Skin irritant-induced cytotoxicity and prostaglandin E2 release in human skin keratinocyte cultures

Pergamon

Toxicology

in Vitro 11 (1997) 627-631

Skin Irritant-induced Cytotoxicity and Prostaglandin E, Release in Human Skin Keratinocyte Cultures J. N. LAWRENCE*, Division

of Chemical

Safety

F. M. DICKSON

and D. J. BENFORD

Evaluation, Robens Institute of Health and Safety. Guildford, Surrey GU2 5XH, UK

University

of Surrey,

Abstract-Damage to the skin induced by chemical irritants is associated with the release of arachidonic acid metabohtes, such as prostaglandin E, (PGE2) which plays an important role in epidermal inflammation. This study investigated cytotoxicity and release of PGEl in human epidennal keratinocytes following an 18 hr exposure of confluent cultures to various skin irritants. The concentration-dependent release of PGEz into the extracellular medium appeared to fall into two categories, which was reflective of possible mechanisms of action. Potent skin irritants, such as phorbol-12-myristate-13-acetate, benzalkonium chloride and tributyltin chloride, elicited an increase in extracellular PGE? levels at concentrations that did not produce overt cell damage (uptake of neutral red at these concentrations was comparable to control levels). Non-irritants (2-methoxyethanol and 2-butoxyethyl acetate) and two less severe irritants (sodium dodecyl sulfate and acetic acid) stimulated release of PGE, only at concentrations that compromised cellular integrity (uptake of neutral red was at least 50% lower than that of control cultures). ,c 1997 Published by Elswier Science Lfd Abbreviations: ELISA = enzyme-linked immunosorbent assay: LTB, = leukotriene B4; NR = neutral red: PGE, = prostaglandin E?; PMA = phorbol-12-myristate-13-acetate; SDS = sodium dodecyl sulfate.

INTRODUCTION

Skin

irritation

and

the

associated

inflammatory

response is a complex process involving interaction between resident skin cells, vascular tissue and infiltrating cells. The keratinocyte, which is the major cell type of the epidermis, is capable of producing a wide range of inflammatory mediators, including membrane lipid-derived factors, in response to a cellular insult (Marks and Furstenberger, 1993). These mediators influence the activity of other inflammatory cell types (Parish, 1992) and evidence suggests that keratinocyte-derived inflammatory mediators play an important role in initiation of an epidermal inflammatory response (Barker et a/., 1991). Keratinocytes are a principal epidermal source of membrane-derived arachidonic acid metabolites such as prostaglandin E, (PGE,) and leukotriene B, (LTB,), which may be formed by the action of various phospholipases or extensive membrane damage. Experimental studies have demonstrated that these two arachidonic acid metabolites are intimately associated with epidermal (proliferation) and vascular (erythema, oedema, cellular infiltration) inflammatory changes (Ruzicka, 1990). Human epidermal keratinocyte cultures represent a suitable in vitro model to investigate the ability of skin irritants to stimulate release of PGEz and LTB,. *Author for correspondence. 0887-2333/97/%17.00 + 0.00 0 SSDI 0887-2333(97)00044-l

1997 Published

Previous studies in our laboratory have demonstrated that keratinocyte cytotoxicity data provided an indication of the potential of a chemical to induce a severe irritant response, but were unable to discriminate between more subtle degrees of irritancy (Lawrence et al., 1996). Our laboratory has also demonstrated that the surfactant sodium dodecyl sulfate (SDS) stimulated release of PGE, in human keratinocyte cultures (Dickson et al., 1993). The aim of this study was to investigate whether PGE, release was indicative of the ability of various chemicals to induce human skin irritation. Chemicals were selected based upon availability of human skin irritation data; and phorbol-l2-myristate13-acetate (PMA) was included as it is also a well-documented activator of protein kinase C second-messenger pathways. 2Methoxyethanol and 2-butoxyethyl acetate were representative of non-irritants; PMA, tributyltin chloride and benzalkonium chloride represented severe irritants, and acetic acid and SDS less severe irritants. MATERIALS

AND METHODS

Materials Tissue culture plastic and Gibco Ltd (Paisley, UK). PMA (purity 99%), were Chemical Company (Poole,

by Elsevier Science

Ltd. All rights

reserved.

medium was supplied by SDS (purity 99%) and purchased from Sigma UK). Acetic acid (purity Printed

in Great

Britain

J. N. Lawrence a~ ~1.

628 140

Concentration (pg/ml) Fig. I. Neutral red concentration-response curves. NR uptake was determined after 18 hr exposure to test chemicals. NR,, values (concentration producing a 50% decrease relative to control cultures) were calculated from concentration-response curves. Data were means of at least three separate experiments,

each using triplicate cultures per datum point: x = tributyltin chloride; IJ = PMA; A = benzalkonium chloride; l = SDS; n = acetic acid: + = 2-butoxyethyl acetate; 0 = 2-methoxyethanol.

99.8%) was supplied by Fisons Scientific (Loughborough, UK). Benzalkonium chloride (purity data not supplied), tributyltin chloride (purity 99.6%) and 2-methoxyethanol (purity 99.99%) were obtained from Aldrich Chemical Company (Gillingham, UK). 2-Butoxyethyl acetate (purity 98%) was obtained from Merck Ltd (Poole, UK). CeN isolation and treatment LIPkeratinocpe cultures Human keratinocytes were isolated from mastectomy discard tissue and cultured according to the method of Lawrence et al. (1996). Keratinocyte cultures were established from a total of five separate individuals and cell stocks were stored in liquid nitrogen. Culture conditions were based on the original method of Rheinwald and Green (1975) with subsequent modifications as described by Barlow and Pye (1990). Keratinocytes (tertiary cultures) for cytotoxicity determination were seeded at 2 x IO4per well of a 96-well collagen-coated plate and for inflammatory mediator release were seeded at 2 x IO5 per well of a 24-well collagen-coated plate. KeratiTable

1. NR,,

values NRw (gg/ml)

Chemical

Meall

f SD

Tributyltio chloride Benzalkonium chloride PMA SDS Acetic acid 2-Butoxyethyl acetate 2-Methoxyethanol

0.33 3.3 30.0 65.0 I670 4600 57 100

0.01 0.3 2.2 15 300 300 6100

nocytes were grown to confluence, approximately 25 days, prior to exposure to test chemicals. 2-Methoxyethanol, SDS, benzalkonium chloride and acetic acid were added directly to the culture medium. Stock solutions of 2-butoxyethyl acetate, tributyltin chloride and PMA were prepared in dimethyl sulfoxide and the final concentration of solvent in the culture medium was 1% (v/v). Triplicate cultures per treatment group were prepared for the neutral red (NR) uptake assay and duplicate cultures were employed for the determination of PGE? levels. Keratinocyte cultures were exposed to test chemicals for I8 hr for determination of NR uptake and measurement of PGE2 levels. Cytotoxicity assay

NR uptake was measured according to the method of Borenfreund and Puerner (1985). Briefly, treated cultures were washed with phosphate buffered saline and incubated with NR-containing medium (50 pg/ ml) for 3 hr, unincorporated dye was removed by washing in buffer and destain was added. Absorbance was measured at 540 nm using a microplate reader. Uptake of NR into the lysosomes of living cells provided a quantitative analysis of cellular viability and this assay was used to select appropriate concentrations for measurement of PGE, release. The upper concentration(s) selected of each test chemical produced cellular injury as indicated by NRm values. The lower concentrations selected of each test chemical elicited NR uptake at levels comparable to

Irritant-induced cytotoxicity and PGE? release in keratinocytes

629

Table 2. Prostaelandin E, release in human keratinocvte cultures Concentration Chemical

(unlml)

2-Butoxyethyl acetate 2-Butoxyethyl acetate 2-Butoxyethil acetate 2-Butoxyethyl acetate 2-Methoxyethanol 2-Methoxyethanol 2-Methoxyethanol 2-Methoxyethanol

1000 2000 4000 8000 5000 10,000 40,000 80,000

those of control cultures (less than a 10% decrease relative to control). Determination protein

of eicosanoid

release

Prostaglandin EJ release Percentage of control

+ SD

46* 53+ 101 153 34* 42’ 76 80

IS 20 13 22 8 IO I9 14

acetate, benzalkonium shown).

chloride and SDS (data not

and cellular DISCUSSION

The method used has been reported previously from this laboratory (Dickson et al., 1993). PGE, and LTB, were determined using an enzyme-linked immunosorbent assay (ELISA) kit (Amersham International, Amersham, UK). Duplicate determinations were performed for each individual replicate sample. Potential interference with the ELISA by the test chemicals was evaluated: the top concentration of PMA (20 pg/ml) elicited a 25% inhibition of the ELISA, which is accounted for in the data presented and the remaining test chemicals did not interfere with the assay. Cellular protein was solubilized by addition of 0.5 M sodium hydroxide, samples were stored at -2O’C until analysis and protein was determined spectrophotometrically (Bradford, 1976). RESULTS

Each test chemical elicited a concentration-dependent inhibition of NR uptake (Fig. 1) and concentrations producing a 50% decrease relative to control cultures (NR,, values) were calculated (Table 1). 2-Methoxyethanol and 2-butoxyethyl acetate did not produce a statistically significant increase in extracellular PGE2 levels (Table 2), even at concentrations that elicited extensive cell damage as indicated by NR,, values. Each skin irritant stimulated a similar degree of PGE, release (Fig. 2) at concentrations that resulted in least a 50% decrease in NR uptake. Tributyltin chloride, benzalkonium chloride and PMA also stimulated release of PGE2 at concentrations that elicited NR uptake at levels comparable to those of control cultures (Fig. 1). In contrast, acetic acid (in the form of acetate salts in culture medium) and SDS stimulated an increase in PGE2 release only at concentrations that reduced NR uptake to at least 50% of control cultures. An increase in extracellular LTB, levels was observed only following exposure of human keratinocyte cultures to cytotoxic concentrations of 2-methoxyethanol, 2-butoxyethyl

The cytotoxicity data appear to reflect the known in uiuo skin irritation potential of the test chemicals, but available human skin irritation data and rather scant and exposure and evaluation procedures vary between reports. 2-Methoxyethanol and 2-butoxyethyl acetate (neat application) elicit no increase or a marginal increase, respectively, in cutaneous blood flow in human volunteers and are regarded as non-irritant in rabbit skin [Commission of the European Communities (CEC, 1990)]. In the case of the skin irritants, the decreasing order of potency to human skin appears to be croton oil (active ingredient is PMA) > benzalkonium chloride > SDS > acetic acid (Osborne and Perkins, 1994; Willis et al., 1988). Croton oil is reported to induce a greater degree of damage to human epidermis than SDS or benzalkonium chloride (Van de Sandt and Rutten, 1995) and tributyltin chloride is a severe irritant to rabbit skin (CEC, 1990). PGE, release per se in keratinocyte cultures was not indicative of irritancy potential, but the most irritant chemicals evaluated in our human keratinocyte model stimulated release of PGE2 in the absence of overt cell damage. The potential of the test chemicals to elicit PGE, release also varied, as indicated by the difference in the magnitude of the increase produced and the shape of the concentration-response curve, which is suggestive of interaction with different target sites within the cell that lead to PGE, production and release. Interactions with potential cellular target sites, such as protein kinase C or phospholipase AI, may explain the observed decrease in PGE2 release following exposure to low concentrations to 2-butoxyethyl acetate and 2-methoxyethanol, as the response was not due to cytotoxicity or to inhibition of the immunoassay. The data highlight the complexity of the biochemical pathways underlying formation of PGE2, as exposure to higher concentrations of the test chemicals resulted in an increase in PGE2 release. The degree of arachidonic acid release following exposure of human keratinocyte cultures to cytotoxic

J. N. Lawrence

CI ol

phorhol-12-myristate-13-acetate

tributyltin chloride

I

NRm 0.33 +/- 0.01

NRw 30.0 +I- 2.2 b)

Concentration

(&ml)

benzalkonium chloride

NRH,3.3 +I- 0.3

NRm 65.0+I-15.0 0

0.0

1.0

2.0 Concentrdhon

4.0

3.0 (&ml

0

5.0

20

40

1

60

80

cm-

acetic acid 500 ..

loo--

NR,,, 1670+/-300

0

100

Concentration

500

1500

looo Concentration

Fig. 2. [legend

(pglml)

opposite]

2caO

2500

I20 (pghnl)

140

160

180

200

Irritant-induced cytotoxicity and PGE] release in keratinocytes concentrations of various skin irritants has also been reported to be chemical dependent (Miller-Decker et al., 1994). Previous data in our laboratory demonstrated that PMA induced a rapid increase in release of PGE, from rodent keratinocytes (Lawrence and Benford, 1995). This is reported to involve direct activation of protein kinase C, subsequent activation of phospholipase AL and release of arachidonic acid metabolites (Jacobsen et al., 1995). The effects of benzalkonium chloride and tributyltin chloride on stimulation of PGE, release, or protein kinase C second-messenger pathways in keratinocytes is unknown. Stimulation of PGEz release by the two weaker irritants, SDS and acetic acid, probably occurred following non-specific activation of membrane phospholipases due to extensive damage at high cytotoxic concentrations. The stimulation of PGEz release in epidermal keratinocytes in the absence of overt cell damage has implications for initiation of a local skin inflammatory response following low-level exposure to specific chemicals. Such chemicals, even if poorly absorbed into the epidermis, may have the ability to initiate a cutaneous inflammatory response, particularly in cases where the stratum corneum barrier is compromised, such as in cuts and abrasions. Acknoitledgemenfs-The authors gratefully acknowledge the financial sponsorship of the FRAME research programme and the technical excellence of Mrs Sue Starkey.

REFERENCES Barker J. N. W. N., Mitra R. S., Griffiths C. E. M., Dixit V. M. and Nickeloff V. S. S. (1991) Keratinocytes as initiators of inflammation. Lancef 337, 21 i-214. Barlow Y. and Pye R. (1990) Keratinocyte culture. In Mefhods in Molecular Biology. Vol. 5. Edited by J. F. Pollard and J. M. Walker. Vol. 5. pp. 83-97. Humana Press. Clifton, NJ. Borenfreund E. and Puerner J. A. (1985) Toxicity determined in virro by morphological alterations and neutral red absorption. Toxicology Leffers 24, 19-124. Bradford M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilising

631

the principle of protein dye binding. Anal.vfical Biochemisfry 72, 248-254.

Commission of the European Communities (CEC) (1990) Collaborative study on relationship between in t~it~o primary irritation and in uifro experimental models. CEC/V/E/LUX/l57/88. Revision I. Dickson F. M., Lawrence J. N. and Benford D. J. (1993) Release of inflammatory mediators in human keratinocyte cultures following exposure to a skin irritant. Toxicology in Vifro 7, 385-388.

Jacobsen P. B.. Kuchera S. L., Metz A., Schachtele C., Imre K. and Schrier D. J. (1995) Anti-inflammatory properties of GO 6850: a selective inhibitor of protein kinase C. Journal of Pharmacology and E..uperimenfal Therapeufics 275, 995-1002.

Lawrence J. N. and Benford D. J. (1995) Prostaglandin E? release in keratinocyte cultures following exposure to various tumour promoters. Toxicology in Vitro 9, 205-2 1 I Lawrence J. N., Starkey S., Dickson F. M. and Benford D. J. (1996) The use of human and rat keratinocyte cultures to assess skin irritation potential. Toxicology in Vifro 10, 331-340.

Marks F. and Fustenberger G. (1993) Proliferative responses of the skin to external stimuli. Enaironmental Healfh Perspecfitjes 101 Suppl. (5), 95-101. Miiller-Decker K., Fursternberger G. and Marks F. (1994) Keratinocyte derived pro-inflammatory key mediator and cell viability as in oizro parameters of irritancy: a possible alternative to the Draize skin irritation test. Toxicology and Applied Pharmacology 127, 99-108.

Osborne R. and Perkins M. A. (1994) An approach for development of alternative test methods based on mechanisms of skin irritation. Food and Chemical Toxicology 32, 133-142.

Parish W. E. (1992) Chemical mediators in inflammation. In Texfbook of Dermatology. Vol. I. 5th Ed. Edited bv R. H. Champion, J. L. Burton and F. J. G. Ebling. Vol. I, 5th Edn. vv. 219-252. Blackwell. Oxford. Rheinwald J. i;‘. and Green H. (1975) Serial cultivation of strains of human epidermal keratinocytes: the formation of keratinising colonies from single cells. Cell 6, 331-343. Ruzicka T. (1990) Arachidonic acid metabolism in normal skin. In Eicosanoids and fhe Skin. Edited by T. Ruzicka. pp. 23-31. CRC Press, Boca Raton, FL. Van de Sandt J. J. M. and Rutten A. A. J. J. L. (1995) Differential effects of chemical irritants in rabbit and human skin organ cultures. Toxicology in Vitro 9, 157-168. Willis C. M., Stephens C. J. M. and Wilkinson J. D. (1988) Experimentally induced irritant contact dermatitis Determination of optimum irritant concentrations. Confacf Dermafifis 18, 20-24.

Fig. 2. Prostaglandin El release in human keratinocyte cultures after exposure to tributyltin chloride, PMA, benzalkonium chloride, SDS or acetic acid. PGE2 was determined in culture supernatants after I8 hr exposure, calculated as pg/pg protein and expressed as percentage of controls. Control levels of extracellular PGE2 were 2.61 + 0.8 pg/pg protein (n = 10 experiments, using five different keratinocyte strains). Data were means of three separate experiments per chemical, each using duplicate cultures per datum point and individual samples were analysed in duplicate. SDS data were the means of three separate experiments each using keratinocytes prepared from a different individual. Data for PMA and acetic acid were the means of two separate experiments (bars represent range of values).Asterisks indicate significant differences from controls (*P i 0.05; Student’s t test).