Effects of temperature, surfactants and skin location on the dermal penetration of haloacetonitriles and chloral hydrate

Abstract

Dermal exposure has been recognized as an important contributor to the total internal dose to disinfection-by-products (DBPs) in water. However, the effect of the use of surfactants, water temperature and area of the body exposed to DBPs on their dermal flux has not been characterized and was the focus of the present study using an in-vitro system. The dermal flux of mg/l concentrations of haloacetonitriles and chloral hydrate (CH), important cytotoxic DBPs, increased by approximately 50% to 170% with increasing temperature from 25 °C to 40 °C. The fluxes for the torso and dorsum of the hand were much higher than that of palm and scalp skin. An increase in flux was observed for chloroacetonitrite and dichloroacetonitrile, two less lipophilic HANs, but not for trichloroacetonitrile or CH, with the addition of 2% sodium lauryl sulfate or 2% sodium laureth sulfate, two surfactants commonly used in soaps and shampoos used in showering and bathing. Thus, factors such as temperature, surfactants and skin location affect dermal penetration and should be considered when evaluating dermal absorption.

INTRODUCTION

Dermal exposure to disinfection-by-products (DBPs) occurs during showering, bathing and swimming. DBPs such as trihalomethanes (THMs), haloacetic acids, haloacetonitriles (HANs) and haloketones are formed by reactions of the disinfectant chlorine with humic substances and other organic material and are commonly found in tap and swimming pool water.^1–4 This study focuses on several factors that can alter the dermal absorption of the DBPs: HANs and chloral hydrate (CH), among the most abundant DBPs in chlorinated drinking and swimming pool water.^5,6 The sum of their concentrations in tap water can be as high as 24 μg/l,^6,7 with much higher concentrations of dichloroacetonitrile (DCAN) and CH (45 μg/l and 265 μg/l, respectively) found in swimming pool water.⁸ HANs and CH have been reported to be mutagenic and teratogenic and are of potential health concerns.^9–15 Epidemiological studies have linked DBPs, often using chloroform or THMs as surrogates for the DBPs, to cancer and adverse reproductive outcomes (as reviewed in Cantor,^16,17 Hrudey¹⁸ and Tardiff et al.¹⁹). Dermal and/or inhalation exposure to THMs, as surrogates for DBPs, from showering and bathing had higher odds ratios for bladder cancer than was calculated for ingestion exposure in a case-control study, with an additional contribution to the odds ratio from THM exposure associated with swimming pool use.²⁰ Increased levels of genotoxicity biomarkers were also associated with exposure to brominated compounds in swimming pools, although distinguishing whether the exposure was through a dermal or inhalation route was not done.²¹

Based on biomarker measurements, dermal absorption has been documented to be an important contributor to the total daily dose of chloroform and other THMs, the DBPs with the highest pool and tap water concentrations^22–32 estimated that dermal exposure accounted for an average 64% of the total dose of lipophilic compounds in tap water during showering. Dermal absorption of several HANs and CH has been suggested to contribute 30–75% of their daily ingestion dose,³³ less than the THMs, but more than the haloacetic acids, the second most abundant DBPs in tap water.^34–37 Dermal exposure in swimming pools was estimated to be one-third of the dose for THMs under moderate exercise determined by comparing individuals wearing scuba gear who therefore only had dermal exposure, compared with individual swimming normally or breathing air next to the pool.³⁸ For highly competitive swimmers, it was reported that dermal exposure could contribute as much as 80% of the THM blood levels.²⁹

Variations in the dermal absorption of an order of magnitude have been reported for hydrocortisone, pesticides and ³²P-labeled organic liquid phosphorous compounds across different areas of the body.^39–41 Thus, contact of different areas of the body with water could result in differential absorption of DBPs. Torso skin is the largest surface area exposed during showering, bathing and swimming, while the hand is exposed to water most frequently. An increase in dermal absorption of chloroform during showering with increasing water temperature has been reported,²⁵ which may result in lower dermal transport during swimming than showering as the temperature of pool water is often lower than used for showering and bathing. Surfactants, which are common components of soaps and shampoos used during showering and bathing, has been reported to affect dermal permeation⁴² and induce skin irritation.⁴³ Two of the most common surfactants used are sodium laureth sulfate (SLES) and sodium lauryl sulfate (SLS). The current study evaluates the dermal flux of aqueous solution of HANs and CH at mg/l concentrations using in-vitro techniques with skin from different locations of the body, at 25 °C, 37 °C and 40 °C and with the use of surfactants.

METHODS

Skin Preparation

Human cadaver skin, obtained from the National Disease Research Interchange (Philadelphia, PA), was frozen before use, thus no metabolic activity was expected. Full-thickness human cadaver skin sections from different areas of the body were prepared by the removal of the subcutaneous tissue, leaving the dermis and epidermis. For some of the torso skin, the epidermis was separated from the dermis after submerging full-thickness skin in water heated to 60 °C for 1 min.⁴⁴ All skin sections were stored at −20 °C, a process not expected to affect their permeability characteristics.⁴⁵

Skin Integrity

The physical integrity of a skin section was tested by determining the dermal flux for tritiated water.⁴⁴ If >0.29% of the applied tritiated water penetrated the skin after a 20-min exposure, the skin was considered to be damaged and not used. One hundred microliters of 10 μCi tritiated water (American Radiolabeled Chemicals, St. Louis, MO) was used for epidermal skin. Three hundred microliters of 10 μCi tritiated water was used to test whole skin. At the end of each time point, 100 μl of the receptor solutions were pipetted into 3 ml of Fisher Scientific ScintiVerse scintillation fluid in 5-ml scintillation vials. The vials were counted for radioactivity using a Packard TRI-CARB 2100TR liquid scintillation analyzer (Meridan, CT).

Experimental Procedures

The skin was thawed at room temperature in phosphate-buffered saline (PBS) solution and placed between two DC-100B side-by-side diffusion cells. The cells were completely mixed using Teflon-coated magnetic stirring bars spun at 600 r.p.m. The area of skin exposed in the diffusion cells was 0.636 cm². The receptor solution was a PBS solution at pH 7.4 maintained at 37 °C representing the body’s blood compartment. The donor solution was water containing the target DBP with or without a surfactant at 37 °C or without the surfactant at 25 °C, 37 °C or 40 °C. The range of temperatures encompasses those commonly used in bathing/showering and heated swimming pools. Heat-separated epidermal skin from the torso was used to study 1-hr exposures at differing temperatures and the effect of 2% surfactant solutions of SLES or SLS with donor cell HANs concentrations of 1 mg/l and 10 mg/l for CH. A 1-hr time frame was used as representative of the contact time during bathing or swimming. Studies comparing the effect of skin location examined dermal fluxes using full-thickness skin because the dermis could not be separated from the epidermis from the scalp section without tearing the skin. The whole-skin experiment used 3-hr exposures with donor cell HAN concentrations of 5 mg/l for CAN and DCAN, 15 mg/l for BCAN, 25 mg/l for DBAN and 100 mg/l for TCAN and CH. The 3-hr time frame was selected owing to the longer lag time for the compounds to penetrate the thicker whole skin to the receptor cell compared with the epidermis skin.

Sample Analysis

The HANs and CH were extracted from the saline or water with methyl-tert-butyl-ether and analyzed using US EPA Method 551.1. with iodoacetonitrile (98%) as an internal standard (Aldrich, Milwaukee, WI).⁴⁶ The extracts were injected into an HP5890 gas chromatograph (Hewlett-Packard, Santa Clara, CA) equipped with a 60-m Restek Rtx-624 capillary column (Bellefonte, PA), 0.25 mm i.d., 1.4 μm film thickness and an electron capture detector. Both the donor with the target DBPs and the receptor PBS solutions were sampled and replaced with fresh solutions at the end of the 1-hr experiment or every hour from the 3-hr experiments. This prevented the build-up of solutes in the receptor and maintained a constant donor concentration within ±20% of the expected value throughout the experiment. The flux (μg/hr-cm²) was calculated directly from the amount of each compound present in the receptor solution, the time between the replacement of the receptor solution and the cross-sectional area of the exposed skin (0.636 cm²).

Test Statistics

An ANOVA analysis was used to determine, for each compound separately, whether there were significant differences in the dermal flux with temperature, for skin from different locations in the body, and if surfactants were present. Newman–Keuls tests were used to compare the means for each condition.

RESULTS AND DISCUSSION

The stratum corneum (about 10–40 μm thick) is the main barrier against penetration of hydrophilic compounds, while the viable epidermis (about 100 μm thick) along with the dermis (about 10–40 μm thick) provide a barrier against lipophilic compounds.⁴⁷ Most compounds that penetrate the skin are quickly absorbed into the bloodstream through a capillary network above the dermis. Thus, using dermatomed skin to estimate dermal absorption provides a more accurate flux estimate than using whole skin. Whole-skin studies are useful in situations where dermatoming or using other methods to separate the epidermis from the dermis is ineffective. This is the case for experiments that compare skin absorption from different areas of the body. Removing the epidermis from scalp skin would leave holes in place of the hair follicles and result in erroneously high dermal flux estimates. While it may take longer for the compounds to penetrate the whole skin, permeability differences across experiments can still be compared as long as identical conditions are used. Therefore, whole skin was used for the in-vitro experiments involving skin obtained from different locations on the body.

Skin Location

Skin sections from the torso and the dorsum hand had higher dermal fluxes than the palm and scalp skin (Table 1). Generally, a thinner stratum corneum and a greater number of epidermal appendages, for example, hair follicles, results in greater skin permeability.⁴⁸ On most parts of the body, hair follicles have minimal effects on dermal absorption as they occupy only a small percentage of the surface area, <1–2% for abdominal skin.⁴⁸ However, this is not the case with scalp skin. Feldmann and Maibach³⁹ assessed hydrocortisone penetration through skin in vivo and observed that the scalp had the greatest permeation followed by the back and palm. Maibach et al.⁴⁰ measured pesticide permeation in vivo and observed that the order of permeability was the scalp > dorsum hand > abdomen > palm. The lower dermal fluxes measured in the current study for the scalp using in-vitro techniques likely reflects the fact that penetration was measured across the full thickness (5 mm) of the skin, while for the human in-vivo studies, the compounds do not need to pass through the entire thickness of the scalp skin before entering the blood capillaries just below the hair follicles. This difference in permeation is consistent with the flux of tritiated compounds in hairless rats being 2–5 times lower than for normal rats.⁴⁹ The difference between hairless and normal rats is expected to be less than the difference between in-vitro and in-vivo human skin studies as rodent skin is far thinner (<1 mm) than full-thickness human scalp skin. The dermal flux of palm skin is much lower than other parts of the body owing to a thicker stratum corneum (0.4 mm to several mm), compared with the 10–40 μm thickness of stratum corneum of torso skin.^39,40,42,48 The total thickness of the skin samples used in these experiments was approximately 2 mm based on visual examination.

Table 1.

Dermal flux for different parts of the body calculated for 3-hr exposures with both side-by-side cells heated to 37 °C±1 °C with the whole skin exposed to 5 mg/l of CAN and DCAN, 15 mg/l of BCAN, 25 mg/l of DBAN, and 100 mg/l of TCAN and CH solutions (n=3).

Surface area of body (%)a		Flux (mg/hr-cm2), (mean and SD)
		CAN	DCAN	TCAN	BCAN	DBAN	CH
Scalp	5.10	0.015±0.028	ND	ND	ND	0.022±0.0053	0.0023±0.00044
Palm	2.60	0.044±0.038	0.068±0.017	ND	0.0018±0.00040	0.038±0.0075b	0.012±0.0022b
Torso	35c	0.27±0.064b	0.25±0.066b	0.63±0.13b	0.54±0.085b	0.97±0.21b	0.089±0.16b
Dorsum	2.60	0.42±0.097b	0.32±0.080b	0.91±0.20b	0.53±0.076b	0.82±0.16b	0.098±0.014b

ND, not detected.

^aSource: US EPA.⁴⁷

^bThese means were found to be significantly higher than next lower group at the 5% level using the Newman–Keuls test that tests multiple means.

^cExcludes extremities (upper 19%, lower 38%.

Temperature Effects

A variety of water temperatures are used during showering, bathing and in swimming pools. Based on exhaled breath measurements, it was observed that the apparent chloroform dose from bath water increased 30-fold when water temperature increased from 30 °C to 40 °C.^25,50 The authors suggested that the temperature effects were the result of changes in blood flow to the skin at different temperatures. An in-vitro permeability study using N-nitrosodiethanolamine showed an increase when the receptor cell temperature was increased from 32 °C to 37 °C.⁵¹ Changes in the skin itself with temperature can also be responsible for increased permeability. In the current study, the skin permeability of HANs and CH was lowest when the donor-side temperature was 25 °C and increased as the donor temperature increased to 37 °C, and then to 40 °C (Table 2). The flux increased from approximately 50% to 170% as the temperature was increased from 25 °C to 40 °C for the compounds evaluated and were statistically different in the ANOVA and the Newman–Keuls post-hoc test at the 5% level for all compounds. One explanation for the change in skin permeability is that an increase in temperature increases the fluidity of the lipophilic layers between the corneocytes.^25,52 It has also been suggested that an increase in temperature affects the lipid viscosity by causing a transition of the lipid in the stratum corneum from a gel to a liquid-crystalline phase.⁵³ These results suggest that the higher in-vivo dermal dose of DBPs reported by Gordon et al.²⁵ from water at warmer temperatures may not only be due to an increase in blood flow to the skin, but also to an increase in the permeability characteristics of the skin. Thus, when estimating the contribution of dermal absorption the temperature of the water should be considered. Swimming pools typically have colder temperatures than used for showering or bathing, which could reduce the dose due to dermal penetration in swimming pools, compared with showering or bathing even though DBP concentrations are often higher in swimming pools.

Table 2.

Dermal flux as a function of temperature after a 1-hr exposure using heat-separated epidermal torso skin exposed to 1 mg/l of the HANs and 10 mg/l of CH solutions (n=3).

Temperature	Flux (mg/hr-cm2), (mean and SD)
	CAN	DCAN	TCAN	BCAN	DBAN	CH
25 °C±1 °C	0.052±0.0092	0.064±0.010	0.053±0.0082	0.076±0.011	0.075±0.0098	0.020±0.0045
37 °C±1 °Ca	0.083±0.012	0.10±0.022	0.079±0.014	0.12±0.018	0.15±0.024	0.037±0.0061
40 °C±1 °C	0.091±0.014	0.12±0.025	0.083±0.013	0.13±0.020	0.16±0.026	0.053±0.0081
% Increase between 25 °C and 40 °C	75	88%	57	71	110	170

The fluxes for all compounds were statistically different at the 5% level using the Newman–Keuls test that tests multiple means.

^aSame data as in Table for no surfactant.

Surfactants

Surfactants are commonly used while showering and bathing and may affect dermal permeability. The surfactants SLES and SLS, which are commonly used in soaps and shampoos, increased the permeation for CAN and DCAN, but not TCAN, BCAN or DBAN (Table 3). Only SLS increased the flux of CH. Surfactants can disrupt the stratum corneum structure, the principal barrier to penetration of environmental contaminates. SLS was the only one of the four detergents (SLS, lutensol AP10, nonyl phenol ethoxylate and ethanol) to increase in-vitro penetration of tritiated water through human skin owing to its ability to compromise the integrity of the skin over a contact time of 4 hrs at concentrations of 0.2% and 2%.⁵⁴ SLS has also been reported to increase the permeability of compounds with lower lipophilicities (log Kow <3).⁵⁵ Surfactants are added to soaps and shampoos as they have both a polar and a non-polar end, this assists the dissolution of lipophilic compounds, such as grease and dirt, into an aqueous solution. This functionality will also facilitate the penetration of hydrophilic compounds through lipid membranes.⁵⁴ Because the stratum corneum includes a lipid matrix, the presence of a surfactant appeared to facilitate the permeability of the least lipophilic HANs evaluated: CAN and DCAN.

Table 3.

Dermal flux calculated after a 1-hr exposure using heat separated epidermal skin with both cells heated to 37 °C±1 °C.

Surfactant	Flux (mg/hr-cm2) (mean and SD)
	CAN	DCAN	TCAN	BCAN	DBAN	CH
No surfactanta	0.083±0.012	0.10±0.022	0.079±0.014	0.12±0.018	0.15±0.024	0.037±0.0061
SLES	0.18±0.029b	0.25±0.071b	0.094±0.013	0.10±0.016	0.18±0.027	0.025±0.0042
SLS	0.21±0.054b	0.71±0.182b,c	0.085±0.011	0.14±0.019	0.17±0.026	0.082±0.012b

The skin was exposed to 1 mg/l of the HANs and 10 mg/l of CH solutions with or without the surfactants sodium laureth sulfate (SLES) and sodium lauryl sulfate (SLS; n=3).

^aSame data as in Table 2 for 37 °C.

^bThese means were found to be significantly higher at the flux with no surfactant at the 5% level using the Newman–Keuls test that tests multiple means.

^cThis mean was found to be significantly higher than the SLES mean at the 5% level using the Newman–Keuls test.

Extrapolation to Environmental Concentrations

It is important to recognize that the mg/l concentrations used in this study greatly exceed the levels measured in either tap water or swimming pool water. As recently discussed, the dermal flux could be dependent upon the concentration, with lower concentrations having higher dermal fluxes.⁵⁶ The flux also changes with time until steady state is reached which can exceed the 1–3 hrs used.³³ Thus, if the dermal flux for these compounds varies with concentration as suggested by Kissel,⁵⁶ then the magnitude of the effect observed may be different at lower concentrations but is expected that direction of the effects would be the same across concentration ranges.

CONCLUSIONS

An increase in temperature increased skin permeability. The two surfactants, SLS and SLES, increased the permeability of CAN and DCAN, the HANs with the greatest solubility in water and the lowest lipophilic nature. The dorsum hand and torso skin were much more permeable than the palm skin because of the greater thickness of the stratum corneum in the palm and the scalp. However, these values may not be applicable in vivo as compounds can enter the blood stream without traversing the entire thickness of the epidermis. These data indicated that an individual’s activities, which part of the body contacts water, the use of shampoos and the water temperature during bathing, showering and swimming alter the degree of dermal absorption of the DBPs and other contaminants in tap water. The magnitude of the effect at environmentally relevant water concentrations still needs to be determined to accurately predict the actual flux of DBP present in chlorinated water to assess their risk.