Reaction mass of 1,4-bis(methylamino)anthraquinone and 1,4-bis[(2-ethylhexyl)amino]anthraquinone and 1-[(2-ethylhexyl)amino]-4-(methylamino)anthraquinone and 9,10-Anthracenedione, 1,4-bis(pentylamino)-, branched and linear and 9,10-Anthracenedione, 1-(methylamino)-4-(pentylamino)-, branched and linear and 9,10-Anthracenedione, 1-[(2-ethylhexyl)amino]-4-(pentylamino)-, branched and linear

Biodegradation in water: screening tests

Link to relevant study record(s)

Reference

Endpoint:

biodegradation in water: ready biodegradability

Type of information:

(Q)SAR

Adequacy of study:

key study

Study period:

2017

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

accepted calculation method

Justification for type of information:

1. SOFTWARE: EPI Suite, v4.10

2. MODEL: BIOWIN, v4.10, issued by US Environmental Protection Agency in 2010 (US EPA, 2012; Boethling et al., 2004; Howard et al., 1992). BIOWIN contains the following six separate models relevant for assessing ready biodegradability:
Biowin1 = linear probability model
Biowin2 = nonlinear probability model
Biowin3 = expert survey ultimate biodegradation model
Biowin4 = expert survey primary biodegradation model
Biowin5 = Japanese MITI (Ministry of International Trade and Industry) linear model
Biowin6 = Japanese MITI (Ministry of International Trade and Industry) nonlinear model

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL:

i. C.I. Solvent Blue 98 (3Amine) Component 1: O=C2c1ccccc1C(=O)c3c2c(ccc3NC)NC
ii. C.I. Solvent Blue 98 (3Amine) Component 2: O=C2c1ccccc1C(=O)c3c2c(ccc3NCCCCCCCCC)NC
iii. C.I. Solvent Blue 98 (3Amine) Component 3: O=C2c1ccccc1C(=O)c3c2c(ccc3NCCCCCCCCC)NCCCCCCCCC

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL

a. Endpoint: Ready biodegradability
i. Biowin1: Fast biodegradation probability
ii. Biowin2: Fast biodegradation probability
iii. Biowin3: Ultimate biodegradation timeframe
iv. Biowin4: Primary biodegradation timeframe
v. Biowin5: Ready biodegradation probability
vi. Biowin6: Ready biodegradation probability
b. Dependent variable: Biodegradation probability or degradation timeframe score

Biodegradability estimates are based upon fragment constants that were developed using multiple linear or non-linear regression analyses against experimental data from curated biodegradation databases or expert survey results.

Model equations are described below:
i. Biowin1: Linear Probability Model and Biowin2: Nonlinear Probability Model

The linear equation for Biowin1 is defined as: Yj=ao +a1* f1+ a2*f2+...+a36*f36+ am*MWj+ej; where Yj is the probability that chemical j will biodegrade fast (based on experimental data); fn is the number of the nth substructure in the jth chemical; a0 is the intercept; an is the regression coefficient for the nth substructure; Mwj the molecular weight; am the regression coefficient for Mw and ej the error term. The training set was used to select 36 chemical fragments plus a molecular weight fragment that have a potential effect on biodegradability. A matrix of 295 chemicals by 37 fragments was formulated. The number of each fragment occurring in each chemical was entered into the matrix along with the chemical's molecular weight. A biodegradation matrix of dimensions 295 chemicals by 1 evaluation was also formulated. The evaluation was either 1 (the chemical biodegrades fast; probability of 1.0) or 0 (the chemical does not biodegrade fast; probability of 0.0). The matrices were then subjected to multiple linear regression analyses to determine probability coefficients for each fragment. The coefficients were then used to determine the biodegradation probability for each compound. A biodegradation probability greater than 0.5 is considered as "biodegrades fast". A biodegradation probability less than 0.5 is considered as "does not biodegrades fast".

A logistic equation is used as the basis for the nonlinear model (Biowin2), according to the following expression: Yj = exp(a0+a1*f1+a2*f2+...a36*f36 +am*Mwj)/ 1 + exp(a0+a1*f1+a2*f2+...a36*f36 +am*Mwj). The regression coefficients are determined by fitting the above nonlinear equation with experimental data in the training set. The fragments and training set compounds are the same as those for Biowin1.

ii. Biowin3: expert survey ultimate biodegradation model and Biowin4: expert survey primary biodegradation model

Primary biodegradation is the transformation of a parent compound to an initial metabolite. Ultimate biodegradation is the transformation of a parent compound to carbon dioxide and water, mineral oxides of any other elements present in the test compound, and new cell material. It is normally determined experimentally by measuring carbon dioxide production, oxygen uptake or disappearance of dissolved organic carbon attributable to the test substance. The models are based upon a survey of 17 biodegradation experts conducted by EPA, in which the experts were asked to evaluate 200 compounds in terms of the time required to achieve ultimate and primary biodegradation in a typical or "evaluative" aquatic environment (Boethling et al., 1994). In the survey, each expert rated the ultimate and primary biodegradation of each compound on a scale of 1 to 5. The ratings correspond to the following time units: 5 - hours; 4 - days; 3 - weeks; 2 - months; 1 - longer. It should be noted that the ratings are only semi-quantitative and are not half-lives. Thus for example if the averaged expert rating for ultimate degradation of a compound is 2.5, it means the experts considered that the compound would biodegrade completely in a time frame somewhere between "a matter of weeks" and "a matter of months", with no exact time or half-life being applied. The ratings for each compound were averaged to obtain a single value for modeling. Matrices were then formulated for both ultimate and primary biodegradation using the same 36 fragments and molecular weight parameter as used in the Linear/Non-Linear Models (Biowin1 and 2). Finally, linear regressions were performed on the matrices using the averaged expert ratings as the solution matrices. The regression equation is the same as for Biowin1.

The following explains how the numerical ratings calculated by the models for a given compound are used to assign the words that are given in parentheses in the results.

Calculated rating (value from model) degradation timeframe
>4.75 - 5 hours
>4.25 - 4.75 hours to days
>3.75 - 4.25 days
>3.25 - 3.75 days to weeks
>2.75 - 3.25 weeks
>2.25 - 2.75 weeks to months
>1.75 - 2.25 months
<1.75 recalcitrant

iii. Biowin5: Japanese MITI (Ministry of International Trade and Industry) linear model and Biowin6: Japanese MITI (Ministry of International Trade and Industry) nonlinear model

A total dataset of 884 chemicals from Japanese MITI (Ministry of International Trade and Industry) MITI-1 test (OECD 301C) was compiled to derive the fragment probability values that are applied in this MITI Biodegradability method. The dataset consists of 385 chemicals that were critically evaluated as "readily degradable" and 499 chemicals that were critically evaluated as "not readily biodegradable". A journal article describing the development and methodology of the MITI Biodegradation Probability Model is published (Tunkel et al., 2000). The 884 compound dataset was divided into a training dataset (589 compounds) and a validation dataset (295 compounds). A compound that is "readily degradable" was assigned a numeric value of 1 and "not readily degradable" was assigned a numeric value of 0 (0 to 1 is the full probability range). The basic approach for deriving the fragment values is very similar to the approach used for Biowin1 and Biowin2. The majority of fragments in the Biowin5 and Biowin6 (MITI models) are identical to fragments in the models for Biowin1 and Biowin2. MITI models incorporate various changes or additional fragments. For example, to provide a fuller characterization of alkyl chain length and branching, the C4 terminal alkyl group fragment was replaced with a fragment set consisting of -CH3, -CH2 (both linear and ring types), -CH (both linear and ring types), and -C=CH (alkenyl hydrogen). The final MITI models contain 42 fragments and molecular weight as independent variables. A biodegradation probability greater than 0.5 is considered as "Readily biodegradable. A biodegradation probability less than 0.5 is considered as "not readily biodegradable".

- Appropriate measures of goodness-of-fit and robustness and predictivity: Estimation results from Biowin1 through Biowin6 are evaluated in a weight-of-evidence approach in determining the ready biodegradability of an unknown compound in question. BIOWIN determines that a compound is readily biodegradable (i.e., Ready Biodegradability Prediction: YES) if Biowin3 (ultimate survey model) result of biodegradation timeframe is "weeks" or faster (i.e. days, days to weeks, or weeks) and Biowin5 (MITI linear model) result of ready biodegradation probability is greater than or equal to 0.5. A compound is determined to be not readily biodegradable (i.e., Ready Biodegradability Prediction: No) if that criteria is not met.

- Mechanistic interpretation: BIOWIN estimates the probability of rapid aerobic biodegradation of an organic compound in the presence of mixed populations of environmental microorganisms. The biodegradation of organic compounds by mixed populations of environmental microorganisms may involve multiple enzymes and multiple species of microorganisms for carrying through the ultimate biodegradation. These models do not differentiate
specific microbial metabolic mechanism.

5. APPLICABILITY DOMAIN

- Descriptor domain:
Biowin1 and Biowin2: 31 ≤ MW ≤ 698
Biowin3 and Biowin4: 53 ≤ MW ≤ 698
Biowin5 and Biowin6: 30 ≤ MW ≤ 959
C.I. Solvent Blue 98 (3Amine) Component 1 has a molecular weight of 266.29, which is in domain for all submodels of BIOWIN v4.10.
C.I. Solvent Blue 98 (3Amine) Component 2 has a molecular weight of 378.51, which is in domain for all submodels of BIOWIN v4.10.
C.I. Solvent Blue 98 (3Amine) Component 3 has a molecular weight of 490.72, which is in domain for all submodels of BIOWIN v4.10

- Structural and mechanistic domains: Structural fragment domain: in domain: The unknowns consist of Linear C4 terminal chain, Aromatic amine and Ketone. These structure features are all represented in the training set of the models.

- Mechanism domain: NA

-Metabolic domain: BIOWIN estimates the probability of rapid aerobic biodegradation of an organic compound in the presence of mixed populations of environmental microorganisms. The biodegradation of organic compounds by mixed populations of environmental microorganisms may involve multiple enzymes and multiple species of microorganisms for carrying through the ultimate biodegradation. These models do not differentiate specific microbial metabolic mechanism.

- Similarity with analogues in the training set: Data on structural analogues are shown in Table 1 for assessing the performance of the QSAR model. The biodegradability of these compounds is reliably predicted by these models.

6. ADEQUACY OF THE RESULT

The QSAR models for BIOWIN v4.10 are considered reliable based on the numbers (N) of analogues in the training set, and the accuracy of the prediction for compounds in the training set and validation set.

Qualifier:

according to guideline

Guideline:

other: QSAR

GLP compliance:

no

Specific details on test material used for the study:

C.I. Solvent Blue 98 (3Amine) Component 1 MW: 266.29
C.I. Solvent Blue 98 (3Amine) Component 2 MW: 378.51
C.I. Solvent Blue 98 (3Amine) Component 3 MW: 490.72

Input for prediction: SMILES Code:
i. C.I. Solvent Blue 98 (3Amine) Component 1: O=C2c1ccccc1C(=O)c3c2c(ccc3NC)NC
ii. C.I. Solvent Blue 98 (3Amine) Component 2: O=C2c1ccccc1C(=O)c3c2c(ccc3NCCCCCCCCC)NC
iii. C.I. Solvent Blue 98 (3Amine) Component 3: O=C2c1ccccc1C(=O)c3c2c(ccc3NCCCCCCCCC)NCCCCCCCCC

Key result

Parameter:

other: Each of the three components of C.I. Solvent Blue 98 (3Amine) is estimated to be not readily biodegradable.

Remarks on result:

not readily biodegradable based on QSAR/QSPR prediction

C.I. Solvent Blue 98 (3Amine) Component 1:

i. Biowin1(Fast biodegradation probability by linear model): 0.17 (Does not biodegrade fast)

ii. Biowin2 ( Fast biodegradation probability by nonlinear model): 0.004 (Does not biodegrade fast)

iii. Biowin3 (Ultimate biodegradation timeframe): 2.30 (Weeks-Months)

iv. Biowin4 (Primary biodegradation timeframe): 3.20 (Weeks)

v. Biowin5 (Ready biodegradation probability): -0.11 (Does not biodegrade fast)

vi. Biowin6 (Ready biodegradation probability): 0.0056 (Does not biodegrade fast)

C.I. Solvent Blue 98 (3Amine) Component 2:

i. Biowin1(Fast biodegradation probability by linear model): 0.22 (Does not biodegrade fast)

ii. Biowin2 ( Fast biodegradation probability by nonlinear model): 0.0053 (Does not biodegrade fast)

iii. Biowin3 (Ultimate biodegradation timeframe): 2.35 (Weeks-Months)

iv. Biowin4 (Primary biodegradation timeframe): 3.31 (Days-Weeks)

v. Biowin5 (Ready biodegradation probability): -0.05 (Does not biodegrade fast)

vi. Biowin6 (Ready biodegradation probability): 0.0067 (Does not biodegrade fast)

C.I. Solvent Blue 98 (3Amine) Component 3:

i. Biowin1(Fast biodegradation probability by linear model): 0.28 (Does not biodegrade fast)

ii. Biowin2 ( Fast biodegradation probability by nonlinear model): 0.0067 (Does not biodegrade fast)

iii. Biowin3 (Ultimate biodegradation timeframe): 2.40 (Weeks-Months)

iv. Biowin4 (Primary biodegradation timeframe): 3.42 (Days-Weeks)

v. Biowin5 (Ready biodegradation probability): 0.01 (Does not biodegrade fast)

vi. Biowin6 (Ready biodegradation probability): 0.0082 (Does not biodegrade fast)

Validity criteria fulfilled:

yes

Interpretation of results:

not readily biodegradable

Conclusions:

Each of the three components of C.I. Solvent Blue 98 (3Amine) is estimated to be not readily biodegradable based on the Biowin outputs. These estimations are deemed reliable because the estimated biodegradabilities of structural analogues agree well with existing experimental data.

Executive summary:

The ready biodegradability of C.I. Solvent Blue 98 (3Amine) was evaluated using quantitative structure-activity relationship (QSAR). BIOWIN (v4.10, a module in EPI Suite, v4.10) was chosen because it uses training sets based on mostly public data in a transparent empirical model. The substance C.I. Solvent Blue 98 (3Amine) is a mixture of three main analogous components: C.I. Solvent Blue 98 (3Amine) Component 1, C.I. Solvent Blue 98 (3Amine) Component 2, and C.I. Solvent Blue 98 (3Amine) Component 3.

Simplified molecular input line entry system (SMILES) codes were used as input for the BIOWIN models. BIOWIN estimates the probability of rapid aerobic and anaerobic biodegradation of an organic compound in the presence of mixed populations of environmental microorganisms. BIOWIN contains the following six separate models relevant for assessing ready biodegradability: Biowin1 (linear probability model), Biowin2 (nonlinear probability model), Biowin3 (expert survey ultimate biodegradation model), Biowin4 (expert survey primary biodegradation model), Biowin5 (MITI linear model), and Biowin6 (MITI nonlinear model). Biodegradability estimates are based upon fragment constants that were developed using multiple linear or non-linear regression analyses against experimental data from curated biodegradation databases or expert survey results.

Estimation results from Biowin1 through Biowin6 are evaluated in a weight-of evidence approach in determining the ready biodegradability of an unknown compound in question. BIOWIN determines that a compound is readily biodegradable (i.e., Ready Biodegradability Prediction: YES) if Biowin3 (ultimate survey model) result of biodegradation timeframe is "weeks" or faster (i.e. days, days to weeks, or weeks) and Biowin5 (MITI linear model) result of ready biodegradation probability is greater than or equal to 0.5. A compound is determined to be not readily biodegradable (i.e., Ready Biodegradability Prediction: No) if that criteria is not met.

Each of the three components of C.I. Solvent Blue 98 (3Amine) is estimated to be not readily biodegradable based on the Biowin outputs listed in the table below. These estimations are deemed reliable because the estimated biodegradabilities of structural analogues agree well with existing experimental data.

Compound	Biowin3	Biowin5	Overall Biodegradability considering both Biowin3 and 5
C.I. Solvent Blue 98 (3Amine) Component 1	2.30	-0.11	Not readily biodegradable
C.I. Solvent Blue 98 (3Amine) Component 2	2.35	-0.05	Not readily biodegradable
C.I. Solvent Blue 98 (3Amine) Component 3	2.40	0.01	Not readily biodegradable

Note: Biowin3 assigns the degradation timeframe as "weeks" or faster if the output value of Biowin3 is ≥ 2.75. Biowin5 classifies a compound as "readily biodegradable” if the output value of Biowin5 is ≥ 0.5.

Description of key information

Each of the three evaluated components of C.I. Solvent Blue 98 (3 Amine) is estimated to be not readily biodegradable based on the Biowin outputs ( BIOWIN, v4.10). These estimations are deemed reliable because the estimated biodegradabilities of structural analogues agree well with existing experimental data.

Key value for chemical safety assessment

Biodegradation in water:

under test conditions no biodegradation observed