Reply

We read carefully the letter from Dr. Landi et al. regarding our article published recently on polymorphic inducibility of CYP1B1 mRNA (1). Dr. Landi et al. insisted in their letter that both the concentrations of dioxins and the expression levels of CYP1B1 mRNA in human lymphocytes were log-normally distributed in 71 subjects from Seveso, Italy, which is a major point to be discussed in this reply. Our data in Japanese (1) are different from their observations. When we reexamined our original data set previously published (1), we found that the concentrations of dioxins were log-normally distributed (P = 0.31, Shapiro-Wilk test) as in their suggestion, whereas the expression levels of CYP1B1 mRNA (P = 0.02) in 72 Japanese subjects were not. The latter results indicated that an unknown factor(s) might be responsible for the expression levels of CYP1B1 mRNA in 72 Japanese subjects.

In our article (1), it was suggested that the inducibility of CYP1B1 mRNA, which was defined as the ratio of the amounts of CYP1B1 mRNA to the concentrations of dioxins in plasma, trimodally distributed in a probit-linear plot on the original scale. This individual variation in the inducibility of CYP1B1 mRNA, therefore, would be the factor for the nonnormal distribution of the expression level of CYP1B1 mRNA in Japanese subjects. Dr. Landi et al. suggested in their letter that the inducibility of CYP1B1 mRNA in 72 Japanese subjects might be a simple log-normal distribution as well as in their 71 subjects from Seveso. We do not know how they drew the curve. According to our calculation, the curve was not a clear simple log-normal distribution (Fig. 1). The curve certainly showed a trimodal distribution in a probit-log plot (this Fig. 1), as well as in a probit-linear plot [original Fig. 2 in our article (1)]. Therefore, the 72 Japanese subjects in our study could be divided into three groups according to the inducibility of CYP1B1 mRNA according to either normal- or log-normal distributions. In agreement with the previous report by Kellermann et al. (2), we found a similar trimodal distribution in the inducibility of CYP1B1 mRNA based on the calculated ratio of the amounts of CYP1B1 mRNA to the concentrations of dioxins in plasma. At present, we are evaluating the aryl hydrocarbon hydroxylase (AHH) activity of CYP1B1 in human lymphocytes. We are also carrying out studies for identifying the genetic factor(s) for trimodal distribution of CYP1B1 inducibility in human white cells in Japanese.