Teleost fish were selected as organisms for this study since lots of toxic effects data related to this fish are available. Teleost fish are key components of both marine and freshwater food chains and serve as one of the important source of food for human [14] [15]. Thus teleost fish have a key role to play in the trophic structure of aquatic ecosystems, vital for energy transfer between the trophic levels [16]. It would be significant to evaluate the effects of exposure time on such a key component of aquatic systems. Zebrafish, fathead minnow, brown trout, and rainbow trout are the organisms selected for study (Table 1). Toxicants having similar toxic mechanism are selected for this study and they included known endocrine disruptors (EE₂, BPA and NP) and natural hormones (E₂). Among these toxicants enter the environment through human activities [2] [17].

Toxic effects data related to teleost fish with these particular endocrine disruptors were obtained from the literature published (Table 1). These data sets include EC₅₀ for hepatic biomarkers (vtg genes or VTG protein) of teleost fish at different exposure times. The liver is an organ mainly characterized by metabolic function in the teleost fish [18] [19], so it is of significance to select hepatic biomarkers as research targets. The data obtained from the literature are in various units for concentration such as ng/L-μg/L. For consistency all units were converted into g/L. Similarly exposure time was also expressed in various units (hours and days) of time and all were converted into days (d). The NLT data of each organism was also obtained from literature published.

The linear RLE model [20] was developed with the use of the concept of reduction in life expectancy and the model equation is given below:

LC 50 = − a ln LT 50 + b (1)

Equation (1) is where LC₅₀ is the lethal concentration, LT₅₀ is the exposure time, NLT₅₀ is the normal life expectancy of the organism, d is a constant, a is 1/d and b is lnNLT₅₀/d.

Consistent with previous studies, LC₅₀ and EC₅₀ are frequently used in various tissues as toxicity endpoints, which is related to nominal concentration and logKow (octanol partition coefficient) [21]. Consequently, the relationship could be extended from the LC₅₀ to the EC₅₀ in the RLE model. Based on this extension a RLE model [5] for estrogenic effect was proposed and may be described by the equation given below:

EC 50 = − a ln ET 50 + b (2)

Equation (2) is where EC₅₀ is the effective concentration, ET₅₀ is the exposure time, NLT₅₀ is the normal life expectancy of the organism, and a and b are constants as prefiously defined.

Equation (3) is when EC₅₀ is zero the organism will have a normal life expectancy and it is related to the model constants as follows:

ln NLT 50 = b / a (3)

The data sets for each teleost fish were used to evaluate the relationship between EC₅₀ and lnET₅₀ with the RLE model expressed in Equation (2). Thus EC₅₀ was plotted against lnET₅₀ and linear regression analysis was used to obtain the regression equation and the correlation coefficient (R²) using Origin software (Microcal Software Inc., Northampton, Massachusetts, USA). The values of the slope (a) and intercept (b) were obtained from the regression equation (Table 2). These values were then used to obtain the calculated NLT of each organism by the use of Equation (3) (Table 3).

The plots of lnET₅₀ against EC₅₀ based on Equation (2) are shown in Figure 1,

The slope (a) and intercept (b) were obtained from the regression equations. ¹For EC₅₀ = −alnET₅₀ + b. ²For lnNLT₅₀ = b/a.

¹For lnNLT₅₀ = b/a.

Figure 2. These plots utilize data from Table 2 on teleost fish for EC₅₀ due to short-term exposure to endocrine disruptors. Plots for the E₂ and EE₂ are shown in Figure 1 and the corresponding plots for BPA and NP are shown in Figure 2. The characteristics of the relationships established using the regression equation can be used to compare the toxic effects of the E2, EE2, BPA and NP to teleost fish based on EC50 of hepatic biomarkers (Table 2). Table 2 indicates that slopes of E₂, EE₂, BPA and NP between −9.18E−05 and −3.91E−08 and R² value between 0.5530 and 0.9264. The slopes of the regression equations reflect the strength of the toxic effects on the various teleost fish. Since the range of slopes obtained for the E₂ and EE₂ is from −4.43E−08 to −3.91E−08 while for the BPA

and NP is from −9.18E−05 to −1.37E−05. The difference between the slopes is larger. Different toxic mechanisms involved with phenols and steroids may be responsible for this difference [22] [23]. Firstly, the observed EC₅₀ of E₂ and EE₂ are often at the level of ng/L, and the observed EC₅₀ of BPA and NP are often at the level of μg/L (Table 1). Secondly, the interaction potential of E₂ was the most potent, followed by BPA and NP [5]; the estrogenic activity of E₂ was much higher than that of BPA and NP [24]; the estrogenic potential of EE₂ was much higher than that of E₂, and the lowest observed effect concentration (LOEC) of E₂ and EE₂ is lower than BPA and NP [25]. And thus, the slops is comparatively higher for the BPA and NP than for the E₂ and EE₂, and the slops of E₂ and EE₂ is extremely small.

All the regression relationships have negative slopes (Table 2) which indicates that the EC₅₀ of hepatic biomarkers is related to the exposure time and declines as the exposure time increases. These results are in accord which were studying the time dependent effects of E₂, BPA and NP to zebrafish hepatic vtg1 gene [5]. Other researchers also reached the same conclusion while studying the toxic effects of organic pollutants to fish [12]. The transfer of compounds from water to fish is the first step in the development of toxic effects [26]. In fish the routes of compounds uptake are from gills, food and outer body surface but uptake of compounds takes place mainly via gills [27]. In Figure 3, compounds in water

can transfer from respiratory surfaces and gastrointestinal tract through circulatory fluid to liver in fish. In the process, some effect loss of the compounds, the compound continues to accumulate in the fish liver, and the increased metabolism of the hepatocytes and then the liver injury became exacerbated, these factors could be used to explain the decrease in the toxic effects over time. This was consistent with in previously studies [5] [28].

Calculated NLT is compared with reported NLT and calculated NLT differs from reported NLT (Table 3). Reported NLT and calculated NLT of teleost fish are in the range of 540 - 1400 d and 13 - 84 d respectively. The ratio of averages of reported NLT and calculated NLT is 1006/35 and the standard deviation is 311/33. The correlation coefficient (R²) obtained from the plots of reported NLT against calculated NLT was greater than 0.91 in a general accord (Table 3). The results show that the correlation between reported NLT and calculated NLT is better. The NLT introduces a fixed limiting point for a teleost fish and it is a reference point for the reduced life expectancy in fish exposed to the compound [12]. In Figure 1, Figure 2, from the cross point of regression line and x-axis to the roported NLT range, the trend changes of toxic effects based on the EC₅₀ of hepatic biomarkers can be predicted by the RLE model when no data available for these extended exposure times. In other words, the toxic effects of a compound can be described using the RLE model and be corrected the experimental data to a specific time point. It is noteworthy that the days from the each exposure time point to the roported NLT range is the days in which fish life expectancy reduced after exposure to the compound (Figure 1, Figure 2 x-axis). These analyses were also referred in the previously RLE model studies [13] [29].