3-methyl-5-(2,2,3-trimethyl-3-cyclopenten-1-yl)pent-4-en-2-ol

Bioaccumulation: aquatic / sediment

Description of key information

The bioaccumulation potential of Ebanol has been assessed using a weight of evidence approach. An in vitro study showing metabolism supports a lower fish BCF than predicted from Kow. Methods corroborated with analogue data and models that take biotransformation into account give similar results. The BCF estimates range from 102 to 366 L/kg. A cautious approach has been taken and the highest, reliable estimate has been chosen as the key value for chemical safety assessment. 

Key value for chemical safety assessment

BCF (aquatic species):

366 L/kg ww

Additional information

Reliable measured aquatic bioaccumulation data is not available for Ebanol. 

 

According to REACH Annex IX, information on bioaccumulation in aquatic species, preferable fish, is required for substances manufactured or imported in quantities of 100 t/y or more unless the substance has a low potential for bioaccumulation (for instance a log Kow ≤ 3). However,REACH Annex XI encourages the use of alternative information before a new vertebrate test, including fish, is conducted.

 

Therefore, a weight of evidence approach has been used to assess the bioaccumulation potential of Ebanol. Bioconcentration factors in fish have been estimated using:

 

i) three commonly used and scientifically valid QSARs based on Kow

ii) a category read-across approach using the OECD QSAR Toolbox and

iii) in vitro metabolism data to refine the estimation of a partitioning based BCF.

 

The methods and estimated BCF values are summarised in the following table.

 

Method	log Kowof training set / category	Chemical domain	Estimated BCF (L/kg)
Linear regression, Veithet al. (1979)	1 to 6.9	neutral, non-ionized chemicals (based on 59 independent measurements, corresponding to 56 chemicals).	741
BCFBAF (v3.01) regression based QSAR	1 to 7.0	Non-ionic (based on 396 chemicals)	274
BCFBAF (v3.01) Arnot-Gobas BCF & BAF method.	0.3 to 8.7	Wide applicability to non-ionic organics substances. Calibrated to large BCF and BAF database.	751 (upper trophic, assuming a biotransformation rate of zero)   190 (upper trophic, including biotransformation rate estimates)   321(middle trophic, including biotransformation rate estimates)   366 (lower trophic, including biotransformation rate estimates)
QSAR Toolbox v2.2 Category definition and read-across.	3.4 to 5.5	11 Non-ionic chemicals containing the bioaccumulation alerts: Aliphatic alcohol, alkenyl hydrogen, Carbon with 4 single bonds & no hydrogens, CH - [cyclic], -CH- [linear], -CH2 -[cyclic], Methyl [-CH3].	240
In vitro – in vivo extrapolation model (Nichols et al)	-	-	793 (Partitioning based BCF, assuming no metabolism)   275 (Refined BCF for Ebanol peak 1, fu calculated)   267 (Refined BCF for Ebanol peak 2, fu calculated)     102 (Refined BCF for Ebanol peak 1, fu = 1)    102 (Refined BCF for Ebanol peak 1, fu = 1)

 

Ebanol is a non-ionic organic substance with a measure log Kow of 4.2. Therefore, it falls within the general applicability domain of all the used models. The experimentally determined high quality Log Kow value was used as an input term in the BCF estimation models.

 

There is evidence that Ebanol is metabolised in fish. In an in vitro metabolism assay using trout liver S9 fractions, Ebanol demonstrated a metabolic turnover of 73-75% of the starting concentration within a 60 minute exposure period. Furthermore, Ebanol is readily biodegradable and it is generally accepted that readily biodegradable chemicals have a higher probability of being metabolised in exposed organisms to a significant extent than less biodegradable chemicals.  Thus Ebanol is expected to have a lower BCF than that predicted using models that are based only on log Kow.

 

The Arnot-Gobas BCF QSAR (assuming a biotransformation rate of zero) and the partitioning based model (assuming no metabolism) were developed to fit upper bound BCF observations. The linear model developed by Veith et al (1979) is also considered to predict maximum log BCF values. It is based on a limited data set of 56 chemicals, which are not expected to be metabolised. In addition, three chemicals from the original data set of 62 were omitted from the correlation because bioconcentration tests indicated that they bioconcentrate very little despite their comparatively large log Kow values. Since there is evidence that Ebanol is metabolised in fish, the estimated BCF values using these 3 models (751, 793 and 741 L/kg respectively) are considered worst-case and unrealistic.

 

The BCFBAF v3.01 regression model (for non-ionic substances with a log Kow of 1.0 to 7.0) was developed from a large data set of 396 diverse chemicals. As such it has a tendency to arrive at an “average” BCF value. Log BCF is estimated from log Kow and a series of correction factors if applicable. No correction factors are applied to Ebanol. Therefore the estimated BCF of 274 L/kg for Ebanol is based only on log Kow. The model training set contains three non-halogenated aliphatic alcohols which are considered structural analogues of Ebanol. For all three analogues the predicted BCF is higher than the experimental BCF. The over-prediction for this group of chemicals may be attributed to the alcohol functional group, which may be a site for metabolic attack. Taking into account the aforementioned analogue data, the predicted BCF value of 274 L/kg for Ebanol is considered reliable and conservative.

The BCFBAF v3.01 non-ionic training set, which is a quality reviewed BCF database, was downloaded from http://esc.syrres.com/interkow/EpiSuiteData.htm and imported into the OECD QSAR Toolbox v2.2. A sub-category of 11 chemicals was selected based on the hypothesis of “common bioaccumulation metabolism alerts”. The read-across was based on the 5 closest analogues in the log Kow descriptor space. The estimated BCF value of 240 L/kg for Ebanol is based on interpolation between category members and is considered reliable and a reasonable realistic estimate.  

Since there is evidence that Ebanol is metabolised in fish, model predictions (including biotransformation rates) are also considered relevant. The Arnot-Gobas method predicts the whole body primary biotransformation rate (kM) for fish using a multiple-linear regression derived equation. The prediction is considered reliable for Ebanol as it falls within the descriptor domain and structural fragment domain of the kM model. The predicted kM is converted to a kM value for the typical body size of an upper trophic, middle trophic and lower trophic fish in the respective BCF algorithms. The resulting refined BCF estimates for Ebanol are 190, 321 and 366 L/kg respectively.

In the “in vitro - in vivo” extrapolation model developed by Nichols, the rate of substance depletion from an in vitro metabolism assay is used to estimate the kM term. A binding term, fu, is used to correct for the difference in free chemical concentration between blood and the in vitro system. Two assumptions are possible: 1) fu can be calculated as the ratio of predicted free fractions in plasma and in the in vitro system using logKow-based algorithms, or 2) binding in vitro and in vivo can be assumed to be equal ( fu = 1.0). The refined BCF has been estimated using both approaches. A GC-MS method was used to monitor the disappearance of Ebanol in fish liver S9 fractions. The test item gives a chromatographic profile consisting of two major peaks which were quantified separately. The refined BCF was estimated for both peaks. The Refined BCF estimates were 102 (peak 1 and 2, fu = 1), 267 (peak 2, fu calculated) and 275 L/kg (peak 1, fu calculated).

Based on the above information, assessed together as part of an overall WoE, it is concluded that there is no need for further investigation of aquatic bioaccumulation with fish.

For classification purposes, an experimentally derived high quality Log Kow value is suitable when a measured BCF on an aquatic organism is not available. Ebanol has a measured log Kow of 4.2. This exceeds the CLP cut-off value of ≥ 4 and the DSD value of ≥ 3. Thus Ebanol is considered to have the potential to bioconcentrate for classification purposes. Predicted BCFs are not relevant for classification purposes.

For PBT/vPvB assessment purposes, predicted BCF estimates and log Kow may be used to provisionally assess bioaccumulation potential. Ebanol has a measured log Kow of 4.2, which is less than the B screening criteria of ≤ 4.5. Estimated “worst-case” maximum fish BCF values are in the range of 741-793 L/kg. This is well below the B definitive criterion of BCF > 2000 L/kg.

There is in vitro evidence that Ebanol is metabolised by fish supporting a lower BCF than predicted using models based only on log Kow. Methods corroborated with analogue data and models that take biotransformation into account give similar results. The BCF estimates range from 102 to 366 L/kg. The highest value of 366 L/kg has been chosen as a relevant and reliable conservative estimate for the purposes of risk assessment.