ethyl 2-acetyl-4-methyltridec-2-enoate

Bioaccumulation: aquatic / sediment

Link to relevant study record(s)

Reference

Endpoint:

bioaccumulation in aquatic species: fish

Type of information:

experimental study

Adequacy of study:

supporting study

Study period:

Experimental : November - December, 2016
Final Report : 24 August, 2017

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

guideline study with acceptable restrictions

Qualifier:

according to guideline

Guideline:

other: OECD 319B - Determination of in vitro intrinsic clearance using Rainbow trout liver S9 subcellular fraction (RT-S9)

Version / remarks:

June, 2018

Deviations:

no

Principles of method if other than guideline:

Bioaccumulation refers to an increase of chemical concentration in an organism through all environmental sources like water, food and sediment. Thus, bioaccumulation can be considered as the net result of absorption, distribution, metabolism and excretion (ADME). The bioaccumulation potential of chemicals is scrutinised on a global basis by regulatory agencies in their risk assessment of chemicals. The Bioconcentration factor (BCF) is routinely used to evaluate the bioaccumulation potential of chemicals (i.e. the ratio of concentration of a substance in an organism like fish to the concentration of the water in a steady state). BCF values of chemicals are usually predicted with computation models that are based mainly on the hydrophobicity of the molecule which is either estimated or measured as n-octanol-water partition coefficient (Kow) of the chemical. The usefulness of the computer models is limited for the estimation of BCFs due to the broad variety of chemical classes and structures. Xenobiotic metabolism in fish is not well understood. Even models which consider e.g. metabolism using a screening-level quantitative structure-activity relationship (QSAR) model for estimation of biotransformation rate constants (kMet) can produce inaccurate estimates of bioaccumulation potential. Definitive determination of the BCF value involves the valid implementation of an OECD 305 fish bioconcentration test.

After absorption or ingestion of a substance by the fish, the chemical may be distributed to various tissues where it may be metabolized by enzymes. Metabolism is considered to be the dominant mechanism of elimination of hydrophobic substances which can significantly reduce their bioaccumulation potential. The first phase of biotransformation (Phase I) is usually the introduction of a polar group e.g. catalysed by Cytochrome P450 monooxygenases, which increases water solubility and renders it a suitable substrate for Phase II reactions. In Phase II reactions, xenobiotics are conjugated to endogenous substrates such as carbohydrates, amino acids, glutathione, or inorganic sulfate. The general trend of these metabolic transformation processes is the enzymatic conversion of lipophilic compounds to more polar hydrophilic metabolites which are usually less toxic and are normally readily excreted. The primary site of xenobiotic metabolism is typically the liver in most fish species and in mammalian systems. Therefore, determination of the in vitro metabolism of chemicals using liver cellular/subcellular fractions can provide an indication of their bioaccumulation potential. Furthermore, in vitro metabolism data can be used as an indication of the in vivo hepatic intrinsic clearance and may be utilized in in vitro to in vivo BCF extrapolation models.

The most commonly used in vitro methods to assess metabolism involve either liver S9 fractions or primary hepatocytes. A pre-validation study with trout liver S9 fractions has been done by a consortium under the coordination of HESI/ILSI (Health and Environmental Sciences Institute) and the protocol was published. The intra- and interlaboratory reliability of a cryopreserved trout hepatocyte assay was compared by three laboratories using six chemicals. A ring trial to compare intra- and interlaboratory reliability of trout liver S9 fractions and hepatocytes by several laboratories from academia, governmental institutions and industry which was led by HESI was recently performed to provide information for OECD guidelines (OECD Project 3.13). The study showed good intra- and interlaboratory reliability and resulted into the preparation of two draft OECD guidelines on the in vitro substrate depletion assays using trout liver hepatocytes or S9 fractions and a guidance document.

In order to determine hepatic metabolism of GR-86-6599, fish liver S9 fractions (Rainbow trout, Oncorhynchus mykiss) have been chosen as a model system. Fish liver S9 fractions contain both Phase I and Phase II enzyme systems. The primary objective of this study was to determine the in vitro intrinsic clearance (CLINT, IN VITRO) of GR-86-6599 in trout liver S9 fractions. Furthermore, in vitro metabolic rates were incorporated into an in vitro – in vivo extrapolation model to predict the BCF for a standardised fish (one that weighs 10g, has a 5% lipid content, and is living at 15 °C). The latest model developed by J. Nichols et al (version: “S9spreadsheet_Public_062713.xlsx”) was applied.

GLP compliance:

no

Radiolabelling:

no

Details on sampling:

Range-Finding Experiment :
Samples taken, reactions stopped and analysis performed at : 0, 30, 60, 90 and 120 minutes

Two Main Experiments :
Samples taken, reactions stopped and analysis performed at 1st Main Experiment : 0, 5, 10, 15, 20 and 30 minutes;
Samples taken, reactions stopped and analysis performed at 2nd Main Experiment : 0, 10, 20, 30, 45 and 60 minutes.

Reactions were stopped at the various sampling time-points by addition of acetonitrile (200 µL) containing methyl laurate (1 µM) as internal standard to the Hirschmann tubes.

Vehicle:

yes

Remarks:

0.1M Potassium Phosphate buffer (pH 7.8)

Details on preparation of test solutions, spiked fish food or sediment:

A stock solution of GR-86-6599 (10 mM) was prepared freshly in methanol and diluted in 0.1 M potassium phosphate buffer, pH 7.8 resulting in 10 µM solutions. Stock solutions of cofactors were prepared freshly in 0.1 M potassium phosphate buffer, pH 7.8. Alamethicin was dissolved in methanol (10 mg/mL; aliquots stored at -80°C) and freshly diluted in buffer (250 µg/mL).
Rainbow Trout liver S9 fractions were thawed on ice. All incubations were performed in potassium phosphate buffer at pH 7.8 (0.1 M) in Hirschmann glass tubes in duplicate or triplicate incubated at 12°C in a Thermomixer block with shaking capabilities (Ditabis Model MKR 23, 400 rpm). Active S9 fractions protein or heat inactivated protein as control (1 mg/mL) was preincubated on ice with alamethicin (final concentration: 25 µg/mL). Alamethicin is a pore-forming peptide antibiotic which permeabilises microsomal membranes and activates glucuronidation by allowing free transfer of UDPGA and glucuronide product across the membrane. After addition of cofactors for Phase I (NADPH, Nicotinamide adenine dinucleotide 2′-phosphate reduced) and Phase II enzymes (UDPGA, Uridine 5′-diphosphoglucuronic acid; PAPS, Adenosine 3′-phosphate 5′-phosphosulfate; GSH, reduced L-glutathione), the reaction was initiated by addition of the test substance. Final concentrations of cofactors, protein and test substance are listed in the table below. The detailed methodology is described by Johanning et al. and in the draft OECD guideline with the following exceptions: A final concentration of 0.5 mM GSH instead of 5 mM GSH was used. The final solvent concentration in the assay was 0.35% methanol.

Range-Finding :
In the range finding experiment, GR-86-6599 (1 µM) was incubated in the presence of 1 mg/mL active S9 protein and cofactors in duplicate for up to 120 minutes. As controls, the test chemical
(1 µM) was incubated in presence of heat inactivated S9 protein (1 mg/mL) and cofactors or with active S9 protein in absence of any cofactors added. Reactions were stopped at 0, 30, 60, 90 and 120 minutes incubation by addition of acetonitrile (200 µL) containing methyl laurate (1 µM) as internal standard to the Hirschmann tubes. Samples were extracted with MTBE (200 µL) in the same tubes by vortexing for 30 seconds, centrifuged to allow a better phase separation and separation of protein (Heraeus Fresco 17 Centrifuge, 13800  g, 5 min, 4°C) and subjected to GC-MS analysis.

In the two main experiments, GR-86-6599 (1 µM) was incubated in the presence of 1 mg/mL active S9 protein and cofactors in triplicate for up to 30 and 60 minutes, respectively, as described above. Reactions were stopped at time 0, 5, 10, 15, 20 and 30 minutes (1st main experiment) and at time 0, 10, 20, 30, 45 and 60 minutes (2nd main experiment). As control, the test chemical (1 µM) was incubated in the presence of heat inactivated S9 protein (1 mg/mL) and cofactors. Furthermore, incubations in the presence of active S9 protein and in absence of any cofactors added were carried out for 30 and 60 minutes, respectively. Reactions were stopped and samples extracted.

Test organisms (species):

Oncorhynchus mykiss (previous name: Salmo gairdneri)

Details on test organisms:

Rainbow Trout (Oncorhynchus mykiss) liver S9 fractions were prepared from four fish at the Veterinary Institute of the University Bern, Switzerland and stored at -80°C. The average body weight of the fish used for the preparation of S9 fractions (batch II) was 366 g. The enzymatic activity of the S9 fractions was characterized to determine the activity of CYP1A (Cytochrome P450 monooxygenase; EROD), and glutathione transferase (GST). The enzymatic activity of newly received S9 fractions was typically compared in house using testosterone, 7-hydroxycoumarin, and pyrene as test chemicals. In addition, Cyclohexyl Salicylate as internal fragrance reference molecule which is biotransformed by different enzyme systems (Phase I and Phase II) was used. S9 fractions were only used for metabolism studies if significant enzymatic conversion of the reference substance was observed. Aliquots of S9 fractions were prepared to prevent several thawing and freezing cycles to avoid inactivation of enzymes. Heat inactivated S9 fractions were prepared by heating 100 µL aliquots at 100°C using a Biometra Thermocycler and stored at -80°C.

Route of exposure:

aqueous

Justification for method:

other: Exposure of Rainbow Trout S9 liver cells to determine In vitro metabolic transformation rate of test item as an indication of bioconcentration potential.

Test type:

static

Water / sediment media type:

other: 0.1M Potassium Phosphate buffer (pH 7.8)

Total exposure / uptake duration:

>= 5 - <= 120 min

Details on test conditions:

Substrate depletion assay of GR-86-6599 in fish liver S9 Fractions :

Initially, a range finding experiment was performed to determine the optimal incubation times to be used in the main experiments.
A stock solution of GR-86-6599 (10 mM) was prepared freshly in methanol and diluted in 0.1 M potassium phosphate buffer, pH 7.8 resulting in 10 µM solutions. Stock solutions of cofactors were prepared freshly in 0.1 M potassium phosphate buffer, pH 7.8. Alamethicin was dissolved in methanol (10 mg/mL; aliquots stored at -80°C) and freshly diluted in buffer (250 µg/mL).
Rainbow Trout liver S9 fractions were thawed on ice. All incubations were performed in potassium phosphate buffer at pH 7.8 (0.1 M) in Hirschmann glass tubes in duplicate or triplicate incubated at 12°C in a Thermomixer block with shaking capabilities (Ditabis Model MKR 23, 400 rpm). Active S9 fractions protein or heat inactivated protein as control (1 mg/mL) was preincubated on ice with alamethicin (final concentration: 25 µg/mL). Alamethicin is a pore-forming peptide antibiotic which permeabilises microsomal membranes and activates glucuronidation by allowing free transfer of UDPGA and glucuronide product across the membrane. After addition of cofactors for Phase I (NADPH, Nicotinamide adenine dinucleotide 2′-phosphate reduced) and Phase II enzymes (UDPGA, Uridine 5′-diphosphoglucuronic acid; PAPS, Adenosine 3′-phosphate 5′-phosphosulfate; GSH, reduced L-glutathione), the reaction was initiated by addition of the test substance. Final concentrations of cofactors, protein and test substance are listed in the table below. The detailed methodology is described by Johanning et al and in the draft OECD guideline with the following exceptions: A final concentration of 0.5 mM GSH instead of 5 mM GSH was used. The final solvent concentration in the assay was 0.35% methanol.

Assay conditions using trout liver S9 fractions:
Addition of: Final concentration
90 µL K-Phosphate buffer, pH 7.8 (0.1 M) 0.1 M K-Phosphate buffer, pH 7.8
10 µL S9 protein (20 mg/mL) 1 mg/mL protein
Preincubation for 5 min at 12°C (400 rpm)
20 µL Alamethicin (250 µg/ml) 25 µg/mL Alamethicin
Incubation on ice for 15 min; pre-incubation at 12°C for 10 min
Addition of cofactors:
20 µL NADPH (10 mM) 1 mM NADPH
20 µL UDPGA (20 mM) 2 mM UDPGA
20 µL GSH (5 mM) 0.5 mM GSH
20 µL PAPS (1 mM) 0.1 mM PAPS
20 µL test substance in buffer (10 µM) 1 µM test substance
Final volume in Hirschmann tubes: 200 µL
Incubation at 12 ± 1 °C (400 rpm) for different time points

In the range finding experiment, GR-86-6599 (1 µM) was incubated in the presence of 1 mg/mL active S9 protein and cofactors in duplicate for up to 120 minutes. As controls, the test chemical
(1 µM) was incubated in presence of heat inactivated S9 protein (1 mg/mL) and cofactors or with active S9 protein in absence of any cofactors added. Reactions were stopped at 0, 30, 60, 90 and 120 minutes incubation by addition of acetonitrile (200 µL) containing methyl laurate (1 µM) as internal standard to the Hirschmann tubes. Samples were extracted with MTBE (200 µL) in the same tubes by vortexing for 30 seconds, centrifuged to allow a better phase separation and separation of protein (Heraeus Fresco 17 Centrifuge, 13800 x g, 5 min, 4°C) and subjected to GC-MS analysis.
In the two main experiments, GR-86-6599 (1 µM) was incubated in the presence of 1 mg/mL active S9 protein and cofactors in triplicate for up to 30 and 60 minutes, respectively, as described above. Reactions were stopped at time 0, 5, 10, 15, 20 and 30 minutes (1st main experiment) and at time 0, 10, 20, 30, 45 and 60 minutes (2nd main experiment). As control, the test chemical (1 µM) was incubated in the presence of heat inactivated S9 protein (1 mg/mL) and cofactors. Furthermore, incubations in the presence of active S9 protein and in absence of any cofactors added were carried out for 30 and 60 minutes, respectively. Reactions were stopped and the samples extracted.

Nominal and measured concentrations:

The depletion rates of test item in the active incubates were, as follows :

Range-Finding Experiment :
Starting concentration of GR-86-6599 was 1 microM
Time 0 100%
30 minutes 51.2% remaining of Time 0
60 minutes 27.2% remaining of Time 0
90 minutes 17.7% remaining of Time 0
120 minutes 11.7% remaining of Time 0

1st main Experiment :
Starting concentration of GR-86-6599 was 1 microM
Time 0 100%
5 minutes 57.7% remaining of Time 0
10 minutes 50.8% remaining of Time 0
15 minutes 41.3% remaining of Time 0
20 minutes 32.4% remaining of Time 0
30 minutes 24.4% remaining of Time 0

2nd main Experiment :
Starting concentration of GR-86-6599 was 1 microM
Time 0 100%
10 minutes 47.3% remaining of Time 0
20 minutes 30.4% remaining of Time 0
30 minutes 23.4% remaining of Time 0
45 minutes 17.4% remaining of Time 0
60 minutes 15.5% remaining of Time 0

Reference substance (positive control):

yes

Remarks:

Enzymatic turnover of the reference chemicals testosterone, 7-hydroxycoumarine, pyrene and Cyclohexyl salicylate was determined to validate the S9 liver fractions. The in vitro intrinsic clearance rates of the reference chemicals were similar.

Details on estimation of bioconcentration:

The in vitro intrinsic clearance (CLint, in vitro) was calculated from the log-transformed measured concentrations of parent compound as a function of time for GR-86-6599 from the two main experiments: 2.14 mL/h/mg protein and 2.24 mL/h/mg protein. There was a slow, but significant abiotic decrease observed in the negative control using inactive S9 protein. Hence, the in vitro intrinsic clearance rate was corrected: 1.97 mL/h/mg protein (2nd main experiment). Since no linear decrease was observed for the negative control in the 1st main experiment, no correction was done. However, these rates were similar to the preliminary rates estimated in the range finding experiment.

Prediction of BCFs for GR-86-6599 :
The metabolic turnover rates of GR-86-6599 in fish liver S9 fractions were similar in the range finding experiment with limited number of time points and duplicates and in the two main experiments with six time points and triplicates. The corrected reaction rate determined in the 2nd main experiment were used for the BCF prediction: 1.97/h.
The log Kow for GR-86-6599 had been measured using OECD TG 117 (3 individual peaks detected using HPLC under study conditions) as 5.3 (21.1% abundance), 5.9* (52.4% abundance) and 6.6* (26.5% abundance) (*indicative values, outside calibration domain; Givaudan report no.15-E364), thus BCFs were predicted based on the in vitro metabolic turnover rate comparing all three values. In the current study, two peaks were detected but were not completely separated under the analytical conditions applied. The partitioning based BCFs assuming no metabolism for GR-86-6599 based on the measured log Kow values (log Kow = 5.3, 5.9, and 6.6) were calculated as part of the in vitro-in vivo extrapolation model (version: “S9spreadsheet_Public_062713.xlsx”) by setting the reaction rate to zero: BCF= 8075 L/kg – 20,907 L/kg. Including the in vitro metabolism rate of GR-86-6599 in trout S9 fractions and other parameters, the predicted BCFs were 195 L/kg - 308 L/kg using an assumed fU of 1.0 and 1341 L/kg - 1934 L/kg assuming different binding to serum in vivo vs. in vitro (fU calc). These predicted BCFs based on in vitro enzymatic turnover rates were significantly lower compared to the BCF value calculated with an assumed turnover rate of 0.

Key result

Conc. / dose:

1 other: microM

Temp.:

12 °C

pH:

7.8

Type:

BCF

Value:

308 L/kg

Basis:

other: Based on fu = 1 (i.e. assumes no differential effect of binding to serum In vivo vs In vitro), LogKow set to 5.9 in IVIVE extrapolation model

Calculation basis:

other: Extrapolated from CLint using IVIVE Model (Nichols, J. W. et al., 2013 (version: "S9spreadsheet_Public_032713.xlsx")

Remarks on result:

other: Nichols, J. W. et al. Towards improved models for predicting bioconcentration of well-metabolized compounds by rainbow trout using measured rates of in vitro intrinsic clearance. Environmental Toxicology and Chemistry, 2013. 32(7): p. 1611 - 22

Conc. / dose:

1 other: microM

Temp.:

12 °C

pH:

7.8

Type:

BCF

Value:

285 L/kg

Basis:

other: Based on fu = 1 (i.e. assumes no differential effect of binding to serum In vivo vs In vitro), LogKow set to 5.3 in IVIVE extrapolation model

Calculation basis:

other: Extrapolated from CLint using IVIVE Model (Nichols, J. W. et al., 2013 (version: "S9spreadsheet_Public_032713.xlsx")

Remarks on result:

other: Nichols, J. W. et al. Towards improved models for predicting bioconcentration of well-metabolized compounds by rainbow trout using measured rates of in vitro intrinsic clearance. Environmental Toxicology and Chemistry, 2013. 32(7): p. 1611 - 22

Conc. / dose:

1 other: microM

Temp.:

12 °C

pH:

7.8

Type:

BCF

Value:

195 L/kg

Basis:

other: Based on fu = 1 (i.e. assumes no differential effect of binding to serum In vivo vs In vitro), LogKow set to 6.6 in IVIVE extrapolation model

Calculation basis:

other: Extrapolated from CLint using IVIVE Model (Nichols, J. W. et al., 2013 (version: "S9spreadsheet_Public_032713.xlsx")

Remarks on result:

other: Nichols, J. W. et al. Towards improved models for predicting bioconcentration of well-metabolized compounds by rainbow trout using measured rates of in vitro intrinsic clearance. Environmental Toxicology and Chemistry, 2013. 32(7): p. 1611 - 22

Conc. / dose:

1 other: microM

Temp.:

12 °C

pH:

7.8

Type:

BCF

Value:

1 341 L/kg

Basis:

other: Based on fu calc (i.e. assumes different binding to serum In vivo vs In vitro), LogKow set to 6.6 in IVIVE extrapolation model

Calculation basis:

other: Extrapolated from CLint using IVIVE Model (Nichols, J. W. et al., 2013 (version: "S9spreadsheet_Public_032713.xlsx")

Remarks on result:

other: Nichols, J. W. et al. Towards improved models for predicting bioconcentration of well-metabolized compounds by rainbow trout using measured rates of in vitro intrinsic clearance. Environmental Toxicology and Chemistry, 2013. 32(7): p. 1611 - 22

Conc. / dose:

1 other: microM

Temp.:

12 °C

pH:

7.8

Type:

BCF

Value:

1 934 L/kg

Basis:

other: Based on fu calc (i.e. assumes different binding to serum In vivo vs In vitro), LogKow set to 5.9 in IVIVE extrapolation model

Calculation basis:

other: Extrapolated from CLint using IVIVE Model (Nichols, J. W. et al., 2013 (version: "S9spreadsheet_Public_032713.xlsx")

Remarks on result:

other: Nichols, J. W. et al. Towards improved models for predicting bioconcentration of well-metabolized compounds by rainbow trout using measured rates of in vitro intrinsic clearance. Environmental Toxicology and Chemistry, 2013. 32(7): p. 1611 - 22

Conc. / dose:

1 other: microM

Temp.:

12 °C

pH:

7.8

Type:

BCF

Value:

1 548 L/kg

Basis:

other: Based on fu calc (i.e. assumes different binding to serum In vivo vs In vitro), LogKow set to 5.3 in IVIVE extrapolation model.

Calculation basis:

other: Extrapolated from CLint using IVIVE Model (Nichols, J. W. et al., 2013 (version: "S9spreadsheet_Public_032713.xlsx").

Remarks on result:

other: Nichols, J. W. et al. Towards improved models for predicting bioconcentration of well-metabolized compounds by rainbow trout using measured rates of in vitro intrinsic clearance. Environmental Toxicology and Chemistry, 2013. 32(7): p. 1611 - 22.

Details on results:

The in vitro intrinsic clearance (CLint, in vitro) was calculated from the log-transformed measured concentrations of parent compound as a function of time for GR-86-6599 from the two main experiments: 2.14 mL/h/mg protein and 2.24 mL/h/mg protein. There was a slow, but significant abiotic decrease observed in the negative control using inactive S9 protein. Hence, the in vitro intrinsic clearance rate was corrected: 1.97 mL/h/mg protein (2nd main experiment). Since no linear decrease was observed for the negative control in the 1st main experiment, no correction was done. However, these rates were similar to the preliminary rates estimated in the range finding experiment.

It was used as inputs into an in vitro - in vivo extrapolation model to predict the BCF using the measured log Kow value 5.9 for the major peak. The predicted BCF (BCFTOT) was 308 L/kg wet weight using an assumed fU = 1.0, i.e. no effect of differential binding to serum and 1934 L/kg wet weight assuming different binding to serum in vivo vs. in vitro (fU calc). BCFs were also predicted using the measured log Kow values determined for two minor peaks of GR-86-6599 (log Kow = 5.3 and 6.6, respectively). The predicted BCFs were 285 L/kg wet weight and 195 L/kg wet weight using an assumed fU = 1.0 and 1548 L/kg wet weight and 1341 L/kg wet weight using fU calc.




Benchmarking of the in Vitro Model Versus Other Fragrance Molecules with Measured in Vivo BCF :

In order to evaluate the validity of the in vitro – in vivo extrapolation model in vivo and in vitro data were compared for fragrance molecules with higher log Kow values (log Kow  3.9-6.0). The in vitro reaction rates of Herbanate, Peonile, d-Damascone, Ebanol, Polysantol,  Sandela, Cashmeran, Cyclohexyl Salicylate, Agrumex, Javanol, Nectaryl, Musk Xylene, Piconia (Isolonigfolanone), Ambrox, Spirogalbanone, Opalal, Iso E Super, Methyl Cedryl Ketone, Cyclohexadecenone, and Galaxolide were determined in trout liver S9 fractions. In addition, Pentachlorobenzene was used as a known bioaccumulative reference chemical. BCFs were predicted using the recently refined in vitro – in vivo extrapolation model by Nichols et al and compared to the BCF values which had been measured in fish.

There is good correlation between the refined BCF estimates calculated with the latest in vitro – in vivo extrapolation model version (“S9spreadsheet_Public_062713.xlsx”) and measured BCFs. Furthermore, a specifically good correlation between predicted BCFs and measured BCFs was observed using an assumed fU of 1.0 for the model (i.e. no effect of differential binding to serum) which is in agreement with other studies. Especially for the fragrance molecules with higher log Kow values (log Kow 4.8-5.9) like Javanol, Nectaryl, Piconia, Ambrox, Spirogalbanone, Opalal, Iso E Super, Cyclohexadecenone, and Sandela (calculated with log Kow 5.9), there was a better correlation of the predicted BCFs calculated with the in vitro-in vivo extrapolation model using an assumed fU of 1.0.

Escher et al. determined the binding (i.e. fU) for four hydrophobic chemicals (log Kow 4.88 – 5.76) in fish plasma and S9 fractions. The use of experimentally derived fU values resulted in predicted BCFs that were much higher than the measured values. In contrast, setting fU = 1.0 (i.e. the approach by Cowan-Ellsberry et al.), resulted in a better prediction of the measured BCFs. The reason for this apparent discrepancy is not known. One possible explanation may be that the assumption of the extrapolation model that only freely dissolved chemicals are available for metabolic turnover is incorrect. Chemicals bound to proteins may desorb rapidly and thus contribute to the metabolic turnover of the chemicals. Presently, it seems to be reasonable according to Nichols et al. to use the two different binding assumptions to estimate upper and lower limits on hepatic clearance. However, when intrinsic clearance values are moderate to low, these two competing assumptions result in substantial differences in calculated hepatic clearance rates and predicted BCFs. This is the case for GR-86-6599. The intrinsic clearance rate is moderate and the measured log Kow values are relatively high (5.3-6.6). Therefore, the difference in predicted BCFs calculated for both assumptions of fU are rather high (~5- to ~7-fold).

Validation of enzymatic activity of rainbow trout liver S9 fraction (prepared at University of Bern, batch II) with reference chemicals. Enzymatic activity of the trout liver S9 batch used in this study was characterised using Testosterone, 7-hydroxycoumarin (glucuronosyl transferase (UGT; UDPGA as single cofactor) and sulfotransferase (SULT; PAPS as single cofactor) activity, pyrene and Cyclohexyl Salicylate :

 

In vitro instrinsic clearance rates of S9 fractions batch II with reference chemicals.

 

Test chemical	Test concentration (µM)	In vitro intrinsic clearance rate (mL/h/mg protein)	In vitro intrinsic clearance rate (mL/h/mg protein) AVG
Testosterone	1	1.35 1.15	1.25
7-Hydroxycoumarin- UGT	1	62.4 65.74	64.07
7-Hydroxycoumarin- SULT	1	2.32 4.10	3.21
Pyrene	0.025	30.87 27.16	29.02
Cyclohexyl Salicylate	1	25.32 28.42	26.87

 

Incubation mixtures contained active S9 protein (1 mg/mL), all cofactors (except for UDPGA or PAPS as single cofactor with 7-hydroxycoumarin) and 1 µM test chemical

Validity criteria fulfilled:

yes

Conclusions:

Moderate turnover of GR-86-6599 was observed over 60 minutes. GR-86-6599 demonstrated a metabolic turnover of 85% of the starting concentration within a 60 minute exposure period.

The BCF values calculated using the Nichols et al. IVIVE Excel spreadsheet, assuming a LogPow input value of 5.9 (LogPow determined for the main component, and the LogPow which results in the most conservative BCF values) are, as follows :

BCF = 308 L/kg (fu = 1.0, i.e. assumes no effect of differential binding to serum In vivo vs In vitro)

BCF = 1934 L/kg (fu calc, i.e. assumes differential binding to serum In vivo vs In vitro)

Further research on fragrance molecules has demonstrated that there is a better correlation of the predicted BCFs calculated with the in vitro-in vivo extrapolation model using an assumed fU of 1.0, particularly for the fragrance molecules with higher log Kow values (log Kow 4.8-5.9).

As such, the IVIVE calculated BCF value of 308 L/kg (wet wt.) is considered the Key Result from this study.

Description of key information

Moderate turnover of GR-86-6599 was observed over 60 minutes. GR-86-6599 demonstrated a metabolic turnover of 85% of the starting concentration within a 60 minute exposure period. There was a significantly slower decrease of GR-86-6599 observed in presence of heat inactivated S9 protein (25 % decrease within a 60 minute exposure period). A similar, slow decrease of GR-86-6599 was observed with active S9 protein in absence of any cofactors added (27% decrease within a 60 minute exposure period; 2^nd main experiment).

The in vitro intrinsic clearance (CL_{int, in vitro}) was calculated from the log-transform measured concentrations of the parent compound as a function of time and corrected for abiotic decrease: 1.97 mL/h/mg protein (2^nd main experiment). It was used as inputs into an in vitro - in vivo extrapolation model to predict the BCF using the measured log K_ow value 5.9 for the major peak. The predicted BCF (BCF_TOT) was 308 L/kg wet weight using an assumed f_U = 1.0, i.e. no effect of differential binding to serum and 1934 L/kg wet weight assuming different binding to serum in vivo vs. in vitro (f_U calc). BCFs were also predicted using the measured log K_ow values determined for two minor peaks of GR-86-6599 (log K_ow = 5.3 and 6.6, respectively). The predicted BCFs were 285 L/kg wet weight and 195 L/kg wet weight using an assumed f_U = 1.0 and 1548 L/kg wet weight and 1341 L/kg wet weight using f_U calc.

Comparison of existing data for fragrance speciality chemicals possessing both in vivo and in vitro fish BCF data sets, indicate that there is a significantly greater correlation between the derived results when a f_U of 1.0 is employed for the calculations, and, in particular, for compounds with high logK_ow values.

Key value for chemical safety assessment

BCF (aquatic species):

308 L/kg ww