Structure of alpha6 beta3 delta GABA(A) receptors and their lack of ethanol sensitivity

Delta (delta) subunit containing GABA(A) receptors are expressed extra-synaptically and mediate tonic inhibition. In cerebellar granule cells, they often form a receptor together with alpha(6) subunits. We were interested to determine the architecture of these receptors. We predefined the subunit arrangement of 24 different GABA(A) receptor pentamers by subunit concatenation. These receptors (composed of alpha(6), beta(3) and delta subunits) were expressed in Xenopus oocytes and their electrophysiological properties analyzed. Currents elicited in response to GABA were determined in presence and absence of 3alpha, 21-dihydroxy-5alpha-pregnan-20-one and to 4,5,6,7-tetrahydroisoxazolo[5,4-c]-pyridin-3-ol. alpha(6)-beta(3)-alpha(6)/delta receptors showed a substantial response to GABA alone. Three receptors, beta(3)-alpha(6)-delta/alpha(6)-beta(3), alpha(6)-beta(3)-alpha(6)/beta(3)-delta and beta(3)-delta-beta(3)/alpha(6)-beta(3), were only uncovered in the combined presence of the neurosteroid 3alpha, 21-dihydroxy-5alpha-pregnan-20-one with GABA. All four receptors were activated by 4,5,6,7-tetrahydroisoxazolo[5,4-c]-pyridin-3-ol. None of the functional receptors was modulated by physiological concentrations (up to 30 mM) of ethanol. GABA concentration response curves indicated that the delta subunit can contribute to the formation of an agonist site. We conclude from the investigated receptors that the delta subunit can assume multiple positions in a receptor pentamer composed of alpha(6), beta(3) and delta subunits.

Directed differentiation and functional maturation of cortical interneurons from human embryonic stem cells

Human pluripotent stem cells are a powerful tool for modeling brain development and disease. The human cortex is composed of two major neuronal populations: projection neurons and local interneurons. Cortical interneurons comprise a diverse class of cell types expressing the neurotransmitter GABA. Dysfunction of cortical interneurons has been implicated in neuropsychiatric diseases, including schizophrenia, autism, and epilepsy. Here, we demonstrate the highly efficient derivation of human cortical interneurons in an NKX2.1::GFP human embryonic stem cell reporter line. Manipulating the timing of SHH activation yields three distinct GFP+ populations with specific transcriptional profiles, neurotransmitter phenotypes, and migratory behaviors. Further differentiation in a murine cortical environment yields parvalbumin- and somatostatin-expressing neurons that exhibit synaptic inputs and electrophysiological properties of cortical interneurons. Our study defines the signals sufficient for modeling human ventral forebrain development in vitro and lays the foundation for studying cortical interneuron involvement in human disease pathology.

Chronic administration of methylphenidate produces neurophysiological and behavioral sensitization.

The electrophysiological properties of acute and chronic methylphenidate (MPD) on neurons of the prefrontal cortex (PFC) and caudate nucleus (CN) have not been studied in awake, freely behaving animals. The present study was designed to investigate the dose-response effects of MPD on sensory evoked potentials recorded from the PFC and CN in freely behaving rats previously implanted with permanent electrodes, as well as their behavioral (locomotor) activities. On experimental day 1, locomotor behavior of rats was recorded for 2 h post-saline injection, and sensory evoked field potentials were recorded before and after saline and 0.6, 2.5, and 10 mg/kg, i.p., MPD administration. Animals were injected for the next five days with daily 2.5 mg/kg MPD to elicit behavioral sensitization. Locomotor recording was resumed on experimental days 2 and 6 after the MPD maintenance dose followed by 3 days of washout. On experimental day 10, rats were connected again to the electrophysiological recording system and rechallenged with saline and the identical MPD doses as on experimental day 1. On experimental day 11, rat's locomotor recording was resumed before and after 2.5 mg/kg MPD administration. Behavioral results showed that repeated administration of MPD induced behavioral sensitization. Challenge doses (0.6, 2.5, and 10.0 mg/kg) of MPD on experimental day 1 elicited dose-response attenuation in the response amplitude of the average sensory evoked field potential components recorded from the PFC and CN. Chronic MPD administration resulted in attenuation of the PFC's baseline recorded on experimental day 10, while the same treatment did not modulate the baseline recorded from the CN. Treatment of MPD on experimental day 10 resulted in further decrease of the average sensory evoked response compared to that obtained on experimental day 1. This observation of further decrease in the electrophysiological responses after chronic administration of MPD suggests that the sensory evoked responses on experimental day 10 represent neurophysiological sensitization. Moreover, two different response patterns were obtained from PFC and CN following chronic methylphenidate administration. In PFC, the baseline and effect of methylphenidate expressed electrophysiological sensitization on experimental day 10, while recording from CN did not exhibit any electrophysiological sensitization.

Highly pathogenic adapted HIV-1 strains limit host immunity and dictate rapid disease progression.

OBJECTIVE: The study of HIV-1 rapid progressors has been limited to specific case reports. Nevertheless, identification and characterization of the viral and host factors involved in rapid progression are crucial when attempting to uncover the correlates of rapid disease outcome. DESIGN: We carried out comparative functional analyses in rapid progressors (n = 46) and standard progressors (n = 46) early after HIV-1 seroconversion (≤1 year). The viral traits tested were viral replicative capacity, co-receptor usage, and genomic variation. Host CD8 T-cell responses, humoral activity, and HLA immunogenetic markers were also determined. RESULTS: Our data demonstrate an unusual convergence of highly pathogenic HIV-1 strains in rapid progressors. Compared with standard progressors, rapid progressor viral strains show higher in-vitro replicative capacity (81.5 vs. 67.9%; P = 0.025) and greater X4/DM co-receptor usage (26.3 vs. 2.8%; P = 0.006) in early infection. Limited or absent functional HIV-1 CD8 T-cell responses and neutralizing activity were measured in rapid progressors. Moreover, the increase in common HLA allele-restricted CD8 T-cell escape mutations in rapid progressors acts as a signature of uncontrolled HIV-1 replication and early impairment of adaptive cellular responses. CONCLUSION: Our data support a dominant role for viral factors in rapid progressors. Robust HIV-1 replication and intrinsic viral properties limit host adaptive immune responses, thus driving rapid disease progression.

The antimalarial drugs quinine, chloroquine and mefloquine are antagonists at 5-HT 3 receptors

Background and Purpose: The antimalarial compounds quinine, chloroquine and mefloquine affect the electrophysiological properties of Cys-loop receptors and have structural similarities to 5-HT3 receptor antagonists. They may therefore act at 5-HT3 receptors. Experimental Approach: The effects of quinine, chloroquine and mefloquine on electrophysiological and ligand binding properties of 5-HT3A receptors expressed in HEK 293 cells and Xenopus oocytes were examined. The compounds were also docked into models of the binding site. Key Results: 5-HT3 responses were blocked with IC50 values of 13.4 μM, 11.8 μM and 9.36 μM for quinine, chloroquine and mefloquine. Schild plots indicated quinine and chloroquine behaved competitively with pA2 values of 4.92 (KB=12.0 μM) and 4.97 (KB=16.4 μM). Mefloquine displayed weakly voltage-dependent, non-competitive inhibition consistent with channel block. On and off rates for quinine and chloroquine indicated a simple bimolecular reaction scheme. Quinine, chloroquine and mefloquine displaced [3H]granisetron with Ki values of 15.0, 24.2 and 35.7 μM. Docking of quinine into a homology model of the 5-HT3 receptor binding site located the tertiary ammonium between W183 and Y234, and the quinoline ring towards the membrane, stabilised by a hydrogen bond with E129. For chloroquine, the quinoline ring was positioned between W183 and Y234 and the tertiary ammonium stabilised by interactions with F226. Conclusions and Implications: This study shows that quinine and chloroquine competitively inhibit 5-HT3 receptors, while mefloquine inhibits predominantly non-competitively. Both quinine and chloroquine can be docked into a receptor binding site model, consistent with their structural homology to 5-HT3 receptor antagonists.

p.L1612P, a Novel Voltage-gated Sodium Channel Nav1.7 Mutation Inducing a Cold Sensitive Paroxysmal Extreme Pain Disorder

BACKGROUND Mutations in the SCN9A gene cause chronic pain and pain insensitivity syndromes. We aimed to study clinical, genetic, and electrophysiological features of paroxysmal extreme pain disorder (PEPD) caused by a novel SCN9A mutation. METHODS Description of a 4-generation family suffering from PEPD with clinical, genetic and electrophysiological studies including patch clamp experiments assessing response to drug and temperature. RESULTS The family was clinically comparable to those reported previously with the exception of a favorable effect of cold exposure and a lack of drug efficacy including with carbamazepine, a proposed treatment for PEPD. A novel p.L1612P mutation in the Nav1.7 voltage-gated sodium channel was found in the four affected family members tested. Electrophysiologically the mutation substantially depolarized the steady-state inactivation curve (V1/2 from -61.8 ± 4.5 mV to -30.9 ± 2.2 mV, n = 4 and 7, P < 0.001), significantly increased ramp current (from 1.8% to 3.4%, n = 10 and 12) and shortened recovery from inactivation (from 7.2 ± 5.6 ms to 2.2 ± 1.5 ms, n = 11 and 10). However, there was no persistent current. Cold exposure reduced peak current and prolonged recovery from inactivation in wild-type and mutated channels. Amitriptyline only slightly corrected the steady-state inactivation shift of the mutated channel, which is consistent with the lack of clinical benefit. CONCLUSIONS The novel p.L1612P Nav1.7 mutation expands the PEPD spectrum with a unique combination of clinical symptoms and electrophysiological properties. Symptoms are partially responsive to temperature but not to drug therapy. In vitro trials of sodium channel blockers or temperature dependence might help predict treatment efficacy in PEPD.

Slow conduction in mixed cultured strands of primary ventricular cells and stem cell-derived cardiomyocytes

Modern concepts for the treatment of myocardial diseases focus on novel cell therapeutic strategies involving stem cell-derived cardiomyocytes (SCMs). However, functional integration of SCMs requires similar electrophysiological properties as primary cardiomyocytes (PCMs) and the ability to establish intercellular connections with host myocytes in order to contribute to the electrical and mechanical activity of the heart. The aim of this project was to investigate the properties of cardiac conduction in a co-culture approach using SCMs and PCMs in cultured cell strands. Murine embryonic SCMs were pooled with fetal ventricular cells and seeded in predefined proportions on microelectrode arrays to form patterned strands of mixed cells. Conduction velocity (CV) was measured during steady state pacing. SCM excitability was estimated from action potentials measured in single cells using the patch clamp technique. Experiments were complemented with computer simulations of conduction using a detailed model of cellular architecture in mixed cell strands. CV was significantly lower in strands composed purely of SCMs (5.5 ± 1.5 cm/s, n = 11) as compared to PCMs (34.9 ± 2.9 cm/s, n = 21) at similar refractoriness (100% SCMs: 122 ± 25 ms, n = 9; 100% PCMs: 139 ± 67 ms, n = 14). In mixed strands combining both cell types, CV was higher than in pure SCMs strands, but always lower than in 100% PCM strands. Computer simulations demonstrated that both intercellular coupling and electrical excitability limit CV. These data provide evidence that in cultures of murine ventricular cardiomyocytes, SCMs cannot restore CV to control levels resulting in slow conduction, which may lead to reentry circuits and arrhythmias.

Efectos electrofisiológicos del ácido acetilsalicílico en la injuria por isquemia/reperfusión

La reperfusión, luego de un período de isquemia miocárdica breve, puede desencadenar un daño paradojal, dentro del cual, se destacan las arritmias ventriculares. Existen estudios que reportan un efecto beneficioso del ácido acetilsalicílico (AAS) a nivel cardiovascular, pero se desconocen los efectos electrofisiológicos en el proceso de injuria por isquemia/reperfusión. El objetivo de este estudio es evaluar las propiedades electrofisiológicas del AAS, en especial si puede evitar las arritmias de reperfusión (AR) en forma independiente de su efecto antiplaquetario. Se trabajó con corazones aislados de rata Sprague Dawley según la técnica de Langendorff sometidos a 10 minutos de isquemia regional. Se realizaron 3 series esperimentales: 1) control (C, n=10); 2) , corazones perfundidos durante todo el protocolo con AAS 0.14 mM (AAS, n=10) y 3) corazones que recibieron la misma dosis de AAS sólo en los 3 primeros minutos de la reperfusión (AASR, n=9). Se analizaron la incidencia y severidad de las AR y su relación con el ECG y los potenciales de acción registrados simultáneamente. El 82% del grupo control presentó AR sostenidas, el 30 % con AAS y el 22% con AASR (ambas p<0.05 por χ2). En la reperfusión se observó que luego de los primeros tres minutos la duración del potencial de acción (DPA) fue mayor en el grupo AASR (81,5 ± 23,1) que en el grupo AAS (55,2 ± 10,0) p<0.05 por ANOVA I. Por lo tanto, la menor incidencia de AR en los grupos tratados podría asociarse al efecto de la aspirina sobre la DPA y que la droga estudiada tendría efectos sobre esta variable sólo al momento de reperfusión.

Elements and isotopic measurements of magmatic rock samples from ODP Hole 104-642E

Trace element and isotopic signatures of magmatic rock samples from ODP Hole 642E at the Vøring Plateau provide insight into the interaction processes of mantle melt with crust during the initial magma extrusion phases at the onset of the continental breakup. The intermediate (basaltic-andesitic) to felsic (dacitic and rhyolitic) Lower Series magmas at ODP Hole 642E appear to be produced by large amounts of melting of upper crustal material. This study not only makes use of the traditional geochemical tools to investigate crust-mantle interaction, but also explores the value of Cs geochemistry as an additional tool. The element Cs forms the largest lithophile cation, and shows the largest contrast in concentration between (depleted) mantle and continental crust. As such it is a very sensitive indicator of involvement of crustal material. The Cs data reinforce the conclusion drawn from isotopic signatures that the felsic magmas are largely anatectic crustal melts. The down-hole geochemical variation within ODP Hole 642E defines a decreasing continental crustal influence from the Lower Series into the Upper Series. This is essential information to distinguish intrinsic geochemical properties of the mantle melts from signatures imposed by crustal contamination. A comparison with data from the SE Greenland margin highlights the compositional asymmetry of the crust-mantle interactions at both sides of the paleo-Iapetus suture. While Lower Series and Middle Series rocks from the SE Greenland margin have isotopic signatures reflecting interactions with lower and middle crust, such signatures have not been observed at the mid-Norwegian margin. The geochemical data either point to a dissimilar Caledonian crustal composition and/or to different geodynamic pre-breakup rifting history at the two NE Atlantic margin segments.

Molecular dissolved organic matter composition in lakes across Sweden as relative intensities of FT-ICR-MS peaks and PARAFAC components and optical indices

Whether intrinsic molecular properties or extrinsic factors such as environmental conditions control the decomposition of natural organic matter across soil, marine and freshwater systems has been subject to debate. Comprehensive evaluations of the controls that molecular structure exerts on organic matter's persistence in the environment have been precluded by organic matter's extreme complexity. Here we examine dissolved organic matter from 109 Swedish lakes using ultrahigh-resolution mass spectrometry and optical spectroscopy to investigate the constraints on its persistence in the environment. We find that degradation processes preferentially remove oxidized, aromatic compounds, whereas reduced, aliphatic and N-containing compounds are either resistant to degradation or tightly cycled and thus persist in aquatic systems. The patterns we observe for individual molecules are consistent with our measurements of emergent bulk characteristics of organic matter at wide geographic and temporal scales, as reflected by optical properties. We conclude that intrinsic molecular properties are an important control of overall organic matter reactivity.

Biogenic silica and silicic acid benthic fluxes of a North-eastern Atlantic Abyssal Locality

Within the framework of the EU-funded BENGAL programme, the effects of seasonality on biogenic silica early diagenesis have been studied at the Porcupine Abyssal Plain (PAP), an abyssal locality located in the northeast Atlantic Ocean. Nine cruises were carried out between August 1996 and August 1998. Silicic acid (DSi) increased downward from 46.2 to 213 µM (mean of 27 profiles). Biogenic silica (BSi) decreased from ca. 2% near the sediment-water interface to <1% at depth. Benthic silicic acid fluxes as measured from benthic chambers were close to those estimated from non-linear DSi porewater gradients. Some 90% of the dissolution occurred within the top 5.5 cm of the sediment column, rather than at the sediment-water interface and the annual DSi efflux was close to 0.057 mol Si/m**2/yr. Biogenic silica accumulation was close to 0.008 mol Si/m**2/yr and the annual opal delivery reconstructed from sedimentary fluxes, assuming steady state, was 0.065 mol Si/m**2/yr. This is in good agreement with the mean annual opal flux determined from sediment trap samples, averaged over the last decade (0.062 mol Si/m**2/yr). Thus ca. 12% of the opal flux delivered to the seafloor get preserved in the sediments. A simple comparison between the sedimentation rate and the dissolution rate in the uppermost 5.5 cm of the sediment column suggests that there should be no accumulation of opal in PAP sediments. However, by combining the BENGAL high sampling frequency with our experimental results on BSi dissolution, we conclude that non-steady state processes associated with the seasonal deposition of fresh biogenic particles may well play a fundamental role in the preservation of BSi in these sediments. This comes about though the way seasonal variability affects the quality of the biogenic matter reaching the seafloor. Hence it influences the intrinsic dissolution properties of the opal at the seafloor and also the part played by non-local mixing events by ensuring the rapid transport of BSi particles deep into the sediment to where saturation is reached.

A common mechanism of defective channel trafficking underlying DFNA2 hearing loss result in different cell surface expression levels of KCNQ4 mutants

KCNQ4 mutations underlie DFNA2, a subtype of autosomal dominant hearing loss. We had previously identified the pore-region p.G296S mutation that impaired channel activity in two manners: it greatly reduced surface expression and abolished channel function. Moreover, G296S mutant exerted a strong dominant-negative effect on potassium currents by reducing the channel expression at the cell surface representing the first study to identify a trafficking-dependent dominant mechanism for the loss of KCNQ4 channel function in DFNA2. Here, we have investigated the pathogenic mechanism associated with all the described KCNQ4 mutations (F182L, W242X, E260K, D262V, L274H, W276S, L281S, G285C, G285S and G321S) that are located in different domains of the channel protein. F182L mutant showed a wild type-like cell-surface distribution in transiently transfected NIH3T3 fibroblasts and the recorded currents in Xenopus oocytes resembled those of the wild-type. The remaining KCNQ4 mutants abolished potassium currents, but displayed distinct levels of defective cell-surface expression in NIH3T3 as quantified by flow citometry. Co-localization studies revealed these mutants were retained in the ER, unless W242X, which showed a clear co-localization with Golgi apparatus. Interestingly, this mutation results in a truncated KCNQ4 protein at the S5 transmembrane domain, before the pore region, that escapes the protein quality control in the ER but does not reach the cell surface at normal levels. Currently we are investigating the trafficking behaviour and electrophysiological properties of several KCNQ4 truncated proteins artificially generated in order to identify specific motifs involved in channel retention/exportation. Altogether, our results indicate that a defect in KCNQ4 trafficking is the common mechanism underlying DFNA2

Contributions to Bayesian network learning with applications to neuroscience

Neuronal morphology is a key feature in the study of brain circuits, as it is highly related to information processing and functional identification. Neuronal morphology affects the process of integration of inputs from other neurons and determines the neurons which receive the output of the neurons. Different parts of the neurons can operate semi-independently according to the spatial location of the synaptic connections. As a result, there is considerable interest in the analysis of the microanatomy of nervous cells since it constitutes an excellent tool for better understanding cortical function. However, the morphologies, molecular features and electrophysiological properties of neuronal cells are extremely variable. Except for some special cases, this variability makes it hard to find a set of features that unambiguously define a neuronal type. In addition, there are distinct types of neurons in particular regions of the brain. This morphological variability makes the analysis and modeling of neuronal morphology a challenge. Uncertainty is a key feature in many complex real-world problems. Probability theory provides a framework for modeling and reasoning with uncertainty. Probabilistic graphical models combine statistical theory and graph theory to provide a tool for managing domains with uncertainty. In particular, we focus on Bayesian networks, the most commonly used probabilistic graphical model. In this dissertation, we design new methods for learning Bayesian networks and apply them to the problem of modeling and analyzing morphological data from neurons. The morphology of a neuron can be quantified using a number of measurements, e.g., the length of the dendrites and the axon, the number of bifurcations, the direction of the dendrites and the axon, etc. These measurements can be modeled as discrete or continuous data. The continuous data can be linear (e.g., the length or the width of a dendrite) or directional (e.g., the direction of the axon). These data may follow complex probability distributions and may not fit any known parametric distribution. Modeling this kind of problems using hybrid Bayesian networks with discrete, linear and directional variables poses a number of challenges regarding learning from data, inference, etc. In this dissertation, we propose a method for modeling and simulating basal dendritic trees from pyramidal neurons using Bayesian networks to capture the interactions between the variables in the problem domain. A complete set of variables is measured from the dendrites, and a learning algorithm is applied to find the structure and estimate the parameters of the probability distributions included in the Bayesian networks. Then, a simulation algorithm is used to build the virtual dendrites by sampling values from the Bayesian networks, and a thorough evaluation is performed to show the model’s ability to generate realistic dendrites. In this first approach, the variables are discretized so that discrete Bayesian networks can be learned and simulated. Then, we address the problem of learning hybrid Bayesian networks with different kinds of variables. Mixtures of polynomials have been proposed as a way of representing probability densities in hybrid Bayesian networks. We present a method for learning mixtures of polynomials approximations of one-dimensional, multidimensional and conditional probability densities from data. The method is based on basis spline interpolation, where a density is approximated as a linear combination of basis splines. The proposed algorithms are evaluated using artificial datasets. We also use the proposed methods as a non-parametric density estimation technique in Bayesian network classifiers. Next, we address the problem of including directional data in Bayesian networks. These data have some special properties that rule out the use of classical statistics. Therefore, different distributions and statistics, such as the univariate von Mises and the multivariate von Mises–Fisher distributions, should be used to deal with this kind of information. In particular, we extend the naive Bayes classifier to the case where the conditional probability distributions of the predictive variables given the class follow either of these distributions. We consider the simple scenario, where only directional predictive variables are used, and the hybrid case, where discrete, Gaussian and directional distributions are mixed. The classifier decision functions and their decision surfaces are studied at length. Artificial examples are used to illustrate the behavior of the classifiers. The proposed classifiers are empirically evaluated over real datasets. We also study the problem of interneuron classification. An extensive group of experts is asked to classify a set of neurons according to their most prominent anatomical features. A web application is developed to retrieve the experts’ classifications. We compute agreement measures to analyze the consensus between the experts when classifying the neurons. Using Bayesian networks and clustering algorithms on the resulting data, we investigate the suitability of the anatomical terms and neuron types commonly used in the literature. Additionally, we apply supervised learning approaches to automatically classify interneurons using the values of their morphological measurements. Then, a methodology for building a model which captures the opinions of all the experts is presented. First, one Bayesian network is learned for each expert, and we propose an algorithm for clustering Bayesian networks corresponding to experts with similar behaviors. Then, a Bayesian network which represents the opinions of each group of experts is induced. Finally, a consensus Bayesian multinet which models the opinions of the whole group of experts is built. A thorough analysis of the consensus model identifies different behaviors between the experts when classifying the interneurons in the experiment. A set of characterizing morphological traits for the neuronal types can be defined by performing inference in the Bayesian multinet. These findings are used to validate the model and to gain some insights into neuron morphology. Finally, we study a classification problem where the true class label of the training instances is not known. Instead, a set of class labels is available for each instance. This is inspired by the neuron classification problem, where a group of experts is asked to individually provide a class label for each instance. We propose a novel approach for learning Bayesian networks using count vectors which represent the number of experts who selected each class label for each instance. These Bayesian networks are evaluated using artificial datasets from supervised learning problems. Resumen La morfología neuronal es una característica clave en el estudio de los circuitos cerebrales, ya que está altamente relacionada con el procesado de información y con los roles funcionales. La morfología neuronal afecta al proceso de integración de las señales de entrada y determina las neuronas que reciben las salidas de otras neuronas. Las diferentes partes de la neurona pueden operar de forma semi-independiente de acuerdo a la localización espacial de las conexiones sinápticas. Por tanto, existe un interés considerable en el análisis de la microanatomía de las células nerviosas, ya que constituye una excelente herramienta para comprender mejor el funcionamiento de la corteza cerebral. Sin embargo, las propiedades morfológicas, moleculares y electrofisiológicas de las células neuronales son extremadamente variables. Excepto en algunos casos especiales, esta variabilidad morfológica dificulta la definición de un conjunto de características que distingan claramente un tipo neuronal. Además, existen diferentes tipos de neuronas en regiones particulares del cerebro. La variabilidad neuronal hace que el análisis y el modelado de la morfología neuronal sean un importante reto científico. La incertidumbre es una propiedad clave en muchos problemas reales. La teoría de la probabilidad proporciona un marco para modelar y razonar bajo incertidumbre. Los modelos gráficos probabilísticos combinan la teoría estadística y la teoría de grafos con el objetivo de proporcionar una herramienta con la que trabajar bajo incertidumbre. En particular, nos centraremos en las redes bayesianas, el modelo más utilizado dentro de los modelos gráficos probabilísticos. En esta tesis hemos diseñado nuevos métodos para aprender redes bayesianas, inspirados por y aplicados al problema del modelado y análisis de datos morfológicos de neuronas. La morfología de una neurona puede ser cuantificada usando una serie de medidas, por ejemplo, la longitud de las dendritas y el axón, el número de bifurcaciones, la dirección de las dendritas y el axón, etc. Estas medidas pueden ser modeladas como datos continuos o discretos. A su vez, los datos continuos pueden ser lineales (por ejemplo, la longitud o la anchura de una dendrita) o direccionales (por ejemplo, la dirección del axón). Estos datos pueden llegar a seguir distribuciones de probabilidad muy complejas y pueden no ajustarse a ninguna distribución paramétrica conocida. El modelado de este tipo de problemas con redes bayesianas híbridas incluyendo variables discretas, lineales y direccionales presenta una serie de retos en relación al aprendizaje a partir de datos, la inferencia, etc. En esta tesis se propone un método para modelar y simular árboles dendríticos basales de neuronas piramidales usando redes bayesianas para capturar las interacciones entre las variables del problema. Para ello, se mide un amplio conjunto de variables de las dendritas y se aplica un algoritmo de aprendizaje con el que se aprende la estructura y se estiman los parámetros de las distribuciones de probabilidad que constituyen las redes bayesianas. Después, se usa un algoritmo de simulación para construir dendritas virtuales mediante el muestreo de valores de las redes bayesianas. Finalmente, se lleva a cabo una profunda evaluaci ón para verificar la capacidad del modelo a la hora de generar dendritas realistas. En esta primera aproximación, las variables fueron discretizadas para poder aprender y muestrear las redes bayesianas. A continuación, se aborda el problema del aprendizaje de redes bayesianas con diferentes tipos de variables. Las mixturas de polinomios constituyen un método para representar densidades de probabilidad en redes bayesianas híbridas. Presentamos un método para aprender aproximaciones de densidades unidimensionales, multidimensionales y condicionales a partir de datos utilizando mixturas de polinomios. El método se basa en interpolación con splines, que aproxima una densidad como una combinación lineal de splines. Los algoritmos propuestos se evalúan utilizando bases de datos artificiales. Además, las mixturas de polinomios son utilizadas como un método no paramétrico de estimación de densidades para clasificadores basados en redes bayesianas. Después, se estudia el problema de incluir información direccional en redes bayesianas. Este tipo de datos presenta una serie de características especiales que impiden el uso de las técnicas estadísticas clásicas. Por ello, para manejar este tipo de información se deben usar estadísticos y distribuciones de probabilidad específicos, como la distribución univariante von Mises y la distribución multivariante von Mises–Fisher. En concreto, en esta tesis extendemos el clasificador naive Bayes al caso en el que las distribuciones de probabilidad condicionada de las variables predictoras dada la clase siguen alguna de estas distribuciones. Se estudia el caso base, en el que sólo se utilizan variables direccionales, y el caso híbrido, en el que variables discretas, lineales y direccionales aparecen mezcladas. También se estudian los clasificadores desde un punto de vista teórico, derivando sus funciones de decisión y las superficies de decisión asociadas. El comportamiento de los clasificadores se ilustra utilizando bases de datos artificiales. Además, los clasificadores son evaluados empíricamente utilizando bases de datos reales. También se estudia el problema de la clasificación de interneuronas. Desarrollamos una aplicación web que permite a un grupo de expertos clasificar un conjunto de neuronas de acuerdo a sus características morfológicas más destacadas. Se utilizan medidas de concordancia para analizar el consenso entre los expertos a la hora de clasificar las neuronas. Se investiga la idoneidad de los términos anatómicos y de los tipos neuronales utilizados frecuentemente en la literatura a través del análisis de redes bayesianas y la aplicación de algoritmos de clustering. Además, se aplican técnicas de aprendizaje supervisado con el objetivo de clasificar de forma automática las interneuronas a partir de sus valores morfológicos. A continuación, se presenta una metodología para construir un modelo que captura las opiniones de todos los expertos. Primero, se genera una red bayesiana para cada experto y se propone un algoritmo para agrupar las redes bayesianas que se corresponden con expertos con comportamientos similares. Después, se induce una red bayesiana que modela la opinión de cada grupo de expertos. Por último, se construye una multired bayesiana que modela las opiniones del conjunto completo de expertos. El análisis del modelo consensuado permite identificar diferentes comportamientos entre los expertos a la hora de clasificar las neuronas. Además, permite extraer un conjunto de características morfológicas relevantes para cada uno de los tipos neuronales mediante inferencia con la multired bayesiana. Estos descubrimientos se utilizan para validar el modelo y constituyen información relevante acerca de la morfología neuronal. Por último, se estudia un problema de clasificación en el que la etiqueta de clase de los datos de entrenamiento es incierta. En cambio, disponemos de un conjunto de etiquetas para cada instancia. Este problema está inspirado en el problema de la clasificación de neuronas, en el que un grupo de expertos proporciona una etiqueta de clase para cada instancia de manera individual. Se propone un método para aprender redes bayesianas utilizando vectores de cuentas, que representan el número de expertos que seleccionan cada etiqueta de clase para cada instancia. Estas redes bayesianas se evalúan utilizando bases de datos artificiales de problemas de aprendizaje supervisado.

A computational model coupling mechanics and electrophysiology in spinal cord injury

Traumatic brain injury and spinal cord injury have recently been put under the spotlight as major causes of death and disability in the developed world. Despite the important ongoing experimental and modeling campaigns aimed at understanding the mechanics of tissue and cell damage typically observed in such events, the differenti- ated roles of strain, stress and their corresponding loading rates on the damage level itself remain unclear. More specif- ically, the direct relations between brain and spinal cord tis- sue or cell damage, and electrophysiological functions are still to be unraveled. Whereas mechanical modeling efforts are focusing mainly on stress distribution and mechanistic- based damage criteria, simulated function-based damage cri- teria are still missing. Here, we propose a new multiscale model of myelinated axon associating electrophysiological impairment to structural damage as a function of strain and strain rate. This multiscale approach provides a new framework for damage evaluation directly relating neuron mechanics and electrophysiological properties, thus provid- ing a link between mechanical trauma and subsequent func- tional deficits.

A computational model coupling mechanics and electrophysiology in spinal cord injury

Traumatic brain injury and spinal cord injury have recently been put under the spotlight as major causes of death and disability in the developed world. Despite the important ongoing experimental and modeling campaigns aimed at understanding the mechanics of tissue and cell damage typically observed in such events, the differentiated roles of strain, stress and their corresponding loading rates on the damage level itself remain unclear. More specifically, the direct relations between brain and spinal cord tissue or cell damage, and electrophysiological functions are still to be unraveled. Whereas mechanical modeling efforts are focusing mainly on stress distribution and mechanistic-based damage criteria, simulated function-based damage criteria are still missing. Here, we propose a new multiscale model of myelinated axon associating electrophysiological impairment to structural damage as a function of strain and strain rate. This multiscale approach provides a new framework for damage evaluation directly relating neuron mechanics and electrophysiological properties, thus providing a link between mechanical trauma and subsequent functional deficits