Hydroquinone

Endpoint summary

Additional information

IN VITRO STUDIES

 

The assessment of the mutagenic potential of HQ in vitro is based on five key studies (Details presented in Table 1) and test results from further supporting studies (Details see Table 2 and 3).

 

HQ showed no mutagenic activity in a bacterial test system in investigations comprising all Salmonella typhimurium strains required by OECD Guideline 471 and including also a strain sensitive to oxidising or crosslinking agents (Key studies: Haworth et al., 1983; NTP, 1989; Watzinger, 2007a).

 

HQ was positive in the mouse lymphoma mutation assay both in the absence and presence of metabolic activation at relative total growths of 32% and 87%,respectively) (test protocol similar to Guideline study; key study: McGregor et al., 1988; NTP, 1989).

 

In vitro chromosome aberration tests withprimary cultures of human lymphocytes from single healthy nonsmoking male donorswere negative up to cytotoxic test concentrations (Guideline and GLP studies; key study: de Vogel, 2000; Roza et al., 2003;supporting study: Van Delft and de Vogel, 1997). In chromosome aberration tests with mammalian cell lines positive effects were reported. With Chinese hamster ovary cells an increased frequency of structural chromosomal aberrations occurred only in the presence of a metabolic activation system and at the upper dose range with simultaneous cytotoxic effects (no guideline study; key study: Galloway et al., 1987; NTP, 1989). In contrast,in V79 cells HQ induced a significant increase of structural chromosomal aberrations only without metabolic activation. Presence of S9 or SOD reduced the clastogenic activity. An association of the clastogenicity of HQ with the formation of the semiquinone radical or quinone, or with the generation of superoxide anion or peroxide was discussed (Supporting study: Do Ceu Silva et al., 2003).

 

In a micronucleus assay in mouse lymphoma L5178Y cells, HQ induced the formation of micronuclei both in the absence and presence of metabolic activation by liver or kidney S9 mix (Supporting study: Watzinger, 2007). HQ was found to be negative in a micronucleus assay with V79 cells in the absence of an exogenous metabolising system. However, the formation of micronuclei in the presence of arachidonic acid indicated that HQ can be activated by prostaglandin-H-synthase to reactive compounds (Supporting study: Dobo and Eastmond, 1994). Micronucleus assays with primary cultures of peripheral blood human leukocytes yielded both positive test results (single donor and 15 donors, respectively; supporting studies: Yager et al., 1990; Do Ceu Silva et al., 2004) and negative test results (8 donors; supporting study: Doepker et al., 2000). Formation of HQ-induced MN depended both on chromosome breakage and chromosome loss indicating both a clastogenic activity and an aneugenic activity by disturbing microtubule assembly and spindle formation (Dobo and Eastmond, 1994; Yager et al., 1990).

 

Significant increases of the frequencies of SCE (test conditions similar to OECD Guideline 479) were observed in CHO cells (Supporting study: Galloway et al., 1987) and in primary cultures of peripheral blood human leukocytes (donor numbers 15, 1, and 1, respectively, supporting studies: Do Ceu Silva et al., 2004; Erexson et al., 1985; Morimoto et al., 1980). Effects in CHO cells were found to be lower after metabolic activation by liver S9 compared to values without activation. In the study with lymphocytes from 15 donors a possible influence of GST polymorphism (glutathione-S-transferase) on the induction of MN and SCE was investigated. While the presence of GSTT1 and GSTP1 had no effect on induction of micronuclei, metabolism of HQ by GSTM1 significantly reduced the formation of MN. In contrast, there was no influence of GST polymorphism on SCE induction (Do Ceu Silva et al., 2004).

 

Several studies indicate that HQ may induce oxidative DNA damage or covalent DNA binding in peroxidase-proficient cells in vitro but not in vivo (see also “IN VIVO STUDIES” below).

The formation of oxidative DNA damage could be demonstrated in cultures of human HL60 cells, a promyelocytic cell line with significant activity of myeloperoxidase. After short 30 min incubations with 10 µM HQ a doubling of the frequency of 8OHdG compared to controls was observed which returned to background levels on extended incubation presumably due to the rapid repair of this DNA damage. In contrast, in vivo no increase of 8OHdG was found in bone marrow of B6C3F1 mice after i.p. application of HQ (Supporting study: Kolachana et al., 1993). In a further study with human HL60 cells, HQ specific DNA adducts were identified by P1 enhanced³²P-postlabeling. There was a single DNA adduct which was identical with the adduct formed by exposure to BQ, but higher concentrations of HQ and longer treatment times were required to achieve same adduct levels as by BQ exposure (e.g. 7.8 HQ specific adducts per 10⁷nucleotides at 500 µM HQ for 8 hrs). As different adducts are formed in the cells compared to direct reaction of HQ with purified DNA, either chromatin structure or enzymatic processes, probably by activation via myeloperoxidase, are influencing DNA adduct formation within cells (Supporting study: Levay et al., 1991). The same HQ-specific adduct as in HL60 cells was also formed in cultures of human bone marrow cells and in mouse bone marrow macrophages in vitro. There was a significant correlation between cellular activity of peroxidase and adduct levels observed in the cells. No adducts were detected in peroxidase deficient leukaemia cell lines U-937 and Raji. The formation of DNA adducts by HQ may reflect a balance between activation by cellular peroxidase with formation of the semiquinone radical, and inactivation by cellular antioxidants (Supporting study: Levay et al., 1993). In cultured rat Zymbal glands from female Sprague-Dawley rats, the formation of several HQ-derived adducts was detected with total adduct levels of 10.8 and 12.5 per 10⁷nucleotides (750 and 1500 µM HQ for 24 hrs). Rat Zymbal gland cells contain peroxidases which can oxidize HQ to reactive metabolites. However, since only a minor fraction (<20%) of HQ-derived adducts corresponded to BQ-derived DNA modifications, it appears that in Zymbal gland cultures BQ is a minor reactive intermediate of HQ, while the semiquinone radical may predominate (Supporting study: Reddy et al., 1989). In contrast, HQ-specific adducts were not detected in the Zymbal gland, bone marrow, liver and spleen in vivo after oral exposure of rats to a mixture of HQ and phenol (Supporting study: Reddy et al., 1990).

 

 

Table 1: Overview on key studies for assessment of in vitro genotoxicity

 

Test system	Cells	Metabolic activation	Test concentration	Test result	Remarks	Reference
Salmonella typhimurium reverse mutation assay (OECD GL 471)	S typh. TA98, TA100, TA1535, TA1537	-/+ ^a	10 - 666 µg/plate	neg/neg	Cytotoxicity at³ 333 µg/plate	Haworth et al., 1983; NTP, 1989
Salmonella typhimurium reverse mutation assay (screening micromethod assay)	S. typh.TA98, TA100, TA102, TA1537	-/+ ^b	0.38 - 5000 µg/assay	neg/neg	Cytotoxicity at³2500 µg/assay and with strain TA102, -S9 at³278 µg/assay	Watzinger, 2007a
Chromosome aberration assay (OECD GL 473, GLP)	Primary cultures of human lymphocytes from a healthy nonsmoking male donor	-/+ ^c	2-100 µg/mL for 3 h treatment; 5-15 µg/mL for 24 h treatment 10 µg/mL for 48 h treatment	neg/neg	Up to 48 hrs of continuous treatment; cytotoxicity at higher test concentrations of the individual assays (mitotic indices 57 - 43%)	de Vogel, 2000; Roza et al., 2003
Chromosome aberration assay (no guideline study^#)	CHO cells	-/+ ^c	-S9: 150-600 µg/mL +S9: 5-20 µg/mL	neg/pos^§	Treatment for 8.5 h -S9 and 2 h +S9; cytotoxicity at higher test concentrations	Galloway et al., 1987; NTP, 1989
Mouse lymphoma mutation assay induction of trifluorothymidine resistance (similar to OECD GL 476)	L5178Y/TK cells	-/+ ^c	-S9: 0.625-50 µg/mL +S9: 0.625-10 µg/mL	pos/pos^*	Treatment time 4 h; cytotoxicity at³1.25 µg/mL, -S9, and at³5 µg/mL, +S9	McGregor et al., 1988; NTP, 1989

 

^a  S9 mix from livers of Aroclor 1254 treated hamsters or rats

^b S9 mix from liver and kidneys of Aroclor 1254 treated rats

^c S9 mix from liver of Aroclor 1254 treated rats

 

^#  single trial with only 100 metaphases evaluated per test concentration, only 50 metaphases at 20 µg/mL, –S9; cytotoxicity was reported to occur at the higher test concentrations (no further data)

^§ -S9: statistically significant increase of the frequency of structural chromosomal aberrations at 20 µg/mL (only 50 metaphases evaluated), negative due to lack of a significant dose-response relationship; +S9: statistically significant increase of the frequency of structural chromosomal aberrations at 450 and 600 µg/mL (P ≤ 0.05, cytotoxic concentrations)

^* significant and dose-dependant increase of mutant fraction at³1.25 µg/mL -S9 (relative total growth 32%), and³2.5 µg/mL +S9 (relative total growth 87%)

 

 

Table 2: Overview of genotoxicity studies to be used as supporting information (Reliability 2)

 

Test system	Cells	Test conditions	Test results	Remarks	Reference
Chromosome aberration assay (Similar to OECD GL 473)	V79	0, 20, 40, 60, 80 µM at pH 6.0, 7.4, 8.0 without S9 0, 80 µM at pH 7.4 with S9, SOD, CAT or SOD + CAT	pH 6.0: -S9 neg pH 7.4: -/+ S9 pos pH 8.0: -S9 pos at 80 µM	Presence of S9 or SOD, or SOD + CAT significantly decreased CA at 80 µM HQ compared to assays without metabolic activation MI at 80 µM (% of control): 72% at pH 6.0, 53% at pH 7.4, 38% at pH 8.0 Cytotoxicity: at 80 µM at pH 8.0	Do Ceu Silva et al., 2003
Chromosome aberration assay (OECD GL 473, GLP)	human lymphocytes (1 male donor)	0, 8, 16 µg/mL without additional metabolic activation	Neg (no additional metabolic activation)	Cytotoxicity: MI 64% and 54% at 8 and 16 µg/mL	Van Delft and de Vogel, 1997
Micronucleus assay (Micromethod screening assay similar to standard conditions)	mouse lymphoma L5178Y	-S9: 0, 1.25, 2.5, 5.0 µg/mL without and with a 20-hr recovery period +S9: 0, 2.44, 4.88, 9.77 µg/mL -/+ liver or kidney S9 (no recovery)	Pos: both –S9 and +S9	-S9: sign. increase of MN at ≥ 1.25 µg/mL without recovery and at ≥ 2.5 µg/mL with recovery; + liver S9: sign. increase of MN at 9.77 µg/mL + kidney S9: sign. increase of MN at ≥ 4.88 µg/mL Cytotoxicity: -S9/without recovery: only slight at 5 µg/mL with 78% survival -S9/with recovery: moderate toxicity at 5 µg/mL with 61% survival + liver S9: 86% survival at 9.77 µg/mL + kidney S9: 50% survival at 9.77 µg/mL	Watzinger, 2007
Micronucleus assay with differentiation of CREST positive and negative MN (Standard conditions)	V79	0, 17.5, 35, 70, 105, 140 µM without additional metabolic activation Specific investigations on role of oxidative metabolism: + arachidonic acid (AA) to permit activity of PHS + catalase and -/+ AA + GSH and -/+ AA	Without additions: neg + AA: pos + AA + catalase: neg + AA + GSH: neg	Without additions: no sign. dose-related increase of MN In the presence of AA sign. dose-related increase of MN Sign. inhibition of MN induction by AA/PHS in the presence of catalase or GSH	Dobo and Eastmond, 1994
Micronucleus assay with identification of kinetochore-positive MN (Standard conditions)	human lymphocytes (1 male donor)	0, 2 – 150 µM without additional metabolic activation	Pos (no additional metabolic activation)	Sign increase of total MN and kinetochore positive MN at ≥ 75 µM Cytotoxicity: cell viability 68% at 75 µM and 42% at 150 µM	Yager et al., 1990
Micronucleus assay (Standard conditions)	human lymphocytes from 15 donors with different GST genotypes	0, 40, 80 µM without additional metabolic activation	Pos (no additional metabolic activation)	Sign. increase of overall frequency of MN; absence of the GSTM1 gene induced a significantly higher frequency of MN than when this gene was present	Do Ceu Silva et al., 2004
Micronucleus assay (Standard assay conditions)	human lymphocytes from 8 male and female donors	0, 12.5 – 200 µM without additional metabolic activation	Neg (no additional metabolic activation)	Cytotoxicity: reduction of replication index was maximally about 30% at 200 µM HQ	Doepker et al., 2000
SCE (Similar to OECD GL 479)	CHO	-S9: 0, 0.5, 1.67, 5.0 µg/mL +S9: 0, 50, 167, 500, 700, 800 µg/mL	-/+S9: pos	Dose-related increase of SCE at all doses of > 20% compared to controls cytotoxicity at higher test concentrations	Galloway et al., 1987
SCE (Standard conditions)	human lymphocytes from 15 donors with different GST genotypes	0, 80 µM without additional metabolic activation	Pos (no additional metabolic activation)	Sign. increase of overall frequency of SCE no influence of GST genotype on SCE frequency	Do Ceu Silva et al., 2004
SCE ((Similar to OECD GL 479)	human lymphocytes (1 male donor)	0, 5, 50, 70, 100, 300 µM without additional metabolic activation	Pos (no additional metabolic activation)	Sign. dose-related increase of SCE at 50 – 100 µM cytotoxic at 300 µM	Erexson et al., 1985
SCE (Similar to OECD GL 479)	human lymphocytes (1 male donor)	0, 1.6, 8, 40, 200, 1000 µM without additional metabolic activation	Pos (no additional metabolic activation)	Positive response: increase of SCE at 40 and 200 µM cytotoxicity: decrease of MI > 50% at ≥ 40µM	Morimoto et al., 1980

AA: arachidonic acid; CAT: catalase; CHO: Chinese hamster ovary cells; CREST assay: labelling of micronuclei with CREST antibody indicates micronuclei formed by loss of whole chromosomes; GST: glutathione-S-transferase; MI: mitotic index; MN: micronuclei; PHS: prostaglandin-H-synthase; SCE: sister chromatid exchange; Sign.: significant; SOD: superoxide dismutase

 

 

Table 3: Overview of studies on formation of DNA adducts to be used as supporting information (Reliability 2)

 

Type of DNA lesion	Assay test system,	Test conditions	Result^a	Remarks	Reference
Oxidative DNA damage	8OHdG adducts (HPLC detection) Human HL60 cells	10 µM HQ without additional metabolic activation incubation time: 0.5 hrs	Pos	Frequency of 8OHdG was doubled compared to background after 30 min of exposure: 0.160 vs. 0.080 pmol 8OHdG/µg DNA. For longer exposures, adduct levels returned to background. Cytotoxicity: none up to 6 hrs exposure	Kolachana et al., 1993
DNA damage by formation of HQ-specific DNA adducts	DNA adducts by P1 enhanced^32P-postlabeling Human HL60 cells	Standard technique 0, 50, 100, 250, 500 µM without additional metabolic activation incubation time: 1, 2, 4, 8, 16 hrs	Pos	Identification of a single DNA adduct being identical with adduct formed by exposure to BQ; adduct levels: 0.5 – 7.8 per 10^7nucleotides (500 µM HQ for 8 hrs) Cytotoxicity: cell viability reduced to 25% at 500 µM HQ for 8 hrs	Levay et al., 1991
	DNA adducts by P1 enhanced^32P-postlabeling Human HL60 cells human bone marrow cells (HBM) mouse bone marrow macrophages (MBMM) leukaemia cell lines U-937 and Raji	Standard technique (optimised) 0 - 500 µM HQ without additional metabolic activation incubation time: 2 - 7 hrs	Pos in HL60, HBM and MBMM Neg in leukaemia cell lines)	Identification of a single identical DNA adduct in HL60, HBM and MBMM, no adduct in leukaemia cell lines; adduct level: HL60 > HBM > MBMM: In contrast, BQ formed two adducts in all cell types.	Levay et al., 1993
	DNA adducts by P1 enhanced^32P-postlabeling cultured rat Zymbal glands from female Sprague-Dawley rats	Standard technique 750, 1500 µM HQ without additional metabolic activation incubation time: 48 hrs	Pos	Formation of several HQ-derived adducts, total adduct level 10.8 and 12.5 per 10^7nucleotides (750 and 1500 µM HQ for 24 hrs)	Reddy et al., 1989
	DNA adducts by P1 enhanced^32P-postlabeling cultured rat Zymbal glands from female Sprague-Dawley rats	Standard technique 750 µM HQ without additional metabolic activation incubation time: 48 hrs	Pos	Formation of several HQ-derived adducts, total adduct level 13.4 per 10^7nucleotides (750 µM HQ for 24 hrs)	Reddy et al., 1990

 

^aAll assays without additional metabolic activation system as e.g. S9

Cells: Human HL60 cells: human promyelocytic cell line; HBM: freshly isolated human bone marrow cells; MBMM: primary cultures of mouse bone marrow macrophages

 

 

 

IN VIVO STUDIES

 

The assessment of the mutagenic potential of HQ in vivo is based on key studies (see Table 4) and supporting studies (see Table 5) with either intraperitoneal or oral application by gavage and both with somatic and germ cells as target cells.

 

HQ, at lowest effective doses of 50 - 80 mg/kg bw, induced both micronuclei and structural and numerical chromosomal aberrations in the bone marrow of male or female mice after intraperitoneal application while tests for induction of polyploidy or SCE were negative. Effects were found to be highly dependant on the sampling time of the bone marrow (Key study: Adler et al., 1990; supporting studies: Xu and Adler, 1990; Paccheriotti et al., 1991). There was also an increase of structural chromosomal aberrations in mouse spermatocytes with a lowest effective intraperitoneal dose of 40 mg/kg bw (Key study: Ciranni and Adler, 1991). HQ was negative in a dominant lethal assay with oral exposure by gavage for 10 weeks (comparable to OECD GL 478) (Key study: Krasavage et al., 1984).

 

In studies, investigating the formation of oxidative or specific DNA damage in somatic cells in vivo, HQ treatment was without effect.

Measurement of levels of 8-hydroxydeoxyguanosine adducts and of HQ-specific or BQ-specific DNA adducts in kidneys of HQ treated F344 rats indicated that HQ does not induce oxidative DNA damage or covalent binding to DNA after multiple gavage doses (0, 50 mg HQ/kg bw/d,6 wks, on 5 d/wk). On the contrary, there were significant reductions in levels of some endogenous adducts, possibly by virtue of the antioxidant properties of HQ. Consequently, there was a lack of formation of any HQ-treatment related DNA adducts, at exposures conditions up to dose levels producing mild kidney toxicity and significant cell proliferation in F344 male rats, thus favouring the expression of damage due to the compromised condition of the kidneys (Key studies: English et al., 1994, 1997). There was no increase of8-OHdG adducts in bone marrow of male B6C3F1 mice after intraperitoneal injection of 75 mg/kg bw, although increased adduct levels were found in vitro in cultures of human HL60 cells, a promyelocytic cell line (Supporting study: Kolachana et al., 1993). Also, there was no in vivo formation of typical HQ-derived adducts inliver, spleen, bone marrow and Zymbal gland after oral treatment (gavage of 75 mg/kg bw) offemale Sprague-Dawley rats which is in contrast to findings after in vitro exposure of Zymbal gland cultures to HQ (Supporting study: Reddy et al., 1990).

Finally, in the in vivo alkaline comet assay conducted in Fischer F344 rats, there was no DNA strand breaks in cells isolated from the liver (major site of HQ metabolisation), duodenum (secondary site of absorption and metabolisation), kidney (as target tissue in male Fischer F344 rats carcinogenicity studies), and male gonads as a surrogate for germ cells (Beevers, 2016). There was no increase in % tail DNA in these tissues sampled at peak plasma time and up to the maximum tolerated dose, although there was evidence of systemic toxicity and tissue exposure as shown by the clinical and biochemical effects, and histopathological alterations in liver and kidney. The early sampling time after the 2nd administration, consistent with peak plasma concentration before elimination (and 24 hours after the initial dose) ensures appropriate conditions of detection, and it is unlikely that potential damages would have been repaired within this short time-frame. Organ exposure was consistent with previous toxicokinetic data that had showed 14C labelling in all the tissues examined

The study was conducted in the same rat strain tested in long-term repeated dose toxicity studies, and the results indicate that the renal tubular tumors observed in Fischer F344 rats are not likely originating from direct DNA damage occurring in this tissue.

In another in vivo mutation assay, no mutagenic effects were observed in a transgenic rodent assay (TGR) in male Muta mice treated by oral gavage for 28 days, followed by an expression period of 3 days, with analysis of liver, stomach, kidney, and lung (Matsumoto et al., 2014). The model responds to mutation at the lacZgene and allows the detection primarily of base pair substitution mutations, frameshift mutations and small insertions/deletions.

 

 

Table 4: Overview on key studies for assessment of in vivo genotoxicity

Target cells	Test system	Test animals and treatment	Route HQ dose duration	Test result	Remarks	Reference
Somatic cells	Micronucleus assay in polychromatic erythrocytes from bone marrow (comparable to OECD 474)	Mouse (101/E1 X C3H/E1) F1 male, female	Intraperitoneal injection single 0, 30, 50, 75, or 100 mg/kg bw, or 1, 2 or 3 days with doses of 15 or 75 mg/kg bw/d; sampling times 6, 18, 24 or 30 h	Pos^a	Lowest effective dose 50 mg/kg bw or 3times 15 mg/kg bw/d (45 mg in total)	Adler et al., 1990
Somatic cells	Micronucleus assay in polychromatic erythrocytes from peripheral blood (comparable to OECD 474)	Mouse (101/E1 X C3H/E1) F1 male	Intraperitoneal injection single 0, 12.5, 25, 50, or 75 mg/kg bw, sampling times 24, 40, 48 or 72 h	Pos^b	Lowest effective dose 25 mg/kg bw Peak frequency at 40 h Possible threshold between 12.5 and 25 mg/kg bw	Grawé et al., 1997
Somatic cells	Micronucleus assay in polychromatic erythrocytes from bone marrow	Mouse (Swiss CD-1)	Intraperitoneal injection or oral gavage single 80 mg/kg bw, sampling times 18, 24, 42, 48 h	Pos	Weakly positive by oral gavage (2x); significant effects by ip route Peak frequency at 18 h	Ciranni et al., 1988
Germ cells	Assay for structural chromosomal aberrations in spermatocytes (similar to OECD 483)	Mouse (102/E1 X C3H/E1) F1 male	Intraperitoneal injection single 0, 40, 80 and 120 mg/kg bw; sampling time 24 h	Pos^c	Lowest effective dose 40 mg/kg bw	Ciranni and Adler, 1991
Germ cells	Dominant lethal assay (comparable to OECD 478)	Rat CD male	Gavage 0, 30, 100, 300 mg HQ/kg bw/d 10 wks, 5 d/w, prior to mating	Neg^d	Toxicity at 300 mg/kg (mortalities 2/25, clinical signs of intoxication)	Krasavage et al., 1984
Somatic cells	Formation of DNA adducts in kidney cells by Nuclease P1-enhanced 32P-postlabeling assay	F344 rat male, female	Gavage 0, 50 mg HQ/kg bw/d 6 wks, on 5 d/wk	Neg^e	Evidence of concurrent kidney toxicity indicated by enzymuria, cell proliferation and histopathologic changes	English et al., 1994
Somatic cells	Oxidative DNA damage in kidney cells formation of 8-OHdG DNA adducts	F344 rat male, female	Gavage 0, 50 mg HQ/kg bw/d 6 wks, on 5 d/wk	Neg^e	Evidence of concurrent kidney toxicity indicated by enzymuria, cell proliferation and histopathologic changes	English et al., 1997
Somatic cells	lacZ transgenic mutation study in somatic cells from liver, stomach, kidney, and lung (OECD 488)	LacZ Muta^(TM)mouse male	Gavage 0, 25, 50, 100, 200 mg/kg bw 4 wks, once a day	Neg	Expression period: 3 days before tissue sampling No clinical signs but decrease in body weight gain in all treatment groups.	Matsumoto et al., 2014
Somatic cells	Single cell gel/comet assay in rodents for detection of DNA damage in liver, kidney, and duodenum (OECD 489)	F344 rat male, female	Gavage 0, 105, 210, 420 mg/kg bw 2 administrations 24 hours apart Sampling time: 30 min after 2^nddosing	Neg	Systemic exposure evidenced by urine coloration and signs of acute tubular necrosis at 210 and 420 mg/kg (MTD), increased hepatocyte mitosis and ALAT, ASAT activities, bilirubin.	Beevers, 2016
Gonads/ germ cells	Single cell gel/comet assay in rodents for detection of DNA damage in male gonads (OECD 489)	F344 rat male	Gavage 0, 105, 210, 420 mg/kg bw 2 administrations 24 hours apart Sampling time: 2 hours after 2^nddosing	Neg	No microscopic findings in testes at MTD	Beevers, 2016

^a Single application: statistically significant increases in micronucleus frequency for sampling times of 24 h at 50-100 mg/kg with a dose-dependant increase, and at sampling times of 18-30 h at 75 mg/kg, maximum responses for 24 h sampling time; multiple dosing: significant increases of micronuclei frequency for 3 doses of 15 mg/kg, and 1, 2 or 3 doses of 75 mg/kg.

^b Single application: statistically significant increase in frequency of micronucleated polychromatic erythrocytes for sampling times 24 and 40 hours at 25 -75 mg/kg, and for sampling time 48 hours at 50-75 mg/kg. The peak response was 40 hours in all dose groups. No significant effect at 12.5 mg/kg. The plot of the Area below the curve versus doses showed a biphasic response suggesting a no-effect threshold.

^c Statistically significant increase of frequencies of aberrant cells (chromatide aberrations exclusive gaps)

^d  no effects on insemination rate, pregnancy rate, counts of corpora lutea, of implantation sites with categorization as early deaths, late deaths or viable embryos, % pre-implantation loss, % post-implantation loss; the positive control substance showed a positive test result

^e no increased or modified formation of HQ-specific or BQ-specific DNA adducts compared to control rats; on the contrary, there were significant reductions in levels of some endogenous adducts

 

 

Table 5: Overview of in vivo genotoxicity studies to be used as supporting information (reliability 2)

 

Target cells	Test system	Test animals and treatment	Route HQ dose duration	Test result	Remarks	Reference
Somatic cells	Oxidative DNA damage in mouse bone marrow formation of 8-OHdG adducts (HPLC detection)	Mouse B6C3F1 male	Intraperitoneal injection single 0, 75 mg/kg bw sampling time 1 h	Neg	No significant increase compared to background frequency; while also 75 mg/kg phenol was also negative, a combination of 75 mg/kg HQ and 75 mg/kg phenol induced a significant increase of8-OHdG	Kolachana et al., 1993
Somatic cells	Formation of DNA adducts in liver, spleen, bone marrow and Zymbal gland by Nuclease P1 enhanced^32P-postlabeling	Sprague-Dawley rat female	Gavage Other test substance: 75 mg HQ + 75 mg phenol4 d sampling time 24 h	Neg	No in vivo formation of the typical HQ-derived adducts identified after in vitro incubation with Zymbal gland cultures	Reddy et al., 1990
Somatic cells	Assay for structural chromosomal aberrations in bone marrow (comparable to OECD GL 474)	Mouse(101/E1 X C3H/E1) F1 male, female	Intraperitoneal injection single 0, 45, 75, or 100 mg/kg bw; sampling times 6, 12, 18, 24 or 36 h	Pos	Lowest effective dose 75 mg/kg bw, 24 h sampling time at 100 mg/kg increase of CA at sampling times of 6 – 24 h	Xu and Adler, 1990
Somatic cells	Assays for numerical chromosomal aberrations, micronuclei and SCE in bone marrow	Mouse (C57B1/Cne X C3H/Cne) F1 male	Intraperitoneal injection single 0, 40, 80, or 120 mg/kg bw; sampling times 18 or 24 h	CAnum:  pos   MN: pos   SCE: neg	At 80 mg/kg and 18 h sampling time sign. increase of hyperploidy and MN, no dose related effect no effect on polyploidy or SCE. A significant increase of average generation time was observed at a sampling time of 18 h and doses of 80 and 120 mg/kg. HQ was discussed to affect a chromosomal component rather than a spindle component of chromosomal segregation.	Paccheriotti et al., 1991

 

CA_num: numerical chromosomal aberrations; MN: micronuclei; SCE: sister chromatid exchange

 

  

SUMMARY AND DISCUSSION

 

In the key in vitro studies, HQ showed no mutagenic activity in a bacterial test system in investigations comprising all Salmonella typhimurium strains required by OECD Guideline 471 and including also a strain sensitive to oxidising or crosslinking agents. An in vitro chromosome aberration test with primary cultures of human lymphocytes from a healthy nonsmoking male donor was negative up to cytotoxic test concentrations (Guideline and GLP study), while in a further test with Chinese hamster ovary cells (no guideline study) an increased frequency of structural chromosomal aberrations occurred only in the presence of a metabolic activation system and at the upper dose range with simultaneous cytotoxic effects. HQ was positive in the mouse lymphoma assay both in the absence and presence of metabolic activation at relative total growths of 32% and 87%, respectively) (test protocol similar to Guideline study).

Based on the total evidence including supporting studies, HQ was found to be a direct acting genotoxic agent in cultures of mammalian cell lines and in human peripheral lymphocytes inducing mutations, DNA adducts, chromosome aberrations, micronuclei and sister chromatid exchange. Effects mostly occurred at test concentrations associated with moderate to significant cytotoxic effects. There are clues that the outcome of genotoxicity assays with HQ is dependent on the routes of metabolisation available in the applied test system. HQ can be activated by prostaglandin-H-synthase and peroxidases to reactive compounds as e.g. semiquinone radical and benzoquinone. A metabolizing system based on hepatic enzymes (liver S9) can favour deactivation of the genotoxic activity of HQ while induction of genotoxic activity in the presence of S9 was reported in other assays. Metabolism by pathways yielding the semiquinone radical and/or reactive oxygen species was found to be a condition supporting clastogenic effects. In test systems applying human peripheral lymphocytes, test results may be dependent on the individual genotypes of metabolizing enzymes, as e.g. GST. The important effect of metabolic routes on genotoxicity testing results in vitro is further supported by the outcome of genotoxicity testing in vivo.

 

In vivo, at lowest effective doses of 50 - 80 mg/kg bw, HQ induced both micronuclei and structural and numerical chromosomal aberrations in the bone marrow of male or female mice after intraperitoneal application while tests for induction of polyploidy or SCE were negative. Effects were found to be highly dependent on the sampling time of the bone marrow. Micronucleus frequency was only weakly positive in mice administered HQ by oral gavage, compared to intraperitoneal injection. In another study investigating micronucleus formation in polychromatic erythrocytes from the peripheral blood of male mice treated intraperitoneally, the lowest effective dose was found at 25 mg/kg, with a peak frequency at sampling time 40 hrs. Integration of the dose-dependent frequencies over 72 hours and plotting area under the curve as a function of the dose showed a biphasic shape which suggested a threshold of effect between 12.5 mg/kg and 25 mg/kg. There was also an increase of structural chromosomal aberrations in mouse spermatocytes with a lowest effective intraperitoneal dose of 40 mg/kg bw. In contrast, no in vivo genotoxicity was found after oral HQ exposure (bolus application via gavage) for extended periods of 6 to 10 weeks in germ cells, or in liver, spleen, kidney, bone marrow and the Zymbal gland of rats or mice. A dominant lethal assay with oral treatment of male CD rats was negative. Measurement of levels of 8-hydroxydeoxyguanosine adducts and of HQ-specific or BQ-specific DNA adducts in kidneys of HQ treated F344 rats indicated that HQ does not induce oxidative DNA damage or covalent binding to DNA after multiple gavage doses. On the contrary, there were significant reductions in levels of some endogenous adducts, possibly by virtue of the antioxidant properties of HQ. Also, there was no in vivo formation of typical HQ-derived adducts in liver, spleen, bone marrow and Zymbal gland after oral treatment off male Sprague-Dawley rats which is in contrast to findings after in vitro exposure of Zymbal gland cultures to HQ. There was no increase of 8-OHdG adducts in bone marrow of male B6C3F1 mice after intraperitoneal injection, although increased adduct levels were found in vitro in cultures of human HL60 cells, a promyelocytic cell line.

Consequently, there was a lack of formation of any HQ-treatment related DNA adducts, at exposures conditions up to dose levels producing mild kidney toxicity and significant cell proliferation in F344 male rats, thus favouring the expression of damage due to the compromised condition of the kidneys. Additionally, the upper dose of 50 mg/kg bw/d was the high dose level of the NTP rat bioassay. These observation suggest that the kidney toxicity observed with HQ in male F344 rats is not likely due to direct (i.e. adducts) or indirect (i.e. oxidative) DNA damage, and that benign kidney tumours observed in the 2-year carcinogenesis bioassay with male F344 rats are produced via a non-genotoxic mechanism. 

This is further supported by two complementary in vivo guideline studies (reliability 1) investigating genotoxic effects in tissues: a Transgenic rodent assay in lacZ transgenic male Muta mice exposed for 28 days, and an in vivo alkaline Comet assay in Fischer F344 rats treated by oral gavage (2 administrations 24-hr apart with sampling at peak plasma time).

The in vivo alkaline comet assay was used to investigate the potential formation of DNA strand breaks in cells isolated from liver, duodenum, kidney in male Fischer, and male gonads as a surrogate for germ cells. There was no increase in % tail DNA in these tissues sampled at peak plasma time and up to the maximum tolerated dose, although there was evidence of systemic toxicity and tissue exposure as shown by the clinical and biochemical effects, and histopathological alterations in liver and kidney. The early sampling time after the 2nd administration, consistent with peak plasma concentration before elimination (and 24 hours after the initial dose) ensures appropriate conditions of detection.

The study was conducted in the same rat strain tested in long-term repeated dose toxicity studies, and the results indicate that the renal tubular tumors observed in Fischer F344 rats are not likely originating from direct DNA damage occurring in this tissue.

Matsumoto et al. also reported no mutagenic effects in the lacZ gene a transgenic rodent assay in male Muta mice treated by oral gavage for 28 days following analysis of liver, stomach, kidney, and lung.

 

Intraperitoneal injection is a mode of application through which unmodified HQ may reach targets of in vivo genotoxicity studies, as e.g., the bone marrow and germ cells, as biotransformation and detoxification of HQ taking place in the liver after oral uptake is bypassed. That hepatic detoxification is substantially reflected in the lower acute toxicity of orally vs. intraperitoneally applied HQ. The comprehensive investigations on the metabolism and toxicokinetics of HQ (see IUCLID Section 7.1.1) have shown, that HQ is quickly metabolised after uptake via the gastro-intestinal tract. A different spectrum of metabolites is available after intraperitoneal application than after oral application presumably with a higher amount of the most reactive metabolites, in particular glutathione-conjugates. Additionally, pharmacokinetic modelling showed a fundamental difference in metabolism between humans and rats as a representative for other experimental animals. Deactivation steps predominate in human liver cells, and bioactivation steps predominate in rat liver cells. Body burdens of higher substituted glutathione HQ conjugates, which are considered to be the reactive metabolites, will be much higher in rats than in humans. Additionally, as oral or dermal exposure are the relevant exposure routes for humans, a finding of in vivo genotoxicity after intraperitoneal application only, is expected to be of no biological significance for the human exposure situation.

Furthermore, in vitro investigations show that the outcome of in vitro genotoxicity studies investigating the genotoxic potential of HQ is dependent on the activity of metabolic pathways leading either to activation or deactivation of genotoxic effects. Consequently, the biological relevance of positive genotoxicity findings in vitro is questionable with regard to the human exposure situation.

 

Results of toxicokinetics studies by intratracheal instillation and PBPK models refined to predict metabolite formation from inhalation exposure (using intratracheal data as input parameters) also showed a rapid absorption and metabolisation of hydroquinone, despite the absence of local pulmonary metabolic activity.

Refinement of PBPK models for dermal exposure indicate a slower absorption due to passage through the multiple skin layers. However, the major metabolites detected were also primarily glucuronide conjugates with only very low levels of glutathione conjugates.

Both for inhalation and dermal exposure, the major HQ metabolites identified were glucuronide conjugates, and sulfate conjugates similar to the metabolites detected by the oral route, and contrasting with the high levels of glutathione conjugates observed in intraperitoneal studies.

Justification for selection of genetic toxicity endpoint
Genotoxic effects were observed in several in vitro assays, and in vivo, in mouse micronucleus assays but essentially following intraperitoneal administration. In contrast, in vivo TGR assay in mouse, and an in vivo Comet in rats treated by oral gavage showed no DNA damage (mutations or DNA strand breaks) in all organs examined, including male rat gonads (as a surrogate to germ cells). The dataset information suggest the positive effects are route-specific and toxicokinetic-dependent.

Short description of key information:
IN VITRO
- S. typhimurium reverse mutation assay (OECD471 or screening micromethod assay): negative in the absence and presence of metabolic activation (Haworth et al., 1983; NTP, 1989; Watzinger, 2007)
- chromosome aberration assay in human lymphocytes (OECD473, GLP): negative with or without metabolic activation (de Vogel, 2000; Roza et al., 2003)
- chromosome aberration assay in CHO cells (no guideline study): negative in the absence of metabolic activation, positive in the presence of metabolic activation (Galloway et al., 1987; NTP, 1989)
- mouse lymphoma mutation assay (similar to OECD476): positive with or without metabolic activation (McGregor et al., 1988; NTP, 1989)

IN VIVO
- micronucleus assay in polychromatic erythrocytes of mouse bone marrow (comparable to OECD474): positive after single intraperitoneal application (Adler et al., 1990)
- micronucleus assay in polychromatic erythrocytes from mouse peripheral blood (comparable to OECD474 with modifications) by flow cytometry analysis: positive after single intraperitoneal application, with indication of a threshold (Grawé et al., 1997)
- structural chromosomal aberrations in mouse spermatocytes (similar to OECD483): positive after single intraperitoneal application (Ciranni and Adler, 1991)
- dominant lethal assay with treatment of male rats (comparable to OECD478): negative after oral application for 10 w (Krasavage et al., 1984)
- formation of DNA adducts in kidney cells, F344 rats: negative after oral application for 6 w (English et al., 1994)
- formation of 8-OHdG adducts in kidney cells, F344 rats: negative after oral application for 6 w (English et al., 1997)
- Transgenic mutation assay in male Muta(TM) mice (OECD488): negative in liver, kidney, stomach, lung after oral gavage for 4 w (Matsumoto et al., 2014)
- Alkaline Comet assay in male and female F344 rats (OECD489): negative in liver, kidney, intestine, male gonads, after 2 oral treatments (Beevers, 2016)

Endpoint Conclusion: Adverse effect observed (positive)

Justification for classification or non-classification

Based on the criteria of the CLP Regulation (EC) 1272/2008, HQ has been classified to Germ cell mutagenicity category 2, H341 suspected of causing genetic effects (genotoxic effects observed in animal experiments with intraperitoneal application or in in vitro studies).

Additional studies in rats (in vivo alkaline comet assay) and mice (transgenic rodent assay) treated by oral gavage showed no DNA damages in the main tissues investigated (liver, kidney, stomach or duodenum, lung and male gonads) indicating that the tissue findings in long-term repeated dose toxicity studies were not directly related to genotoxic activity in those tissues. Information provided by toxicokinetics data and PBPK models indicate that the main metabolites resulting from dermal and inhalation exposure would be essentially similar (identity and proportions) to those observed following oral administration, and which substantially differ from the metabolic profile detected after intraperitoneal injection.

Overall, it can be concluded that the adverse effects reported in repeated oral dose toxicity studies, in particular in rat kidneys, are occuring through a non-genotoxic mechanism.

The available data provide evidence that a more severe GHS hazard category is not justified.