Toutes les molécules en développement préclinique (ICH S7B) doivent être testées pour évaluer leur potentiel torsadogène. Le but de ce travail est d’établir le profil électrophysiologique de molécules torsadogènes connues afin de mieux comprendre le mécanisme de déclenchement des Torsades de Pointes et de déterminer des points clés nous permettant de prédire les molécules à risque. Il existe une base de données, TdPScreen®, combinant données cliniques et tests réalisés sur fibres de Purkinje de chien, qui permet d’attribuer un score pro-arythmique aux molécules testées. Treize molécules connues ont été choisies dans cette base de données, et testées en patch-clamp sur des cellules HEK293 exprimant le canal hERG (IKR), le canal KvLQT1+MinK (IKS), le canal Kir2.1 (IK1), le canal NaV1.5 (INa), ou le canal CaV1.2+? (ICaL). Des investigations in vivo ont également été réalisées, afin de mettre en évidence l’impact du système nerveux autonome sur l’allongement de l’intervalle QT lors d’études de pharmacologie de sécurité. / According to the ICH S7B guidelines, the torsadogenic risk of new drug candidates must be evaluated before clinical trials. The aim of this work was to establish the electrophysiological profile of known torsadogenic drugs to better understand the mechanism triggering the Torsades de Pointe and defined key points for prediction of proarrhythmic risk. TdPScreen®, a predictive tool, based on clinical data and the model of isolated canine Purkinje fibres allows determination of a proarrhythmic score. Thirteen drugs were chosen in this data base, and tested in patch-clamp on HEK293 cells expressing different channels: hERG (IKR), KvLQT1+MinK (IKS), Kir2.1 (IK1), NaV1.5 (INa), or CaV1.2+? (ICaL). In vivo investigations were also performed, to bring to light the impact of the autonomic nervous system on QT interval prolongation in safety pharmacology.

The  rapidly  activating  delayed  rectifier  potassium   channel  (IKr)  encoded  by  the  human  ether-­a-­go-­go  related  gene  (hERG)  is  a  crucial   component  of cardiac  repolarization  of  the  action  potential.  The   gating  kinetics  of  the  channel  and  the  number  of  channels  on  the  membrane  determine  the   degree  of  current  flowing  through  hERG  channels.  Membrane  channel  density  is  directly   related  to  a  balance  between  protein  trafficking  and  degradation.  While  hERG  channel   trafficking  has  been  studied  in  the  past  decade,  recently  our  lab  and  others  have  made  new   discoveries  regarding  hERG  degradation.  These  studies  show  that  Neural  Down  Regulated   Gene  4  subtype  2  (Nedd4-­‐2)  is  the  ligase  that  covalently  attaches  ubiquitin  to  the  hERG   channel  and  facilitates  degradation  (Albesa  et  al.,  2011;Guo  et  al.,  2012).  Given  these   findings,  my  goal/objective  was  to  study  the  molecular  mechanisms  that  control  the   destructive  effects  of  Nedd4-­‐2  on  the  hERG  channel.  In  the  present  study,  I  demonstrated   that  overexpression  of  the  stress-­‐responsive  serum-­‐  and  glucocorticoid-­‐inducible  kinase   (SGK)  isoforms  SGK1  and  SGK3  increase  the  current  and  expression  level  of  the  membrane-­‐ localized  mature  proteins  of  hERG  channels  stably  expressed  in  HEK  293  (hERG-­‐HEK)  cells.   I  have  found  that  the  overexpression  of  SGK1  and  SGK3  increased  Nedd4-­‐2   phosphorylation,  which  is  known  to  inhibit  Nedd4-­‐2  activity.  Furthermore,  the  synthetic   glucocorticoid,  dexamethasone,  increased  the  current  and  abundance  of  mature  ERG   proteins  in  both  hERG-­‐HEK  cells  and  neonatal  rat  cardiac  myocytes  through  the   enhancement  of  SGK1  but  not  SGK3  expression.  Additionally,  disruption  of  Rab11  proteins led  to  a  complete  elimination  of  SGK-­‐mediated  increase  in  hERG  expression.  These  results   indicate  that  SGK  enhances  the  expression  level  of  mature  hERG  channels  by  inhibiting   Nedd4-­‐2  as  well  as  by  promoting  Rab11-­‐mediated  hERG  recycling. / Thesis (Master, Physiology) -- Queen's University, 2013-08-14 16:01:43.951

The cardiac rapidly activating delayed rectifier potassium channel (IKr) is encoded by the human ether-a-go-go related gene (hERG), which is important for repolarization of the cardiac action potential. Reduction in hERG expression levels due to genetic mutations or drugs causes Long QT Syndrome (LQTS). Recently, we demonstrated that ubiquitination of hERG channels is involved in low K+ induced hERG endocytic degradation. Since homeostatic degradation is an important pathway in maintaining hERG membrane expression levels, we investigated the molecular mechanisms for hERG degradation by focusing on the role and consequence of overexpressing the ubiquitin (Ub) ligase, Nedd4-2 (Neural Precursor Cell- Expressed Developmentally Downregulated Gene 4- 2) (Yang & Kumar, 2010). Previous work in the lab demonstrated that Ub plays a role in the internalization of cell-surface hERG channels, and I hypothesized that ubiquitination of hERG channels is facilitated through Nedd4-2. To study the effects of Nedd4-2 on hERG channels, I overexpressed Nedd4-2 in human embryonic kidney (HEK) 293 cells that stably express the hERG channels. Electrophysiological recordings, Western blot, co-immunoprecipitation analysis, and confocal microscopy were performed to identify Nedd4-2’s role in hERG expression. The data from whole-cell patch clamp recordings demonstrated that, among hEAG, Kv1.5 and hERG, Nedd4-2 specifically eliminates the hERG channel current. Western blot and confocal imaging analyses showed that Nedd4-2 overexpression led to a significant reduction in mature hERG channels in the plasma membrane. Data obtained using co-immunoprecipitation indicated that Nedd4-2 significantly increases ubiquitinated hERG channels. These data indicate that Nedd4-2 may play a role in hERG homeostatic degradation. / Thesis (Master, Physiology) -- Queen's University, 2011-08-15 18:17:00.452

The HERG gene encodes for the delayed rectifier K+ channel in human cardiac tissue and contributes to the repolarization phase of the ventricular action potential. Defects in its activity underlies a cardiac disorders linked to a prolongation in the QT interval known as acquired long QT syndrome. The channel has structural properties that lead to its unintentional inhibition by various classes of drugs and is a source of drug induced cardiac toxicity. To date, no assay has been set as a standard due to variability across laboratories and the use of animals providing variable results due to differences in the ion channels involved in repolarisation. This thesis focuses on the development of testing assays for HERG using animal-free methodology. In Chapter 2, a human embryonic kidney (HEK293) cell line was cultured and transfected with the human HERG gene using animal-free methodologies. The success of the transfection was confirmed using PCR, patch clamp electrophysiology and a non-radioactive rubidium assay. Using a non-radioactive rubidium assay, drug inhibition on the transfected cell line was measured. The IC50 values obtained for a range of drugs were compared to those obtained using electrophysiological studies in the literature and there was a high correlation (r2 = 0.76). In Chapter 3, a human neuroblastoma cell line (SH-SY5Y) was tested for its validity for testing the effect of drugs on the endogenously expressed HERG K+ channel. The drug IC50 values obtained using the Rb+ assay were well correlated (r2= 0.82) with patch clamp studies in HERG transfected HEK293 cells in the literature. Clomipramine a clinically used antidepressant causes prolongation in the QT interval, however its mechanism of action on cardiac cells leading to this cardiotoxic effect is unclear. In this study, clomipramine was tested using HERG transfected HEK293 cells and the neuroblastoma cell line (SH-SY5Y) using a rubidium assay and whole cell patch clamp. Clomipramine inhibited HERG with an IC50 value of 8.35 µM and 2.18 µM in HERG transfected HEK293 cells and the neuroblastoma cell line (SH-SY5Y) using the rubidium assay respectively. Clomipramine inhibited HERG currents with an IC50 value of 0.50 µM using the patch clamp technique in HEK293 cells. The results indicate that the prolongation in the QT interval caused by clomipramine may involve HERG inhibition. The HERG K+ channel is regulated by several protein kinases including protein kinase A and protein kinase B. In Chapter 5, the specific PKC activator and phorbol ester PDA was used to study HERG regulation by PKC in HERG transfected HEK293 cells. PDA caused a reduction in HERG currents in HEK293 cells. The PKC pseudo substrate inhibitor PKC [19-36] did not inhibit the effect of PDA on HERG currents. The results of the study suggest that (1) PDA could be acting directly on the channel and inhibiting its function or (2) PDA is activating other proteins which are affecting HERG currents in the HERG transfected HEK293 cells.

国立情報学研究所で電子化したコンテンツを使用している。

The human ether-à-go-go-related gene (hERG) encodes the pore-forming α-subunits of the Kv11.1 channel that is responsible for the cardiac rapidly activating delayed rectifier K+ current (IKr), which plays a critical role in cardiac repolarization. Dysfunction of hERG causes long QT syndrome (LQTS), a cardiac electrical disorder that can lead to severe cardiac arrhythmias and sudden death (Mitcheson et al., 2000a; Roden, 2004; Maier et al., 2006; Misner et al., 2012). The overall function of hERG channels is dependent on the channel density at the plasma membrane as well as proper channel gating. Previous work from our lab demonstrated that degradation of hERG protein in the lysosome is regulated by ubiquitin ligase Nedd4-2-mediated monoubiquitination (Sun et al., 2011; Guo et al., 2012). However, whether the internalized hERG proteins can be recycled back to the plasma membrane remains to be determined. In the present study, we investigated the regulatory effects of various Rabs on hERG channels using Western blot analysis, co-immunoprecipitation (Co-IP), whole-cell patch clamp and immunofluorescence microscopy. The data revealed that, among hERG, human Kv1.5 (cardiac ultra-rapidly activating delayed rectifier K+ channel), and human EAG (ether-à-go-go gene) potassium channels, Rab4 selectively decreased the mature hERG protein expression on the plasma membrane. Mechanistically, Rab4 did not directly target the internalized hERG protein for recycling. Instead, Rab4 increased the expression level of the E3 ubiquitin ligase Nedd4-2 (Neural Precursor Cell-expressed Developmentally Downregulated Protein 4-2), which has been shown to mediate hERG ubiquitination and degradation (Guo et al., 2012). Nedd4-2 binding site mutations ∆1073 (binding site is removed) and Y1078A (binding site is modified) in hERG completely abolished the effect of Rab4. It has been shown that Nedd4-2 undergoes self-ubiquitination after targeting substrates (Bruce et al., 2008). My data further demonstrated that Rab4 decreased the degradation rate of Nedd4-2 and increased the rate of recycling. The increased Nedd4-2 then decreases hERG expression at the plasma membrane by targeting the PY-motif in the C-terminus of hERG channels. In summary, the present study showed that Rab4 decreases the expression and function of hERG potassium channels on the plasma membrane through enhancing the recycling of the ubiquitin ligase Nedd4-2. / Thesis (Master, Physiology) -- Queen's University, 2013-08-09 12:11:27.938

<div>The human ether-a-go-go-related gene (hERG, KCNH2) potassium channel has been implicated in diverse physiological and pathological processes. The KCNH2 gene encodes a rectifier voltage-gated potassium channel (Kv 11.1) that governs the chief repolarizing current, IKr, which is essential for normal electrical activity in excitable cells such as cardiomyocytes. It is also involved in cell growth and apoptosis regulation in non-excitable cells, such as tumor cells. Dysfunction of hERG is associated with potentially lethal complications, including diseases and sudden death under certain circumstances. While the mechanisms regulating KCNH2 expression remain unclear, recent data suggested that microRNAs (miRNAs) are involved, particularly in the context of several pathologic effects. </div><div>miRNA is a class of RNA defined by its conserved, short, non-coding nature. miRNAs are important regulators of gene expression at the post-transcriptional level that bind through complimentary annealing to the 3’ untranslated regions (3’ UTRs) of target mRNAs, resulting in mRNA destabilization and translational repression. The primary objectives of this research were to 1) identify miRNAs regulating KCNH2 expression in cancer, 2) investigate the potential association between miR-362-3p expression and risk of drug-induced QT interval lengthening, and 3) identify miRNAs potentially regulating KCNH2 expression and function in cardiac cells. </div><div>Through bioinformatics approaches, five miRNAs were identified to potentially regulate KCNH2 expression and function in breast cancer cells. The five identified miRNAs were validated through a Dual-Luciferase Assay using the KCNH2 3′ UTR. Only miR-362-3p was validated to bind to the KCNH2 3’ UTR, decreasing luciferase activity by 10% ± 2.3 (P < 0.001, n = 3) when compared to cells transfected with luciferase plasmid alone. miR-362-3p was also the only miRNA that its expression positively correlated with overall survival of patients with breast cancer from The Cancer Genome Atlas-Cancer Genome (TCGA) database by log-rank test (HR: 0.39, 95% CI: 0.18 to 0.82, P = 0.012). Cell proliferation was assessed by MTS assay (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium) 48 hours following transfection in breast cancer cell lines, including SK-BR-3 and MCF-7. miR-362-3p significantly decreased proliferation of SK-BR-3 and MCF-7 cells by 23% ± 8.7 (P = 0.014, n = 3) and 11.7% ± 1.0 (P < 0.001, n = 3), respectively. Cell cycle phases in SK-BR-3 and MCF-7 cells were differentiated by flow cytometry 48 hours following transfection. miR-362-3p and hERG siRNA (positive control) significantly increased the accumulation of cells in G0/G1 phase in MCF-7 by 11.7% (from 51.1% ± 0.64 to 57.1 ± 0.96, P = 0.002, n = 3) and 10% (from 51.1% ± 0.64 to 56.8 ± 0.96, P < 0.001, n = 3), respectively. </div><div>The demonstrated ability of miR-362-3p to regulate hERG in breast cancer cells coupled with previously published data that indicated an alteration of miR-362-3p expression during HF and a potential association between its expression and QT interval prolongation suggesting an important role for this miRNA in regulation of hERG function during HF. Therefore, the contribution of miR-362-3p to hERG function was investigated in patients administered the QT prolonging drug ibutilide, known to inhibit hERG. A total of 22 patients completed a prospective, parallel-group comparative study during which they received subtherapeutic doses (0.003 mg/kg) of ibutilide. The study was originally designed to investigate the influence of heart failure with preserved ejection fraction (HFpEF) on response to drug-induced QT prolongation. Blood for determination of serum Ibutilide concentrations and miR-362-3p expression, along with electrocardiograms (ECGs) were serially collected over a span of 12 hours. ΔΔ-Fridericia-heart rate corrected QT (ΔΔ QTF) intervals were utilized for all analyses to account for baseline and diurnal variation. </div><div>To assess the ability of miR-362-3p to predict ibutilide QT-induced ΔΔQTF changes, nonlinear mixed effects pharmacokinetic/ pharmacodynamic (PKPD) modeling was performed to assess the contribution of miR-362-3p to drug-induced QT interval lengthening. The model that best fit serum ibutilide concentrations versus time was a 3-compartment model with first order elimination and proportional residual errors, while the model that best described the ibutilide concentration- ΔΔQTF relationship was an Emax model with an effect compartment. In addition to miR-362-3p expression, several demographic and clinical data were evaluated as potential covariates on PK and PD parameter estimates. Of tested covariates, heart failure (HF) status on Emax (ΔOFV = -4.1; P < 0.05), and miR-362-3p expression on EC50 (ΔOFV = -9.9; P < 0.05) were incorporated in the final PKPD model. The mean individual Emax was significantly higher in HF patients when compared to non-HF patients (P = 0.015), while EC50 was negatively correlated with miR-362-3p expression (P < 0.0001, R2 0.93). </div><div>Previous evidence indicates that miR-362-3p is altered in patients with HF. In addition, several miRNAs commonly regulate the same ion channel. Therefore, we have developed a large-scale high-throughput bioassay (HT-bioassay) to explore and identify other miRNAs potentially involved in KCNH2 expression and function in human induced pluripotent stem cell-derived cardiomyocytes (hiPS-CM) during sustained β-adrenergic receptor (βAR) stimulation or overexpression of activated calcium/calmodulin-dependent protein kinase 2 (CaMKII), which are classical consequences of HF. </div><div>Through bioinformatic approaches, putative miRNA binding sites (n=327) were identified in the KCNH2 3′ UTR. Fragments containing these putative binding sites were synthesized, cloned into linearized plasmids, and amplified. The plasmid pool was transfected into hiPS-CM cells either treated with βAR stimulation or overexpressing CaMKII. Next-generation sequencing was performed to identify: 1) expression of putative miRNA binding sites and 2) endogenous miRNAs versus control. Eight predicted binding sites were found to be significantly downregulated in the CAMKII group (P <0.05, log fold change -0.287 to -0.59), and six significantly downregulated in the sustained βAR group (P <0.05, log fold change -0.29 to -0.72). Two binding sites were significantly reduced in both treatment groups (P < 0.05, log fold change between -0.38 and -0.61).</div><div>Thirty-one miRNAs were predicted to bind to the 16 binding sites identified from the bioassay. Of these, seven were selected for further screening using dual luciferase assays. None of the putative miRNAs reduced luciferase activity. However, hERG expression was assessed by immunoblot analysis following transfection of the seven miRNAs into HEK293 cells stably expressing hERG (HEK293-hERG). Six of the seven miRNA mimics reduced hERG protein expression. An additional validation step was performed by assessing hERG-related current density by whole cell electrophysiology, in which three of the six miRNAs inhibited hERG protein transfected into HEK293-hERG cells. Those same three miRNA mimics significantly decreased Ikr current (P <0.05). </div><div>Finally, expression of the miRNAs identified by HT-bioassay was examined in the patients enrolled in the clinical trial in which genome-wide next generation sequencing was performed on miRNAs extracted from whole blood samples. Of the 31 miRNAs identified from HT-bioassay, six were found to be expressed in patients (n = 12). A correlation analysis was performed between levels of the expressed miRNAs and corresponding QTF interval lengthening with ibutilide. Of the six miRNAs, only miR-4665-5p was significantly associated with QTF interval (P = 0.0379). </div><div>In summary, miR-362-3p was identified to regulate hERG, and reduces proliferation of breast cancer cells through a mechanism that may be partially mediated by hERG inhibition. While miR-362-3p may have modest effects in cancer, in Aim 2 we demonstrated that it along with HF status accounts for a significant amount of variability in QTF prolongation following ibutilide administration. However, it is common for several miRNAs to regulate a single ion channel. Therefore, an HT-bioassay was developed to identify all miRNAs that potentially regulate KCNH2 during HF. In addition to miR-362-3p, thirty-one miRNAs were predicted to regulate KCNH2; one miRNA (miR-4665-5p) was significantly associated with QTF prolongation. The potential for miR-362-3p and HT-bioassay-identified miRNAs to reduce hERG-related current and influence susceptibility to drug-induced QT interval prolongation warrants further investigation. </div><div><br></div>

Le récepteur sigma-1 est une protéine chaperonne active dans des tissus lésés. Le récepteur sigma-1 est principalement exprimé dans le cerveau et joue un rôle neuroprotecteur dans l’ischémie ou les maladies neurodégénératives. Le récepteur sigma-1 est également exprimé dans des lignées cellulaires cancéreuses et des travaux récents suggèrent sa participation dans la prolifération et l’apoptose. Cependant, son rôle dans la carcinogenèse reste à découvrir. Les canaux ioniques sont impliqués dans de nombreux processus physiologiques (rythme cardiaque, influx nerveux, …). Ces protéines membranaires émergent actuellement comme une nouvelle famille de cibles thérapeutiques dans les cancers. Au cours de ma thèse, j’ai montré que le récepteur sigma-1 régule l’activité du canal potassique voltage-dépendent hERG et du canal sodique voltage-dépendent Nav1.5 respectivement dans des cellules leucémiques et des cellules issues de cancer du sein. J’ai également montré que le récepteur sigma-1, à travers son action sur l’adressage du canal hERG, augmente l’invasivité des cellules leucémiques en favorisant leur interaction avec le microenvironnement tumoral. Ces résultats mettent en évidence le rôle du récepteur sigma-1 sur la plasticité électrique des cellules cancéreuses et suggèrent l’intérêt de cette protéine chaperonne comme cible thérapeutique potentielle pour limiter la progression tumorale. / The sigma-1 receptor is a chaperone protein active in damaged tissues. The sigma-1 receptor is mainly expressed into brain and have a neuroprotective role in ischemia and neurodegenerative diseases. The sigma-1 receptor is also expressed into cancer cell lines and recent investigations suggest its involvement into proliferation and apoptosis. However, its role in carcinogenesis remains to delineating. Ion channels are involved in numerous physiological processes (heart beating, nervous influx, …). These membrane proteins currently emerge as a new class of therapeutic targets in cancer. During my thesis, I observed that the sigma-1 receptor regulates voltage-dependent potassium channel hERG and voltage-dependent sodium channel Nav1.5 activities respectively into leukemic and breast cancer cell lines. I also demonstrated that the sigma-1 receptor, through its action on hERG channel, increases leukemia invasiveness by promoting interaction with tumor microenvironment. These results highlight the role of the sigma-1 receptor on cancer cell electrical plasticity and suggest this chaperone protein as a potential therapeutic target to limit tumor progression.

L'unité de recherche EA4612 de l'Université Claude Bernard Lyon 1 s'intéresse à la physiopathologie des troubles du rythme cardiaque, en particulier d'origine héréditaire. Nous avons étudié des mutations de gène de canaux ioniques découvertes chez des patients hétérozygotes atteints d'un syndrome du QT long ou de bradycardie sinusale et de fibrillation atriale. La mutation R148W du gène hERG diminue le courant maximal de 29%. Dans un modèle mathématique, ceci allonge la durée du potentiel d'action ventriculaire, ce qui pourrait rendre compte du phénotype QT long des porteurs. La mutation F627L du gène hERG se situe au centre du motif de sélectivité ionique (GFG) de la protéine hERG. Elle cause une perte de la sélectivité ionique du courant, de la propriété d'inactivation et de la sensibilité aux bloqueurs spécifiques. Ainsi, la présence du groupement aromatique de la chaîne latérale semble essentielle au maintien des propriétés du canal. La mutation Q1476R du gène SCN5A provoque un gain de fonction du courant sodique persistant. Dans un modèle de cellule cardiaque ventriculaire humaine, nous montrons une surcharge sodique intracellulaire pouvant protéger de l'allongement de la durée du potentiel d'action ventriculaire. La mutation D600E du gène HCN4 accélère la désactivation, ce qui pourrait causer une bradycardie. Par ailleurs, la mutation abolit la réponse à la suppression de l'adénosine monophosphate cyclique (AMPc) intracellulaire. La mutation V501M du gène HCN4 cause une perte totale de courant à l'état homozygote. A l'état hétérozygote, l'amplitude moyenne du courant est inchangée par rapport au WT. Cependant, un décalage négatif de la courbe d'activation rendrait compte de la bradycardie des patients porteurs / The EA4612 unit of the University Lyon 1 focuses on the pathophysiology of heart rhythm disorders, especially hereditary. We studied ion channel gene mutations discovered in heterozygote patients with long QT syndrome or sinus bradycardia and atrial fibrillation.The R148W mutation of the hERG gene decreases the maximum current by 29%. In a mathematical model, this lengthens the duration of the ventricular action potential, which could account for long QT phenotype of the patients. The F627L mutation of the hERG gene is in the center of the ion selectivity filter (GFG) of the hERG protein. It causes a loss of the ionic selectivity of the current, the inactivating property and sensitivity to specific blockers. Thus, the presence of this aromatic group of the side chain seems to be essential to the maintenance of the channel properties. The mutation Q1476R in the SCN5A gene causes a gain-of-function of the persistent sodium current. In a model of human ventricular heart cells, we show an intracellular sodium overload that can protect against the lengthening of the duration of the ventricular action potential. The D600E mutation of the HCN4 gene accelerates deactivation, which could cause bradycardia. Moreover, the mutation abolishes the response to the suppression of intracellular cyclic adenosine monophosphate (cAMP). The V501M mutation of the HCN4 gene causes a total loss of current in the homozygous state. In the heterozygous state, the average amplitude of the current is unchanged from the WT. However, a negative shift of the activation curve would account for bradycardia in patients

The toxicological screening of the numerous chemicals that we are exposed to requires significant cost and the use of animals. Accordingly, more efficient methods for the evaluation of toxicity are required to reduce cost and the number of animals used. Computational strategies have the potential to reduce both the cost and the use of animal testing in toxicity screening. The ultimate goal of this thesis is to develop computational models for the prediction of toxicological endpoints that can serve as an alternative to animal testing. In Paper I, an attempt was made to construct a global quantitative structure-activity relationship (QSAR)model for the acute toxicity endpoint (LD50 values) using the Munro database that represents a broad chemical landscape. Such a model could be used for acute toxicity screening of chemicals of diverse structures. Paper II focuses on the use of acute toxicity data to support the prediction of chronic toxicity. The results of this study suggest that for related chemicals having acute toxicities within a similar range, their lowest observed effect levels (LOELs) can be used in read-across strategies to fill gaps in chronic toxicity data. In Paper III a k-nearest neighbor (k-NN) classification model was developed to predict human ether-a-go-go related gene (hERG)-derived toxicity. The results suggest that the model has potential for use in identifying compounds with hERG-liabilities, e.g. in drug development.