Reaction mass of didodecyl 3,3'-thiodipropionate and dioctadecyl 3,3'-thiodipropionate and octadecyl 3-[[3-(dodecyloxy)-3-oxopropyl]thio]propionate

Toxicity to reproduction

Administrative data

Endpoint:

reproductive toxicity, other

Type of information:

read-across from supporting substance (structural analogue or surrogate)

Adequacy of study:

supporting study

Justification for type of information:

REPORTING FORMAT FOR THE ANALOGUE APPROACH
A brief overview of the read-across study is reported below. Detailed information on the read-across justification is included in the read-across study report available in the "Attached justification" field. Please also refer to this report for the list of tools used in the assessment.

1. HYPOTHESIS FOR THE ANALOGUE APPROACH
The present read-across study falls within the RAAF scenario 2, i.e. one-to-one analogue approach based on the hypothesis that different compounds are supposed to cause the same type of effects as a result of structural similarity.
In the current study, dilauryl thiodipropionate (E12) was selected as source chemical for the read-across prediction of reproductive and developmental toxicity for the target E1218. The two substances are members of the same intramolecular-S family (derived from Thiodiglycolic acid, Thiodipropionic acid, Dithiodiglycolic acid or Dithiodipropionic acid) . In particular, the source E12 is a mono-constituent substance, while the target E1218 is a multi-constituent substance, being a mixture of three esters E12, E18 (distearyl thiodipropionate) and E1218 (dilauryl stearyl thiodiproprionate). It is highlighted that one of the constituents of the target (i.e., E12) is identical to the mono-constituent substance E12; the remaining two constituents of the target (i.e., E18 ad E1218) belong to the same “pool” of structurally related constituents. Thus, according to the RAAF guidance for UVCB and MCSs , structural similarity as a basis for the read-across may be assumed.
Supporting information to further justify the analogue approach is provided by comparing the source and target substances in terms of mechanistic similarity, physico-chemical properties and ADME profile.

2. SOURCE AND TARGET CHEMICAL(S) (INCLUDING INFORMATION ON PURITY AND IMPURITIES)
Please refer to section 1.2 (general information/composition) of both target and source's IUCLID dossiers, and to the attached read-across study report.

3. ANALOGUE APPROACH JUSTIFICATION
Structural similarity: No significant structural differences in terms of functional groups and chemical class were identified, since both compounds share the 3,3'-sulfanediyldipropanoate fragment, which is esterified with aliphatic alkyl chain of different length (C12 and/or C18). For all the analysed structures, two functional groups, i.e. Carboxylic acid ester and Sulfide, and the ECOSAR chemical class of Esters was identified. The main structural difference among the source and target is represented by the carbon chain length of the two lateral chains (C12 for the source E12 and for the target constituent E12, C12 for the target constituent E18 and C12/C18 for the target constituent E1218). This difference is reflected in minimal differences in molecular structure data (i.e., different no. of atoms and bonds).
Concluding, the target and source substances exhibit high structural similarity. Uncertainty associated with structural similarity was assessed to be low.
The structural differences identified in the target compound with respect to the source were critically analysed to understand if they might cause any difference in the properties/effects to be predicted by read-across. To this aim, the target and source compounds were analysed for their similarity and differences in terms of the underlying mechanism, physico-chemical properties, ADME profile and metabolism (i.e., potential metabolic products).

Mechanistic similarity: The target and source compounds exhibited high mechanistic similarity. As it can be noted from the table above, no structural alerts were identified in both target and source constituents by any of the profilers under consideration. This suggests that both chemicals did not match the structural criteria specified in the boundaries of each profiling schemes. In particular, an unlikely binding to the estrogen-receptors (ER) interacting site was predicted for both chemicals, based on their high molecular weight (> 500 Da). Furthermore, no protein and retinoid acid receptor binding alerts were found in both target and source compounds by the OECD QSAR Toolbox profilers.
Finally, no known DART potential was predicted by the DART scheme for target and source chemicals, and no chemotype alerts for cleft palate were identified.
Concluding, the target and source substances exhibited high mechanistic similarity. However, the uncertainty associated with mechanistic similarity was assessed as medium, due to limitations associated with the endpoint type.
In fact, reproductive and developmental toxicity is a high-tier and complex endpoint. Although a variety of modes of action have been established for DART chemicals and some AOPs have been described , the total number of possible adverse outcome pathways is unclear, and an understanding of how chemical structures map to mode(s) of action has yet to be established .

Additional considerations for multi-constituents and UVCB:
No reactive groups/structural alerts relevant for reproductive and developmental toxicity were identified in the target and source substances. Thus, based on the mechanistic profile of the target and source substances, no interactions (e.g., no additive effects) are expected among the target constituents. Thus, it is reasonable to conclude that the combined exposure of the three constituents would not have a significant impact on the type and strength of the predicted effects.

Physico-chemical similarity: From the analysis and comparison of experimental and in silico predicted physico-chemical data, no major differences are noted between target and source substances. As expected from their chemical structure characterised by rather long chains of carbon atoms, both substances were predicted to be highly hydrophobic and insoluble, with LogKow values greater than 10 and water solubility (experimental and predicted) below 1 mg/L. Since no dissociating groups are found in both structures, pH has no influence on LogKow, water solubility, and ADME properties. Lastly, volatilisation is not expected to occur for both substances. The majority of physico-chemical data used in the present assessment are characterised by good reliability (Klimish 2 experimental studies as well as moderately reliable QSPR predictions), with the exception of few predicted data (i.e., LogKow predictions for the constituents E1218 and E18 and water solubility predictions for both target and source) characterised by borderline reliability. In addition, a good consistency between experimental and predicted data was observed for the source and target substances.
Concluding, the target and source compounds exhibited high similarity in terms of their physico-chemical profile. Specifically, very high hydrophobicity, low water solubility, no volatilisation potential were observed in both substances. The uncertainty associated with PC similarity was assessed to be low.

ADME similarity: Permeability depends on lipophilicity, degree of ionization and molecular size. Moreover, passive transport generally increases with increasing LogKow values up to 3.5, but declines thereafter (Liu and Hunt 2005). As discussed in the paragraph above, target and source substances are highly lipophilic and insoluble molecules with a molecular weight greater than 500 Da, and with no dissociating properties.
Experimental toxicokinetic studies were available for E12 (source substance and one of target main constituents) as well as for another target constituent E18. Both studies were assessed to be reliable with restrictions (Klimisch 2) by the data owner, and were based on three publications (Reynolds et al. 1974, Liebert 1992, and Diamante et al. 2011). The studies were conducted according to guideline equivalent or similar to the OECD TG 417 (Toxicokinetics), where the fate of the two substances after oral application to rats was investigated. The main results are reported in the table above and following summarised:
o both E12 and E18 were found to be almost entirely absorbed from the gastrointestinal tract.
o distribution in tissues was reported to be close to normal values except the value of radioactivity in the fat, which was found to be elevated at 4 days after dosing and remained so at 8 and 34 days.
o the metabolite dithiopropionic acid was recovered in urine free or as an acid-labile conjugate.
o elimination occurred mostly via urine (up to 88 and 90%, for E12 and E12, respectively); more than 90% of the dose was eliminated within 24 h.
As far as concern the similarity analysis based on in-silico ADME properties, a similar ADME profile was noted for the target E1218 and the source E12. In particular, both substances were predicted to cross a biological membrane by transcellular mechanisms, resulting in an optimal passive absorption across intestinal epithelium (i.e., HIA 100%). The good absorption from the gastrointestinal tract predicted for E12 and E18 find confirmation in the experimental studies mentioned above.
A very poor oral bioavailability, and thus a very low amount of substance that reaches systemic circulation after oral administration (%F), was predicted for both substances. This was due to the estimated chemicals’ low solubility in the gastro-intestinal tract, low susceptibility to the stomach decomposition, and significant first-pass metabolism, which represents the probability that substances will be metabolically transformed by enzymes in liver and intestine. In general, passive absorption does not affect the %F.
Target and source chemicals were also characterized by volumes of distribution above the volume of total body water of 0.7 L/kg (Vd increasing with the increase of carbon chain length). This suggests that both substances have affinity to extravascular tissues. Affinity for fat tissue was found experimentally for both E12 and E18. Predictions of plasma protein binding resulted not reliable for all the structures.
Prediction of metabolic stability, i.e. t1/2 in Human Liver Microsomes, resulted to be undefined for target and source compounds, thus preventing assessment of metabolism rate. Prediction of potential metabolites is discussed in the following paragraph.
Finally, no in silico data are available for estimating the elimination rate. However, experimental data available for E12 and E18 indicate that more than 90% of the dose is eliminated within 24h.
Concluding, target and source compounds exhibited high similarity in terms of their predicted ADME profile. In addition, the predicted ADME profile for the target is in agreement with available toxicokinetic data. Overall, the uncertainty associated with ADME similarity was assessed to be low.

Additional considerations for multi-constituents and UVCB:
No interactions are expected among the target constituents (no reactive groups/alerts were identified in the target and source structures). Thus, no variations in physico-chemical properties and toxicokinetics are expected due to interactions among target’s constituents.

Potential metabolic products: Two phase I reactions are predicted for the structures E1218, E12 and E18:
• hydrolysis of the carboxylic acid ester, leading to the formation of the acids 3-{[3-(dodecyloxy)-3-oxopropyl]sulfanyl}propanoic acid (for E1218 and E12) and 3-{[3-(octadecyloxy)-3-oxopropyl]sulfanyl}propanoic acid (for E1218 and E18), and the related alcohols dodecan-1-ol (for E1218 and E12) and octadecan-1-ol (for E1218 and E18);
• sulphide oxidation, leading to the formation of dodecyl octadecyl 3,3'-sulfinyldipropanoate (for E1218), didodecyl 3,3'-sulfinyldipropanoate (for E12) and dioctadecyl 3,3'-sulfinyldipropanoate (for E18).
If a 2nd level of metabolism was included in the Liver BioPath simulation (i.e., reaction rules are applied to the first generation of metabolites), the metabolite thiodipropionic acid was among the predicted metabolites with higher odds ratio via hydrolysis of the carboxylic acid ester. The presence of this metabolite is confirmed by the toxicokinetics studies performed with E12 and E18.
Finally, the predicted metabolites of target and source compounds were profiled within the OECD QSAR toolbox (structural- and mechanistic-based profilers), ACD/Percepta (ACD/Estrogen receptor binding) and ChemTunes (Cleft palate chemotypes). The assessment of their structural and mechanistic similarity was performed in order to possibly understand whether some metabolites of the target might have an influence on the prediction of reproductive and developmental toxicity. The results of this profiling are reported in Appendix A of the report, and are following summarised:
• No relevant structural differences were identified among the target and source metabolites in terms of functional groups and chemical classes.
• No mechanistic differences were identified among the target and source metabolites concerning protein binding, ER and RAR binding, and DART potential (including cleft palate). In particular, no structural alerts for protein binding, ER binding and RAR binding were identified in any of the target and source metabolites. Finally, no known DART potential was predicted by the DART scheme for all target and source metabolites, and no chemotype alerts for cleft palate were identified.

Concluding, the target and source substances exhibited high similarity in terms of their potential metabolites. In particular, the fact that both substances are likely to undergo to ester hydrolysis find confirmation in experimental toxicokinetic studies. The predicted metabolites showed high structural and mechanistic similarity. Thus, the uncertainty associated with the identification of potential metabolic products was assessed to be low.

Source experimental data:
The reproductive toxicity study based on the source substance E12 was waived by the registrant, based on the following justification:
In accordance with Annex IX (8.7.3) of the REACH legislation, the test does not need to be conducted if there are no adverse effects on reproductive organs observed in the 90 days study. Systemic exposure to this substance is limited because it is handled in a non-dusty form during production and is incorporated in a matrix in articles. None of the available toxicity studies indicated reactivity such as protein binding or cytotoxicity. Furthermore, developmental toxicity studies did not indicate disturbances of fertility and reproductive performance. Therefore, no further hazard for reproductive toxicity is expected and testing is not considered necessary.

A key developmental toxicity/teratogenicity experimental study was available for the source substance E12. The study report has been reported by the registrant to be reliable with restrictions (Klimisch score 2), since limited information on test item, animal husbandry and results were provided. Based on this study, no classification as a developmental toxicant at either Category 1B or 2 may be warranted because of no effects were seen in adults without consequential effects on fetuses.

Two additional developmental toxicity/teratogenicity experimental studies are available for the source substance E12, and were reported as supporting material by the registrant. Details are reported in the attached report.

4. DATA MATRIX
Data matrix is included in the attached report.

Cross-reference

Cross-reference 1

Reason / purpose for cross-reference:

read-across source

Remarks:

Source E12 (CAS No. 123-28-4)

Reference

Endpoint:

developmental toxicity

Type of information:

experimental study

Adequacy of study:

supporting study

Study period:

1972

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

other: Acceptable study report with restrictions (limited information on test item, animal husbandry and results).

Qualifier:

equivalent or similar to guideline

Guideline:

OECD Guideline 414 (Prenatal Developmental Toxicity Study)

Deviations:

yes

Remarks:

(limited information on test item and animal husbandry, no data on maternal toxicity/food and water consumption)

GLP compliance:

no

Limit test:

no

Species:

rat

Strain:

Wistar

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: no data
- Age at study initiation: no data, virgin
- Weight at study initiation: average 217 g
- Fasting period before study: no data
- Housing: individually housed in mesh bottom cages
- Diet: ad libitum
- Water (fresh): ad libitum
- Acclimation period: no data

ENVIRONMENTAL CONDITIONS
- Temperature (°C): controlled, but no further details given
- Humidity (%): controlled, but no further details given
- Air changes (per hr): no data
- Photoperiod (hrs dark / hrs light): no data

Route of administration:

oral: gavage

Vehicle:

corn oil

Details on exposure:

VEHICLE
- Amount of vehicle (if gavage): 1mL/kg bw

Analytical verification of doses or concentrations:

no

Details on mating procedure:

Females were mated with young adult males, and observation of the vaginal sperm plug was considered Day 0 of gestation.

Duration of treatment / exposure:

Day 6 and continuing daily through Day 15 of gestation

Frequency of treatment:

daily

Duration of test:

During pregnancy until caesarian on day 20.

No. of animals per sex per dose:

20 or 21, depending on the group

Control animals:

yes, sham-exposed

Details on study design:

250 mg/kg of aspirin was used as a postive control.
Attention was called to the fact that this is the seventeenth of a series of reports performed for the US FDA. Eventually, a total of at least 42 compounds were planned to be tested in 21 pairs; each pair being run concurrently against one sham-treated control and one positive control group. Because
of the inherent variability of biological data of the type dealt with. Here, the accumulation and pooling of sequential sets of control values would greatly enhance the statistical value of the findings and the ultimate reliability of the test results.

Maternal examinations:

Body weights were recorded on Days 0, 6, 11, 15, and 20 of gestation. All animals were observed daily for appearance and behaviour with particular attention to food consumption and weight, in order to rule out any abnormalities which may have occurred as a result of anorexic effects in the pregnant female animal.

Ovaries and uterine content:

On Day 20 all dams were subjected to caesarean section under surgical anesthesia, and the numbers of implantation site and resorption sites were recorded. The urogenital tract of each dam was examined in detail for anatomical normality. The number of live and dead fetuses were recorded.

Fetal examinations:

The body weights of the live pups were recorded.
All foetuses were examined grossly for the presence of extemal congenital abnormalities. One-third of the foetuses of each litter underwent detailed visceral examinations employing 10x magnification. The remaining two-thirds were cleared in potassiiam hydroxide (KOH), stained with alizarin red dye and examined for skeletal defects.

Statistics:

not performed

Historical control data:

Not available at the time of reporting.

Details on maternal toxic effects:

Maternal toxic effects:no effects

Details on maternal toxic effects:
The method part states that clinical signs were assessed, but no further details were given in the results part.
No mortalitiy occured and the body weight development was not affected.

Dose descriptor:

NOAEL

Effect level:

1 600 mg/kg bw/day (actual dose received)

Based on:

test mat.

Basis for effect level:

other: other:

Details on embryotoxic / teratogenic effects:

Embryotoxic / teratogenic effects:no effects

Abnormalities:

not specified

Developmental effects observed:

not specified

Treatment with aspirin at 250 mg/kg bw resulted in an increased number of resorptions, dead foetuses, missing, bipartitie or incompletely ossified sternebrae, incompletely ossified vertebrae and incompletely closed skulls.

Cross-reference 2

Reason / purpose for cross-reference:

read-across source

Remarks:

Source E12 (CAS No. 123-28-4)

Reference

Endpoint:

two-generation reproductive toxicity

Remarks:

based on test type (migrated information)

Data waiving:

other justification

Justification for data waiving:

other:

Reproductive effects observed:

not specified

Materials and methods

Test guideline

Test guideline 1

Qualifier:

according to guideline

Guideline:

other: ECHA, 2015: Read-Across Assessment Framework (RAAF).

Test guideline 2

Qualifier:

according to guideline

Guideline:

other: ECHA, 2017: Read-Across Assessment Framework (RAAF). Considerations on multi-constituent substances and UVCBs.

Test material

Test material information

Constituent 1

Reference substance name:

Octadecyl 3-[[3-(dodecyloxy)-3-oxopropyl]thio]propionate

EC Number:

236-025-0

EC Name:

Octadecyl 3-[[3-(dodecyloxy)-3-oxopropyl]thio]propionate

Cas Number:

13103-52-1

Molecular formula:

C36H70O4S

IUPAC Name:

Dodecyl octadecyl 3,3'-sulfanediyldipropanoate

Test material form:

solid: pellets

Specific details on test material used for the study:

SMILES: O=C(OCCCCCCCCCCCC)CCSCCC(=O)OCCCCCCCCCCCCCCCCCC
InChI=1S/C36H70O4S/c1-3-5-7-9-11-13-15-16-17-18-19-20-22-24-26-28-32-40-36(38)30-34-41-33-29-35(37)39-31-27-25-23-21-14-12-10-8-6-4-2/h3-34H2,1-2H3

Results and discussion

Results: P0 (first parental generation)

Effect levels (P0)

Key result

Dose descriptor:

NOAEL

Effect level:

1 600 mg/kg bw/day (actual dose received)

Based on:

other: Source

Remarks on result:

other:

Remarks:

Read-across

Applicant's summary and conclusion

Conclusions:

Based on the read-across study presented, it is concluded that the reproductive and developmental toxicity data available for the source compound E12 could be used to support the same conclusion of no classification of reproductive and developmental toxicity for the target E1218.

Executive summary:

This study was designed to generate read-across predictions of reproductive and developmental toxicity for the multi-constituent substance Lauryl stearyl thiodiproprionate (E1218) to be used for its safety assessment in the regulatory framework of REACH. The target E1218 (trade name Evanstab™ 1218) is a mixture of 3 esters: dilauryl thiodipropionate (E12, or DLTDP), distearyl thiodipropionate (E18, or DSTDP), and the mixed ester dilauryl/stearyl thiodipropionate.

In theread-across analysis,E12(mono-constituent substance) was selected as source chemical for the target E1218. One of the constituents of the target (i.e., E12) is identical to the mono-constituent substance E12; the remaining two constituents of the target (i.e., E18 ad E1218) belong to the same “pool” of structurally related constituents. Thus, according to the RAAF guidance for UVCB and MCSs, structural similarity as a basis for the read-across may be assumed.The similarity assessment, which consisted of a comparison of structural, mechanistic, physicochemical, ADME and metabolism profiles, showedthat the sourceE12and the targetE1218are sufficiently similar to justify the read-across approach.

Three experimental developmental toxicity studies were available for the source E12, one of which was used as key study, and the remaining two as supporting data. All of these studies were assessed as adequate for the read-across prediction for the target E1218.

QSAR predictions for endpoints relevant for reproductive and developmental toxicity were performed for the target substance E1218 (considering the three main constituents E1218, E12 and E18). No reliable predictions were obtained for reproductive toxicity, whereas negative predictions of mostly moderate reliability were achieved for developmental toxicity. These QSAR predictions were in agreement with, and further supported, the read-across prediction for developmental toxicity, thus contributing to reducing the read-across uncertainty.

In the present read-across study, an overall low uncertainty was associated with similarity justification. However, the following issue was highlighted: the endpoint to read-across does not have yet well-defined mechanistic bases (high-tier endpoint). The main uncertainties associated with the read-across argument were related to the following issues: i) endpoint-type, ii) analogue set limited to one source compound; iii) no bridging studies among the source and target substances.

Concluding, the read-across from the developmental toxicityexperimental studies in rats, mice and hamsters, which are available for the source substance E12, are likely to predict the reproductive and developmental toxicity potential of the target substance E1218 and are considered as adequate to fulfil the information requirement of Annex VIII 8.7 of the REACH Regulation.