Estudios “In-Silico” del Transporte Epitelial Intestinal de Fármacos en Cultivo de Células Caco-2 Aplicando Descriptores TOMOCOMD-CARDD
Archivos
Fecha
2005-06-25
Autores
Vera González, Edith M.
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad Central "Marta Abreu" de las Villas
Resumen
Nuevos descriptores moleculares han sido aplicados a la estimación del transporte intestinal de fármacos en cultivos de células Caco-2. Los índices cuadráticos totales y locales fueron calculados por el programa TOMOCOMD-CARDD. Utilizando la Regresión Lineal Múltiple, dos modelos QSPerR para la predicción del coeficiente de permeabilidad en células Caco-2 en ambas direcciones de absorción {P [apical a basolateral (AP→BL) y basolateral a apical (BL→AP)]) fueron obtenidos. El procedimiento de validación cruzada dejando uno fuera evidenció que los modelos de regresión tuvieron una adecuada estabilidad y un buen poder predictivo.
En orden de desarrollar una validación más exhaustiva del poder predictivo de los modelos obtenidos, una serie de predicción externa fue utilizada; mostrándose un comportamiento similar al obtenido en la serie de exploración. El uso combinado de ambos modelos permite estimar el Radio Direccional de Permeabilidad (PDR, BL→AP/ AP→BL) y obtener conclusiones preliminares sobre la interacción del compuesto con la glicoproteína-P. Además, el análisis discriminante lineal fue usado para obtener una función de clasificación que discrimina los compuestos de elevada absorción ( ≥8 x10-6cm/s) de los de moderada o pobre absorción (P<8x10-6cm/s). Una base de datos de 134 compuestos con una gran diversidad estructural y dos series de
compuestos tipo-fármacos (12 compuestos) fueron usados como series de entrenamiento y de predicción, respectivamente. Los modelos obtenidos clasificaron correctamente el 81.13% de compuestos con moderada o pobre absorción y el 96.30% de compuestos con elevada absorción, mostrando una buena clasificación general
de 90.30% en la serie de entrenamiento. Los procesos de validación interna y externa demostraron la robustez y el poder predictivo de los modelos obtenidos. En este sentido, estos modelos clasificaron correctamente el 79.24% y el 92.52% de los compuestos de la serie de entrenamiento pertenecientes a los grupos de moderada o
pobre absorción y a los de elevada absorción, respectivamente, en el proceso de validación cruzada dejando uno fuera; siendo la clasificación global de 87.31%. Este modelo también muestra un 87.5, 85.6, 84.7, 85.0, 85.3, 83.5, 84.1, 86.2, 85.9 y 85.9% de buena clasificación cuando n varía de 2 a 11 en el procedimiento de validación cruzada dejando n-compuestos fuera en cada ciclo. El modelo se estabilizó alrededor de un 85.9% de buena clasificación global, para n >9. Siete inhibidores de la proteasa del VIH (4 peptidomiméticos lineales y 3 ureas cíclicas) y 5 nuevos derivados 6-fluoroquinolínicos se usaron como serie de predicción externa. El modelo
QSAR mostró un 83.33% de buena clasificación global en esta serie de predicción. Esta aproximación nos permite obtener una buena explicación en términos estructurales del proceso de absorción, evidenciando el rol fundamental de los H-unidos a heteroátomos y del tamaño molecular. Finalmente, el modelo desarrollado fue usado en la evaluación in silico de 241 fármacos con valores reportados de porcentaje de absorción intestinal experimental (Abs %). Se estableció una relación (extrapolación) entre los resultados teóricos in vitro (Caco-2) y los experimentales in vivo Abs % (145 compuestos con datos de buena calidad), con porcentaje de buena
relación mayor del 82%. Una comparación con resultados derivados de otros tres estudios teóricos mostraron un satisfactorio comportamiento del presente método. Todos estos resultados muestran que los índices cuadráticos pueden predecir exitosamente la permeabilidad intestinal y sugieren que la metodología propuesta será una gran herramienta para el estudio de la absorción oral de candidatos a fármacos durante los procesos de desarrollo.
Novel topologic molecular descriptors have been applied to estimate the intestinal epithelial transport of drug in Caco-2 cell culture. Total and local quadratic indices used in this study were calculated by TOMOCOMDCARDD software. Two QSPR models, through a multiple linear regression, were obtained to predict the P [apical to basolateral (AP→BL) and basolateral to apical (BL→AP)]. A leave-one-out cross-validation procedure revealed that the regression models, had a fairly good predictability. Furthermore, others 18 drugs were selected as a test set in order to assess the predictive power of the models and the accuracy of the final prediction was similar to achieve for the data set. Besides, the use of both regression models, in a combinative way, is possible to predict the Permeability Directional Ratio (PDR, BL→AP/ AP→BL) value. Linear discriminant analysis was used to obtain a quantitative model that discriminates the high absorption compounds (P ≥8 x10-6cm/s) from those with moderate-poor absorption (P<8 x10-6cm/s). A data set of 134 diverse structure drugs and two series of drugs-like compounds (12 compounds) were used as training and test set, respectively. The obtained model classified correctly 81.13% of compounds with moderate-poor absorption properties and the 96.30% of compounds with high absorption, showing a global good classification of 90.30% in the training set. Internal and external validation processes to demonstrate the robustness and predictive power of the obtained model were carried out. In this sense, this model classified correctly 79.24% and 92.52% of compounds of the training set belonging to the moderate-poor and high absorption groups in the leave-one-out cross-validation procedure, respectively; being the global classification of 87.31%. Also this model shown an 87.5, 85.6, 84.7, 85.0, 85.3, 83.5, 84.1, 86.2, 85.9 and 85.9% of global good classification when n varied from 2 to 11 in the leave-n-out cross validation procedure. The model was stabilized around 85.9% when n was > 9. A data set of 7 HIV protease inhibitors (4 linear peptidomimetic and 3 new cyclic urea) and 5 new 6- fluoroquinolones derivatives was used as test set. The QSAR model classified correctly the 83.33% of compounds in this external prediction set. This approach permits us to obtain a good explanation of the experiment based on the molecular structural features, evidencing the main role of H-bonding and size properties in permeability process. Finally, the model developed was used in the virtual screening of 241 drugs with the percentage of human intestinal absorption (Abs %) values reported. A relationship between the predicted permeability coefficients in Caco-2 and the Abs % (145 compounds with good data quality) was established, with a percentage of good relation greater than 82 %. A comparison with results derived from other three theoretical studies shown a quite satisfactory behavior of the present method. All these results shown that quadratic indices can successfully predict intestinal permeability and suggest that the proposed methodology will be a good tool for studying the oral absorption of drug candidates during the drug development process.
Novel topologic molecular descriptors have been applied to estimate the intestinal epithelial transport of drug in Caco-2 cell culture. Total and local quadratic indices used in this study were calculated by TOMOCOMDCARDD software. Two QSPR models, through a multiple linear regression, were obtained to predict the P [apical to basolateral (AP→BL) and basolateral to apical (BL→AP)]. A leave-one-out cross-validation procedure revealed that the regression models, had a fairly good predictability. Furthermore, others 18 drugs were selected as a test set in order to assess the predictive power of the models and the accuracy of the final prediction was similar to achieve for the data set. Besides, the use of both regression models, in a combinative way, is possible to predict the Permeability Directional Ratio (PDR, BL→AP/ AP→BL) value. Linear discriminant analysis was used to obtain a quantitative model that discriminates the high absorption compounds (P ≥8 x10-6cm/s) from those with moderate-poor absorption (P<8 x10-6cm/s). A data set of 134 diverse structure drugs and two series of drugs-like compounds (12 compounds) were used as training and test set, respectively. The obtained model classified correctly 81.13% of compounds with moderate-poor absorption properties and the 96.30% of compounds with high absorption, showing a global good classification of 90.30% in the training set. Internal and external validation processes to demonstrate the robustness and predictive power of the obtained model were carried out. In this sense, this model classified correctly 79.24% and 92.52% of compounds of the training set belonging to the moderate-poor and high absorption groups in the leave-one-out cross-validation procedure, respectively; being the global classification of 87.31%. Also this model shown an 87.5, 85.6, 84.7, 85.0, 85.3, 83.5, 84.1, 86.2, 85.9 and 85.9% of global good classification when n varied from 2 to 11 in the leave-n-out cross validation procedure. The model was stabilized around 85.9% when n was > 9. A data set of 7 HIV protease inhibitors (4 linear peptidomimetic and 3 new cyclic urea) and 5 new 6- fluoroquinolones derivatives was used as test set. The QSAR model classified correctly the 83.33% of compounds in this external prediction set. This approach permits us to obtain a good explanation of the experiment based on the molecular structural features, evidencing the main role of H-bonding and size properties in permeability process. Finally, the model developed was used in the virtual screening of 241 drugs with the percentage of human intestinal absorption (Abs %) values reported. A relationship between the predicted permeability coefficients in Caco-2 and the Abs % (145 compounds with good data quality) was established, with a percentage of good relation greater than 82 %. A comparison with results derived from other three theoretical studies shown a quite satisfactory behavior of the present method. All these results shown that quadratic indices can successfully predict intestinal permeability and suggest that the proposed methodology will be a good tool for studying the oral absorption of drug candidates during the drug development process.
Descripción
Palabras clave
Descriptores Moleculares, Molecular Descriptors