EMPIRICAL MODELING OF AN IN-VITRO ACTIVITY OF POLYCHLORINATED BIPHENYL CONGENERS AND MIXTURES

The goal of this research is to predict an in vitro activity of polych lorinated biphenyl (PCB) congeners and their mixtures and to describe the relationship between this activity and chemical structure. The tes t system used multiple PCB concentrations on each cell culture plate i n a repeated measures design, which improved precision for comparing b etween concentration levels. A weighted regression that accounted for this experimental design feature was used in fitting a nonlinear dose- response exponential model to the PCB concentration-activity data from an in vitro test system in which H-3-phorbol ester binding was measur ed in cerebellar granule cells exposed to different PCB congeners to t est for their effects on protein kinase C translocation. The model all owed for the minimum level to be less than control, a common slope, an d the estimation of the log of the concentration that produces an acti vity 50% above the control activity (E50) for 36 congeners and 3 comme rcial mixtures. Next, weighted logistic regression using a second orde r response model in the and Cl-ortho, Cl-para, and Cl-meta was used to relate the estimated log E50s to indicators of chemical structure. Th is model was preferred over models that might seem more mechanisticall y based because in internal validation, it attained a smaller PRESS st atistic (the sum of squares between all observed and predicted observa tions) than other models. Evidently, this second order model makes mor e efficient use of parameters than other models considered. Plots of t he predictions of the logistic second order response model versus log K-ow confirm the usual pattern that congeners with intermediate levels of log K-ow are the more active. The data of three commercial mixture s were included in this regression by assuming a common combination in dex (ratio of observed E50 to predicted E50, assuming dose addition). The logistic model suggests that congeners with one, two, or three chl orine substitutions at the ortho position are more active than other c ongeners. Also, congeners with log K-ow between 5.2 and 6.6 are genera lly more active. The estimated combination index indicated that the jo int action of PCB congeners in the three commercial mixtures was less than dose additive. The error sum of squares was significantly large, which may indicate a lack of fit of the logistic model. Empirical Baye s estimates (EBE) are weighted averages of model predictions and obser vations of E50s and can be better estimates than the fitted model when there is a lack of fit. The PRESS statistic for the EBE indicated lar ger prediction error than for the logistic model, but the EBE provided better estimates of commercial mixture E50s based on dose addition. T his may indicate that the logistic model is not incorporating all the information in the single congener data needed to predict mixtures.