux aisément absorbé dans le tractus gastro-intestinal et atteignent leur niveau sérique maximal dans la première heure suivant l’ingestion. Les concentrations d’isopropanol mesurées à l’arrivée des patients intoxiqués à l’hôpital varient considérablement. Certains individus peuvent survivre à des niveaux sériques de cet alcool aussi élevé que 560 mg/dL (0,56% ou 93 mM), mais on considère que des concentrations plus grandes que 400 mg/dL sont potentiellement mortelles et nécessitent généralement la dialyse (Lacouture et al., 1983; Emadi and Coberly, 2007). La

173 concentration sérique moyenne sublétale d’isopropanol chez des patients sévèrement intoxiqués est toutefois de 310 mg/dL, 7 heures après l’ingestion (Ekwall, 1997). La concentration sérique de méthanol à l’admission à l’urgence varie de 20 à 1290 mg/dL (0,02-1,29% ou 6,24-402 mM). La valeur moyenne pour des individus sévèrement intoxiqués était toutefois de 179 mg/dL (0,179% ou 56 mM) dans une étude rétrospective canadienne et de 170 mg/dL (0,17% ou 53 mM) chez les patients qui ont participé à l’étude clinique du fomepizole pour le traitement de l’empoisonnement au méthanol (Liu et al., 1998; Brent et al., 2001). Une revue de la littérature à propos de l’exposition humaine à l’isopropanol et au méthanol révèle que la plupart des mesures des concentrations sériques de ces substances ont été faites plusieurs heures après l’ingestion. Ces mesures sous-estiment sans doute les concentrations réelles de ces alcools présentes lors des premiers moments de l’intoxication (Daniel et al., 1981; Gaudet and Fraser, 1989; Kostic and Dart, 2003; Mueller-Kronast et al., 2003). Des études faites avec des patients volontaires et un modèle pharmacocinétique ont montré que les concentrations sériques d’isopropanol mesurées trois heures après l’ingestion étaient précédées par un niveau maximal jusqu’à 7 fois plus élevé dans la première heure d’intoxication (Lacouture et al., 1983; Monaghan et al., 1995; Clewell et al., 2001). Cette disparité entre la concentration mesurée à l’admission à l’hôpital et le niveau maximal est particulièrement évidente pour les cas d’empoisonnement au méthanol. De fait, cet alcool a une toxicité intrinsèque relativement faible et les patients intoxiqués recherchent souvent de l’aide beaucoup plus tard, parfois jusqu’à deux jours après l’ingestion, lorsque la concentration de formate devient plus importante et cause des symptômes d’intoxication plus apparents (Liu et al., 1998; Barceloux et al., 2002). En support de cette affirmation, une étude rétrospective de 173 patients intoxiqués a révélé que ceux qui s’étaient présenté à l’urgence de 24 à 48 heures après l’ingestion avaient une concentration sérique médiane de 110 mg/dL (0,11% ou 34mM), ce qui contraste avec les quelques patients admis en deçà des 6 heures suivant l’ingestion qui avaient une concentration médiane de 300 mg/dL (0,3% ou 94 mM) (Kostic and Dart, 2003). Dans nos modèles d’intoxication aigue, les niveaux d’alcools sériques obtenus 30 minutes après l’injection intrapéritonéale étaient de 200 mg/dL (0,2% ou 33 mM) pour

174 l’isopropanol et de 238 mg/dL (0,238% ou 74 mM) pour le méthanol. Ces valeurs se situent à l’intérieur de la gamme de concentration qui possède une activité biologique in vitro (commençant de 13 à 26 mM pour l’isopropanol et à 25 mM pour le méthanol). Plus important encore, ces concentrations sont cliniquement atteignables pour les deux alcools lors des premières heures d’une intoxication aigue. Par exemple, pour illustrer le contexte temporel, la concentration sérique de 238 mg/dL de méthanol utilisée dans nos expériences animales se compare au niveau maximal atteint chez des humains qui présente une concentration sérique de 42,5 mg/dL (0,04% ou 13mM) 24 heures après l’ingestion (Jacobsen et al., 1988; Kostic and Dart, 2003). Ainsi, en accord avec la pharmacocinétique du méthanol et de l’isopropanol, nous pouvons assumer que les dérèglements transcriptionnels décrit précédemment et leurs conséquences immunologiques seraient les plus intenses dans les premières heures suivant un empoisonnement. Les résultats des chapitres 2, 3 et 4 indiquent que des animaux qui subissent une intoxication aigue ont un système immunitaire dysfonctionnel caractérisé par un état d’immunosuppression dans le cas de l’isopropanol et par un état d’hyperréactivité dans le cas du méthanol. Il est donc possible que les patients intoxiqués à l’isopropanol soient plus susceptibles à des infections locales ou systémiques dans les premières heures suivant l’ingestion. Cette probabilité pourrait augmenter encore davantage en présence de lésions de la peau ou d’autres formes de traumatismes. En fait, l’état d’ébriété qui résulte de la consommation excessive des alcools à courte-chaîne est souvent associé à des traumatismes (Saxena et al., 1987; Fitzgerald et al., 2007). De plus, l’isopropanol est souvent employé comme désinfectant topique. Lorsqu’utilisé à cet effet, les impacts immunologiques de cet alcool sur la peau adulte normale sont fort probablement limités, transitoires et confinés aux cellules intra dermiques à cause de la faible absorption par cette voie. Par contre, il serait avisé d’être prudent en ce qui concerne l’emploi sur la peau malade jusqu’à ce que les conséquences potentielles soient mieux connues. À cet effet, JunB est sous-exprimé dans des kératinocytes de patients atteint de lupus érythémateux systémique ce qui cause un dérèglement de la production des cytokines sécrétées par la peau; ce dysfonctionnement pourrait possiblement être accentué par une exposition topique à l’isopropanol (Pflegerl et al., 2009).

175

À l’opposé, la production massive de cytokines pro-inflammatoires pourrait jouer un certain rôle dans la pathophysiologie de l’empoisonnement au méthanol. Par exemple, la quantité anormalement élevée de TNF-α produite chez des patients en réponse à des produits bactériens, présents à cause d’une blessure ou d’une infection préalable, pourraient contribuer au développement de complications telle que la pancréatite aigue, une affection qui peut se retrouver chez la moitié des cas d’intoxication aigue au méthanol (Hantson and Mahieu, 2000). Si confirmées, ces prédictions pourraient mener à l’établissement de mesures de précaution afin de limiter le risque d’infection pour les cas d’empoisonnement à l’isopropanol et à l’instauration de traitements anti-inflammatoires pour contrer les propriétés pro- inflammatoires du méthanol lors d’une intoxication aigue.

La toxicité aigue des alcanols aliphatiques en dehors du méthanol, de l’éthanol et de l’isopropanol est passablement moins bien documentée que celle de ces alcools d’usage courant. Cependant, il est fort intéressant d’apprendre que dès 1869, Richardson avait noté que le potentiel toxique des alcanols aliphatiques croissait avec leur poids moléculaire (Richardson, 1869). Cette observation a par la suite été confirmée dans plusieurs modèles : des « germes » jusqu’aux chats et aux chiens, en passant par les paramécies et même les plantes (Kamm, 1921; Gibbs, 1934; Wallgren, 1960; Macgregor et al., 1964; Lachenmeier et al., 2008). Cette corrélation s’apparente étrangement à la règle de Meyer-Overton pour l’effet anesthésiant et à nos résultats obtenus pour la fonction effectrice des lymphocytes T. Il semble donc que la relation qu’entretiennent les n-alcanols avec le vivant obéit à une règle très générale qui lie leur « potentiel d’activité biologique » à leur taille ou leur hydrophobicité. Les cas d’expositions aigues à d’autres alcools aliphatiques sont moins fréquents que ceux documentés pour l’isopropanol et le méthanol, mais on en a dénombré tout de même un peu plus de 1200 en 2009 en Amérique du Nord (Bronstein et al., 2010). Cependant, les rapports détaillés de cas cliniques demeurent assez rare. Par exemple, le rapport d’un cas d’empoisonnement volontaire au n-butanol en Slovénie réfère à un seul autre cas rapporté en Russie en 1965 (Bunc et al., 2006). La concentration sérique de butanol chez le patient

176 slovène n’a cependant pas été mesurée à son admission en service hospitalier, car les auteurs ignoraient la nature de la substance qu’il avait ingérée. Toutefois, puisque celui-ci était dans un état comateux, il est probable que la concentration sérique de butanol ait atteint un niveau au moins comparable à ceux qui inhibent la sécrétion d’IFN-γ in vitro dans les lymphocytes T (à partir de 2,4 mM), et ce puisque les valeurs de ED50 rapporté pour les effets neurobiologiques/anesthétiques de cet alcool tournent autour de 15 mM (Miller et al., 1973; Franks and Lieb, 1990). Il est donc possible que, tout comme pour une intoxication à l’isopropanol, les patients empoisonnés à d’autres alcools se retrouvent dans un état d’immunosuppression temporaire.

L’intérêt toxicologique de la série des n-alcanols ne se restreint pas seulement aux cas d’intoxications aigues, il touche aussi aux potentiels usages thérapeutiques des membres de cette famille de molécules et/ou à leurs dérivés. À titre d’exemple, examinons le cas du n- octanol et de son métabolite l’acide octanoïque. Ces deux molécules, et plus particulièrement l’acide octanoïque, ont affiché une efficacité prometteuse pour le traitement des symptômes du tremblement essentiel (Bushara et al., 2004; Shill et al., 2004; Nahab et al., 2011). Cette maladie est le trouble du mouvement le plus répandu et elle se manifeste par des tremblements musculaires, généralement des mains, du visage, du cou et des cordes vocales (Vignola, 2011). Les niveaux sériques d’acide octanoïque obtenus qui montraient le plus d’efficacité variaient de 9,14 à 30,07 µg/mL. D’autre part, les données réunies dans cette thèse suggèrent l’existence sur des cibles protéiques de pochettes hydrophobes auxquelles les n-alcanols peuvent se lier, altérant conséquemment la fonction de ces protéines. Or, d’autres molécules ayant des caractéristiques structurales similaires pourraient agir de la même manière. Donc, puisque nous avons montré que le n-octanol avait un effet important sur une fonction effectrice des lymphocytes T, il est logique de croire que l’acide octanoïque possède un pouvoir immunosuppresseur similaire. C’est d’ailleurs ce qu’indiquent nos résultats qui sont présentés dans la figure ci-dessous.

177

10000

9000

8000

7000

6000

5000

(pg/mL)

IFN 4000

3000

2000

1000

0 (+) 0,00015 0,0003 0,0015 0,003 0,015 0,03 (-) (-) OA [Acide Octanoïque] (%m/v)

Figure 6.2 – L’acide octanoïque inhibe la sécrétion d’IFN-γ par les lymphocytes T.

Des lymphocytes T purifiés ont été stimulés avec des billes tapissées d’anticorps anti-CD3/CD28 pour 6h en présence ou non des concentrations (% m/v) d’acide octanoïque indiquées. Les quantités d’IFN-γ sécrétée dans les surnageants ont été mesurées par ELISA. La moyenne de la quantité d’IFN-γ ± l’erreur standard de trois expériences est montré. Symboles : (+), cellules stimulées en absence d’acide octanoïque; (-), cellules non-stimulées en absence d’acide octanoïque; (-) Oct Acid, cellules non-stimulées en présence de 0,03% d’acide octanoïque.

Ces données indiquent donc que l’acide octanoïque a aussi la capacité d’inhiber la sécrétion d’IFN-γ par les lymphocytes T, cependant il le fait à des concentrations substantiellement plus élevées que le n-octanol. L’alcool a un IC50 d’environ 0,15 mM alors que l’acide inhibe d’un peu moins de 50% la sécrétion d’IFN-γ à 1,04 mM (0,015%). Le n-octanol est donc environ 7 fois plus puissant que son acide apparenté pour l’inhibition de la fonction effectrice des lymphocytes T.

178

Si l’on compare les concentrations d’acide octanoïque qui induisent un effet immunomodulateur (à partir de 0,015%) aux concentrations sériques maximales obtenues dans une étude clinique pour le tremblement essentiel (0,0009% à 0,003%) (Nahab et al., 2011), on constate que ces dernières concentrations tombent en dehors de la gamme potentiellement immunosuppressive. Par contre, la prudence pourrait devenir de mise dans l’éventualité que des doses plus élevées d’acide octanoïque soient testées sur des patients. D’autant plus que dans une étude récente sur un modèle murin de tremblement essentiel, des doses d’acide octanoïque qui pourraient atteindre des concentrations sériques aussi élevé que 0,03 à 0,1% ont été utilisées (Nahab et al., 2012) et se sont avérés très efficaces. D’un autre côté, nos résultats ne concernent qu’une exposition aigue à l’acide octanoïque. Il pourrait être intéressant de s’attarder aux potentiels effets chroniques de plus faibles doses de cette substance, et ce puisque les patients qui pourraient être traités pour le tremblement essentiel pourraient recevoir un régime chronique de celle-ci.

Nous avons montré aux chapitres 2 et 4 que l’isopropanol pouvait protéger des animaux d’un choc toxique autrement létal. Ces résultats permettent d’envisager un usage thérapeutique direct des alcanols pour le traitement de conditions pathologiques impliquant la sécrétion massive et néfaste de cytokines pro-inflammatoires, comme le TNF-α. Les chocs toxiques causés par différents superantigènes bactériens ou viraux représentent de bons exemples de telles conditions (Cohen, 2002). D’ailleurs, en se basant sur nos résultats, des chercheurs du laboratoire national de microbiologie de Santé Canada ont testé l’efficacité de l’isopropanol et d’autres substances, en combinaison avec un vaccin candidat, pour le traitement de l’infection au virus Ébola dans des rongeurs (Richardson et al., 2011). L’intérêt de l’utilisation d’un alcanol pour le traitement de l’infection à l’Ébola réside dans le fait que ce virus engendre une réponse immunitaire innée dérégulée. Cette réponse est caractérisée par une inflammation non-productive qui ne limite pas la propagation virale et pourrait jouer un rôle clé dans la progression des symptômes de type choc toxique, les anomalies de coagulation et les multiples défaillances d’organes qui causent la mort de l’hôte (Richardson et al., 2011). En limitant l’amplitude de cette réaction inflammatoire, il pourrait logiquement être possible de diminuer la sévérité des symptômes

179 et d’améliorer la survie des patients. Dans cette expérience, des cobayes contrôles infectés par Ébola ont rapidement perdu du poids et sont tous morts en 7 jours. Un retard de 5 jours dans le temps de la mort a été obtenu chez les rongeurs traités avec le vaccin candidat et l’isopropanol. Ce résultat est semblable à celui que les auteurs ont obtenu pour d’autres traitements potentiels testés (décès des animaux entre 10 et 15 jours). L’isopropanol n’a pas induit la survie des animaux dans ce modèle, mais il se pourrait qu’un alcanol plus puissant dans son action immunodépressive, de C7 à C9 par exemple, obtienne de meilleurs résultats. Évidemment, ces molécules sont plus hydrophobes que l’isopropanol et donc beaucoup moins soluble dans des solutions aqueuses. Leur formulation et la voie par laquelle elles pourraient être administrées deviennent conséquemment très importantes afin d’assurer une dispersion efficace dans le corps de l’hôte.

Conclusion

Les données présentées dans cette thèse confirment nos hypothèses de départ, à savoir qu’à l’instar de l’éthanol, les autres alcools aliphatiques primaires ainsi que l’isopropanol possèdent tous des propriétés immunomodulatrices. Ces impacts immunologiques sont analogues aux effets anesthésiques de ces molécules puisqu’ils sont corrélés à leur hydrophobicité, mais ils n’impliquent aucun constituant de la membrane plasmique dans leur mode d’action au moins en ce qui concerne les lymphocytes T. La coupure distinctive de l’activité immunosuppressive des alcanols à C11 sur ces cellules ainsi que la forme générale de la courbe de réponse en fonction du nombre d’atomes de carbones suggèrent la liaison des alcanols à des pochettes hydrophobes situés dans des cibles protéiques intracellulaires communes. Les données à propos des facteurs de transcription permettent d’émettre l’hypothèse de la présence d’une pochette pouvant accommoder 10 carbones sur NFAT et d’une plus petite cavité (d’une capacité d’environ 8 carbones) sur une molécule de la voie d’activation de NF-κB. Les effets biologiques des alcanols, comprenant leurs capacités anesthésiantes et immunomodulatrices, pourraient résulter de leurs interactions avec plusieurs cibles moléculaires potentielles, en autant qu’elles possèdent des cavités hydrophobes adéquates. Bien sûr, ces effets biologiques dépendent probablement des

180 profils d’expression protéique définis par le type cellulaire et l’activation de ces cellules. Ainsi, les résultats présentés pour la série des n-alcanols aliphatiques primaires ne devraient être interprétés que dans le contexte de l’activation de lymphocytes T primaires humain via leur récepteur à l’antigène. Il en va de même pour les effets de l’isopropanol sur les monocytes primaires humains activés par le lipopolysaccharide. En définitive, mes travaux contribuent à une meilleure compréhension des activités biologiques des alcanols et ils apportent un éclairage nouveau sur les mécanismes responsables de leurs effets immunomodulateurs, principalement sur la fonction des lymphocytes T.

Bibliographie

Alifimoff, J. K., L. L. Firestone and K. W. Miller (1989). "Anaesthetic potencies of primary alkanols: implications for the molecular dimensions of the anaesthetic site." Br J Pharmacol 96(1): 9-16. Arbabi, S., I. Garcia, G. J. Bauer and R. V. Maier (1999). "Alcohol (ethanol) inhibits IL-8 and TNF: role of the p38 pathway." J Immunol 162(12): 7441-5. Aryal, P., H. Dvir, S. Choe and P. A. Slesinger (2009). "A discrete alcohol pocket involved in GIRK channel activation." Nat Neurosci 12(8): 988-95. Attema, J. L., R. Reeves, V. Murray, I. Levichkin, M. D. Temple, D. J. Tremethick and M. F. Shannon (2002). "The human IL-2 gene promoter can assemble a positioned nucleosome that becomes remodeled upon T cell activation." J Immunol 169(5): 2466-76. Ballinger, P. a. L., F.A. (1960). "Acid ionization constants of alcohols. II. Acidities of some substituted methanols and related compounds." J. Am. Chem. Soc. 82: 795-798. Barceloux, D. G., G. R. Bond, E. P. Krenzelok, H. Cooper and J. A. Vale (2002). "American Academy of Clinical Toxicology practice guidelines on the treatment of methanol poisoning." J Toxicol Clin Toxicol 40(4): 415-46. Ben-Eliyahu, S., G. G. Page, R. Yirmiya and A. N. Taylor (1996). "Acute alcohol intoxication suppresses natural killer cell activity and promotes tumor metastasis." Nat Med 2(4): 457-60. Boe, D. M., S. Nelson, P. Zhang, L. Quinton and G. J. Bagby (2003). "Alcohol-induced suppression of lung chemokine production and the host defense response to Streptococcus pneumoniae." Alcohol Clin Exp Res 27(11): 1838-45. Brent, J., K. McMartin, S. Phillips, C. Aaron and K. Kulig (2001). "Fomepizole for the treatment of methanol poisoning." N Engl J Med 344(6): 424-9. Bronstein, A. C., D. A. Spyker, L. R. Cantilena, Jr., J. L. Green, B. H. Rumack and S. L. Giffin (2010). "2009 Annual Report of the American Association of Poison Control Centers' National Poison Data System (NPDS): 27th Annual Report." Clin Toxicol (Phila) 48(10): 979-1178. Bunc, M., T. Pezdir, H. Mozina, M. Mozina and M. Brvar (2006). "Butanol ingestion in an airport hangar." Hum Exp Toxicol 25(4): 195-7. Bushara, K. O., S. R. Goldstein, G. J. Grimes, Jr., A. H. Burstein and M. Hallett (2004). "Pilot trial of 1-octanol in essential tremor." Neurology 62(1): 122-4. Chiappelli, F., M. Kung, P. Lee, L. Pham, E. Manfrini and P. Villanueva (1995). "Alcohol modulation of human normal T-cell activation, maturation, and migration." Alcohol Clin Exp Res 19(3): 539-44. Clewell, H. J., 3rd, P. R. Gentry, J. M. Gearhart, T. R. Covington, M. I. Banton and M. E. Andersen (2001). "Development of a physiologically based pharmacokinetic model of isopropanol and its metabolite acetone." Toxicol Sci 63(2): 160-72. Cohen, J. (2002). "The immunopathogenesis of sepsis." Nature 420(6917): 885-91. Dai, Q., J. Zhang and S. B. Pruett (2005). "Ethanol alters cellular activation and CD14 partitioning in lipid rafts." Biochem Biophys Res Commun 332(1): 37-42. Daniel, D. R., B. H. McAnalley and J. C. Garriott (1981). "Isopropyl alcohol metabolism after acute intoxication in humans." J Anal Toxicol 5(3): 110-2. 182

Dolganiuc, A., G. Bakis, K. Kodys, P. Mandrekar and G. Szabo (2006). "Acute ethanol treatment modulates Toll-like receptor-4 association with lipid rafts." Alcohol Clin Exp Res 30(1): 76-85. Dou, X., C. E. Menkari, S. Shanmugasundararaj, K. W. Miller and M. E. Charness (2011). "Two alcohol binding residues interact across a domain interface of the L1 neural cell adhesion molecule and regulate cell adhesion." J Biol Chem 286(18): 16131-9. Dwyer, D. S. and R. J. Bradley (2000). "Chemical properties of alcohols and their protein binding sites." Cell Mol Life Sci 57(2): 265-75. Eckenhoff, R. G. (2001). "Promiscuous ligands and attractive cavities: how do the inhaled anesthetics work?" Mol Interv 1(5): 258-68. Ekwall, B., Clemedson, C. (1997). Time-Related Lethal Blood Concentrations from Acute Human Poisoning of Chemicals Part 2: The Monographs No. 10 Isopropyl Alcohol. Stanga, Sweden, CTLU. Emadi, A. and L. Coberly (2007). "Intoxication of a hospitalized patient with an isopropanol-based hand sanitizer." N Engl J Med 356(5): 530-1. Environmental Health Criteria 102 - 1-Propanol (1990). I. P. o. C. Safety. Genève, United Nations Environment Programme Environmental Health Criteria 103 - 2-Propanol (1990). I. P. o. C. Safety. Genève, United Nations Environment Programme. Environmental Health Criteria 196 - Methanol (1997). I. P. o. C. Safety. Genève, United Nations Environment Programme. Fitzgerald, D. J., K. A. Radek, M. Chaar, D. E. Faunce, L. A. DiPietro and E. J. Kovacs (2007). "Effects of acute ethanol exposure on the early inflammatory response after excisional injury." Alcohol Clin Exp Res 31(2): 317-23. Forman, S. A., K. W. Miller and G. Yellen (1995). "A discrete site for general anesthetics on a postsynaptic receptor." Mol Pharmacol 48(4): 574-81. Franks, N. P. (2006). "Molecular targets underlying general anaesthesia." British Journal of Pharmacology 147: S72-S81. Franks, N. P., A. Jenkins, E. Conti, W. R. Lieb and P. Brick (1998). "Structural basis for the inhibition of firefly luciferase by a general anesthetic." Biophys J 75(5): 2205- 11. Franks, N. P. and W. R. Lieb (1982). "Molecular mechanisms of general anaesthesia." Nature 300(5892): 487-93. Franks, N. P. and W. R. Lieb (1984). "Do general anaesthetics act by competitive binding to specific receptors?" Nature 310(5978): 599-601. Franks, N. P. and W. R. Lieb (1985). "Mapping of general anaesthetic target sites provides a molecular basis for cutoff effects." Nature 316(6026): 349-51. Franks, N. P. and W. R. Lieb (1990). "Mechanisms of general anesthesia." Environ Health Perspect 87: 199-205. Franks, N. P. and W. R. Lieb (2004). "Seeing the light: protein theories of general anesthesia. 1984." Anesthesiology 101(1): 235-7. Gallo, E. M., K. Cante-Barrett and G. R. Crabtree (2006). "Lymphocyte calcium signaling from membrane to nucleus." Nat Immunol 7(1): 25-32. Gaudet, M. P. and G. L. Fraser (1989). "Isopropanol ingestion: case report with pharmacokinetic analysis." Am J Emerg Med 7(3): 297-9.

183

Ghare, S., M. Patil, P. Hote, J. Suttles, C. McClain, S. Barve and S. Joshi-Barve (2011). "Ethanol inhibits lipid raft-mediated TCR signaling and IL-2 expression: potential mechanism of alcohol-induced immune suppression." Alcohol Clin Exp Res 35(8): 1435-44. Gibbs, R. D. (1934). "THE TOXICITY OF NORMAL ALIPHATIC ALCOHOLS. I." AmericanJournal of Botany. 21(10): 645-650. Goldstein, D. B. (1984). "The effects of drugs on membrane fluidity." Annu Rev Pharmacol Toxicol 24: 43-64. Goral, J., J. Karavitis and E. J. Kovacs (2008). "Exposure-dependent effects of ethanol on the innate immune system." Alcohol 42(4): 237-47. Goral, J. and E. J. Kovacs (2005). "In vivo ethanol exposure down-regulates TLR2-, TLR4- , and TLR9-mediated macrophage inflammatory response by limiting p38 and ERK1/2 activation." J Immunol 174(1): 456-63. Hantson, P. and P. Mahieu (2000). "Pancreatic injury following acute methanol poisoning." J Toxicol Clin Toxicol 38(3): 297-303. Harder, T. and D. Sangani (2009). "Plasma membrane rafts engaged in T cell signalling: new developments in an old concept." Cell Commun Signal 7: 21. Harris, R. A., J. R. Trudell and S. J. Mihic (2008). "Ethanol's molecular targets." Sci Signal 1(28): re7. Hasinoff, B. B. and J. P. Davey (1989). "The inhibition of a membrane-bound enzyme as a model for anaesthetic action and drug toxicity." Biochem J 258(1): 101-7. Hebert, P. and S. B. Pruett (2002). "Ethanol suppresses polyinosinic:polycytidylic acid- induced activation of natural killer cells primarily by acting on natural killer cells, not through effects on other cell types." Alcohol 28(2): 75-81. Hoefer, F., Ed. (1869). Histoire de la Chimie, 2. Hogan, P. G., L. Chen, J. Nardone and A. Rao (2003). "Transcriptional regulation by calcium, calcineurin, and NFAT." Genes Dev 17(18): 2205-32. Jacobsen, D., R. Webb, T. D. Collins and K. E. McMartin (1988). "Methanol and formate kinetics in late diagnosed methanol intoxication." Med Toxicol Adverse Drug Exp 3(5): 418-23. Jammalamadaka, D. and S. Raissi (2010). "Ethylene glycol, methanol and isopropyl alcohol intoxication." Am J Med Sci 339(3): 276-81. Janeway, C. (2009). Immunobiologie. Bruxelles, De Boeck. Jung, S., M. H. Akabas and R. A. Harris (2005). "Functional and structural analysis of the GABAA receptor alpha 1 subunit during channel gating and alcohol modulation." J Biol Chem 280(1): 308-16. Kamm, O. (1921). "The Prediction of the Physiological Action of Alcohols." Science 54(1385): 55. Kawai, T. and S. Akira (2010). "The role of pattern-recognition receptors in innate immunity: update on Toll-like receptors." Nat Immunol 11(5): 373-84. Kawakami, M., B. R. Switzer, S. R. Herzog and A. A. Meyer (1991). "Immune suppression after acute ethanol ingestion and thermal injury." J Surg Res 51(3): 210-5. Kishore, R., J. R. Hill, M. R. McMullen, J. Frenkel and L. E. Nagy (2002). "ERK1/2 and Egr-1 contribute to increased TNF-alpha production in rat Kupffer cells after chronic ethanol feeding." Am J Physiol Gastrointest Liver Physiol 282(1): G6-15.

184

Kishore, R., M. R. McMullen and L. E. Nagy (2001). "Stabilization of tumor necrosis factor alpha mRNA by chronic ethanol: role of A + U-rich elements and p38 mitogen-activated protein kinase signaling pathway." J Biol Chem 276(45): 41930- 7. Klemm, W. R. (1998). "Biological water and its role in the effects of alcohol." Alcohol 15(3): 249-67. Kobayashi, K., L. D. Hernandez, J. E. Galan, C. A. Janeway, Jr., R. Medzhitov and R. A. Flavell (2002). "IRAK-M is a negative regulator of Toll-like receptor signaling." Cell 110(2): 191-202. Kok, S. H., C. Y. Hong, M. Y. Kuo, C. C. Wang, K. L. Hou, Y. T. Lin, D. L. Galson and S. K. Lin (2009). "Oncostatin M-induced CCL2 transcription in osteoblastic cells is mediated by multiple levels of STAT-1 and STAT-3 signaling: an implication for the pathogenesis of arthritis." Arthritis Rheum 60(5): 1451-62. Kostic, M. A. and R. C. Dart (2003). "Rethinking the toxic methanol level." J Toxicol Clin Toxicol 41(6): 793-800. Kraut, J. A. and I. Kurtz (2008). "Toxic alcohol ingestions: clinical features, diagnosis, and management." Clin J Am Soc Nephrol 3(1): 208-25. Kruse, S. W., R. Zhao, D. P. Smith and D. N. Jones (2003). "Structure of a specific alcohol- binding site defined by the odorant binding protein LUSH from Drosophila melanogaster." Nat Struct Biol 10(9): 694-700. Lachenmeier, D. W., S. Haupt and K. Schulz (2008). "Defining maximum levels of higher alcohols in alcoholic beverages and surrogate alcohol products." Regul Toxicol Pharmacol 50(3): 313-21. Lacouture, P. G., S. Wason, A. Abrams and F. H. Lovejoy, Jr. (1983). "Acute isopropyl alcohol intoxication. Diagnosis and management." Am J Med 75(4): 680-6. Lee, A. G. (1976). "Interactions between anesthetics and lipid mixtures. Normal alcohols." Biochemistry 15(11): 2448-54. Liu, J., A. R. Yang, T. Kelly, A. Puche, C. Esoga, H. L. June, Jr., A. Elnabawi, I. Merchenthaler, W. Sieghart, H. L. June, Sr. and L. Aurelian (2011). "Binge alcohol drinking is associated with GABAA alpha2-regulated Toll-like receptor 4 (TLR4) expression in the central amygdala." Proc Natl Acad Sci U S A 108(11): 4465-70. Liu, J. J., M. R. Daya, O. Carrasquillo and S. N. Kales (1998). "Prognostic factors in patients with methanol poisoning." J Toxicol Clin Toxicol 36(3): 175-81. Lobo, I. A., R. A. Harris and J. R. Trudell (2008). "Cross-linking of sites involved with alcohol action between transmembrane segments 1 and 3 of the glycine receptor following activation." J Neurochem 104(6): 1649-62. Macgregor, D. C., E. Schoenbaum and W. G. Bigelow (1964). "Acute Toxicity Studies on Ethanol, Propanol, and Butanol." Can J Physiol Pharmacol 42: 689-96. Macian, F. (2005). "NFAT proteins: key regulators of T-cell development and function." Nat Rev Immunol 5(6): 472-84. Mandrekar, P., S. Bala, D. Catalano, K. Kodys and G. Szabo (2009). "The opposite effects of acute and chronic alcohol on lipopolysaccharide-induced inflammation are linked to IRAK-M in human monocytes." J Immunol 183(2): 1320-7.

185

Mandrekar, P., D. Catalano, A. Dolganiuc, K. Kodys and G. Szabo (2004). "Inhibition of myeloid dendritic cell accessory cell function and induction of T cell anergy by alcohol correlates with decreased IL-12 production." J Immunol 173(5): 3398-407. Mandrekar, P., D. Catalano, L. Girouard and G. Szabo (1996). "Human monocyte IL-10 production is increased by acute ethanol treatment." Cytokine 8(7): 567-77. Mandrekar, P., D. Catalano and G. Szabo (1999). "Inhibition of lipopolysaccharide- mediated NFkappaB activation by ethanol in human monocytes." Int Immunol 11(11): 1781-90. McGovern, P. E., J. Zhang, J. Tang, Z. Zhang, G. R. Hall, R. A. Moreau, A. Nunez, E. D. Butrym, M. P. Richards, C. S. Wang, G. Cheng, Z. Zhao and C. Wang (2004). "Fermented beverages of pre- and proto-historic China." Proc Natl Acad Sci U S A 101(51): 17593-8. Miller, K. W., L. L. Firestone, J. K. Alifimoff and P. Streicher (1989). "Nonanesthetic alcohols dissolve in synaptic membranes without perturbing their lipids." Proc Natl Acad Sci U S A 86(3): 1084-7. Miller, K. W., W. D. Paton, R. A. Smith and E. B. Smith (1973). "The pressure reversal of general anesthesia and the critical volume hypothesis." Mol Pharmacol 9(2): 131- 43. Miyake, K. (2006). "Roles for accessory molecules in microbial recognition by Toll-like receptors." J Endotoxin Res 12(4): 195-204. Monaghan, M. S., K. M. Olsen, B. H. Ackerman, G. L. Fuller, W. H. Porter and A. A. Pappas (1995). "Measurement of serum isopropanol and the acetone metabolite by proton nuclear magnetic resonance: application to pharmacokinetic evaluation in a simulated overdose model." J Toxicol Clin Toxicol 33(2): 141-9. Mueller-Kronast, N., A. A. Rabinstein, L. Voung and A. M. Forteza (2003). "Isopropanol intoxication mimicking basilar artery thrombosis." Neurology 61(10): 1456-7. Nagy, L. E. (2003). "Recent insights into the role of the innate immune system in the development of alcoholic liver disease." Exp Biol Med (Maywood) 228(8): 882-90. Nahab, F. B., A. Handforth, T. Brown, C. Shin, A. Quesada, C. Dong, D. Haubenberger and M. Hallett (2012). "Octanoic Acid suppresses harmaline-induced tremor in mouse model of essential tremor." Neurotherapeutics 9(3): 635-8. Nahab, F. B., L. Wittevrongel, D. Ippolito, C. Toro, G. J. Grimes, J. Starling, G. Potti, D. Haubenberger, D. Bowen, P. Buchwald, C. Dong, D. Kalowitz and M. Hallett (2011). "An open-label, single-dose, crossover study of the pharmacokinetics and metabolism of two oral formulations of 1-octanol in patients with essential tremor." Neurotherapeutics 8(4): 753-62. Nelson, B. K., W. S. Brightwell, A. Khan, E. F. Krieg and A. M. Hoberman (1989). "Developmental Toxicology Evaluation of 1-Pentanol, 1-Hexanol, and 2-Ethyl-1- Hexanol Administered by Inhalation to Rats." Journal of the American College of Toxicology 8(2): 405-410. Nelson, B. K., W. S. Brightwell, A. Khan, E. F. Krieg and A. M. Hoberman (1990). "Developmental Toxicology Assessment of 1-Octanol, 1-Nonanol, and 1-Decanol Administered by Inhalation to Rats." Journal of the American College of Toxicology 9(1): 93-97.

186

Nelson, S., G. J. Bagby, B. G. Bainton and W. R. Summer (1989). "The effects of acute and chronic alcoholism on tumor necrosis factor and the inflammatory response." J Infect Dis 160(3): 422-9. Nizza, D. T. and K. Gawrisch (2009). "A layer model of ethanol partitioning into lipid membranes." Gen Physiol Biophys 28(2): 140-5. Nury, H., C. Van Renterghem, Y. Weng, A. Tran, M. Baaden, V. Dufresne, J. P. Changeux, J. M. Sonner, M. Delarue and P. J. Corringer (2011). "X-ray structures of general anaesthetics bound to a pentameric ligand-gated ion channel." Nature 469(7330): 428-31. Oak, S., P. Mandrekar, D. Catalano, K. Kodys and G. Szabo (2006). "TLR2- and TLR4- mediated signals determine attenuation or augmentation of inflammation by acute alcohol in monocytes." J Immunol 176(12): 7628-35. Ohashi, I., R. Pohoreki, K. Morita and P. M. Stemmer (2004). "Alcohols increase calmodulin affinity for Ca2+ and decrease target affinity for calmodulin." Biochim Biophys Acta 1691(2-3): 161-7. Opdyke, D. L. (1973). "Monographs on fragrance raw materials." Food Cosmet Toxicol 11(1): 95-115. Oukka, M., I. C. Ho, F. C. de la Brousse, T. Hoey, M. J. Grusby and L. H. Glimcher (1998). "The transcription factor NFAT4 is involved in the generation and survival of T cells." Immunity 9(3): 295-304. Pang, K. Y., L. M. Braswell, L. Chang, T. J. Sommer and K. W. Miller (1980). "The perturbation of lipid bilayers by general anesthetics: a quantitative test of the disordered lipid hypothesis." Mol Pharmacol 18(1): 84-90. Pegan, S., C. Arrabit, P. A. Slesinger and S. Choe (2006). "Andersen's syndrome mutation effects on the structure and assembly of the cytoplasmic domains of Kir2.1." Biochemistry 45(28): 8599-606. Peoples, R. W., C. Li and F. F. Weight (1996). "Lipid vs protein theories of alcohol action in the nervous system." Annu Rev Pharmacol Toxicol 36: 185-201. Pflegerl, P., P. Vesely, B. Hantusch, M. Schlederer, R. Zenz, E. Janig, G. Steiner, A. Meixner, P. Petzelbauer, P. Wolf, A. Soleiman, G. Egger, R. Moriggl, T. Kishimoto, E. F. Wagner and L. Kenner (2009). "Epidermal loss of JunB leads to a SLE phenotype due to hyper IL-6 signaling." Proc Natl Acad Sci U S A 106(48): 20423-8. Pringle, M. J., K. B. Brown and K. W. Miller (1981). "Can the lipid theories of anesthesia account for the cutoff in anesthetic potency in homologous series of alcohols?" Mol Pharmacol 19(1): 49-55. Pruett, S. B., C. Schwab, Q. Zheng and R. Fan (2004). "Suppression of innate immunity by acute ethanol administration: a global perspective and a new mechanism beginning with inhibition of signaling through TLR3." J Immunol 173(4): 2715-24. Pruett, S. B., Q. Zheng, R. Fan, K. Matthews and C. Schwab (2004). "Acute exposure to ethanol affects Toll-like receptor signaling and subsequent responses: an overview of recent studies." Alcohol 33(3): 235-9. Ramaswamy, S., H. Eklund and B. V. Plapp (1994). "Structures of horse liver alcohol dehydrogenase complexed with NAD+ and substituted benzyl alcohols." Biochemistry 33(17): 5230-7.

187

Richardson, B. W. (1869). "Physiological research on alcohols." The Medical Times and Gazette 2: 703-706. Richardson, J. S., G. Wong, S. Pillet, S. Schindle, J. Ennis, J. Turner, J. E. Strong and G. P. Kobinger (2011). "Evaluation of Different Strategies for Post-Exposure Treatment of Ebola Virus Infection in Rodents." J Bioterr Biodef S1(007). Roth, S. and P. Seeman (1972). "The membrane concentrations of neutral and positive anesthetics (alcohols, chlorpromazine, morphine) fit the Meyer-Overton rule of anesthesia; negative narcotics do not." Biochim Biophys Acta 255(1): 207-19. Saeed, R. W., S. Varma, T. Peng, K. J. Tracey, B. Sherry and C. N. Metz (2004). "Ethanol blocks leukocyte recruitment and endothelial cell activation in vivo and in vitro." J Immunol 173(10): 6376-83. Saxena, K., K. Lerud and J. J. Cicero (1987). "Methanol intoxication in multiple trauma." Am J Emerg Med 5(1): 60-3. Screening Information Data Set 1-Dodecanol (2002). U. Publications. Berlin, Organization for Economic Cooperation and Development. Screening Information Data Set 2-Propanol (2002). U. Publications. Berlin, Organization for Economic Cooperation and Development Screening Information Data Set Ethanol (2004). U. Publications. Berlin, Organization for Economic Cooperation and Development. Screening Information Data Set Long Chain Alcohols (C6-22 primary aliphatic alcohols) (2006). U. Publications. Berlin, Organization for Economic Cooperation and Development. Screening Information Data Set n-Butyl Alcohol (2001). U. Publications. Berlin, Organization for Economic Cooperation and Development. Screening Information Data Set Primary Amyl Alcohol (2006). U. Publications. Berlin, Organization for Economic Cooperation and Development. Seeman, P. (1972). "The membrane actions of anesthetics and tranquilizers." Pharmacol Rev 24(4): 583-655. Seeman, P., S. Roth and H. Schneider (1971). "The membrane concentrations of alcohol anesthetics." Biochim Biophys Acta 225(2): 171-84. Shahidullah, M., T. Harris, M. W. Germann and M. Covarrubias (2003). "Molecular features of an alcohol binding site in a neuronal potassium channel." Biochemistry 42(38): 11243-52. Shill, H. A., K. O. Bushara, Z. Mari, M. Reich and M. Hallett (2004). "Open-label dose- escalation study of oral 1-octanol in patients with essential tremor." Neurology 62(12): 2320-2. Simmonds, C. (1919). Alcohol. Its Production, Properties, Chemistry, and Industrial Applications. London, England. Slater, S. J., K. J. Cox, J. V. Lombardi, C. Ho, M. B. Kelly, E. Rubin and C. D. Stubbs (1993). "Inhibition of protein kinase C by alcohols and anaesthetics." Nature 364(6432): 82-4. Slater, S. J., M. B. Kelly, J. D. Larkin, C. Ho, A. Mazurek, F. J. Taddeo, M. D. Yeager and C. D. Stubbs (1997). "Interaction of alcohols and anesthetics with protein kinase Calpha." J Biol Chem 272(10): 6167-73.

188

Smith-Garvin, J. E., G. A. Koretzky and M. S. Jordan (2009). "T cell activation." Annu Rev Immunol 27: 591-619. Smothers, C. T. and J. J. Woodward (2006). "Effects of amino acid substitutions in transmembrane domains of the NR1 subunit on the ethanol inhibition of recombinant N-methyl-D-aspartate receptors." Alcohol Clin Exp Res 30(3): 523-30. Streiff, J. H., T. W. Allen, E. Atanasova, N. Juranic, S. Macura, A. R. Penheiter and K. A. Jones (2006). "Prediction of volatile anesthetic binding sites in proteins." Biophys J 91(9): 3405-14. Szabo, G., S. Chavan, P. Mandrekar and D. Catalano (1999). "Acute alcohol consumption attenuates interleukin-8 (IL-8) and monocyte chemoattractant peptide-1 (MCP-1) induction in response to ex vivo stimulation." J Clin Immunol 19(1): 67-76. Szabo, G., A. Dolganiuc, Q. Dai and S. B. Pruett (2007). "TLR4, ethanol, and lipid rafts: a new mechanism of ethanol action with implications for other receptor-mediated effects." J Immunol 178(3): 1243-9. Szabo, G. and P. Mandrekar (2009). "A recent perspective on alcohol, immunity, and host defense." Alcohol Clin Exp Res 33(2): 220-32. Szabo, G., B. Verma and D. Catalano (1993). "Selective inhibition of antigen-specific T lymphocyte proliferation by acute ethanol exposure: the role of impaired monocyte antigen presentation capacity and mediator production." J Leukoc Biol 54(6): 534- 44. Taieb, J., C. Delarche, F. Ethuin, S. Selloum, T. Poynard, M. A. Gougerot-Pocidalo and S. Chollet-Martin (2002). "Ethanol-induced inhibition of cytokine release and protein degranulation in human neutrophils." J Leukoc Biol 72(6): 1142-7. Takeda, K. and S. Akira (2005). "Toll-like receptors in innate immunity." Int Immunol 17(1): 1-14. Takens-Kwak, B. R., H. J. Jongsma, M. B. Rook and A. C. Van Ginneken (1992). "Mechanism of heptanol-induced uncoupling of cardiac gap junctions: a perforated patch-clamp study." Am J Physiol 262(6 Pt 1): C1531-8. Thode, A. B., S. W. Kruse, J. C. Nix and D. N. Jones (2008). "The role of multiple hydrogen-bonding groups in specific alcohol binding sites in proteins: insights from structural studies of LUSH." J Mol Biol 376(5): 1360-76. Trudell, J. R. (1977). "A unitary theory of anesthesia based on lateral phase separations in nerve membranes." Anesthesiology 46(1): 5-10. Vedula, L. S., G. Brannigan, N. J. Economou, J. Xi, M. A. Hall, R. Liu, M. J. Rossi, W. P. Dailey, K. C. Grasty, M. L. Klein, R. G. Eckenhoff and P. J. Loll (2009). "A unitary anesthetic binding site at high resolution." J Biol Chem 284(36): 24176-84. Verma, B. K., M. Fogarasi and G. Szabo (1993). "Down-regulation of tumor necrosis factor alpha activity by acute ethanol treatment in human peripheral blood monocytes." J Clin Immunol 13(1): 8-22. Vignola, J. (2011). Tremblement Essentiel: nouvelles pistes en vue. Salle de Presse - CHUQ. Québec. Wallgren, H. (1960). "Relative intoxicating effects on rats of ethyl, propyl and butyl alcohols." Acta Pharmacol Toxicol (Copenh) 16: 217-22.

189

Westerman, P. W., J. M. Pope, N. Phonphok, J. W. Doane and D. W. Dubro (1988). "The interaction of n-alkanols with lipid bilayer membranes: a 2H-NMR study." Biochim Biophys Acta 939(1): 64-78. Yoshimura, M., S. Pearson, Y. Kadota and C. E. Gonzalez (2006). "Identification of ethanol responsive domains of adenylyl cyclase." Alcohol Clin Exp Res 30(11): 1824-32. Zaman, F., A. Pervez and K. Abreo (2002). "Isopropyl alcohol intoxication: a diagnostic challenge." Am J Kidney Dis 40(3): E12.

Annexe 1 – Short-term immunological effects of non- ethanolic short-chain alcohols

Olivier Désy, Damien Carignan and Pedro O. de Campos-Lima*

Laval University Cancer Research Center, Quebec City, Quebec, G1R 2J6, Canada

Toxicol Lett. 2012 Apr 5; 210(1):44-52.

* Corresponding author: Dr. Pedro O. de Campos-Lima, Laval University Cancer Research Center,

McMahon St 9, Quebec City, QC, G1R 2J6, Canada. Phone: 1 418 525 4444, fax: 1 418 691 5439, e-mail: [email protected]

Keywords: immunomodulation, alcohol, lymphocyte, monocyte, isopropanol, methanol.

______

Abbreviations: NFAT, nuclear factor of activated T cells; AP-1, activator protein-1; NF-B, nuclear factor-B; ERK, extracellular signal-regulated kinase; MAPK, mitogen-activated protein kinase;

TLR4, Toll-like receptor 4; TCR, T-cell receptor; APC, antigen-presenting cell; CRAC channel, Ca++ release-activated Ca++ channel; MyD88, myeloid differentiation primary response gene (88); IRAK1, interleukin-1 receptor-associated kinase 1; ZAP70, zeta-chain-associated protein kinase 70; CK1, casein kinase 1; GSK3, glycogen synthase kinase 3; DYRK1A, dual-specificity tyrosine phosphorylation-regulated kinase 1A; NMDA, N-methyl-D-aspartate; GIRK2, G protein-gated inwardly rectifying potassium channel 2; IRK1, G protein-insensitive inwardly rectifying potassium channel; GLIC, Gloeobacter violaceus pentameric ligand-gated ion channel; GABAA receptor, - aminobutyric acid type A receptor. 191

Abstract

Short-chain alcohols are embedded into several aspects of modern life. The societal costs emanating from the long history of use and abuse of the prototypical example of these molecules, ethanol, has stimulated considerable interest in its general toxicology. A much more modest picture exists for other short-chain alcohols, notably as regards their immunotoxicity. A large segment of the general population is potentially exposed to two of these alcohols, methanol and isopropanol. Their ubiquitous nature and their eventual use as ethanol surrogates are predictably associated to accidental or deliberate poisoning. This review addresses the immunological consequences of acute exposure to methanol and isopropanol. It first examines the general mechanisms of short-chain alcohol-induced biological dysregulation and then provides a tentative model to explain the molecular events that underlie the immunological dysfunction produced by methanol and isopropanol. The time-related context of serum alcohol concentrations in acute poisoning as well as the clinical implications of their short-term immunotoxicity are also discussed.

192

1. Introduction

The use of alcohols is deeply engrained into social, occupational, and domestic aspects of modern life.

The ubiquitousness of these substances creates ample opportunity for acute poisoning. In order to gain perspective on the magnitude of this problem, we shall examine the number of cases of acute alcohol exposure reported to poison centers in North America in a single year (Fig. 1). Not surprisingly, ethanol comes on top of the list with about 85,000 cases (Bronstein et al., 2010). Nevertheless, if one excludes those instances related to beverage consumption, there are close to 34,000 cases of acute exposure to other sources of ethanol such as cleaning products or mouthwash. Isopropanol is a close second with just under 21,000 acute exposure occurrences distributed in three main subgroups (Rubbing alcohol:

7,787; Cleaning products: 5,728; and Non-rubbing/Non-cleaning products: 7,305). Glycols account for about 10,000 exposures, most of them related to automotive products. Finally, exposure of the general population to dietary and environmental sources of the simplest alcohol, methanol, usually does not amount to dangerous levels (Shelby et al., 2004). Yet, methanol is also an important cause of alcohol intoxication and the mortality rate may reach 15-36% in the more severe cases (Brent et al., 2001;

Hunderi et al., 2004; Liu et al., 1998). Altogether, poison centers receive in excess of 2,000 reports of acute methanol exposure each year (Bronstein et al., 2010).

Intentional or accidental systemic exposure to short-chain alcohols has many biological consequences, including neurological and behavioral dysfunction (Jammalamadaka and Raissi 2010; Kraut and Kurtz

2008). It is not surprising that the long history of alcohol use and abuse in our society has stimulated considerable interest in the general toxicology of ethanol (Organization for Economic Co-operation and

Development, OECD-SIDS 2004). A much more modest picture exists for other short-chain alcohols, notably as regards their immunotoxicity. A substantial amount of data from multiple laboratories

193 indicates that ethanol modulates the immune function of T cells, monocytes, macrophages and neutrophils (Goral and Kovacs 2005; Saeed et al., 2004; Szabo et al., 2007; Taieb et al., 2002). Ethanol also inhibits the leukocyte/endothelial cell interaction thereby limiting the inflammatory response

(Saeed et al., 2004). The structural similarity between ethanol and other short-chain alcohols has lent support to the hypothesis that some of the latter molecules may also have immunomodulatory properties. Indeed, recent work on isopropanol and methanol has provided solid evidence that non- ethanolic short-chain alcohols also possess an immunological footprint (Carignan et al., 2012; Désy et al., 2008, 2010); they modulate the effector function of lymphocytes and monocytes through an isoform-specific effect on the activation of the nuclear factor of activated T cells (NFAT) family of transcription factors with or without additional involvement of the activator protein-1 (AP-1). This alcohol-induced dysregulation is restricted to AP-1 when the triggering of the immune cell effector function does not require NFAT (Carignan et al., 2012). Different aspects of the immunotoxicology of ethanol have already been addressed in several authoritative articles (Goral et al., 2008; Pruett et al.,

2004; Szabo et al., 2007, 2011; Szabo and Mandrekar, 2009). In the next sections, we shall focus our attention on a subject that has not been previously reviewed in the alcohol literature, the immunological impact of methanol and isopropanol. We will first examine the general mechanisms of short-chain alcohol-induced biological dysregulation and then provide a contextual insight into the molecular events that underlie the immunological effect of methanol and isopropanol.

194

2. General mechanisms of short-chain alcohol-induced biological dysregulation

The molecular events that underlie the biological effects of alcohols are complex but a more comprehensive picture begins to emerge at least in part by borrowing some key concepts from studies on anesthetic drugs (Eckenhoff, 2001; Streiff et al., 2006; Vedula et al., 2009). Our understanding of how alcohols (and other more efficient and/or less toxic anesthetics) work has been focused on the cell membrane for most of the twentieth century (Overton, 1901; Pringle et al., 1981; Roth and Seeman,

1972; Seeman et al., 1971). The first glimpse into the putative mechanism of action of alcohols came from the observation that they follow the Meyer-Overton rule similarly to other general anesthetics

(McCreery and Hunt, 1978; Meyer, 1899; Overton, 1901; Pringle et al., 1981). This rule correlates the anesthetic potency of a given compound to its oil:water partition coefficient; thus, molecules endowed with longer carbon chains tend to be more powerful as compared to those that are less hydrophobic.

The predominant view has been that the dissolution of lipophilic molecules in the lipid bilayer would change its physical properties thereby compromising indirectly the function of the embedded proteins

(Eckenhoff, 2001; Seeman, 1972). An alternative interpretation that does not involve the alteration of the lipid bilayer but that rather envisages proteins as direct alcohol targets received considerable support after primary alcohols were shown to bind luciferase in vitro in absence of a lipid context

(Franks and Lieb, 1984, 1985); they compete with the luciferin substrate and inhibit the enzymatic activity according to the size of their carbon chain. These findings suggest the existence of a hydrophobic binding pocket with the capacity to accommodate two n-hexanol molecules or a single n- dodecanol molecule, while larger alcohols would remain partially exposed to the solvent (Franks and

Lieb, 2004). Subsequent work involving the mutagenesis of specific amino acid residues helped to

195 define putative alcohol binding sites in several proteins (Table 1). Most of these proteins are implicated in neurobiological processes and do not have a direct immunological impact; it is worth noting, however, that there may be a crosstalk between the alcohol-induced molecular events that produce neurobiological consequences and those that lead to immunological effects. In this regard, it has recently been reported that the Toll-like receptor 4 (TLR4) expression is regulated by the- aminobutyric acid type A (GABAA) receptor in the central amigdala and is associated to binge drinking

(Liu et al., 2011). The GABAA receptor possesses an alcohol-binding cavity formed by amino acids from four transmembrane domains and has long been considered a target for the alcohol effect in the central nervous system (Harris et al., 2008). Most importantly, the mutagenesis data consolidated the findings obtained with the luciferase model to establish the principle that alcohols may interact directly with proteins that belong to very different functional groups provided that they display suitable alcohol- interaction pockets.

The anatomy of such binding pockets may be inferred from the most basic features of the alcohol molecule, the existence of a hydroxyl group bound to a carbon atom; the hydroxyl group allows alcohols to behave as weak acids and hydrogen bond donors while the aliphatic chain is responsible for their hydrophobic properties (Ballinger and Long, 1960; Dwyer and Bradley, 2000). It has been suggested that alcohols would bind a mirror image site in the target proteins (Dwyer and Bradley,

2000). According to this view, one would expect to find in close vicinity: i. Hydrogen bond acceptor site(s); ii. A localized net positive charge to interact with the electronegative atom of the hydroxyl group; and iii. A hydrophobic groove that could result from the packing of -helices. Thus, alcohols would displace water molecules from such pockets and establish atomic interactions with the proteins via hydrogen bonds that would be stabilized by van der Waals forces within the hydrophobic

196

environment (Klemm, 1998). The local distortion caused by this intermolecular interaction could

eventually lead to an alteration in protein function (Fig. 2).

The predicted structural features described above were largely confirmed by crystallographic data on alcohol-protein complexes (Table 1). The odorant binding protein LUSH has a water-filled cavity that binds a single molecule of ethanol, n-propanol and n-butanol at 30-50 mM concentrations. In this case, the alcohol molecule was shown to form hydrogen bonds with Thr57 and Ser52 and to be stabilized by interaction with hydrophobic residues within a pocket that sits between -helices (Thode et al., 2008).

Similar cavities were revealed by the resolution of the X-ray structure of the G protein-insensitive inwardly rectifying potassium channel (IRK1) and Gloeobacter violaceus pentameric ligand-gated ion channel (GLIC) in complex with 2-methyl-2,4-pentanediol and propofol, respectively (Nury et al.,

2011; Pegan et al., 2006). Finally, the structure of the alcohol-binding site of the alcohol dehydrogenase is well defined and, although the pocket has the same hydrophobic nature of those found in other previously described protein targets, it is the only one that uses a zinc atom to form a bond with the hydroxyl group (Ramaswamy et al., 1994). This peculiarity may account for the substantially higher alcohol-binding affinity of the enzyme, which is in the order of 1 mM.

The characterization of the molecular steps that trigger the neurobiological impact of alcohols is far more advanced than what is known about the events mediating their immunomodulatory activity. Thus, while short-chain alcohols were shown by structural and biochemical means to bind targets such as ion channels or adhesion molecules that may plausibly account for the neurobiological effects (Arevalo et al., 2008; Aryal et al., 2009; Harris et al., 2008), the conclusive identification of direct protein targets of alcohols in immune cells is still in its infancy. We do know that alcohols change multiple signal

197 transduction pathways that result in differential activation of transcription factors that are essential for the immune cell effector function (Désy et al., 2008; Mandrekar et al., 2007; Saeed et al., 2004; Szabo et al., 2007). Nevertheless, the bulk of the available information concerns ethanol only and the immunotoxicology of other short-chain alcohols has just begun to be explored. Moreover, an extra level of complexity is added by the observation that the impact of alcohols on immune cells may have different triggering points. Thus, the effect of ethanol on the lipopolysaccharide-induced TLR4 signaling in macrophages seems to initiate at the membrane level, while methanol and isopropanol act downstream of the cell membrane to produce the transcriptional dysfunction that characterizes their immunological impact (Carignan et al., 2012; Dai et al., 2005; Dai and Pruett, 2006; Desy et al., 2008,

2010; Szabo et al., 2007). In the next sections, we shall present a tentative model to explain how methanol and isopropanol interfere with lymphocyte activation; the clinical implications of the immunotoxicity of these alcohols will also be addressed.

3. Mechanisms of non-ethanolic short-chain alcohol-induced immunological dysfunction

Exposure to isopropanol is detrimental to human lymphocyte and monocyte activity in vitro, while methanol synergizes with activating stimuli to increase cytokine production by lymphocytes. This immune dysfunction can also be observed in animal models of acute alcohol intoxication as judged by the abnormal release of proinflammatory cytokines in the serum in response to the staphylococcal enterotoxin B or lipopolysaccharide (Carignan et al., 2012; Désy et al., 2008, 2010). Table 2 summarizes the consequences of acute exposure on the effector function of immune cells in vitro and in vivo.

198

3.1 Mechanisms of alcohol-induced T lymphocyte dysfunction

Both methanol and isopropanol have a defined immunomodulatory impact on lymphocytes, albeit the resulting outcome of acute exposure to these alcohols is clearly different. T lymphocytes constitute a major component of the adaptive immune system and any dysfunction is bound to have profound health implications (Smith-Garvin et al., 2009). They use their T-cell receptor (TCR) to scan the surface of antigen-presenting cells (APCs) for short non-self peptides presented by major histocompatibility complex-encoded molecules (Wucherpfennig et al., 2010). Upon activation, the T cell/APC interface undergoes a dramatic reorganization leading to the clustering of the antigen receptor, co-receptor, and

CD28 on one side and their respective ligands on the other (Monks et al., 1998).

It is conceivable that alcohols could interfere with the profound reorganization of membrane molecules that precede the formation of the immunological synapse. This assumption is based on the known capacity of alcohols to partition into cell membranes (Nizza and Gawrisch, 2009). In fact, the highly choreographed maturation of the immunological synapse depends on the coalescence of lipid rafts of nanometer dimensions into micrometer-sized membrane microdomains that constitute the operational platforms for TCR signaling (Harder and Sangani, 2009). In this scenario, methanol and isopropanol could disturb lipid-protein interactions, disorder the structure of lipid microdomains and interrupt their coalescence. The end result would be a dysfunctional synapse and the abnormal lymphocyte activation.

Nevertheless, TCR early signaling as measured by the activation of the Zeta-chain-associated protein kinase 70 (ZAP70) is preserved in methanol- and isopropanol-treated lymphocytes and both alcohols initiate their effects downstream of the cell membrane (Désy et al., 2008, 2010). This interpretation is

199 in line with the fact that the degree of insertion of an alcohol molecule into the hydrophobic core of the bilayer is directly dependent on the size of its aliphatic chain. While long-chain alcohols partition well into the membrane with their aliphatic chains aligned in parallel to the lipid hydrocarbon chains, short- chain alcohols such as n-butanol are less efficient in reaching the hydrophobic core (Westerman et al.,

1988). Isopropanol and methanol are significantly more hydrophilic than n-butanol and, thus, are expected to interfere poorly with the hydrophobic core of the membrane.

Alcohols could also mediate their immunomodulation through the direct interaction with target proteins. When used in high concentrations (> 500 mM), they induce the formation of -helices and may disrupt tertiary structures, effectively leading to protein denaturation (Dwyer and Bradley, 2000).

However, they induce subtler structural changes at more physiologically relevant concentrations that may be associated to functional dysregulation of specific proteins such as those listed in table 1. As the available experimental evidence makes unlikely the involvement of a membrane protein such as the

TCR itself, the putative target(s) for methanol and isopropanol could conceivably reside in the chain of molecular events that transmit the activation signal from the T cell receptor to the nucleus. Once

ZAP70 becomes activated it phosphorylates the cytoplasmic tail of LAT, a transmembrane protein that localizes to the lipid rafts, and triggers a phosphorylation cascade that is followed by Ca++ release from intracellular stores and by the extra-cellular influx regulated by Ca++ release-activated Ca++ (CRAC) channels (Gallo et al., 2006; Smith-Garvin et al., 2009). The increase in intracellular Ca++ is sensed by calcineurin that, in presence of calmodulin, activates cytoplasmic NFAT and permits its nuclear translocation (Smith-Garvin et al., 2009). It is noteworthy that although alcohols may interfere with the function of ion channels in other experimental systems (Jung et al., 2005; Shahidullah et al., 2003), isopropanol does not change the CRAC-regulated Ca++ influx in activated lymphocytes. In addition, the

200 phosphatase activity of calcineurin is preserved in presence of methanol or isopropanol. This observation is relevant since ethanol was shown to bind calmodulin enhancing its affinity for Ca++ and changing its impact on calcineurin function (Chattopadhyaya et al., 1992; Ohashi et al., 2004).

Moreover, despite their structural similarity to ethanol, methanol and isopropanol do not compromise the nuclear translocation of NF-B, a transcription factor that has been associated to many of the biological effects of ethanol (Desy et al., 2010; Mandrekar et al., 2007; Saeed et al., 2004; Szabo et al.,

2007).

It is now clear that methanol and isopropanol are capable of inducing a selective dysregulation of discrete members of the NFAT family of transcription factors (Désy et al., 2008, 2010). This family is composed of five members, from which NFATc1, c2, and c3 have been well characterized in the immune system (Hogan et al., 2003). NFATc3 is expressed in thymocytes and is important for T cell development (Oukka et al., 1998). NFATc2 is abundantly expressed in the cytoplasm of resting peripheral lymphocytes. NFATc1 is similarly found in peripheral T cell subsets but only one of its isoforms, NFATc1A, is strongly inducible by antigen receptor activation and participates in an autoregulatory loop (Macian, 2005). Although both methanol and isopropanol target the NFAT family, they do so in a different manner. Isopropanol disables the activation of NFATc1, while methanol selectively enhances the NFATc2 nuclear content following TCR engagement (Désy et al., 2008,

2010). The differential methanol effect is not likely to result from an abnormal shuttling of NFATc2 mediated by importin 1 because the nuclear translocation of other substrates remains unaltered. The closely related NFATc1 as well as c-Fos are among the transcription factors that bind importin 1

(Ishiguro et al., 2011; Malnou et al., 2007) and traffic normally in presence of this alcohol. The fact that

201 several enzymes that function as nuclear export kinases and/or cytoplasmic maintenance kinases are capable of targeting simultaneously both NFATc1 and c2 may help to discard other possible mechanistic scenarios for the methanol selective action. Thus, casein kinase 1 (CK1), glycogen synthase kinase 3 (GSK3), and dual-specificity tyrosine phosphorylation-regulated kinase (DYRK) 1A are examples that fit in this category. In contrast to the previous examples, the inducible kinase p38 is a plausible candidate to be a methanol target because it specifically phosphorylates NFATc2 at the first serine of the SRR1 motif (Gómez del Arco et al., 2000). Loss of function of p38 could conceivably be associated to NFAT nuclear retention. However, analysis of the p38 postactivation status revealed no discernable difference in protein expression, phosphorylation pattern, and kinase activity in methanol- treated and untreated samples (Désy et al., 2010). Figure 3, panels A and B, presents a tentative model for isopropanol and methanol action, in which these alcohols interact directly with different members of the NFAT family of transcription factors. Isopropanol would bind to NFATc1 producing a conformational change that favors its nuclear export or cytoplasmic retention; methanol would interact with NFATc2 producing a conformational change that masks the nuclear-export signal or reduces the accessibility of phosphorylation sites.

3.2 Mechanisms of alcohol-induced innate immune cell dysfunction

Isopropanol also modulates the effector function of NK cells, monocytes, and macrophages following

TLR engagement. In particular, these cells lose their ability to secrete inflammatory cytokines in vitro and in vivo (Table 2). The magnitude of the dampening effect on TNF- production is enough to protect alcohol-intoxicated animals from toxic shock induced by otherwise lethal doses of lipopolysaccharide (Carignan et al., 2012). This functional dysregulation is produced by biochemical

202 changes that resemble those found in alcohol-treated T lymphocytes but involve distinct molecular targets. Thus, the recognition of pathogen-associated molecular patterns by the specific Toll-like receptor leads to the triggering of signaling cascades that eventually activate the transcription of immunologically relevant genes (Fig. 3C). Lipopolysaccharide recognition leads to receptor clustering and recruitment of key molecules to the TLR4 cytoplasmic tail (da Silva Correia et al., 2001; Miyake,

2006). Among those, the adapter protein encoded by the Myeloid differentiation primary response gene

(88) (MyD88) is responsible for the recruitment of the Interleukin-1 Receptor-Associated Kinase 1

(IRAK1), which in its turn is central to the activation of NF-B and AP-1 (Kawai and Akira, 2010). In presence of isopropanol, TLR4 early signaling proceeds normally because the kinase activity of IRAK1 remains intact. The existence of a downstream target was subsequently confirmed by the finding that isopropanol compromises the Extracellular signal-Regulated Kinase (ERK) module of the Mitogen-

Activated Protein Kinase (MAPK) signaling cascade in primary monocytes (Carignan et al., 2012).

Specifically, the phosphorylation of ERK2, but not of ERK1, is diminished in response to lipopolysaccharide stimulation. As a consequence, the nuclear content of c-Fos and JunB is reduced leading to a substantial alteration of the effector function of these cells (Fig. 3C).

One alternative interpretation for the alcohol-induced immune dysfunction described in the above paragraphs is that it is secondary to a stress response. It is well established that ethanol acute exposure causes a stress response in mice (Pruett et al. 2009); thus, corticosterone could conceivably mediate indirectly the immunosuppression observed in animals poisoned with isopropanol. Although corticosterone and/or its metabolites have not been measured in animals acutely intoxicated, this possibility is rather unlikely considering that the isopropanol-induced biochemical changes in signal transduction, the transcriptional dysregulation, and ultimately the alterations of immune cell effector

203 function occur in vitro in absence of a systemic endocrine response (Désy et al., 2008).

In conclusion, the immunomodulatory properties of methanol and isopropanol on innate and adaptive immune cells result from the defective activation of NFAT and/or AP-1. Therefore, although methanol does not interfere with the pathway leading to the activation of AP-1, the defective AP-1 activation induced by isopropanol cooperates with the reduced nuclear translocation of NFAT to limit the cytokine release from lymphocytes (Désy et al., 2008, 2010). Instead, in activation conditions that are not reliant on NFAT such as the response of monocytes to lipopolysaccharide, the down-regulation of

AP-1 is sufficient to produce a dose-dependent defective phenotype (Carignan et al., 2012). As a final point, NFAT and/or AP-1 predictably participate in the transcriptional regulation of all chemokine/cytokines whose syntheses are altered by alcohols (IL-2, IFN-, TNF-, and CCL2)

(Attema et al., 2002; Carignan et al., 2012; Durand et al., 1988, Kok et al., 2009). Even the up- regulation of IL-6 in monocytes, which occurs paradoxically despite the overall loss of function induced by isopropanol, can be explained by the down-regulation of JunB. This transcription factor acts as a repressor of the IL-6 promoter (Pflegerl et al., 2009).

4. Clinical considerations and concluding remarks

The alcohol-induced transcriptional dysregulation discussed in the previous section has clear immunological consequences. The most immediate clinical implications of the alcohol-induced immunotoxicity are in the field of emergency medicine. Both alcohols are easily accessible to the general population in relatively large quantities as part of a myriad of diverse household products.

Confirmed alcoholics, children, and suicidal individuals compose the main group of patients who are

204 admitted in the emergency room with suspected acute alcohol intoxication (Barceloux et al., 2002;

Zaman et al., 2002). There are well-documented cases of poisoning caused by inhalation of methanol

(Bebarta et al., 2006; Frenia and Schauben, 1993; LoVecchio et al., 2004; McCormick et al., 1990;

Velez et al., 2003) and by percutaneous absorption of methanol (Aufderheide et al., 1993) and isopropanol (Arditi and Killner, 1987; Dyer et al., 2002; Leeper et al., 2000). Nevertheless, most cases of intoxication occur after deliberate or accidental oral ingestion of either alcohol (Kraut and Kurtz,

2008). Both methanol and isopropanol are readily absorbed by the gastrointestinal tract and reach maximum serum levels within the first hour after ingestion.

4.1 Time-related context of methanol and isopropanol serum concentrations in acute intoxication

The serum alcohol levels measured upon presentation at the hospital vary considerably. Careful analysis of the literature on human exposure to isopropanol and methanol reveals that many of the reported laboratory measurements have been made hours after ingestion and may underestimate the alcohol serum amounts present in the early phase of the intoxication (Daniel et al., 1981; Gaudet and

Fraser, 1989; King et al., 1970; Kostic and Dart, 2003; Mueller-Kronast et al., 2003; Rich et al., 1990;

Rosansky, 1982). The reported average sublethal serum isopropanol concentration in severely intoxicated patients is 310 mg/dl (0.31% or 51.5 mmol/l) with an average time lag between ingestion and first quantification of about 7h (Ekwall and Clemedson, 1997). Studies made with human volunteers and with a pharmacokinetic model showed that serum isopropanol concentrations measured

3h after intake are preceded by a peak that may be seven times higher within the first hour (Clewell et al., 2001; Lacouture et al., 1983; Monaghan et al., 1995). This disparity between the values obtained on

205 hospital admission and peak levels is particularly evident in the case of methanol poisoning because of its relatively low intrinsic toxicity; the patients usually seek help much later, sometimes two days after methanol ingestion, when the formate concentration builds up causing a more obvious clinical picture

(Barceloux et al., 2002; Kostic and Dart, 2003). In support of this scenario, we have reviewed 86 cases published between 1953 and 2011 of sublethal methanol poisoning that included a precise time of ingestion and a serum concentration measured on or after admission. The mean interval between ingestion and serum methanol measurement ± SEM was 26 ± 3h. We have calculated the predicted methanol peak value at 30 min post-ingestion for each patient and found it to be on average 4.5 times higher than the first alcohol measurement. Overall, the mean methanol peak value in this group of patients was 2.2 fold higher than the mean serum measurement at the hospital (Fig. 4).

In agreement with the pharmacokinetics of these two alcohols, we can assume that the transcriptional dysregulation and consequent immune dysfunction would be most intense during the first few hours post-ingestion in the case of isopropanol intoxication and during the first several hours of methanol poisoning.

4.2 Medical implications of the immunological effects of methanol and isopropanol

Acutely intoxicated animals have a dysfunctional immune system characterized by a state of profound immunosupression in the case of isopropanol exposure and by a state of hyperreactivity in the case of methanol exposure (Désy et al., 2008, 2010). Thus, it is conceivable that patients intoxicated with isopropanol are more susceptible to local or systemic infections during the first hours post-ingestion; this possibility may be more pronounced when skin damage or other forms of trauma are present. In

206 fact, the state of inebriation resulting from excessive consumption of any alcohol is often associated to trauma (Fitzgerald et al., 2007; Saxena et al., 1987). In addition, isopropanol is often utilized as a disinfectant. When used in most conventional topical applications, the immunological effect of the alcohol on normal adult skin, if any, is likely to be rather limited, transient, and confined to intradermal cells because of its poor absorption through this route. Nevertheless, a cautious attitude would be justifiable for the use of isopropanol on diseased skin until the potential consequences are better known. In this regard, it is noteworthy that JunB is down-regulated in keratinocytes from patients with systemic lupus erythematosus leading to an abnormal profile of skin-produced cytokines, which may be potentially worsened by isopropanol topical exposure (Carignan et al., 2012; Pflegerl et al., 2009).

From another standpoint, the massive production of inflammatory cytokines may play a role in the pathophysiology of methanol intoxication; it is conceivable that the abnormal TNF- release in response to bacterial products in injured patients or to an underlying infection could contribute to the development of complications such as methanol-induced acute pancreatitis, a condition that may be identified in half of the cases of severe acute intoxication (Hantson and Mahieu, 2000). It has also been speculated that TNF- up-regulation could be involved in the methanol-induced selective toxic injury of the basal ganglia (Désy et al., 2010). If confirmed, these predictions could provide the rationale for the institution of precautionary measures to limit the risk of infection in isopropanol poisoning and for the establishment of anti-inflammatory measures to offset the proinflammatory properties of methanol.

In conclusion, acute exposure to methanol and isopropanol has immunological consequences. The existing body of information on the immunotoxicology of these two alcohols, albeit small, justifies the inclusion of immunological end points into the design of future studies.

207

Conflict of interest

The authors declare that there is no conflict of interest.

Acknowledgements

The authors wish to thank Dr. Manuel Caruso for helpful discussions and critical reading of the manuscript; this work was supported by grants from the National Sciences and Engineering Research

Council of Canada and the Canada Foundation for Innovation.

References

Arditi, M., Killner, M.S., 1987. Coma following use of rubbing alcohol for fever control. Am. J. Dis.

Child. 141, 237-238.

Arevalo, E., Shanmugasundararaj, S., Wilkemeyer, M.F., Dou, X., Chen, S., Charness, M.E., Miller,

K.W., 2008. An alcohol binding site on the neural cell adhesion molecule L1. Proc. Natl. Acad.

Sci. USA, 105, 371-375.

Aryal, P., Dvir, H., Choe, S., Slesinger, P.A., 2009. A discrete alcohol pocket involved in GIRK

channel activation. Nat. Neurosci. 12, 988-995.

Attema, J.L., Reeves, R., Murray, V., Levichkin, I., Temple, M.D., Tremethick, D.J., Shannon, M.F.,

2002. The human IL-2 gene promoter can assemble a positioned nucleosome that becomes

208

remodeled upon T cell activation. J. Immunol. 169, 2466-2476.

Aufderheide, T.P., White, S.M., Brady, W.J., Stueven, H.A., 1993. Inhalational and percutaneous

methanol toxicity in two firefighters. Ann. Emerg. Med. 22, 1916-1918.

Ballinger, P., Long, F.A., 1960. Acid ionization constants of alcohols. II. Acidities of some substituted

methanols and related compounds. J. Am. Chem. Soc. 82, 795-798.

Barbulescu, K., Becker, C., Schlaak, J.F., Schmitt, E., Meyer zum Büschenfelde, K.H., Neurath, M.F.,

1998. IL-12 and IL-18 differentially regulate the transcriptional activity of the human IFN-gamma

promoter in primary CD4+ T lymphocytes. J. Immunol. 160, 3642-3647.

Barceloux, D.G., Bond, G.R., Krenzelok, E.P., Cooper, H., Vale, J.A., 2002. American Academy of

Clinical Toxicology practice guidelines on the treatment of methanol poisoning. J. Toxicol. Clin.

Toxicol. 40, 415-446.

Bebarta, V.S., Heard, K., Dart, R.C., 2006. Inhalational abuse of methanol products: elevated methanol

and formate levels without vision loss. Am. J. Emerg. Med. 24, 725-728.

Bennett Jr, I.L., Cary, F.H., Mitchell Jr, G.L., Cooper, M.N., 1953. Acute methyl alcohol poisoning: a

review based on experiences in an outbreak of 323 cases. Medicine 32, 431-463.

Brahmi, N., Blel, Y., Abidi, N., Kouraichi, N., Thabet, H., Hedhili, A., Amamou, M., 2007. Methanol

poisoning in Tunisia: report of 16 cases. Clin. Toxicol. 45, 717-720.

Brent, J., Lucas, M., Kulig, K., Rumack, B.H., 1991. Methanol poisoning in a 6-week-old infant. J.

Pediatr. 118, 644-646.

Brent, J., McMartin, K., Phillips, S., Aaron, C., Kulig, K., 2001. Fomepizole for the treatment of

methanol poisoning. N. Engl. J. Med. 344, 424-429.

Bronstein, A.C., Spyker, D.A., Cantilena Jr, L.R., Green, J.L., Rumack, B.H., Giffin, S.L., 2010. 2009

Annual Report of the American Association of Poison Control Centers' National Poison Data

209

System (NPDS): 27th Annual Report. Clin. Toxicol. 48, 979-1178.

Brown, M.J., Shannon, M.W., Woolf, A., Boyer, E.W., 2001. Childhood methanol ingestion treated

with fomepizole and hemodialysis. Pediatrics 108(4):E77 Brown MJ, Shannon MW, Woolf A,

Boyer EW (2001) Childhood methanol ingestion treated with fomepizole and hemodialysis.

Pediatrics 108(4), E77.

Bucaretchi, F., De Capitani, E.M., Madureira, P.R., Cesconetto, D.M., Lanaro, R., Vieira, R.J., 2009.

Suicide attempt using pure methanol with hospitalization of the patient soon after ingestion: case

report. Sao Paulo Med. J. 127, 108-110.

Burns, M.J., Graudins, A., Aaron, C.K., McMartin, K., Brent, J., 1997. Treatment of methanol

poisoning with intravenous 4-methylpyrazole. Ann. Emerg. Med. 30, 829-832.

Caravati, E.M., Anderson, K.T., 2010. Breath alcohol analyzer mistakes methanol poisoning for

alcohol intoxication. Ann. Emerg. Med. 55, 198-200.

Carignan, D., Désy, O., Caruso, M., de Campos-Lima, P.O., 2012. The dysregulation of the

monocyte/macrophage effector function induced by isopropanol is mediated by the defective

activation of distinct members of the AP-1 family of transcription factors Toxicol. Sci. 125, 144-

156.

Chattopadhyaya, R., Meador, W.E., Means, A.R., Quiocho, F.A., 1992. Calmodulin structure refined at

1.7 A resolution. J. Mol. Biol. 228, 1177-1192.

Clewell 3rd, H.J., Gentry, P.R., Gearhart, J.M., Covington, T.R., Banton, M.I., Andersen, M.E., 2001.

Development of a physiologically based pharmacokinetic model of isopropanol and its metabolite

acetone. Toxicol. Sci. 63, 160-172.

Coulter, C.V., Isbister, G.K., Duffull, S.B., 2011. The pharmacokinetics of methanol in the presence of

ethanol: a case study. Clin. Pharmacokinet. 50, 245-251.

210 da Silva Correia, J., Soldau, K., Christen, U., Tobias, P.S., Ulevitch, R.J., 2001. Lipopolysaccharide is

in close proximity to each of the proteins in its membrane receptor complex. transfer from CD14 to

TLR4 and MD-2. J. Biol. Chem. 276, 21129-21135.

Dai, Q., Pruett, S.B., 2006. Ethanol suppresses LPS-induced Toll-like receptor 4 clustering,

reorganization of the actin cytoskeleton, and associated TNF-alpha production. Alcohol. Clin. Exp.

Res. 30, 1436-1444.

Dai, Q., Zhang, J., Pruett, S.B., 2005. Ethanol alters cellular activation and CD14 partitioning in lipid

rafts. Biochem. Biophys. Res. Commun. 332, 37-42.

Daniel, D.R., McAnalley, B.H., Garriott, J.C., 1981. Isopropyl alcohol metabolism after acute

intoxication in humans. J. Anal. Toxicol. 5, 110-112.

Désy, O., Carignan, D., Caruso, M., de Campos-Lima, P.O., 2008. Immunosuppressive effect of

isopropanol: down-regulation of cytokine production results from the alteration of discrete

transcriptional pathways in activated lymphocytes. J. Immunol. 181, 2348-2355.

Désy, O., Carignan, D., Caruso, M., de Campos-Lima, P.O., 2010. Methanol induces a discrete

transcriptional dysregulation that leads to cytokine overproduction in activated lymphocytes.

Toxicol. Sci. 117, 303-313.

Dou, X., Menkari, C.E., Shanmugasundararaj, S., Miller, K.W., Charness, M.E., 2011. Two alcohol

binding residues interact across a domain interface of the L1 neural cell adhesion molecule and

regulate cell adhesion. J. Biol. Chem. 286, 16131-16139.

Durand, D.B., Shaw, J.P., Bush, M.R., Replogle, R.E., Belagaje, R., Crabtree, G.R., 1988.

Characterization of antigen receptor response elements within the interleukin-2 enhancer. Mol.

Cell. Biol. 8, 1715-1724.

Dwyer, D.S., Bradley, R.J., 2000. Chemical properties of alcohols and their protein binding sites. Cell.

211

Mol. Life Sci. 57, 265-275.

Dyer, S., Mycyk, M.B., Ahrens, W.R., Zell-Kanter, M., 2002. Hemorrhagic gastritis from topical

isopropanol exposure. Ann. Pharmacother. 36, 1733-1735.

Eckenhoff, R.G., 2001. Promiscuous ligands and attractive cavities: how do the inhaled anesthetics

work? Mol. Interv. 1, 258-268.

Ekins, B.R., Rollins, D.E., Duffy, D.P., Gregory, M.C., 1985. Standardized treatment of severe

methanol poisoning with ethanol and hemodialysis. West. J. Med. 142, 337-340.

Ekwall, B., Clemedson, C., 1997. Time-Related Lethal Blood Concentrations from Acute Human

Poisoning of Chemicals. Part 2: The Monographs. No. 10 Isopropyl Alcohol. CTLU, Stanga,

Sweden.

Fitzgerald, D.J., Radek, K.A., Chaar, M., Faunce, D.E., DiPietro, L.A., Kovacs, E.J., 2007. Effects of

acute ethanol exposure on the early inflammatory response after excisional injury. Alcohol. Clin.

Exp. Res. 31, 317-323.

Fontenot, A.P., Pelak, V.S., 2002. Development of neurologic symptoms in a 26-year-old woman

following recovery from methanol intoxication. Chest 122, 1436-1439.

Forman, S.A., Miller, K.W., Yellen, G., 1995. A discrete site for general anesthetics on a postsynaptic

receptor. Mol. Pharmacol. 48, 574-581.

Franks, N.P., Jenkins, A., Conti, E., Lieb, W.R., Brick, P., 1998. Structural basis for the inhibition of

firefly luciferase by a general anesthetic. Biophys. J. 75, 2205-2211.

Franks, N.P., Lieb, W.R., 1984. Do general anaesthetics act by competitive binding to specific

receptors? Nature 310, 599-601.

Franks, N.P., Lieb, W.R., 1985. Mapping of general anaesthetic target sites provides a molecular basis

for cutoff effects. Nature 316, 349-351.

212

Franks, N.P., Lieb, W.R., 2004. Seeing the light: protein theories of general anesthesia. Anesthesiology

101, 235-237.

Frenia, M.L., Schauben, J.L., 1993. Methanol inhalation toxicity. Ann. Emerg. Med. 22, 1919-1923.

Gallo, E.M., Cante-Barrett, K., Crabtree, G.R., 2006. Lymphocyte calcium signaling from membrane to

nucleus. Nat. Immunol. 7, 25-32.

Gaudet, M.P., Fraser, G.L., 1989. Isopropanol ingestion: case report with pharmacokinetic analysis.

Am. J. Emerg. Med. 7, 297-299.

Ghannoum, M., Haddad, H.K., Lavergne, V., Heinegg, J., Jobin, J., Halperin, M.L., 2010. Lack of toxic

effects of methanol in a patient with HIV. Am. J. Kidney Dis. 55, 957-961.

Gómez del Arco, P., Martínez-Martínez, S., Maldonado, J.L., Ortega-Pérez, I., Redondo, J.M., 2000. A

role for the p38 MAP kinase pathway in the nuclear shuttling of NFATp. J. Biol. Chem. 275,

13872-13878.

Goral, J., Karavitis, J., Kovacs, E.J., 2008. Exposure-dependent effects of ethanol on the innate immune

system. Alcohol 42, 237-247.

Goral, J., Kovacs, E.J., 2005. In vivo ethanol exposure down-regulates TLR2, TLR4, and TLR9-

mediated macrophage inflammatory response by limiting p38 and ERK1/2 activation. J. Immunol.

174, 456-463.

Grufferman, S., Morris, D., Alvarez, J., 1985. Methanol poisoning complicated by myoglobinuric renal

failure. Am. J. Emerg. Med. 3, 24-26.

Hantson, P., Lambermont, J.Y., Mahieu, P., 1997. Methanol poisoning during late pregnancy. J.

Toxicol. Clin. Toxicol. 35, 187-191.

Hantson, P., Wallemacq, P., Brau, M., Vanbinst, R., Haufroid, V., Mahieu, P., 1999. Two cases of

acute methanol poisoning partially treated by oral 4-methylpyrazole. Intensive Care Med. 25, 528-

213

531.

Hantson, P., Haufroid, V., Mahieu, P., 2000. Survival with extremely high blood methanol

concentration. Eur. J. Emerg. Med. 7, 237-240.

Hantson, P., Mahieu, P., 2000. Pancreatic injury following acute methanol poisoning. J. Toxicol. Clin.

Toxicol. 38, 297-303.

Harder, T., Sangani, D., 2009. Plasma membrane rafts engaged in T cell signalling: new developments

in an old concept. Cell Commun. Signal. 7, 21.

Harris, R.A., Trudell, J.R., Mihic, S.J., 2008. Ethanol's molecular targets. Sci. Signal. 1, re7.

Hogan, P.G., Chen, L., Nardone, J., Rao, A. 2003. Transcriptional regulation by calcium, calcineurin,

and NFAT. Genes Dev. 17, 2205-2232.

Hunderi, O.H., Hovda, K.E., Lie, B., Listerud, S., Hartmann, A., Jacobsen, D., 2004.

Metanolforgiftningene i 2002. Tidsskr. Nor. Laegeforen. 124, 3199-3202.

Ingemansson, S.O., 1984. Clinical observations on ten cases of methanol poisoning with particular

reference to ocular manifestations. Acta Ophthalmol. 62, 15-24.

Ishiguro, K., Ando, T., Maeda, O., Watanabe, O., Goto, H., 2011. Cutting edge: tubulin α functions as

an adaptor in NFAT-importin β interaction. J. Immunol. 186, 2710-2713.

Jacobsen, D., Ovrebø, S., Arnesen, E., Paus, P.N., 1983. Pulmonary excretion of methanol in man.

Scand. J. Clin. Lab. Invest. 43, 377-379.

Jacobsen, D., Webb, R., Collins, T.D., McMartin, K.E., 1988. Methanol and formate kinetics in late

diagnosed methanol intoxication. Med. Toxicol. Adverse Drug Exp. 3, 418-423.

Jammalamadaka, D., Raissi, S., 2010. Ethylene glycol, methanol and isopropyl alcohol intoxication.

Am. J. Med. Sci. 339, 276-281.

Jones, B., Chen, J. 2006., Inhibition of IFN-gamma transcription by site-specific methylation during T

214

helper cell development. EMBO J. 25, 2443-2452.

Jung, S., Akabas, M.H., Harris, R.A., 2005. Functional and structural analysis of the GABAA receptor

alpha 1 subunit during channel gating and alcohol modulation. J. Biol. Chem. 280, 308-316.

Kawai, T., Akira, S., 2010. The role of pattern-recognition receptors in innate immunity: update on

Toll-like receptors. Nat. Immunol. 11, 373-384.

Keyvan-Larijarni, H., Tannenberg, A.M., 1974. Methanol intoxication. Comparison of peritoneal

dialysis and hemodialysis treatment. Arch. Intern. Med. 134, 293-296.

King Jr, L.H., Bradley, K.P., Shires Jr, D.L., 1970. Hemodialysis for isopropyl alcohol poisoning.

JAMA 211, 1855.

Klemm, W.R., 1998. Biological water and its role in the effects of alcohol. Alcohol 15, 249-267.

Kok, S.H., Hong, C.Y., Kuo, M.Y., Wang, C.C., Hou, K.L., Lin, Y.T., Galson, D.L., Lin, S.K., 2009.

Oncostatin M-induced CCL2 transcription in osteoblastic cells is mediated by multiple levels of

STAT-1 and STAT-3 signaling: an implication for the pathogenesis of arthritis. Arthritis Rheum.

60, 1451-1462.

Kostic, M.A., Dart, R.C., 2003. Rethinking the toxic methanol level. J. Toxicol. Clin. Toxicol. 41, 793-

800.

Kraut, J.A., Kurtz, I., 2008. Toxic alcohol ingestions: clinical features, diagnosis, and management.

Clin. J. Am. Soc. Nephrol. 3, 208-225.

Kruse, S.W., Zhao, R., Smith, D.P., Jones, D.N., 2003. Structure of a specific alcohol-binding site

defined by the odorant binding protein LUSH from Drosophila melanogaster. Nat. Struct. Biol. 10,

694-700.

Lacouture, P.G., Wason, S., Abrams, A., Lovejoy Jr, F.H., 1983. Acute isopropyl alcohol intoxication.

Diagnosis and management. Am. J. Med. 75, 680-686.

215

Leeper, S.C., Almatari, A.L., Ingram, J.D., Ferslew, K.E., 2000. Topical absorption of isopropyl

alcohol induced cardiac and neurologic deficits in an adult female with intact skin. Vet. Hum.

Toxicol. 42, 15-17.

Liu, J., Yang, A.R., Kelly, T., Puche, A., Esoga, C., June Jr., H.L., Elnabawi, A., Merchenthaler, I.,

Sieghart, W., June Sr., H.L., Aurelian, L., 2011. Binge alcohol drinking is associated with GABAA

alpha2-regulated Toll-like receptor 4 (TLR4) expression in the central amygdala. Proc. Natl. Acad.

Sci. USA 108, 4465-4470.

Liu, J.J., Daya, M.R., Carrasquillo, O., Kales, S.N., 1998. Prognostic factors in patients with methanol

poisoning. J. Toxicol. Clin. Toxicol. 36, 175-181.

Lobo, I.A., Harris, R.A., Trudell, J.R., 2008. Cross-linking of sites involved with alcohol action

between transmembrane segments 1 and 3 of the glycine receptor following activation. J.

Neurochem. 104, 1649-1662.

LoVecchio, F., Sawyers, B., Thole, D., Beuler, M.C., Winchell, J., Curry, S.C., 2004. Outcomes

following abuse of methanol-containing carburetor cleaners. Hum. Exp. Toxicol. 23, 473- 475.

Macian, F., 2005. NFAT proteins: key regulators of T-cell development and function. Nat. Rev.

Immunol. 5, 472-484.

Mahieu, P., Hassoun, A., Lauwerys, R., 1989. Predictors of methanol intoxication with unfavourable

outcome. Hum. Toxicol. 8, 135-137.

Malnou, C.E., Salem, T., Brockly, F., Wodrich, H., Piechaczyk, M., Jariel-Encontre, I., 2007.

Heterodimerization with Jun family members regulates c-Fos nucleocytoplasmic traffic. J. Biol.

Chem. 282, 31046-31059.

Mandrekar, P., Jeliazkova, V., Catalano, D., Szabo, G., 2007. Acute alcohol exposure exerts anti-

inflammatory effects by inhibiting IkappaB kinase activity and p65 phosphorylation in human

216

monocytes. J. Immunol. 178, 7686-7693.

Martens, J., Westhovens, R., Verberckmoes, R., Delooz, H., Daenens, P., 1982. Recovery without

sequelae from severe methanol intoxication. Postgrad. Med. J. 58, 454-456.

McCormick, M.J., Mogabgab, E., Adams, S.L., 1990. Methanol poisoning as a result of inhalational

solvent abuse. Ann. Emerg. Med. 19, 639-642.

McCreery, M.J., Hunt, W.A., 1978. Physico-chemical correlates of alcohol intoxication.

Neuropharmacology 17, 451-461.

Meyer, H.H., 1899. Zur Theorie der Alkoholnarkose. Erste Mittheilung. Welche Eigenschaft der

Anästhetica bedingt ihre narkotische Wirkung? Archiv. f. experiment. Pathol. u. Pharmakol. 42,

109-118.

Miethke, T., Wahl, C., Heeg, K., Echtenacher, B., Krammer, P.H., Wagner, H., 1992. T cell-mediated

lethal shock triggered in mice by the superantigen staphylococcal enterotoxin B: critical role of

tumor necrosis factor. J. Exp. Med. 175, 91-98.

Monaghan, M.S., Olsen, K.M., Ackerman, B.H., Fuller, G.L., Porter, W.H., Pappas, A.A., 1995.

Measurement of serum isopropanol and the acetone metabolite by proton nuclear magnetic

resonance: application to pharmacokinetic evaluation in a simulated overdose model. J. Toxicol.

Clin. Toxicol. 33, 141-149.

Monks, C.R., Freiberg, B.A., Kupfer, H., Sciaky, N., Kupfer, A., 1998. Three-dimensional segregation

of supramolecular activation clusters in T cells. Nature 395, 82-86.

Mowry, J.B., Judge, B.S., Bloch, R., Rusyniak, D.E., 2002. Rapid methanol clearance with high-flux

hemodialysis. J. Toxicol. Clin. Toxicol. 40, 609-610.

Mueller-Kronast, N., Rabinstein, A.A., Voung, L., Forteza, A.M., 2003. Isopropanol intoxication

mimicking basilar artery thrombosis. Neurology 61,1456-1457.

217

Miyake, K., 2006. Roles for accessory molecules in microbial recognition by Toll-like receptors. J.

Endotoxin Res. 12, 195-204.

Nizza, D.T., Gawrisch, K., 2009. A layer model of ethanol partitioning into lipid membranes. Gen.

Physiol. Biophys. 28, 140-145.

Nury, H., Van Renterghem, C., Weng, Y., Tran, A., Baaden, M., Dufresne, V., Changeux, J.P., Sonner,

J.M., Delarue, M., Corringer, P.J., 2011. X-ray structures of general anaesthetics bound to a

pentameric ligand-gated ion channel. Nature 469, 428-431.

Ohashi, I., Pohoreki, R., Morita, K., Stemmer, P.M., 2004. Alcohols increase calmodulin affinity for

Ca2+ and decrease target affinity for calmodulin. Biochim. Biophys. Acta 1691, 161-167.

Organization for Economic Cooperation and Development, Screening Information Data Set, 2004.

Ethanol. SIDS Initial Assessment Report for SIAM 19. UNEP Publications, Berlin, Germany.

Osterloh, J.D., Pond, S.M., Grady, S., Becker, C.E., 1986. Serum formate concentrations in methanol

intoxication as a criterion for hemodialysis. Ann. Intern. Med. 104, 200-203.

Oukka, M., Ho, I.C., de la Brousse, F.C., Hoey, T., Grusby, M.J., Glimcher, L.H., 1998. The

transcription factor NFAT4 is involved in the generation and survival of T cells. Immunity 9, 295-

304.

Overton, E., 1901. Studien über die Narkose, Zugleich ein Beitrag zur allgemeinen Pharmakologie.

Gustav Fischer, Jena, Germany.

Palatnick, W., Redman, L.W., Sitar, D.S., Tenenbein, M., 1995. Methanol half-life during ethanol

administration: implications for management of methanol poisoning. Ann. Emerg. Med. 26, 202-

207.

Pegan, S., Arrabit, C., Slesinger, P.A., Choe, S., 2006. Andersen's syndrome mutation effects on the

structure and assembly of the cytoplasmic domains of Kir2.1. Biochemistry 45, 8599-8606.

218

Pfister, A.K., McKenzie, J.V., Dinsmore, H.P., Edman, C.D., 1966. Extracorporeal dialysis for

methanol intoxication. JAMA 197, 1041-1043.

Pflegerl, P., Vesely, P., Hantusch, B., Schlederer, M., Zenz, R., Janig, E., Steiner, G., Meixner, A.,

Petzelbauer, P., Wolf, P., Soleiman, A., Egger, G., Moriggl, R., Kishimoto, T., Wagner, E. F.,

Kenner, L., 2009. Epidermal loss of JunB leads to a SLE phenotype due to hyper IL-6 signaling.

Proc. Natl. Acad. Sci. USA 106, 20423-20428.

Porter, C.M., Clipstone, N.A., 2002. Sustained NFAT signaling promotes a Th1-like pattern of gene

expression in primary murine CD4+ T cells. J. Immunol. 168, 4936-4945.

Pringle, M.J., Brown, K.B., Miller, K.W., 1981. Can the lipid theories of anesthesia account for the

cutoff in anesthetic potency in homologous series of alcohols? Mol. Pharmacol. 19, 49-55.

Pruett, S.B., Zheng, Q., Fan, R., Matthews, K., Schwab, C., 2004. Acute exposure to ethanol affects

Toll-like receptor signaling and subsequent responses: an overview of recent studies. Alcohol 33,

235-239.

Pruett, S.B., Fan, R., Zheng, Q., Schwab, C., 2009. Patterns of immunotoxicity associated with chronic

as compared with acute exposure to chemical or physical stressors and their relevance with regard

to the role of stress and with regard to immunotoxicity testing. Toxicol Sci. 109, 265-275.

Ramaswamy, S., Eklund, H., Plapp, B.V., 1994. Structures of horse liver alcohol dehydrogenase

complexed with NAD+ and substituted benzyl alcohols. Biochemistry 33, 5230-5237.

Rich J, Scheife RT, Katz N, Caplan LR (1990) Isopropyl alcohol intoxication. Arch. Neurol. 47, 322-

324.

Rosansky, S.J., 1982. Isopropyl alcohol poisoning treated with hemodialysis: kinetics of isopropyl

alcohol and acetone removal. J. Toxicol. Clin. Toxicol. 19, 265-271.

Roth, S., Seeman, P., 1972. The membrane concentrations of neutral and positive anesthetics (alcohols,

219

chlorpromazine, morphine) fit the Meyer-Overton rule of anesthesia; negative narcotics do not.

Biochim. Biophys. Acta 255, 207-219.

Rozenfeld, R.A., Leikin, J.B., 2007. Severe methanol ingestion treated successfully without

hemodialysis. Am. J. Ther. 14, 502-503.

Saeed, R.W., Varma, S., Peng, T., Tracey, K.J., Sherry, B., Metz, C.N., 2004. Ethanol blocks leukocyte

recruitment and endothelial cell activation in vivo and in vitro. J. Immunol. 173, 6376-6383.

Saxena, K., Lerud, K., Cicero, J.J., 1987. Methanol intoxication in multiple trauma. Am. J. Emerg.

Med. 5, 60-63.

Seeman, P., Roth, S., Schneider, H., 1971. The membrane concentrations of alcohol anesthetics.

Biochim. Biophys. Acta 225, 171-184.

Seeman, P., 1972. The membrane actions of anesthetics and tranquilizers. Pharmacol. Rev. 24, 583-

655.

Shahidullah, M., Harris, T., Germann, M.W., Covarrubias, M., 2003. Molecular features of an alcohol

binding site in a neuronal potassium channel. Biochemistry 42, 11243-11252.

Shelby, M., Portier, C., Goldman, L., Moore, J., Iannucci, A., Jahnke, G., Donkin, S., 2004. NTP-

CERHR expert panel report on the reproductive and developmental toxicity of methanol. Reprod.

Toxicol. 18, 303-390.

Smith-Garvin, J.E., Koretzky, G.A., Jordan, M.S., 2009. T cell activation. Annu. Rev. Immunol. 27,

591-619.

Smothers, C.T., Woodward, J.J., 2006. Effects of amino acid substitutions in transmembrane domains

of the NR1 subunit on the ethanol inhibition of recombinant N-methyl-D-aspartate receptors.

Alcohol. Clin. Exp. Res. 30, 523-530.

Streiff, J.H., Allen, T.W., Atanasova, E., Juranic, N., Macura, S., Penheiter, A.R., Jones, K.A., 2006.

220

Prediction of volatile anesthetic binding sites in proteins. Biophys. J. 91, 3405-3414.

Sutton, T.L., Foster, R.L., Liner, S.R., 2002. Acute methanol ingestion. Pediatr. Emerg. Care 18, 360-

363.

Szabo, G., Dolganiuc, A., Dai, Q., Pruett, S.B., 2007. TLR4, ethanol, and lipid rafts: a new mechanism

of ethanol action with implications for other receptor-mediated effects. J. Immunol. 178, 1243-

1249.

Szabo, G., Mandrekar, P., 2009. A recent perspective on alcohol, immunity, and host defense. Alcohol.

Clin. Exp. Res. 33, 220-232.

Szabo, G., Mandrekar, P., Petrasek, J., Catalano, D., 2011. The unfolding web of innate immune

dysregulation in alcoholic liver injury. Alcohol. Clin. Exp. Res. 35, 782-786.

Taieb, J., Delarche, C., Ethuin, F., Selloum, S., Poynard, T., Gougerot-Pocidalo, M.A., Chollet-Martin,

S., 2002. Ethanol-induced inhibition of cytokine release and protein degranulation in human

neutrophils. J. Leukoc. Biol. 72, 1142-1147.

Teixeira, L.K., Fonseca, B.P., Vieira-de-Abreu, A., Barboza, B.A., Robbs, B.K., Bozza, P.T., Viola,

J.P., 2005. IFN-gamma production by CD8+ T cells depends on NFAT1 transcription factor and

regulates Th differentiation. J. Immunol. 175, 5931-5939.

Thode, A.B., Kruse, S.W., Nix, J.C., Jones, D.N., 2008. The role of multiple hydrogen-bonding groups

in specific alcohol binding sites in proteins: insights from structural studies of LUSH. J. Mol. Biol.

376, 1360-1376.

Tsytsykova, A.V., Goldfeld, A.E., 2000. Nuclear factor of activated T cells transcription factor NFATp

controls superantigen-induced lethal shock. J. Exp. Med. 192, 581-586.

Vedula, L.S., Brannigan, G., Economou, N.J., Xi, J., Hall, M.A., Liu, R., Rossi, M.J., Dailey, W.P.,

Grasty, K.C., Klein, M.L., Eckenhoff, R.G., Loll, P.J., 2009. A unitary anesthetic binding site at

221

high resolution. J. Biol. Chem. 284, 24176-24184.

Velez, L.I., Kulstad, E., Shepherd, G., Roth, B., 2003. Inhalational methanol toxicity in pregnancy

treated twice with fomepizole. Vet. Hum. Toxicol. 45, 28-30.

Wenzl, J.E., Mills, S.D., McCall, J.T., 1968. Methanol poisoning in an infant. Successful treatment

with peritoneal dialysis. Am. J. Dis. Child. 116, 445-447.

Westerman, P.W., Pope, J.M., Phonphok, N., Doane, J.W., Dubro, D.W., 1988. The interaction of n-

alkanols with lipid bilayer membranes: a 2H NMR study. Biochim. Biophys. Acta 939, 64-78.

Wucherpfennig, K.W., Gagnon, E., Call, M.J., Huseby, E.S., Call, M.E., 2010. Structural biology of the

T-cell receptor: insights into receptor assembly, ligand recognition, and initiation of signaling.

Cold Spring Harb. Perspect. Biol. 2, a005140.

Yang, D.D., Conze, D., Whitmarsh, A.J., Barrett, T., Davis, R.J., Rincón, M., Flavell, R.A., 1998.

Differentiation of CD4+ T cells to Th1 cells requires MAP kinase JNK2. Immunity 9, 575-585.

Yoshimura, M., Pearson, S., Kadota, Y., Gonzalez, C.E., 2006. Identification of ethanol responsive

domains of adenylyl cyclase. Alcohol. Clin. Exp. Res. 30, 1824-1832.

Zaman, F., Pervez, A., Abreo, K., 2002. Isopropyl alcohol intoxication: a diagnostic challenge. Am. J.

Kidney Dis. 40, E12.

222

Tables

Table 1 Proteins whose functions are affected by alcohols. Inferred binding and interactions confirmed by X-ray structural studies.

TARGETS AS DEFINED BY BIOCHEMICAL CO-CRYSTALLIZED a REFERENCE AND MUTATION ANALYSES MOLECULE

PENTAMERIC LIGAND-GATED ION CHANNELS (pLGIC)

GABAA receptor Jung et al. (2005) Glycine receptor Lobo et al. (2008) Nicotinic acetylcholine receptor Forman et al. (1995) N-methyl-D-aspartate (NMDA)–type glutamate receptors Smothers & Woodward (2006) Bacterial homologue of pLGIC (GLIC) Propofol Nury et al. (2011)

POTASSIUM CHANNELS Shaw2 Shahidullah et al. (2003) G protein-gated inwardly rectifying potassium channel 2 (GIRK2) Aryal et al. (2009) IRK1 2-methyl-2,4-pentanediol, Pegan et al. (2006)

ADHESION MOLECULES L1 Dou et al. (2011)

ODORANT-BINDING PROTEINS LUSH Ethanol, 1-propanol, 1-butanol Kruse et al. (2003)

ENZYMES

b Luciferase (Bromoform) Franks et al. (1998) Alcohol dehydrogenase Pentafluorobenzylalcohol, trifluoroethanol Ramaswamy et al. (1994) Adenylyl cyclase Yoshimura et al. (2006)

a Targets for which there are crystal structure analyses are boxed in gray. b Bromoform is not an alcohol but has general anesthetic properties and binds to the alcohol-binding site in luciferase.

223

Table 2 - Functional impact of acute exposure to methanol and isopropanol on immune cells.

ALTERED FUNCTION MeOH IPA IN VITRO T LYMPHOCYTE a CYTOKINE RELEASE IL-2 ↑ ↓ IFN- ↑ ↓ TNF- ↑ ↓ CYTOTOXIC ACTIVITY ↔ ↓ PROLIFERATIVE CAPACITY ↑ ↓ NK CELL CYTOKINE RELEASE IFN- ND ↓ CYTOTOXIC ACTIVITY ↔ ↓ MONOCYTE CYTOKINE RELEASE TNF- ND ↓ IL-1 ND ↔ IL-6 ND ↑ IL-8 ND ↔ CCL-2 ND ↓ MACROPHAGE CYTOKINE RELEASE TNF- ND ↓ IL-6 ND ↑ CCL-2 ND ↓ PHAGOCYTOSIS ND ↓ IN VIVO T LYMPHOCYTE CYTOKINE RELEASE b IL-2 ↑ ↓ IFN- ↑ ↓ TNF- ↑ ↓

MONOCYTE/MACROPHAGE c CYTOKINE RELEASE TNF- ND ↓ IL-6 ND ↔ CCL-2 ND ↓

224 a Primary human immune cells are responsive to a range of alcohol concentrations in a dose-dependent manner. The lowest effective isopropanol concentration to alter cytokine release in vitro is 13 mM in NK cells and 26 mM in T cells and monocytes/macrophages. The lowest effective methanol concentration to alter cytokine release from T cells in vitro is 25 mM. b Cytokine release in response to staphylococcal enterotoxin B challenge in vivo. c Cytokine release in response to lipopolysaccharide in vivo. Detailed information on cell activation conditions and alcohol exposure in vitro and in vivo can be found in Carignan et al., 2012; Désy et al. 2008, 2010.

Figure symbols: ↔ no effect; ↑ increase; ↓ decrease; ND, not done; MeOH, methanol; IPA, isopropanol.

Figures and Legends

Fig. 1 - Acute alcohol exposure in North America.

The numbers in parenthesis correspond to the exposure cases reported per year to the American Association of Poison Control Centers. Glycol and methanol mixtures (188 cases) were included in the methanol slice (Based on Bronstein et al., 2010).

225

Fig. 2 Basic features of the alcohol-binding pocket in target proteins.

The scheme depicts the interaction of 1-propanol with a hypothetical protein site through hydrogen bonding and van der Waals forces. The hydrophobic residues (black circles) within the boundaries of the cavity are capable of stabilizing the alcohol molecule through van der Waals forces (shading). The Hydrogen bond acceptor site is represented by a crescent and the bond itself by a dashed line. Water displaced by the lodging of the alcohol molecule within the pocket is indicated by the dashed arrows. The alcohol model was generated with the NCBI PubChem 3D Viewer v2.0 (http://pubchem.ncbi.nlm.nih.gov/pc3d/).

226

Fig. 3 Hypothetical models:

The diagram illustrates a simplified model of NFAT activation in lymphocytes in presence of methanol (MeOH) (A) or isopropanol (IPA) (B). T cell receptor triggering increases cytoplasmic Ca++ thereby activating calmodulin and unleashing the phosphatase activity of calcineurin. Dephosphorylation exposes the nuclear-localization signal (nls) and allows the translocation of NFAT into the nucleus via importins. Multiple kinases (GSK3, CK1, DYRK1A, JNK and p38) rephosphorylate the NFAT proteins in the nucleus leading to their export through the nuclear pore in a complex with CRM1/RanGTP; only JNK and p38 are indicated. Methanol may interact directly with NFATc2 producing a conformational change that masks the nuclear-export signal (nes) or reduces the accessibility of phosphorylation sites. Isopropanol may interact with NFATc1 causing a conformational change that limits its nuclear localization. Quantitative differences are indicated by arrow thickness.(C) The diagram depicts a simplified model of activation of the ERK module of the MAPK pathway in monocytes in absence or presence of isopropanol. The engagement of the MD2/TLR4 complex by lipopolysaccharides triggers early signaling events that activate several pathways including the one mediated by MAPKs. The signal gets amplified by the sequential activation of MAP3Ks (not shown), MAPKKs (represented by MEK1/2), MAPKs (represented by ERK1/2), and MAPK-activated protein kinases (represented by MSK1/2, RSK1-4, MNK2). ERK1/2 determine the nuclear content of c-Fos through its direct phosphorylation and by the transcriptional control of the c-fos gene via Elk-1. ERK1/2 also activate the junB gene indirectly via RSK2/MSK1/2/CREB.

227

Fig. 4 Predicted serum methanol peak concentration in sublethal acute intoxication.

Predicted individual values are shown as filled triangles and the highest measured concentration for each surviving patient is represented by an empty triangle. Both predicted and measured data points are depicted in ascending order, not paired. We have assumed zero-order kinetics and an elimination rate of 8.5 mg/dl/h (Jacobsen et al., 1988). The means are indicated by crossing lines and were compared by the Mann-Whiney U test (*** p < 0.0001, n: 86). The information on serum methanol concentration and time of ingestion was obtained from the following sources: Bennett et al. (1953); Brahmi et al. (2007); Brent et al. (1991), (2001); Brown et al. (2001); Bucaretchi et al. (2009); Burns et al. (1997); Caravati and Anderson (2010); Coulter et al. (2011); Ekins et al. (1985); Fontenot and Pelak (2002); Ghannoum et al. (2010); Grufferman et al. (1985); Hantson et al. (1997), (1999), (2000); Ingemansson (1984); Jacobsen et al. (1983); Keyvan-Larijarni and Tannenberg (1974); Mahieu et al. (1989); Martens et al. (1982); Mowry et al. (2002); Osterloh et al. (1986); Palatnick et al. (1995); Pfister et al. (1966); Rozenfeld and Leikin (2007); Sutton et al. (2002); Wenzl et al. (1968). MeOH, methanol.