No / Evaluation of drug candidates in in-vitro assays of action potential duration (APD) is one component of preclinical safety assessment. Current assays are limited by technically-demanding, time-consuming electrophysiological methods. This study aimed to assess whether a voltage-sensitive dye-based assay could be used instead. Methods Optical APs were recorded using di-4-ANEPPS in electrically field stimulated Beagle left ventricular midmyocardial myocytes (LVMMs). Pharmacological properties of di-4-ANEPPS on the main cardiac ion channels that shape the ventricular AP were investigated using IonWorks™ and conventional electrophysiology. Effects of 9 reference drugs (dofetilide, E4031, d-sotalol, ATXII, cisapride, terfenadine, alfuzosin, diltiazem and pinacidil) with known APD-modulating effects were assessed on optically measured APD at 1 Hz. Results Under optimum conditions, 0.1 μM di-4-ANEPPS could be used to monitor APs paced at 1 Hz during nine, 5 s exposures without altering APD. di-4-ANEPPS had no effect on either hIERG, hINa, hIKs and hIto currents in transfected CHO cells (up to 10 µM) or ICa,L current in LVMMs (at 16 µM). di-4-ANEPPS had no effect on APs recorded with microelectrodes at 1 or 0.5 Hz over a period of 30 min di-4-ANEPPS displayed the sensitivity to record changes in optically measured APD in response to altered pacing frequencies and sequential vehicle additions did not affect the optically measured APD. APD data obtained with 9 reference drugs were as expected except (i) d-sotalol-induced increases in duration were smaller than those caused by other IKr blockers and (ii) increases in APD were not detected using low concentrations of terfenadine. Discussion Early in drug discovery, the di-4-ANEPPS-based method can reliably be used to assess drug effects on APD as part of a cardiac risk assessment strategy.

Cardiac arrhythmias are caused by a disruption of the normal initiation or propagation of electrical impulses in the heart. Hundreds of mutations in genes encoding ion channels or ion channel regulatory proteins are linked to congenital arrhythmia syndromes that increase the risk for sudden cardiac death. This dissertation focuses on how mutations in a gene (KCNQ1) that encodes a voltage-gated K+ ion channel (Kv7.1) can disrupt proper channel function and lead to abnormal repolarization of atrial and ventricular cardiomyocytes. In the heart, Kv7.1 coassembles with a regulatory protein to conduct the slowly activating delayed rectifier K+ current (IKs). Loss-of-function KCNQ1 mutations are linked to type 1 long QT syndrome (LQT1), and typically decrease IKs, which can lead to ventricular action potential (AP) prolongation. In patients, LQT1 is often characterized by an abnormally long corrected QT (QTc) interval on an electrocardiogram (ECG), and increases the risk for polymorphic ventricular tachycardias. KCNQ1 mutations are also linked to atrial fibrillation (AF), but cause a gain-of-function phenotype that increases IKs. Surprisingly, patients diagnosed with both LQT1 and AF are increasingly identified as genotype positive for a KCNQ1 mutation. The first aim of this dissertation was to determine a unique functional phenotype of KCNQ1 mutations linked to both arrhythmia syndromes by functional analyses via the whole-cell patch clamp technique in HEK293 cells. A proportion of patients with LQT1-linked KCNQ1 mutations do not have abnormal QTc prolongation known as latent LQT1. Interestingly, exercise can reveal abnormal QTc prolongation in these patients. During exercise, beta-adrenergic activation stimulates PKA to phosphorylate Kv7.1, causing an increase in IKs to prevent ventricular AP prolongation. Therefore, the second aim of this dissertation was to determine a molecular mechanism of latent LQT1 through functional analyses in HEK293 cells while incorporating pharmacological and phosphomimetic approaches to study PKA regulation of mutant Kv7.1 channels. The findings in this dissertation provide new insight into how KCNQ1 mutations disrupt the function of Kv7.1 in a basal condition or during beta-adrenergic activation. Also, this dissertation suggests these approaches will improve patient management by identifying mutation specific risk factors for patients with KCNQ1 mutations.

Clustered regularly interspaced short palindromic repeats (CRISPRs) act to prevent viral infection and horizontal gene transfer in prokaryotes. The genomic CRISPR array contains short sequences (“spacers”) that are derived from foreign genetic elements. The CRISPR array is transcribed and processed into CRISPR RNAs (crRNAs) used in the sequence-specific degradation of foreign nucleic acids. This process is called interference and is mediated by CRISPR-associated (Cas) proteins. This thesis has focused on the structural and functional characterisation of four Cas proteins from the CRISPR/Cas system of Sulfolobus solfataricus. The crystal structure of Cmr7 (Sso1725), a Sulfolobales-specific subunit of the ssRNA-degrading CMR complex, allowed for the identification of a putative protein-binding site, though no specific function could be ascribed to the protein. Cas6 (Sso1437) is the enzyme responsible for crRNA maturation and the characterisation of this protein allowed for the molecular rationalisation of its atypical RNA cleavage mechanism. Csa5 and Cas8a2 are subunits of the aCascade complex that targets dsDNA. Csa5 (Sso1398) was shown to have a putative role in R-loop stabilisation during interference while the role of Cas8a2 (Sso1401) was not determined. The structures of these two proteins were used to define relationships between the subunits of interference complexes from various CRISPR/Cas systems. A second aspect of this work has been the expression and purification of eukaryotic ion channels for structural studies. The acid sensing ion channel (ASIC) and FMRFamide-gated sodium channel (FaNaC) are gated ion channels with unknown mechanisms of channel activation. These ion channels must be expressed in eukaryotic systems and so human embryonic kidney (HEK) cells and baculovirus-insect cell expression systems were developed to express ASIC and FaNaC constructs. The expression and purification protocols have been optimised to allow for the preparation of soluble protein that will in future be used for crystallography and electron paramagnetic resonance (EPR) studies.

There are long-standing hypotheses that endogenous ion currents act to control cell dynamics in development, wound healing and regeneration. However, the mechanisms employed by cells to detect the electric field (EF) and translate it into a discernable message to drive specific cell behaviors, such as migration, proliferation and differentiation, are not well understood. A better understanding of how cells are able to sense EFs and react to them is vital to understanding physiological mechanisms are involved in regeneration. Ion channel signaling provides a reasonable suspect for mediating these effects based on their documented involvement in proliferation, migration and differentiation. To investigate mechanisms underlying ionic regulation of critical cellular processes in non-excitable cells, a novel, in vivo assay was developed to screen multiple pharmacological inhibitors of ion channels during larval A. mexicanum tail regeneration. This assay was used to identify individual channels that were then targeted for further analysis regarding their involvement in the regenerative process. Chapter 2 presents data from a study that indicates that a wound-like response can be generated in an invertebrate model by application of exogenous, low-amplitude sine-wave electrical stimulation. This was characterized by recruitment of hemocytes at the stimulation site which was dependent on voltage-gated potassium channels. Chapter 3 presents data from a comprehensive and systematic screen of pharmacological compounds against larval salamander tail regeneration that indicates 8 specific target ion channels. This chapter also describes results indicating specific mechanisms by which these channels may be perturbing regeneration. Chapter 4 presents data that indicate that the Anoctamin 1 channel identified in the aforementioned screen is a regulator of cellular proliferation. This is shown to be accomplished via amplification of intracellular calcium surges and a subsequent increase in the activity of the p44/42 MAPK signaling cascade.

Swelling-activated Cl- current (ICl,swell) is an outwardly rectifying Cl- current that influences cardiac electric activities and acts as a potential effector of mechanoelectrical feedback that antagonizes the effects of stretch-activated cation channels. Persistent activation of ICl,swell has been observed in multiple models of cardiovascular diseases. Previously we showed that angiotensin II (AngII) signaling and reactive oxygen species (ROS) produced by NADPH oxidase (NOX) are involved in the activation of ICl, swell by both beta1-integrin stretch and osmotic swelling. Because endothelin-1 (ET-1) is a potential downstream mediator of AngII and ETA receptor blockade abrogates AngII-induced ROS generation, we studied how ET-1 signaling regulates ICl,swell and the relationship between AngII and ET-1 signaling. Under isosmotic conditions, ET-1 elicited an outwardly rectifying Cl- current that was fully blocked by the highly selective ICl,swell inhibitor DCPIB and by osmotic shrinkage. Selective ETA blockade (BQ123), but not ETB blockade (BQ788), fully suppressed the ET-1-induced current. ET-1-induced ICl,swell was abolished by blockade of EGFR kinase (AG1478) and PI-3K inhibitors (LY294002 and wortmannin), which also suppress beta1-integrin stretch- and swelling-induced ICl,swell. ET-1-induced ICl,swell was abrogated by ebselen, a membrane-permeant glutathione peroxidase mimetic that dismutates H2O2 to H2O, suggesting that ROS were required intermediates in ET-1-induced activation of ICl,swell. Both NOX and mitochondria are important sources of ROS in cardiomyocytes. Blocking NOX with apocynin or mitochondrial complex I with rotenone both completely suppressed ET-1-induced ROS generation and activation of ICl,swell, indicating that ROS from both NOX and mitochondria were required to activate ICl,swell, and complete block by inhibitors of either ROS source suggests mitochondrial and NOX must act in series rather than in parallel. ICl,swell elicited by antimycin A, which stimulates superoxide production by mitochondrial complex III, was insensitive to NOX inhibitor apocynin and the NOX fusion peptide inhibitor gp91ds-tat. Activation of ICl,swell induced by diazoxide, which stimulates mitochondrial ROS production by opening mitochondrial KATP channels, was not affected by gp91ds-tat. These data suggests that mitochondrial ROS is downstream from NOX in the regulation of ICl,swell. Mitochondrial ROS production that is enhanced by NOX ROS is likely to be responsible for the activation of ICl,swell by ET-1. In order to determine the role of ERK in the proposed signaling pathway that regulates ICl,swell, we examined the effect of ERK inhibitors (PD 98059 and U0216) on the activation of ICl,swell elicited by ET-1, EGF, and H2O2. ERK inhibitors partially blocked ET-1-induced ICl,swell but fully inhibited activation of ICl,swell in response to EGF. However, ERK inhibitors did not affect ICl,swell elicited by exogenous H2O2. We also established the the relationship of ET-1 to AngII and osmotic swelling in the regulation of ET-1 ICl,swell. ETA blockade abolished ICl,swell elicited by both AngII and osmotic swelling, whereas AT1 blockade did not effect ET-1-induced ICl,swell, suggesting that ET-1 signaling is downstream from AngII and osmotic swelling. HL-1 cell is a murine atrial cell line that retain phenotypic characteristics of adult cardiomyocytes. We showed that osmotic swelling and ET-1 turned on DCPIB-sensitive outwardly rectifying Cl- current in HL-1 cells with both physiological and symmetrical Cl- gradients. The swelling-induced current was suppressed by gp91ds-tat and rotenone but insensitive to apocynin. Blockade of ETA receptor (BQ123) and NOX (gp91ds-tat) completely inhibited ET-1-induced ICl,swell in HL-1 cells. These data indicate that ICl,swell is present in HL-1 cell and regulated by similar mechanisms as in native cells. Finally, we confirmed the production of ROS by ET-1 signaling by flow cytometry of HL-1 cells using the nominally H2O2-selective fluorescent probe C H2DCFDA-AM. Exposure to ET-1 increased ROS production, as did H2O2, a positive control. ET-1-induced ROS production was fully suppressed by both gp91ds-tat and rotenone. HL-1 cell ROS production also was stimulated by the mitochondrial complex III inhibitor antimycin A, and antimycin A-induced ROS production was blocked by rotenone but not by gp91ds-tat. These data suggest that ET-1 ETA receptor signaling elicits ICl,swell by sequentially stimulating ROS production by NOX and mitochondria. ETA receptor signaling is down stream from AngII in the osmotic swelling-induced activation of ICl,swell and is upstream from EGFR kinase and PI-3K. Endothelin signaling is likely to be an important means of activating ROS production and ICl,swell in a variety of cardiovascular diseases.

Heterotrimeric guanine nucleotide-binding proteins (G-proteins) consist of an alpha subunit (Gα) and the dimeric beta-gamma subunit (Gβγ). The first example of direct cell signaling by Gβγ was the discovery of its role in activating G-protein regulated inwardly rectifying K+ (GIRK) channels which underlie the acetylcholine-induced K+ current responsible for vagal inhibition of heart rate. Published crystal structures have provided important insights into the structures of the G-protein subunits and GIRK channels separately, but co-crystals of the channel and Gβγ together remain elusive and no specific reciprocal residue interactions between the two proteins are currently known. Given the absence of direct structural evidence, we attempted to identify these functionally important channel-Gβγ interactions using a computational approach. We developed a multistage computational docking algorithm that combines several known methods in protein-protein docking. Application of the docking protocol to previously published structures of Gβγ and GIRK1 homomeric channels produced a clear signal of a favored binding mode. Analysis of this binding mode suggested a mechanism by which Gβγ promotes the open state of the channel. The channel-Gβγ interactions predicted by the model in silico could be disrupted in vitro by mutation of one protein and rescued by additional mutation of reciprocal residues in the other protein. These interactions were found to extend to agonist induced activation of the channels as well as to activation of the native heteromeric channels. Currently, the structural mechanism by which Gβγ regulates the functional conformations of GIRK channels or of any of its membrane-associated effector proteins is not known. This work shows the first evidence for specific reciprocal interactions between Gβγ and a GIRK channel and places these interactions in the context of a general model of intracellular regulation of GIRK gating.

La fonction des canaux ioniques est finement régulée par des changements structuraux de sites clés contrôlant l’ouverture du pore. Ces modulations structurales découlent de l’interaction du canal avec l’environnement local, puisque certains domaines peuvent être suffisamment sensibles à des propriétés physico-chimiques spécifiques. Les mouvements engendrés dans la structure sont notamment perceptibles fonctionnellement lorsque le canal ouvre un passage à certains ions, générant ainsi un courant ionique mesurable selon le potentiel électrochimique. Une description détaillée de ces relations structure-fonction est cependant difficile à obtenir à partir de mesures sur des ensembles de canaux identiques, puisque les fluctuations et les distributions de différentes propriétés individuelles demeurent cachées dans une moyenne. Pour distinguer ces propriétés, des mesures à l’échelle de la molécule unique sont nécessaires. Le but principal de la présente thèse est d’étudier la structure et les mécanismes moléculaires de canaux ioniques par mesures de spectroscopie de fluorescence à l’échelle de la molécule unique. Les études sont particulièrement dirigées vers le développement de nouvelles méthodes ou leur amélioration. Une classe de toxine formeuse de pores a servi de premier modèle d’étude. La fluorescence à l’échelle de la molécule unique a aussi été utilisée pour l’étude d’un récepteur glutamate, d’un récepteur à la glycine et d’un canal potassique procaryote. Le premier volet porte sur l’étude de la stœchiométrie par mesures de photoblanchiment en temps résolu. Cette méthode permet de déterminer directement le nombre de monomères fluorescents dans un complexe isolé par le décompte des sauts discrets de fluorescence suivant les événements de photoblanchiment. Nous présentons ici la première description, à notre connaissance, de l’assemblage dynamique d’une protéine membranaire dans un environnement lipidique. La toxine monomérique purifiée Cry1Aa s’assemble à d’autres monomères selon la concentration et sature en conformation tétramérique. Un programme automatique est ensuite développé pour déterminer la stœchiométrie de protéines membranaires fusionnées à GFP et exprimées à la surface de cellules mammifères. Bien que ce système d’expression soit approprié pour l’étude de protéines d’origine mammifère, le bruit de fluorescence y est particulièrement important et augmente significativement le risque d’erreur dans le décompte manuel des monomères fluorescents. La méthode présentée permet une analyse rapide et automatique basée sur des critères fixes. L’algorithme chargé d’effectuer le décompte des monomères fluorescents a été optimisé à partir de simulations et ajuste ses paramètres de détection automatiquement selon la trace de fluorescence. La composition de deux canaux ioniques a été vérifiée avec succès par ce programme. Finalement, la fluorescence à l’échelle de la molécule unique est mesurée conjointement au courant ionique de canaux potassiques KcsA avec un système de fluorométrie en voltage imposé. Ces enregistrements combinés permettent de décrire la fonction de canaux ioniques simultanément à leur position et densité alors qu’ils diffusent dans une membrane lipidique dont la composition est choisie. Nous avons observé le regroupement de canaux KcsA pour différentes compositions lipidiques. Ce regroupement ne paraît pas être causé par des interactions protéine-protéine, mais plutôt par des microdomaines induits par la forme des canaux reconstitués dans la membrane. Il semble que des canaux regroupés puissent ensuite devenir couplés, se traduisant en ouvertures et fermetures simultanées où les niveaux de conductance sont un multiple de la conductance « normale » d’un canal isolé. De plus, contrairement à ce qui est actuellement suggéré, KcsA ne requiert pas de phospholipide chargé négativement pour sa fonction. Plusieurs mesures indiquent plutôt que des lipides de forme conique dans la phase cristalline liquide sont suffisants pour permettre l’ouverture de canaux KcsA isolés. Des canaux regroupés peuvent quant à eux surmonter la barrière d’énergie pour s’ouvrir de manière coopérative dans des lipides non chargés de forme cylindrique. / The function of ion channels is finely regulated by structural changes of key domains controlling the pore opening. These structural modulations arise from interactions with the local environment, since several domains can be sensitive to specific physico-chemical properties. Movements generated in the structure become notably perceptible when channels open a passage for some ions, thus generating a measurable ionic current according to the electrochemical potential. A detailed description of these structure-function relationships is however difficult to obtain from measurements involving a set of identical channels, since the fluctuations and distributions of different individual properties remain hidden in an average. To differentiate these properties, single-molecule recordings are required. The main purpose of this thesis is to study the structural aspects and molecular mechanisms of ion channels using fluorescence spectroscopy at the single-molecule level. Studies are oriented towards the development or improvement of new methods. A class of pore-forming toxin served as a first study model. Single-molecule fluorescence was also used to study an ionotropic glutamate receptor, a glycine receptor and a prokaryotic potassium channel. The first part focuses on the study of stoichiometry using fluorescent subunit counting. This method allows a direct measure of the number of fluorescent monomers within a single complex by counting the number of step-wise fluorescence intensity decrease following photobleaching events. Here we present the first description, to our knowledge, of the dynamic assembly of a membrane protein in a lipid environment. The purified monomeric Cry1Aa toxin clusters with other monomers depending on the concentration and saturates in a tetrameric conformation. An automated method has been developed to determine the stoichiometry of GFP-tagged membrane proteins expressed on mammalian cell surface. Although this expression system is suitable for the study of proteins of mammalian origin, background fluorescence is particularly important and significantly increases the risk of error in the manual counting process. The presented method allows a fast and automated analysis based on fixed criteria. The algorithm responsible for counting fluorescent monomers was optimized from simulations and adjusts its detection parameters automatically according to the fluorescence trace recording. The composition of two ion channels was successfully verified using this program. Finally, single-molecule fluorescence is measured together with ionic current of KcsA channels using a voltage-clamp fluorometry setup. These combined recordings allowed us to describe the function of ion channels simultaneously to their position and density as they diffuse in a lipid membrane of defined composition. We observed clustering of KcsA channels for various lipid compositions. Clustering does not appear to be caused by protein-protein interaction, but rather by microdomains induced by the shape of reconstructed channels in the lipid bilayer. It seems that clustered KcsA channels could then become coupled, resulting in cooperative gating events with conductance levels multiple to the “normal” unitary channel conductance. Moreover, as opposed to what is currently suggested, KcsA does not require a negatively charged phospholipid for its function. Several of our recordings rather suggest that conically shaped lipids in the lamellar liquid crystalline phase are sufficient to allow single channel opening. Clustered channels can on the other hand overcome the energy barrier to open cooperatively in uncharged cylindrical lipids.

Purinergic receptors are membrane ion channels that are activated by extracellular ATP. In vertebrates, seven genes encode subunits of P2X receptors. The subunits, designated P2X1-7, are 40 - 50% identical in amino acid sequences. P2X receptors are composed of three subunits and are found as homo- and heterotrimers in tissues of vertebrates. P2X receptors have a wide distribution in the organism, functional receptors are found in neurons, glial cells, muscle cells and also in nonexcitable tissues as epithelial, endothelial, and in hemopoietic tissue. Purinergic signalling plays an important role in pain transmission, at CNS injury and immune processes. P2X receptor subunit consists of two transmembrane domains, extracellular domain and intracellular N-and C-termini. Each transmembrane domain contains two amino acids conserved across all P2X subunits. In the first transmembrane domain receptor P2X2 are that Gly30 and Tyr43. In previous experiments performed on P2X2 receptor, electrophysiological measurements demonstrated that substitution of conserved Tyr43 in the first transmembrane domain with alanine prolongs the deactivation time of ion channel after agonist wash out. This work is focused on clarifying the role of conserved tyrosine in the process of opening and closing of ion channel of P2X...

Štěpána Boukalová Mechanisms of activation and modulation of vanilloid TRP channels TRPV1 and TRPV3 are thermosensitive ion channels from the vanilloid subfamily of TRP receptors. TRPV1, which is primarily expressed in nociceptive sensory neurons, is an important transducer of painful stimuli and is also involved in the detection of noxious heat. TRPV3 is expressed mainly in the skin where it regulates proliferation and differentiation of keratinocytes. Similarly to voltage-dependent potassium (Kv) channels, TRP receptors are comprised of four subunits, each with six transmembrane segments (S1-S6). Using mutational approach, we tried to elucidate the role of S1 in TRPV1 functioning. Our results indicate that the extracellular portion of S1 plays a crucial role in TRPV1 gating. TRPV1 channels with a conservative mutation of positively charged residue in this region (R455K substitution) were overactive. However, they were neither activated nor potentiated by low pH; on the contrary, protons stabilized the closed conformation of this mutant channel. Very similar phenotypic properties were found in other TRPV1 mutants with substitution in S4/S5-S5 region and in the pore helix. In Kv channels, extracelular portion of S1 forms a small contact surface with the pore helix, which allows efficient transmission of...

Le segment initial de l'axone (SIA) joue un rôle dans la maintenance de la polarité neuronale et dans l'initiation du potentiel d'action. Il se construit autour de l'ankyrine G (ankG) qui relie les protéines membranaires au cytosquelette d'actine et de microtubules. Cette structure est dynamiquement régulée par des kinases. S'il a été clairement établi que l'ankG est cruciale à la formation et à la maintenance du SIA, les mécanismes responsables de sa concentration dans la partie proximale de l'axone restent encore inconnus. Il en est de même pour la protéine kinase CK2 qui régule l'interaction entre l'ankG et les canaux sodiques (Nav1). Dans un premier temps, nous avons montré, par des approches de mutagénèse, que le domaine serine-rich (SR), notamment ses 73 premiers acides aminés, porte l'information nécessaire à l'entrée de l'ankG dans l'axone. Mais ce domaine n'est pas suffisant pour le confiner dans la partie proximale de l'axone, cette propriété est portée par le domaine Tail. En plus de la coopération de ces domaines, nous avons aussi observé que l'adressage de l'ankG est régulé par les kinases Cdk5 et PKC. Dans un second temps, nous avons montré que l'accumulation de la CK2 au SIA dépend de l'expression des Nav1. L'existence d'un complexe formé par les Nav1 et la CK2 serait donc importante au recrutement de la protéine kinase CK2 au SIA. En outre, le développement des anticorps phosphospécifiques nous a permis de monter que les Nav1 sont phosphorylés in vivo au niveau de leur motif de liaison à l'ankG. L'ensemble de nos résultats ouvre de nouvelles perspectives dans la compréhension de la formation du SIA et des mécanismes de régulation qui peuvent être associés. / The axon initial segment (AIS) is responsible for both the maintenance of neuronal polarity and the generation of action potentials. The scaffolding protein ankyrin G (ankG) is specifically expressed in the AIS where it links transmembrane proteins to the subjacent actin and microtubule cytosqueletons. Moreover, the AIS is dynamically regulated by kinases. Although, it has been clearly established that ankG directs AIS assembly and maintenance, the mechanisms regulating ankG proper transport and tethering remain unclear. Another AIS component, the protein kinase CK2 is also playing an important role via the phosphorylation of the ankG-binding motif (ABM) on sodium channels (Nav1) to strengthen their interaction with ankG. But, the mechanism regulating its targeting and anchoring to the AIS remain still unknown. Here, we report that the first 73 residues of the serine-rich domain are necessary for the targeting of ankG to the axon and the tail domain for the proper positioning along the proximal axon. We also observed that ankG axonal localization is modulated by post-translational modifications. Using phosphospecific antibodies and inhibition/depletion approaches, we also provide evidence that the ABM of Nav1 are phosphorylated in vivo and that CK2 accumulation at the AIS depends on Nav1 expression, with which they form tight complexes. This suggests that CK2-mediated phosphorylation participates in Nav1 clustering in vivo and that its specific localization at the AIS is dependent on Nav1 expression. Altogether, our results open new perspectives in understanding the formation of AIS and regulatory mechanisms that may be involved.