Willow, Salix alba, ext.

Endpoint summary

Description of key information

An in vitro assay for keratinocyte activation (KeratinoSens, OECD442D) is available for white willow bark extract (WWBE) (aqueous extract). According to this assay WWBE is to be considered a skin sensitizer.

In addition, GARD assays for skin sensitization (GARD skin) and sensitization potency (GARD potency) were performed with the registered substance. According to the decision value of the GARD skin assay, WWBE is a skin sensitizer. The GARD potency assay was not able to directly derive a skin sensitization potency, because no defined molecular weight can be given for WWBE. Instead a molecular weight boundary was calculated for WWBE, which indicates that WWBE has to categorized as 1B. See below for justifications.

Key value for chemical safety assessment

Skin sensitisation

Link to relevant study records

Reference

Reference 1

Endpoint:

skin sensitisation: in vitro

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

11 February 2019 - 27 September 2019

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

comparable to guideline study

Qualifier:

no guideline available

Principles of method if other than guideline:

This study followed the procedures indicated by internal Eurofins Munich SOPs and is conducted according to the internal requirements from Senza Gen AB (DB-Alm Protocol GARD – Genomic Allergen Rapid Detection (GARDskin)). For the SOP see Johansson, H., Gradin, R., Johansson, A., Adriaens, E., Edwards, A., Zuckerstätter, V., Jerre, A., Burleson, F., Gehrke, H., Roggen, E., 2019. Validation of the GARD™skin assay for assessment of chemical skin sensitizers – ring trial results of predictive performance and reproducibility. Toxicol. Sci., Volume 170, Issue 2, August 2019, Pages 374–381
- Principle of test: The transcriptional response of a myeloid leukemia cell line after challenge with the test material is measured. A statistical model (GARD potency) was developed based on the transcriptional response of substances with known sensitization potency under the same experimental conditions. This model is used to prodict the potency of the test material.

- Short description of test conditions: A dendritic cell line (SenzaCells) is treated for 24 h with the test item and controls. After exposure, cell viability is measured and RNA extracted.

- Parameters analysed / observed: 2 decision values (DV, 1 for classification of sensitizer/non-sensitizer, 1 for classification of potency), viability, RNA integrity.

GLP compliance:

yes (incl. QA statement)

Remarks:

Only for cell exposure and RNA extraction. Measurements of gene expression were not performed under GLP

Type of study:

other: activation of myeloid leukemia cells

Justification for non-LLNA method:

Obligation to use in vitro testing for classification

Details on the study design:

Details on study design:
Pre-Experiments:

Solubility test:
Solubility of the test item was determined prior to the Input Finder assay. The test item was soluble in DMSO at a concentration of 500 mg/mL. All test item solutions were freshly prepared immediately prior to use.

Phenotypic Quality Control
At the same day, as the chemical stimulation was performed, the cells were quality controlled by a phenotypic analysis. Six times, 2 x 10^5 cells were transferred into FACS tubes. The cells were washed twice with Dulbecco’s phosphate buffered saline (DPBS) containing 0.5-1% bovine serum albumin (BSA; i.e. FACS buffer). After washing the cells were resuspended in 50 µL FACS buffer and stained with surface markers (see acceptance criteria in section 'Any other information of materials and methods') for 15 minutes at 2-8 °C in the dark.
After staining the cells were washed once with 1 mL FACS Buffer and the cell pellets were resuspended in 200 µL FACS buffer for analysis.
If the phenotypic quality control passes all criteria (see section 'Any other information of materials and methods'), the cells were used for stimulation.

Input finder assay:
The Input Finder assay was performed using stock solutions with a concentration of 500 mg/mL (applied concentration 500 µg/mL) and cell harvested in Passage 11. Starting from 500 mg/mL, maximal nine stock solutions (i.e. nine concentrations) down to 1 mg/mL were prepared. (i.e. 500 mg/mL – 400 mg/mL – 300 mg/mL – 200 mg/mL – 100 mg/mL – 50 mg/mL – 10 mg/mL – 5 mg/mL – 1 mg/mL). These stock solutions were further diluted 100-fold into semi-complete medium giving the stock B solutions. For testing, the highest concentration was 500 µg/mL, even if this concentration is non-toxic. The stock B solutions were finally used for treatment by adding the test item to the SenzaCell suspension in complete medium in a 24-well plate to achieve a further 10- fold dilution.
For testing, SenzaCells were pre-cultured for at least 72 h - 96 h in culture flasks at a cell density of 0.2 x 10^6 cells/mL. Prior to test item application, cells were harvested from the cell culture flask by centrifugation and were re-suspended in fresh semi-complete medium at a density of 2.22 x 10^5 cell/mL. Then 1.8 mL of cell suspension was seeded into a 24-well plate-bottom plate (2 x 10^5 cells/well).
The stock B solutions of the solvent control and the dilution range of the test item was mixed 1:10 (v/v) with the cell suspension prepared in the 24-well plate. Treated plates were incubated for 24 h ± 0.5 h at 37°C ± 1 °C and 5% ± 0.5% CO2.
The cells were stained to check the quality of the cells as described above.
After 24 h ± 0.5 h of exposure, cells were transferred and split into two FACS tubes. FACS buffer was added and the cells were collected by centrifugation (approx. 300 x g). The supernatant was discarded and the remaining cells were washed once with FACS buffer. After washing, cells were re-suspended in 50 µL propidium iodide (PI) solution (1 µg/mL) in FACS buffer. Cells were stained ~ 10 minutes at 2-8°C in the dark. After staining cells were washed with FACS buffer again and re-suspended in 200 µL FACS buffer.
The PI uptake of the cells (as an indicator of cytotoxicity) was analysed immediately after the staining procedure by flow cytometry using an excitation wavelength of λ = 488 nm and an emission wavelength of λ = 650 nm. A total of 10.000 living (PI negative) cells were acquired and cell viability was calculated for each test concentration.
Test items that induced cytotoxicity were used for the main stimulation at the concentration that produced 90% ± 5% relative viability (RV90). Test items that induced no or weak cytotoxicity (≥ 95.5% relative viability) were tested at 500 µg/mL final concentration or at the highest soluble concentration.
If cytotoxic effects were determined, but the relative viability is not between the range of 90% ± 5% the input finder was repeated with another dilution range.
The determined concentration had to be 1000-fold higher for the stock solution for the main stimulation experiment (= GARD input concentration).

Main Stimulation:
For determination of the sensitizing potential of the test substance three independent experiments were performed using separate thawed, cultivated and harvested cells at passage 6 (main stimulation 1) and passage 7 (main stimulation 2 and 3) for all three main stimulations. For each experiment separately weighted samples and preparations were used.

A stock solution of the positive control (75 mM) in DMSO and a stock solution 1000-fold higher as the input concentration of the test item in the solvent determined in the solvent finding pre-experiment were freshly prepared immediately before use.
The stock solution of the positive control, pure DMSO as a negative control, the stock solution of the test item and the solvent of the test item were diluted 100-fold into semi-complete medium (= stock B) to achieve the in-well concentration of 0.1% DMSO.

For testing, SenzaCells were pre-cultured for at least 72 h - 96 h in culture flasks at a cell density of 0.2 x 10^6 cells/mL. Prior to test item application, cells were harvested from the cell culture flask by centrifugation and were re-suspended in fresh semi-complete medium at a density of 2.22 x 10^5 cell/mL. Then 3.6 mL of cell suspension were seeded into a 12 well plate-bottom plate (2 x 10^5 cells/well).
The stock B solutions of the positive control, the solvent control of the test item input concentration and the solvent of the test item were mixed 1:10 (v/v) with the cell suspension prepared in the 12-well plate. Treated plates were incubated for 24 h ± 0.5 h at 37°C ± 1 °C and 5% ± 0.5% CO2.
The cells were stained to check the quality of the cells as described in chapter above.
After 24 h ± 0.5 h of exposure, three times 1 mL of the cells were transferred into small reaction tubes and twice further 500 µL were transferred into two FACS tubes. The cells of the reaction tubes were collected by centrifugation (approx. 300 x g), re-suspended in 500 µL TRIzol and frozen at < -20 °C. The cells of the FACS tubes were washed twice with FACS buffer. After washing, cells were re-suspended in 50 µL propidium iodide (PI) solution (1 µg/mL) in FACS buffer. Cells were stained for approx. 10 minutes at 2-8°C in the dark. After staining, cells were washed with FACS buffer again and will be re-suspended in 200 µL FACS buffer.
The PI uptake of the cells (as an indicator of cytotoxicity) was analysed immediately after the staining procedure by flow cytometry using an excitation wavelength of λ = 488 nm and an emission wavelength of λ = 650 nm. A total of 10.000 living (PI negative) cells were acquired and cell viability was be calculated for each sample. If the cell viability passes the criteria (described below), the RNA of the TRIzol samples were collected as described below.

RNA Isolation
The first tube of each TRIzol sample triplicate was thawed slowly on ice. The samples were centrifuged to remove particulate debris and the supernatant was transferred into new tubes. An equal volume of 95% - 100% ethanol was added and the whole mixture was transferred on the column of a Zymo-SpinTM IIC Columns (ZymoResearch Cat. R5052).
The flow through was discarded and the column was washed twice with 400 µL with the Direct-zolTM RNA PreWash. The flow through was discarded again. After the pre-washing step the column was washed with 700 µL RNA wash buffer. The flow through was discarded again and the column was placed in a new RNAse-free tube. The RNA was eluted with 25 µL RNAse-free water. After centrifugation the collected RNA was transferred on the column again to elute the whole RNA. The RNA was stored at -80°C for shipment.

RNA Quality Control
The RNA quality and quantity was measured with an Agilent 2100 BioAnalyser. If the RNA sample did not pass the quality control criteria (see section 11.3) the RNA had to be isolated from the last two TRIzol samples and quality had to be measured again.
RNA samples which had passed the criteria undergo Nanostring endpoint measurement with an appropriate code set for GARD skin.

Nanostring Hybridisation
All hybridization samples used a total RNA concentration of 100 ng.
All RNA samples were diluted to 20 ng/µL with RNA-free water to use 5 µL of each for hybridization.
For hybridization mix 8 µL of the master mix (provided in the Nanostring master kit), 5 µL RNA sample (20 ng/µL) and 2 µL of the capture code set were mixed. The whole approach was incubated in a thermocycler at 65°C for 24 h ± 0.5 h.

Setting up the Nanostring Preparation and Measurement
The nCounter Prepstation (Nanostring) was prepared as described in the manual. After hybridization the samples were removed from the thermocycler and spinned down before opening. The samples were placed in the nCounter Prepstation, too. The preparation of the nanostring cartridges took around 3 hours.
The cartridges were analysed by an nCounter Digital Analyser (Nanostring).

Data Analysis:
Calculation of relative viability
The cell viability was measured with FACS. The FACS data analysis was performed using the software BD FACS DIVA 6.0. The relative viability percentage calculation is based on the following equation:

Rv = Vs/Vc * 100

Rv = Relative viability of the sample in %
Vs = Absolute viability of the sample in %
Vc = Absolute viability of the mean of the two unstimulated control samples in %

Nanostring data analysis
The NanoString raw data files were analysed using the GARD Data Analysis Application provided from SenzaGen AB. If the DV showed sensitisation potential, the RNA samples may be measured again with a different code set to get information on the potency of the substance to categorize as 1A or 1B.

Acceptance criteria:
For acceptance criteria for the phenotypic quality control see table in section 'Any other information of materials and methods'. Cells can be used for treatment if all criteria are fullfilled

Criteria of the relative viability Quality Control
The cell viability of the cell main stimulation experiment has to fulfil the following criteria:
− Substances that are expected to induce cytotoxicity have a relative viability of 84.5% - 95.4%
− Substances that are not expected to induce cytotoxicity have a relative viability of ≥ 95.5%
− The unstimulated control should have an absolute viability of ≥ 84.5%
If the criteria for the relative viability of the test item are not fulfilled because of an incorrect input concentration, the test item has to be retested to find the correct input concentration.
If one of the controls does not fit the criteria because of a technical error, the control has to be treated twice in the next experiment. If it happens in the third main stimulation a fourth main stimulation has to be done.
If the unstimulated cells do not reach the acceptance criteria, a re-run has to be done.

Criteria for the RNA Quality Control
The RNA samples have to fulfil the following criteria:
− The RNA concentration of a qualified sample should be ≥ 20 ng/µL
− The RNA integrity number (RIN) for a qualified sample should be ≥ 8.0

Prediction criteria
The prediction for classification as sensitizer/non-sensitizer (GARD Skin assay) is defined as described below:
− If the mean decision value (DV) of biological replicate samples is ≥ 0, the substance is classified as a sensitizer
− If the mean decision value (DV) of biological replicate samples is < 0, the substance is classified as a non-sensitizer.

The prediction for classification as sensitizing 1A or 1B (GARD Potency assay) is defined as described below:
− If the mean decision value (DV) of biological replicate samples is ≥ 0, the substance is classified as 1A.
− If the mean decision value (DV) of biological replicate samples is < 0, the substance is classified as 1B.

Positive control results:

Positive control results were valid for both GARD skin (mean decision value of 3 experiments = 12.51) and GARD Potency (mean decision value of 3 experiments = 8.62)

Key result

Run / experiment:

other: mean of 3 experiments

Parameter:

other: Decision value

Remarks:

decision value for skin sensitization (GARD Skin)

Value:

1.76

Vehicle controls validity:

valid

Negative controls validity:

not examined

Positive controls validity:

valid

Run / experiment:

other: 1

Parameter:

other: Decision value

Remarks:

decision value for skin sensitization (GARD Skin)

Value:

2.92

Vehicle controls validity:

valid

Negative controls validity:

not examined

Positive controls validity:

valid

Run / experiment:

other: 2

Parameter:

other: Decision value

Remarks:

decision value for skin sensitization (GARD Skin)

Value:

0.94

Vehicle controls validity:

valid

Negative controls validity:

not examined

Positive controls validity:

valid

Run / experiment:

other: 3

Parameter:

other: Decision value

Remarks:

decision value for skin sensitization (GARD Skin)

Value:

1.42

Vehicle controls validity:

valid

Negative controls validity:

not examined

Positive controls validity:

valid

Other effects / acceptance of results:

No decision value for potency could be calculated, because the analysis software needs molecular weight as input parameter, which is not known for this UVCB.
Acceptance criteria for toxicity were met for all conditions (vehicle control DMSO, positive control, test item). See table under 'any other information' for details.

Results of the Input Finder

Sample		Experiment 1
		Concentration applied [µg/mL]	Relative Cell Viability[%]
Medium Control	--	--	100.00
Solvent Control	--	--	99.2
WWBE	A1	500	99.4
	A2	400	99.2
	A3	300	100.0
	A4	200	98.8
	A5	100	99.1
	A6	50	99.3
	A7	10	99.5
	A8	5	99.7
	A9	1	99.9

 

Cytotoxicity Main Stimulation 1

Main Stimulation 1
Sample	Medium Control	DMSO Control	Positive Control	WWBE
Concentration (mM) Test Item (mg/mL)	-	-	75	500
Cell Viability [%]	95.0	96.1	81.1	96.4
	96.1	96.1	83.3	96.0
Mean [%]	95.6	96.1	82.2	96.2
RV [%]	100.0	100.6	86.0	100.7

 

Cytotoxicity Main Stimulation 2

Main Stimulation 2
Sample		Medium Control		DMSO Control		Positive Control		WWBE
Concentration (mM) Test Item (mg/mL)		-		-		75		500
Cell Viability [%]		94.9		95.7		81.7		95.6
		96.4		95.8		81.0		94.6
Mean [%]		95.7		95.8		81.4		95.1
RV [%]		100.0		100.1		85.0		99.4

 

 

Cytotoxicity Main Stimulation 3

Main Stimulation 3
Sample	Medium Control	DMSO Control	Positive Control	WWBE
Concentration (mM) Test Item (mg/mL)	-	-	75	500
Cell Viability [%]	96.6	97.2	83.5	97.6
	96.7	97.3	83.6	97.6
Mean [%]	96.7	97.3	83.6	97.6
RV [%]	100.0	100.6	86.4	101.0

 

Quality Control for the Main Stimulations

Quality Control for the Main Stimulations
Main Stimulation		1		2		3
Surface marker	Acceptance criteria	measured value		measured value		measured value
CD 86	10-40%	31.5	pass	36.0	pass	34.2	pass
CD 54	+	98.8	pass	98.7	pass	100.0	pass
HLA-DR	+	87.7	pass	88.3	pass	93.6	pass
CD 80	< 10%	0.9	pass	0.6	pass	0.3	pass
CD 34	+	60.6	pass	68.8	pass	78.8	pass
CD 14	+	10.0	pass	13.5	pass	13.7	pass
CD 1α	+	18.7	pass	17.5	pass	15.0	pass
Viabilitymarker
PI	> 84.5%	94.9	pass	93.4	pass	91.2	pass

 

Decision Values (GARD Skin assay) of the Test Item and the Controls

Negative Control		Positive Control		WWBE (500 mg/mL)
Main Stimulation	Decision Values	Main Stimulation	Decision Values	Main Stimulation	Decision Values
1 (repetition)	-0.65	1	12.30	1	2.92
2	-1.03	2	12.98	2	0.94
3.1	-1.93	3	12.25	3	1.42
Mean	-1.20	Mean	12.51	Mean	1.76
Predicition
negative		positive		positive

 

Decision Values (GARD Potency assay) of the Test Item and the Positive Control

Positive Control		WWBE
Main Stimulation	Decision Values		Main Stimulation	Decision Values*
1 (repetition)	8.62		1	-
2	8.18		2	-
3	8.66		3	-
Mean	8.62		Mean	-
Prediction
1A			1Aor1B

*No decision values could be calculated.

 

Interpretation of results:

other: Classification possible based on assumptions for molecular weight

Remarks:

see endpoint summary

Conclusions:

In this study under the given conditions the test item did change the genomic profile of the cells for sensitisation in at least three independent experiment runs. According to the decision value calculated by the model of the GARD skin assay, the test item is considered as sensitiser.
The GARD potency assay is not able to make a prediction of the potency of an UVCB, as the molecular weight is not known. If the test item would have a molecular weight > 2090 g/mol the test item would have to be categorized into class 1A, if the test item would have a molecular weight < 2090 g/mol the test item would have to be categorized into class 1B.

Executive summary:

WWBE was tested in the GARD Skin and GARD Potency assay under GLP conditions.

The in vitro GARD (TM )Test enables detection of the sensitising potential of a test item by addressing the third molecular key event of the adverse outcome pathway (AOP), namely human dendritic cell activation, by measuring changes in the genomic profile of the cells after treatment. These changes are used to support discrimination between skin sensitiser and non-sensitisers. Before every treatment the phenotypic profile of the SenzaCells were checked.

In the present study WWBE was dissolved in DMSO at a concentration of 500 mg/mL. For the Input finding assay, stock solutions in a concentration range from 500 mg/mL to 1 mg/mL were prepared by serial dilution (finally applied concentration 500 µg/mL to 1 µg/mL). Cells were incubated with the test item for 24 h at 37°C. After exposure cells were stained with propidium iodide and cell viability was measured by FACS analysis.

No RV <90 ± 5% could be determined within the tested range. The main stimulations were done with 500 µg/mL stock solution (500 µg/mL final concentration).

Cells were incubated with the test item for 24 h at 37 °C. After exposure one part of the cells were stained with propidium iodide and cell viability was measured by FACS analysis and the second part of the cells were transferred into TRIzol and stored at -80°C.

All three main stimulations fulfilled the criteria. Cells treated with 500 µg/mL were used for RNA isolation.

The RNA samples were shipped temperature monitored by World Courier to Test Site 1. All samples arrived without any issues.

The RNA quality was measured first for all single tube TRIzol isolations. All samples fulfilled the quality criteria and no pooled samples were necessary. All single isolations were used for nanostring measurement.

The decision values for the test item are > 0. Therefore the test item has to be classified as a sensitiser.

The decision values of the positive control are > 0 (12.30; 12.98; 12.25) and for the negative control are < 0 (-0.65; -1.03; -1.93).

The controls confirmed the validity of the study for all experiments.

Due to a sensitising potential, the potency of the test item was measured. Because no molecular weight is given and the used model for categorization of the potency depends on the molecular weight just a threshold can be given. If the test item would have a molecular weight > 2090 g/mol the test item could be categorized into class 1A, if the test item would have a molecular weight < 2090 g/mol the test item could be categorized into class 1B. Because no molecular weight is known, no categorization can be done.

The decision values of the positive control are > 0 (8.62, 8.18, 8.66), which is class 1A.

The data generated with this test should be considered in the context of integrated approached such as IATA, combining the result with other complementary information, e.g. derived from in vitro assays addressing other key events of the skin sensitisation AOP.

Reference 2

Endpoint:

skin sensitisation: in vitro

Type of information:

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

Adequacy of study:

weight of evidence

Justification for type of information:

REPORTING FORMAT FOR THE ANALOGUE APPROACH

1. HYPOTHESIS FOR THE ANALOGUE APPROACH
One of the main constitutents of the source and target substance is salicine, which is a known skin sensitizer. The extraction methods for both substances are similar, so that no significant qualitative differences in composition are expected.

2. SOURCE AND TARGET CHEMICAL(S) (INCLUDING INFORMATION ON PURITY AND IMPURITIES)
The source analogue is an aqueous extract of white willow bark.
The target analogue for the read-accross is an ethanolic extract of white willow bark.
The purity is 100% in both cases (as UVCB).

3. ANALOGUE APPROACH JUSTIFICATION
Aqueous white willow bark extract is a skin sensitizer according to an in vitro keratinocyte activation study (OECD442D, see cross-referenced study). In addition, a GARD potency assay to predict skin sensitization potency was performed with the target substance. A WoE approach, taking into account both study results, is used to transfer skin sensitizing properties of the aqueous extract to the ethanolic extract.

4. DATA MATRIX
White willow bark, aquous extract (only source analogue): positive

Reason / purpose for cross-reference:

read-across source

Key result

Run / experiment:

other: 1

Parameter:

other: I_max

Value:

3.99

Vehicle controls validity:

valid

Negative controls validity:

not examined

Positive controls validity:

valid

Remarks on result:

positive indication of skin sensitisation

Key result

Run / experiment:

other: 2

Parameter:

other: I_max

Value:

7.03

Vehicle controls validity:

valid

Negative controls validity:

not examined

Positive controls validity:

valid

Remarks on result:

positive indication of skin sensitisation

Key result

Run / experiment:

other: 1

Parameter:

other: EC1.5

Remarks:

[µg/mL]

Value:

118.1

Vehicle controls validity:

valid

Negative controls validity:

not examined

Positive controls validity:

valid

Remarks on result:

positive indication of skin sensitisation

Key result

Run / experiment:

other: 2

Parameter:

other: EC1.5

Remarks:

[µg/mL]

Value:

179.55

Vehicle controls validity:

valid

Negative controls validity:

not examined

Positive controls validity:

valid

Remarks on result:

positive indication of skin sensitisation

Other effects / acceptance of results:

All validity criteria are met.
In both repetitions, I_max is higher than 1.5, the EC1.5 is lower than 200 µg/mL and at the EC 1.5 concentration, viability is higher than 70%.

Interpretation of results:

Category 1 (skin sensitising) based on GHS criteria

Conclusions:

Under the experimental conditions the source analogue (White willow bark, aquous extract) may be classified as sensitizer.

Executive summary:

The skin sensitisation potential of the source analogue (aqueous white willow bark extract) was tested with an in vitro test measuring keratinocyte activation (OECD 442D, KeratinoSens), performed under GLP conditions. The test method KeratinoSensTM is considered scientifically valid to be used as part of an integrated approaches to testing and assessment, to support the identification of the sensitization potential of test item for hazard classification and labeling purposes.

According to the keratinocyte activation study, the source analogue (aqueous extract of white willow bark) may be classified as a skin sensitizer.

The study results of the source compound were considered applicable to the target compound and were used for classification and labelling acc. to REGULATION (EC) No 1272/2008. The aqueous extract is used as the source analogue to predict the skin sensitizing potential of the target substance (ethanolic extract of white willow bark).Justification and applicability of the read-across approach (structural analogue) is outlined in the field "justification for type of information"

In conclusion, the test material ethanolic extract of white willow bark is considered a skin sensitizer.

Endpoint conclusion

Endpoint conclusion:

adverse effect observed (sensitising)

Additional information:

Skin sensitisation assessment according to the OECD approved "Adverse Outcome Pathway for Skin Sensitisation Initiated by Covalent Binding to ProteinsPart 1: Scientific EvidenceAOP" ( ENV/JM/MONO(2012)10/PART1), which is reflected in the standard information requierements according to Regualtion (EC) No. 2006/1907 column 1, section 8.3.1. (a), (b), and (c) :

(a) KE1 (molecular interaction with skin proteins): The DPRA assay as the only assay covering Key event 1 of the AOP for skin sensitisation is not applicable to UVCBs. Therefore no assessment of KE1 is possible.

(b) KE2 (inflammatory response in keratinocytes): An in vitro assay for keratinocyte activation (KeratinoSens, OECD442D) was performed for white willow bark extract (WWBE) (aqueous extract). According to this assay WWBE is to be considered a skin sensitizer.

(c) KE3 (activation of dendritic cells): A GARD skin assay was performed with the registered substance. The GARD skin assay investigates the activation of dendritic cells by measuring transcriptional changes on a defined set of genes. GARD skin employs a statistical model to derive a decision value (DV) that determines whether a test item is a skin sensitizer or not. According to the decision value of the GARD Skin assay, WWBE is a skin sensitizer.

Determination of the hazard classification for skin sensitization:

In addition to the GARD skin assay, the GARD potency assay was performed with the registered substance. The GARD potency assay follows the same principle as the GARD skin, but uses a different statistical model to determine the sensitizing potency (category 1A or 1B). No decision value could be calculated for the GARD potency assay, because the analysis software needs a molecular weight of the test item for calculation.

As WWBE is a UVCB, no defined molecular weight is available. However boundaries for classification as 1A or 1B can be calculated for WWBE:

If the test material had a MW (molecular weight) > 2090 g/mol the decision value would indicate category 1A. For a molecular weight below 2090 g/mol, the model derives a decision value corresponding to category 1B. WWBE contains salicine, a known sensitizer, as one of the main constituents.

Even though not all constituents of WWBE are known with their exact chemical structure, the major fraction of the mass of WWBE is is from substances with no sensitizing potential.

Salicine has a MW of 286.3 g/mol, which would result in a classification as 1B. The MW of salicine is far away from the classification boundary, so that potential other sensitizing constituents, which have not been identified, would need to have a very large MW to indicate a classification as 1A. In addition, the decision value in the GARD skin and KeratinoSens was rather low, so that the weight of the evidence warrants a classification for WWBE as skin sensitizer category 1B.

Justification for classification or non-classification

Skin sensitization:

As the substance is a UVCB substance the only test available for hey event 1 (i.e. DPRA) is not applicable.

For the other two key events adequate and reliable in vitro studies are available.

An in vitro assay for keratinocyte activation (KeratinoSens, OECD442D, key event 2) is available for white willow bark extract (WWBE) (aqueous extract), indicating that WWBE is a skin sensitizer.

The GARD Skin assay (key event 3), performed with WWBE, also comes to the conclusion that WWBE has to be considered a skin sensitizer.

No assay for direct classification of skin sensitizing potency is availabe, but the weight of the evidence leads to classification as skin sensitizer, category 1B according to Regulation (EC) No. 1272/2008:

As WWBE is a UVCB, no defined molecular weight is available. However boundaries for classification as 1A or 1B can be calculated for WWBE:

If the test material had a MW (molecular weight) > 2090 g/mol the decision value would indicate category 1A. For a molecular weight below 2090 g/mol, the model derives a decision value corresponding to category 1B. WWBE contains salicine, a known sensitizer, as one of the main constituents.

Even though not all constituents of WWBE are known with their exact chemical structure, the major fraction of the mass of WWBE is is from substances with no sensitizing potential.

Salicine has a MW of 286.3 g/mol, which would result in a classification as 1B. The MW which is far away from the classification boundary, so that potential other sensitizing constituents, which have not been identified, would need to have a very large MW to indicate a classification as 1A. In addition, the decision value in the GARD skin and Keratinosens was rather low, so that the weight of the evidence warrants a classification for WWBE as skin sensitizer category 1B.