Paraquat-dichloride

Endpoint summary

Key value for chemical safety assessment

Genetic toxicity in vitro

Description of key information

See 'Additional information' field.

Endpoint conclusion

Endpoint conclusion:

adverse effect observed (positive)

Genetic toxicity in vivo

Description of key information

See 'Additional information' field.

Endpoint conclusion

Endpoint conclusion:

no adverse effect observed (negative)

Additional information

SUMMARY

The test substance has been extensively examined in a wide range of guideline and non-guideline studies both in vitro and in vivo, including end-points for mutagenicity, clastogenicity and aneugenicity. The in vitro assays allow an assessment of the intrinsic ability of the compound to induce genotoxicity whilst the in vivo assays determine whether any such activity will be expressed in the whole animal.

The interpretation of these studies was performed using a weight of evidence judgement with regards to the revised principles for genotoxicity testing by OECD¹, specifically:

 

- Effects seen at high concentration / high cytotoxicity in vitro should be considered to be non-relevant;

- Effects determined in non-mammalian cell systems (with the exception of the OECD 471 bacterial reverse mutation assay) should not be given equal weighting to mammalian cell effects (ideally any effects noted in non-mammalian systems should be corroborated by an appropriate assay in mammalian cell systems);

- Effects observed in test systems for which guidelines have now been deleted due to a lack of robust understanding of the assay, e.g. OECD 479, should be discounted;

- Interpretation of positive effects in assays, either in vitro or in vivo, should show statistical and biological relevance and preferably demonstrate a dose dependent effect and an effect that is greater than the upper value of the historical negative control data range;

- In vivo assessment of an endpoint should be given significantly higher weighting than in vitro assessment of the same endpoint;

- In vivo studies should use a relevant route of exposure, hence multiple studies reporting in vivo effects following intraperitoneal (i.p.) dose administration may be considered to be inappropriate to assessment of human hazard and hence should afforded a lower weighting in the weight of evidence argument than those studies using a relevant route.

In conclusion, based on a thorough review and weight of the evidence assessment and taking into account current OECD guidance on the conduct and interpretation of in vitro and short term in vivo studies for genotoxicity, there is insufficient evidence to conclude any mutagenic potential of the test substance. Moreover the lack of any concern over carcinogenicity or generational effects in reproduction studies further confirms a lack of mutagenicity.

 

 

IN VITRO ASSAYS

 

Mutagenicity (bacteria)

Bacterial gene mutation assays (Ames)

A total of 18 Ames tests have been performed using multiple Salmonella typhimurium and Escherichia coli strains in presence or absence of a metabolic activation system. Conflicting results have been reported for the test substance when using a pre-incubation protocol. Four out of 16 Ames tests gave a positive result for the S. typhimurium strain TA100, TA104, TA98, TA1535 and TA92. The positive incidences for these strains are 3/13, 1/1, 1/10, 1/9 and 1/1 respectively. One study reported very low mutagenic activity of strains TA100 and TA1538.

In the studies considered key in this evaluation, the test substance was assayed in the mutagenicity plate incorporation assay (equivalent to OECD TG 471 or similar protocols) using

-         S. typhimurium TA1535, TA1538, TA98 and TA100 strains (Anderson et al., 1977).

-         S. typhimurium TA102, TA2638 and E. coli strains WP2/pKM101 and WP2 uvra/pKM101 (Watanabe et al., 1988)

-         S. typhimurium TA100, TA98, TA1535, TA1537, TA1538 and E. coli strain WP2 her (Moriya et al., 1983).

The substance was not found to be mutagenic in these assays.

 

Overall, the most robust in vitro assessment of gene mutation in non-mammalian systems is the one that is performed in the bacterial reverse mutation assay. As described above both the test substance and formulations of the test substance have been tested and overall shown to be negative in this assay. Hence, the test substance is concluded to be not mutagenic in bacterial reverse mutation assays. 

 

Other bacterial assays

A number of other bacterial assays were conducted which methods are not included in the current principles for genotoxicity testing following the OECD guidance (SOS chromotest/umu test/rec assay and 8-AG resistance assay). A total of 9 studies were conducted, five of which reported a positive result. The weight of these studies in the WoE approach is limited.

 

Clastogenicity

Chromosome aberration in mammalian cells

A total of 12 Chromosomal aberration tests were conducted using various cell types, such as Chinese hamster cells or human derived cells, in the presence and/or absence of a metabolic activation system. All but one of these assays were reported to be positive under the conditions tested. In the GLP study considered key in this evaluation (Sheldon, 1985) human lymphocytes were investigated in accordance with OECD TG 473.The results indicate that at dose levels beyond the currently recommended maximum concentrations of 10 mM or 2000 µg/mL and where there is marked (>50%) cytotoxicity, the test substance is clastogenic to human lymphocytes in vitro. Overall, when assessed against the current guidelines (OECD 273, 2016) for top test substance concentration and cytotoxicity required in the assays, no relevant effects are noted in this assay. Hence, a current interpretation would likely conclude this study to be negative. In 3 other studies (Tanaka, 1989; Nicotera, 1985; Speit, 1998), chromosomal aberrations were found only at concentrations that induced more than the currently recommended 55 ± 5% cytotoxicity, and hence these results should be interpreted with great caution. When conducted according to the current OECD guideline, they would most likely be concluded to be negative. In two studies the authors conclude a clastogenic response (Tanaka 1989 and 1992), however, the evaluation was conducted including gaps. Since gaps may occur by non-genotoxic modes of action, gaps should be excluded from the evaluation. When gaps are excluded from the analysis, these studies should be considered as negative.

The possible mechanism of induction of chromosomal damage has been examined in several studies and it is concluded that the generation of superoxide anion by the test substance leads to a reactive oxygen-induced clastogenicity. No direct action of the test substance itself on the genetic material was indicated.

 

Overall, chromosome damage in mammalian cells in vitro appears to be associated with high concentrations of the test substance and/or high(er) cytotoxicity. A number of the studies reported use concentrations in excess of those now recommended by OECD and the reliability and relevance of the outcomes can be concluded to be reduced by cytotoxicity. Although there remains some evidence that the test substance may induce chromosomal aberrations in cell culture in vitro, this is only observed under conditions no longer supported by OECD for the proper conduct of these studies. These findings support the mechanism of the test substance to generate reactive oxygen species as it is well known that oxidative stress induces chromosomal instability.

 

In vitro micronucleus test

Three in vitro micronucleus tests were reported in human lymphocytes, one with a positive outcome
and two with a negative outcome. In the study with the positive outcome, frequency of chromosome
aberrations in cells treated with the test substance was dose dependent – with 8.5 % at the lowest
concentration tested (10 μmol/L) up to 20 % in the highest dose (1 mmol/L) (Jovchev 2010). The
chromosome aberrations were at least 5 times more frequent than in control. Furthermore, micronucleus
induction also proved to be dose-dependent - it was three to four times more frequent than in the negative
control.The dose-dependent frequency of chromosome aberrations as well as micronucleus induction
prove the test substance to be genotoxic under the conditions of this study. Cytotoxicity was up to 100
μmol/L well within the established acceptable range.

In one to the two negative studies, a modification of the micronucleus assay was used to investigate the ability of the test substance to induce excision-repair removable DNA lesions. The test substance was tested up to concentrations limited by toxicity to the cells but gave a negative result in the assay (Surrales, 1994). In the other study the test substance did not induce an significant increase in the number of micronucleus cells, although an increase in binucleated cells was found at 500 µg/mL for one donor. This finding was considered to be marginal. A great reduction in CBPI was found at all concentrations, indicating that cytotoxicity was induced (Ribas, 1997).

These results are in good agreement with the ability of the test substance to induce reactive oxygen species, which in turn predominantly result in single strand breaks (detected in chromosomal aberration test and comet assay), whereas micronuclei are formed following the occurrence of double strand breaks.

 

In addition, one in vitro micronucleus study including photocatalytic degradation was performed in
Hamster V79 cells (Canavenera, 2007). The test method of this test is not according to the OECD
standards, but studies the effect of photocatalytic irradiation on the mutagenicity and clastogenicity of
the test substance. It was found that the products of test substance photocatalysis are mutagenic rather than clastogenic. This finding has no relevance for the conclusion on mutagenicity of the substance
itself.

 

Sister chromatid exchange test

A number of sister chromatid exchange studies were performed. However the OECD has recently concluded that due to the lack of understanding on the origin of the formation of SCE and their biological consequences this assay is no longer appropriate to use to assess the intrinsic properties of materials. The OECD deleted the SCE test guideline in 2014 and state that other more relevant tests should be used. As a consequence, the weight of the 8 sister chromatid exchanges studies in the WoE approach is very limited. 

 

Mutagenicity (in mammalian cells)

Comet assay

The in vitro comet assay has reported positive results for all 6 studies conducted in a variety of mammalian cell types, but there is no agreed guideline on the conduct or reporting of this assay type and the assay has not yet been validated. Increased tail length was observed either at cytotoxic concentrations or no details of levels of toxicity induced were given. The lack of cytotoxicity measurements limits the confidence in any conclusion of the observed DNA damage as being a primary event rather than secondary to toxicity. In addition, increases in tail length is a known phenomenon for substances inducing oxidative stress.

 

Mammalian gene mutation test (HPRT, XPRT and MLA (tk))

A total of 6 different gene mutation tests have been conducted in mammalian cells similar to or in accordance with OECD TG 476. One of these studies, a mouse lymphoma (tk) assay in L5178Y cells was reported to be positive at concentrations causing significant cytotoxicity (20% relative total growth, McGregor et al., 1988). In a second mouse lymphoma assay in L5178Y cells, increases in mutation frequency were observed in some experiments, in the presence of S9 mix.These effects were considered to be cytotoxicity-induced and not of biological significance and the test substance was reported as non-mutagenic (Clay et al., 1985). Four other assays (HPRT and XPRT) were negative.

As part of projects examining the effects of oxidative stress, positive results were reported for the test substance and hydrogen peroxide in an assay using microsatellite vector constructs of human lung cancer cell lines HCC-15 and NCI-H2009, using neomycin and hygromycin as markers (Zienolddiny et al., 2000); however, the test substance was reported to be non-mutagenic in the parent CHO-K1-BH4 cell line, and the derived AS52 and XRS-5 sensitive cell lines (Salazar and Hsie, 1999).

 

These results support the findings in the Ames tests (negative) and chromosomal aberration test (induction of aberrations at high concentrations and high cytotoxicity), since the mouse lymphoma assay is capable of detecting both mutagenic and clastogenic effects, whereas the HPRT and XPRT assays detect only mutagenic effects.

 

Unscheduled DNA synthesis

Four studies investigating unscheduled DNA synthesis were conducted. The study conducted in rat hepatocytes was reported to be negative (Trueman et al., 1985), whereas a study in human lymphocytes was ambiguous (no dose response) and two further studies in human epithelial cells and rat thymocytes were reported to be positive. Since the OECD deleted the UDS test guideline (TG 482) in 2014 and stated that other more relevant tests should be used, the weight of the UDS tests in the WoE approach is limited.

 

Other type of in vitro genotoxicity assays

Plant assays

A number of plant and algae assays were performed with the test substance in Nostoc muscorum, Vicia faba, Allium fistulosum and barley root cells. For all the studies in this category, the lack of validation of the test systems, together with their questionable relevance to the mammalian situation means that these data are of minimal value in appraising the likely genetic toxicity of the test substance to humans.

 

Sub-mammalian eukaryotic assays

A number of tests were performed with the test substance on Sacchryromyces cerevisiae and Aspergillus nidulans. The potential for gene conversion and interchromosomal rearrangements were assessed in S. cerevisiae and gene mutation and lethal recessive damage in A. nidulans. For both test systems, no positive controls were reported for any of the studies performed, this together with the conflicting results and high doses used makes the data of little value in an overall genotoxic assessment of the test substance.

 

IN VIVO ASSAYS

 

Clastogenicity

Chromosome aberration assay

In vivo, 3 chromosomal aberration assays were conducted in rats and mice. The studies conducted in rats were reported to be negative following single oral administration (Howard, 1987) or oral administration for 5 consecutive days (Anderson, 1978) as was the study in mice following single intraperitoneal dose (Rios, 1995). When mice were dosed intraperitoneally during 10 consecutive days, the response to the treatment was considered ambiguous by the authors. The induction of aberrations however, was only significant when gaps were included in the analysis. Since gaps may occur by non-genotoxic modes of action, gaps should be excluded from the evaluation. When gaps are excluded from the analysis, the study is considered to be negative.

Overall, it can be concluded that the test substance does not induce chromosomal aberrations in vivo.

 

Erythrocyte and bone marrow micronucleus test

A GLP and OECD TG 474 compliant in vivo study in mouse bone marrow erythrocytes conducted by Sheldon et al. (1985) has reported clear negative results. In a study conducted by Peña et al. (1999) isolated increases in micronuclei were reported but these formed no dose or time response, and their relevance must be seriously questioned in terms of the animal numbers involved and the lethal doses employed. In three other tests for micronucleus formation in vivo (two using intraperitoneal administration and one using oral administration), the test substance gave positive results. In these three tests, the doses used were high; it is thus possible to conclude that the test substance may induce chromosome damage at high doses in assays in bone marrow in vivo.

The hypothesis that these effects are caused by the well-established ability of the test substance to generate reactive oxygen species, which are not detoxified at high doses owing to saturation of cellular defensive mechanisms, is likely to be the explanation for the results discussed above. For such an effect it is likely there would be a threshold as, except at high doses, reactive oxygen species are rapidly detoxified.

Overall, the test substance may induce micronucleus formation when administered by the i.p. route, but does not induce micronucleus formation by the oral route. The OECD 474 (2014) test guideline notes, “Intraperitoneal injection is generally not recommended since it is not an intended route of human exposure, and should only be used with specific scientific justification.” This exposure route is not considered relevant to the assessment of human hazard due to the fact that it is not the route of exposure for humans. Hence, the data generated under such conditions should not carry a significant weight in a genotoxicity evaluation.

D’Souza et al. (2005) examined the test substance administered by the dermal route in the bone marrow micronucleus assay in the rat. The conclusion of this study was that the test substance exerts genotoxic and cytotoxic effects. However, the increased MN frequencies reported at the 24- (MN-PCE and MN-NCE) and 48-h (MN-NCE) sampling times in the presence of increased PCE cell mortality are not plausible, given the kinetics of cell maturation from progenitor cells to form PCE and NCE. In order for an induced micronucleus to be present in a PCE, the cell has to encounter the genotoxin 24 - 48 h prior to cell scoring at the nucleated, proerythrocyte stage, before it’s last division and before it develops into the enucleated PCE. Similarly, because the NCE are derived from the PCE, the cells exposed in their progenitor stages would not evolve from PCE into NCE until, at least, 48 h after exposure to a genotoxin.Another important element in the methodology is the use of May-Gruenwald/Giemsa stain on the cells. This stain is not recommended for use with the rat because it will stain granules from mast cells, and these can overlap erythrocytes and appear as micronuclei. This can lead to erroneous apparent increases in the incidence of the micronucleated polychromatic erythrocytes.Thus, the unlikely findings reported, brings all the data and conclusions of this study into question.

 

Mutagenicity

Dominant lethal assay

A dominant lethal assay conducted by McGregor et al. (1974) was reported to be negative. Results from the dominant lethal test clearly demonstrated that no mutagenic effects could be detected if the compound was administered orally to mice up to a dose of 4.0 mg/kg bw/day for 5 days. There was no increase in the percentage of early deaths or number of early deaths per pregnancy. Neither was there evidence of pre-implantation losses or reduction in fertility of the treated males. Hence, the test substance is not mutagenic to germ cells in vivo. These results were confirmed in another study conducted by Pasi et al. (1974).

 

Unscheduled DNA synthesis in vivo

An in vivo UDS study conducted by Trueman et al. (1987) showed no UDS in response to in vivo DNA damage. Under the assay conditions, the test substance did not induce unscheduled DNA synthesis in the rat liver in vivo when tested up to doses causing hepatocyte toxicity.

It should be noted however, that the test responds positively only to chemicals that induce the type of DNA damage that is repaired by nucleotide excision repair (mainly bulky adducts). This is generally considered to be a limitation of the test and as a consequence it is being less widely used and less favoured by some regulatory authorities than in the past.

 

Other type of in vivo genotoxicity assays

The results of tests for genotoxicity with the test substance in Drosophila melanogaster were conflicting, but are in any case irrelevant to the situation in mammals in vivo. Since the OECD deleted the sex-linked recessive lethal test in Drosophila melanogaster test guideline (TG 477) in 2014 and stated that other more relevant tests should be used, the weight of these tests in the WoE approach is limited.

D’Souza et al. (2006) examined sperm parameters of morphology, number, motility and mortality and characterized the abnormal morphology of spermatozoa as “genotoxic” and the result of point mutations. Although the report concluded that the test substance had a statistically significant effect on sperm morphology at all stages of sperm development, the pattern of statistically significant responses are not supported by the biology of sperm development and maturation. The sperm sampled at 7 days post-treatment for instance, would have been exposed to the test substance as fully formed sperm, in which the chromosomal DNA was in a condensed, inactive state. The data presented also suggest that a cell damaged at the early spermatocyte stage, i.e., the 42-day sperm sample, would be able to successfully transit through mitosis, meiosis, and the sperm maturation stages to produce morphologically aberrant, motile sperm. In addition, it has been published in literature that sperm morphology alterations are not evidence of genotoxicity in the testes (Yauk et al., 2015). This lack of plausibility casts serious doubt on the data and conclusions from the entire study and the interpretation that the observed effects are due to point mutations in the germ cells is clearly not supported by the data. 

Van Osch performed a study to examine the mutations and histology in the skin of harlequin disease mice. The frequency and pattern of spontaneous cII mutations in the skin from adult hq disease and wild-type mice was investigated 15 days after a single intraperitoneal injection of test substance. The test substance was chosen in this new biological model because it is a known reactive oxygen species generator. No effect of test substance administration on the hybrid reporter system and only modest effects on the cII (denoted “wild type”) reporter system was reported. 

 

DISCUSSION

In vitro, the test substance has been found overall to be negative in the most widely conducted genotoxicity assay, the Ames test. It does, however, induce chromosomal aberrations in mammalian cells in vitro at high or cytotoxic concentrations. The test substance was also found to induce the mutation frequency in the mouse lymphoma (tk) assay, but not in the HPRT and XPRT assays. This also hints towards a clastogenic action, since this endpoint is only detected in the mouse lymphoma assay and not in the mammalian gene mutation assays. The clastogenic action of the test substance has been investigated and it was concluded that the effects seen in some studies are caused by the well-established ability of the test substance to generate reactive oxygen species, which are not detoxified at high doses owing to saturation of cellular defensive mechanisms. Following from these observations, a number of other in vitro assays have been conducted using the test substance as a material that can generate reactive oxygen species. These have produced mixed results, but overall provide further support for the conclusion that the test substance can cause DNA damage in vitro through formation of reactive oxygen species, especially at high and/or cytotoxic doses of the test substance. For such an effect it is likely there would be a threshold as, except at high doses, reactive oxygen species are rapidly detoxified.

In vivo, the test substance has been examined in both the rat and the mouse using the oral route of administration and shown to be non-genotoxic when tested to maximum tolerated doses in the bone marrow micronucleus assay, the bone marrow cytogenetic assay, the liver DNA repair assay and the germ cell dominant lethal assay. It is concluded that whilst the test substance has been shown to induce a genotoxic response in some in vitro systems, often at high or cytotoxic dose levels, it has no significant genotoxicity in vivo. In addition, long-term / chronic regulatory studies (rat and mouse carcinogenicity studies, dominant lethal studies and rat multi-generation studies conducted over three generations) provide no evidence for any mutagenic effect of the test substance in vivo. There is no occupational or dietary consumption scenario which would result in human exposure to similarly high doses.

 

In conclusion, based on a thorough review and weight of the evidence assessment and taking into account current OECD guidance on the conduct and interpretation of in vitro and short term in vivo studies for genotoxicity, there is insufficient evidence to conclude any mutagenic potential of the test substance. Moreover the lack of any concern over carcinogenicity or generational effects in reproduction studies further confirms a lack of mutagenicity.

 

References

1.    OECD 2016, Overview of the set of OECD Genetic Toxicology Test Guidelines and updates performed in 2014-2015

 

Justification for classification or non-classification

Based on the available information classification for genetic toxicity is not warranted in accordance with EU Classification, Labelling and Packaging of Substances and Mixtures (CLP) Regulation (EC) No. 1272/2008.