Genotoxicity of oxidative hair dye precursors

Introduction

Oxidative permanent and semi-permanent hair dyes usually contain p-phenylenediamine (PPD), toluene-2,5-diamine (p-toluylenediamine; PTD), substituted para-diamines such as 2-methoxymethyl-p-phenylenediamine (ME-PPD), or ortho- or para-aminophenols. These colourless precursors react with a developer such as hydrogen peroxide in a series of oxidative reactions, eventually yielding large coloured molecules which in the course of polymerisation become trapped in the hair cortex, thus changing the hair colour.1 Many of these chemicals are classified with regard to their toxicological properties and labelled according to the Regulation (EC) No 1272/2008 on classification, labelling and packaging of substances and mixtures (CLP).2 For example, the most common precursors, PPD, PTD and ME-PPD, are classified as acutely toxic in contact with skin, and skin sensitizers (i.e. can also induce allergic reaction (elicitation) in already sensitized individuals).2^,3 However, under the CLP Regulation, none of these precursors are classified as carcinogenic or mutagenic.

In the Scientific Committee on Consumer Safety (SCCS) opinion on PPD published in 2012,4 PPD is considered non genotoxic, but with a note that information submitted was insufficient to allow a final risk assessment to be carried out. As the clastogenic effects found in vitro were not confirmed in in vivo tests, the SCCS considered PTD to have no in vivo genotoxic potential,5 and a similar conclusion was reached for ME-PPD in spite of equivocal results in some of the in vivo assays.6 Animal studies on carcinogenicity of oxidative hair dye ingredients, which were included in the SCCS opinions and in the International Agency for Cancer Research (IARC) opinion on risks of personal hair dye use and occupational exposure,7 mostly dated from 1970s and 1980s, with serious deviations from the current Organization for Economic Co-operation and Development Test Guidelines (OECD TGs) study design.8 More recent in vitro experiments indicated that PPD causes DNA damage, increases intracellular levels of reactive oxygen species leading to activation of apoptotic signalling pathways and induces cytotoxic effects through alteration of microRNA expression levels.9^,10^,11 In the new analysis of mutagenic components of oxidative hair dyes, however, a high rate of positive in vitro results was considered poorly predictive for in vivo assays but also for the outcomes of animal carcinogenicity assays.12

There is a concern that prolonged personal use of oxidative hair dyes might lead to bladder, hematopoietic and breast cancer.13^,14^,15 In addition, hairdressers could also be at increased risk due to higher exposure to hairdressing chemicals at work compared to consumers.16^,17 Increased risk of bladder cancer among hairdressers in comparison to the general population was noted in the study from New Zealand.18 Carcinogenicity could be the result of a genotoxic mode of action. A recent meta-analysis showed increased micronucleus frequencies in buccal swabs from hairdressers compared to non-exposed persons.19

There is a clear need to assess the newer experimental findings, which would facilitate the interpretation of epidemiological studies. Therefore, the aim of the present systematic review was to evaluate recent in vitro and in vivo studies published in the last two decades investigating genotoxicity of oxidative hair dye precursors.

Methods

Publication selection

The process of publication selection was performed in accordance with the PRISMA-P guidelines22 (Figure 1). Two researchers (Ž.B. and R.T.) independently screened publications (2000-2021) listed in search results for any relevant product identifiers and toxicological endpoints. Full text articles were retrieved and assessed for eligibility in the same manner as described above.

Open multimedia modal

Results

A total of 9 publications were analysed with the results of 17 studies published from 2000 onwards that investigated genotoxic effects of hair dye precursors (Table 1). All tested the effects of PPD and only three also included PTD. Only two studies were considered to be of lower quality,27^,28 mainly due to the methodological problems and incomplete result presentation.

Open multimedia modal

In vitro bacterial gene mutation assay was performed in three studies,29^,30^,31 all of good quality, adhering to relevant OECD TGs.32 All studies had comparable protocols for in vitro bacterial assay following the current OECD guideline which was adopted in 1997,33 used similar PPD dose ranges, and provided comparable results. PPD tested positive only in one Salmonella typhimurium strain (TA 98) in a dose-related manner, but only in the presence of metabolic activation with rat liver S-9 mix. In other strains tested no increase in the number of revertant colonies was seen with or without metabolic activation. PPD metabolites MAPPD (N-monoacetyl-p-phenylenediamine (N-acetyl-p-phenylenediamine, 4-aminoacetanilide) and DAPPD (N,N’-diacetyl-p-phenylenediamine (N,N’-p-phenylenebisacetamide), formed by N-acetylation in human and mammalian skin and in human hepatocytes, were negative in the Ames test.29 PTD exhibited a weak increase in revertant colony counts without activation in TA98 and very strong increase in TA98, TA100, TA1535, and TA1537 following metabolic activation with S9.30 N-acetylated PTD metabolites 2-mono-Ac-PTD (4-amino-2-methylacetanilide), 5-mono-Ac-PTD (4-amino-3-methyl-acetanilide) and di-Ac-PTD (2,5-diacetamino-toluene), were also negative in the Ames test.30 Both PPD and PTD tested positive in an additional bacterial test in Salmonella typhimurium TA1535/pSK1002 strain with metabolic activation.28 The method, however, was not fully validated and the study did not follow OECD guidelines.

Three studies also included an in vitro mammalian gene mutation assay with PPD29^,31 and PTD.30 For PPD, the results were clearly positive in the chromosomal aberration assay with or without S-9, in Chinese hamster lung cell line31 and in compliance with a more recent OECD guideline.34 However, in the mouse lymphoma Hprt assay statistically significant increase in mutation frequency was observed following treatment with PPD at 100 μg/mL only with metabolic activation. The authors argued that there was no evidence of a dose-response and that the increase in mutation frequency was small in magnitude and within the historical control range. The chance event of no biological relevance should therefore be considered.29 PTD, when tested in Chromosomal aberration assay with V79 cells, showed clastogenic potential but no evidence of aneugenicity.30

A newly-developed rodent in vivoPig-a gene mutation assay in mutant red blood cells or reticulocytes (defined as RBC^CD59- or RET^CD59-, respectively) conducted in line with recently adopted OECD TGs 470 also showed the mutagenic effect of PPD31 in accordance with classical genotoxicity test battery (gene mutations, structural and numerical chromosomal aberrations) recommended by the OECD.32

PPD tested positive in mammalian cell micronucleus test in vitro in cultured human lymphocytes,29 performed in line with the 2003 draft OECD TGs 487,35 which was confirmed by in vivo murine erythrocyte micronucleus test31 conducted in line with the most recent guideline.36 N-acetylated derivatives of PPD and PTD, did not show a genotoxic potential in the human lymphocyte micronucleus assay in vitro29^,30

Four in vitro studies investigated DNA breaks caused by PPD,30^,37^,38^,39 and PTD,30 all by alkaline Comet assay, and one of them additionally also by a terminal deoxynucleotidyl transferase dUTP nick end labelling (TUNEL) assay.39 The test systems were a human uroepithelial cell line,37 cultured human lymphocytes,38 a canine kidney cell line,39 Chinese hamster V79/V79NAT1*4 cell lines and human keratinocyte HaCaT cells.30 The protocols were designed by the authors with reference to their previous publications, and although mostly described in detail, a reference to standardized procedures was lacking. For example, the test substance incubation periods before the comet assay analysis varied: 24 h in Huang et al., 2007,37 2 hours in Chye et al., 2008,38 3 hours in Chen et al., 201039 and three and 48 h in Zeller et al., 2014.30 Nevertheless, in all four studies, DNA damage scores were significantly increased in a dose-dependent manner in cells treated with PPD and PTD in comparison to the controls. None of the N-acetylated PPD or PTD metabolites induced a biologically relevant increase in any of the Comet parameters.30 The results of the TUNEL assay after the incubation of canine kidney cell line with 37.5 μg/mL PPD for 3–12 h also showed increasing DNA damage compared to control, in a time-dependent manner.39

Two studies used the in vivo rodent alkaline comet assay to detect DNA damage in mice and rat isolated tissues, following administration of PPD and PTD at doses of 80 and 60 mg/kg b.w., respectively, as a single i.p. dose27 or 3 consecutive daily doses of 25–100 mg/kg/day of PPD by oral gavage.40 The study design broadly followed the later OECD TG 486 which was not formally adopted until 2016.41 PPD did not induce DNA damage in any of the organs studied in both species, while PTD tested positive in rat stomach only.27

Discussion

Findings from in vitro and in vivo experimental studies included in this review point to genotoxic potential of PPD and PTD, which are common precursors of oxidative hair dyes. Although the number of reviewed studies is limited, this finding indicates that a risk of genotoxicity after exposure to PPD and PTD can hitherto not be ruled out.

Nine publications were identified describing 17 assays, covering main genotoxicity endpoints with 10 different testing protocols (Table 1). Except for the bacterial reverse mutation assays for which the OECD TGs were not revised since 1997, only one mammalian cell chromosomal aberration test in vitro and one erythrocyte MN test in vivo were performed according to the updated OECD TGs,31 while two more in vivo assays closely followed the current TGs on the alkaline comet assay and Pig-a-gene mutations,40^,31 respectively.

An adequate chemical characterization of samples tested, which is lacking in several studies, is also critical for reliable results, as impurities may significantly affect the genotoxic response both in vitro and in vivo.

The studies of good quality designed in compliance with the OECD TGs29^,30^,31 showed positive results of in vitro bacterial mutation assays for both PPD and PTD, depending on the bacterial strain and mainly with previous metabolic activation. Compared to mammalian cells, high levels of reductive enzymes of bacteria can efficiently activate nitro and azo compounds to electrophilic metabolites. A lack of Phase II enzymes in liver S-9 preparations for metabolic activation can furthermore lead to misleading positive test results in standard in vitro genotoxicity assays. According to the conclusions of the EURL ECVAM Workshop42 an Ames positive result may be due to the conditions that favour high levels of oxidative metabolism. Similar experiments in mammalian cells and/or in vivo assays may show lower genotoxic potential, unless the doses of chemical used exceed the capacity of detoxication pathways, leading to metabolic activation of toxic pathways and resulting in genotoxicity but with a possible threshold dose-response.43 However, when tested in a newly developed Rodent Pig-a assay in vivo PPD also induced gene mutations in rat peripheral blood after 28-day oral exposure.31 Additionally, positive results indicating clastogenic potential of PPD and PTD were obtained by in vitro chromosomal aberration assay, although PPD was negative in the mouse lymphoma assay. PTD tested positive in all in vitro assays identified in this search,28^,30 however in the thymidine kinase gene mutation study (TK6 assay) using human lymphoblastoid TK6 cells, which express human metabolic enzymes, no increase in mutation frequency was observed after 4-h treatment in the absence or presence of S9 metabolic activation nor after 24-h treatment without S9.44 PPD was positive in one in vitro micronucleus assay in cultured human lymphocytes,29 and showed increased micronucleus frequencies in mice erythrocytes following high dose acute oral exposure in vivo31 Another indicator of genotoxicity, namely DNA breaks, were investigated in four in vitro studies37^,38^,39^,30 by an alkaline comet assay, and significantly higher DNA damage scores after PPD and PTD exposure was noted in all of them. The in vitro comet assay is considered to provide complementary information on genotoxicity and mechanism of action of chemicals, but it is not a standard battery test for mutagenic, clastogenic and aneugenic potential, and consequently has not been implemented into official regulatory testing guidelines.45 High concentrations of test agents may cause cytotoxicity or cell death, and give rise to false positive results, but further research is needed to establish a threshold value to distinguish between true and potentially false positive genotoxic effects detected by the comet assay.46 The effect of cytotoxicity may explain why in vitro comet assay results for PPD and PTD were not fully confirmed by the in vivo comet assay where PPD was clearly negative while PTD exhibited positive result in rat stomach only, following i.p. application.27^,40 Nevertheless, it should be noted that in vivo comet assay also has important limitations: measurement of transient DNA lesion instead of irreversible gene and chromosomal mutations; sensitivity to indirect genotoxicity mechanisms linked to toxicity or cellular stress but low sensitivity to detect aneugenicity and DNA crosslinking.47

Considering previous Scientific Committee on Consumer Safety (SCCS) reviews, the results of the present systematic review may help in the interpretation of epidemiological studies on cancer risk among hairdressers. Overall, SCCS did not consider PTD or PPD alone as genotoxic, but pointed out positive findings from in vivo and in vitro genotoxicity studies of PPD in combination with couplers or hydrogen peroxide.4^,5 Recent studies reported increased micronucleus frequencies in hairdressers,19^,48 a population heavily exposed to hairdressing products including oxidative hair dyes.49 In addition, significantly higher DNA damage scores were found among hairdressers compared to controls, based on the comet assay performed on blood samples.50 Nevertheless, linking experimental results with epidemiological studies is difficult owing to, among other things, the kinetic properties of oxidative hair dye precursors.7 Dermal absorption of PPD and PTD is poor.4^,5 No reliable data exist on inhalation exposure uptake, which may be relevant in occupationally exposed population. There are indications that PPD, PTD and related compounds undergo N-acetylation in the skin which is a detoxification mechanism resulting in non-genotoxic N- acetylated metabolites,30^,51 although at higher exposure levels, the detoxification capabilities of the skin may become a limiting factor, exposing the organism to the parent molecules.52 Lastly, there is no evidence of oxidative metabolism of absorbed PPD or PTD by human hepatic cytochrome P450 enzymes, a reaction that is presumably an activation step of carcinogenic aromatic amines.53^,54

On the other hand, as demonstrated in several experimental studies, DNA damage induced by PPD and PTD is obviously dose-dependent which may be important in occupational conditions, as significantly higher exposure frequency in hairdressers than in clients and personal hair dye users were already confirmed.17 Little is known about the level of occupational external or systemic exposure of hairdressers to hair dye ingredients. Occupational exposure of hairdressers can be quantified by the use of biological monitoring of urinary PPD and PTD. However, the extent of exposure in hairdressers was found to be 2–3 orders of magnitude lower than after personal application of hair dyes, and the use of protective gloves did not influence the results.55 Another biomonitoring study showed that proper use of gloves in hairdressers may significantly reduce systemic exposure to PTD but not to PPD.56 To our knowledge, only one complete investigation of the occupational exposure of hairdressers to oxidative hair dyes was published, but it was performed under controlled conditions with a small number of subjects and only focused on PPD exposure.57 The authors concluded that occupational use of oxidative hair dyes produces negligible local or systemic exposure to PPD thus posing no risk to human health and that current protective measures were sufficient to limit exposure to acceptable levels. However, such definite conclusions evidently need further confirmation, as the problem of genotoxic potential of PPD and its derivatives is still not fully resolved. It was suggested that well designed epidemiological studies with high quality measurements of personal exposure levels in hair salons and possible biological exposure indices in hairdressers, as well as using for example the Pig-a assay for the detection of genetic effects, may further elucidate the suspected genotoxic potential of PPD and its derivatives.31 As animal testing of cosmetic ingredients is not allowed under the current EU Cosmetics Regulation, the use of micronucleus and comet assays in three-dimensional reconstituted human skin models is proposed as an alternative.58

An increased incidence of tumour formation was already reported in some animal studies with exposure to hair dye precursors.7 Several meta-analyses found significant 20%–30% higher risk for bladder cancer in hairdressers compared to the general population or non-exposed occupations, and significantly increased frequency of other types of cancers in hairdressers, specifically laryngeal cancer and multiple myeloma.59^,60^,61^,62 Hairdressers and barbers were therefore classified under occupational exposure group 2A as probable human carcinogens by the International Agency for Research on Cancer, but personal use of hair dyes was categorized as not classifiable as to its carcinogenicity to humans (Group 3).7 The SCCS also considered PPD, PTD and ME-PPD in hair dyes safe for consumers and concluded that the weight of evidence supports PPD and derivatives not being mutagenic and carcinogenic in humans.4^,5^,6

The results of this review add further knowledge about hairdresser’s exposure to potentially genotoxic agents which may initiate a sequence of events that leads to cancer formation. Because of the high relevance of genotoxicity information for the mode of action of any potential carcinogen, it is important to critically review recent publications to find more evidence whether PPD and its derivatives are mutagenic or clastogenic. Also, the OECD recently made significant revisions to the genotoxicity test guidelines, regarding the conduct of assays and the interpretation of test results, thus rendering older studies less reliable than the more recent ones conducted in accordance with modern standards.8^,63

The main limitation of our review is a small number of identified eligible publications, often reporting studies not conducted in line with recent OECD guidelines or with significant deviations from the new OECD recommendations.8

To conclude, although an interest for this research area is limited in the scientific community, a number of classical genotoxicity assays recommended by OECD TGs indicate that PPD and its derivatives have genotoxic potential and thus may pose an important public health concern. A significant segment of general population is exposed to hair dyes during adulthood, also including occupational exposure. More in-depth experimental studies conducted in line with recent OECD guidelines, as well as high quality epidemiological studies are imperative to further elucidate the risks associated with exposure to hair dye precursors, particularly in occupationally exposed populations such as hairdressers but also workers in the plastic and chemical industries where these compounds are also used.