Highly efficient star formation in NGC 5253 possibly from stream-fed accretion.

Gas clouds in present-day galaxies are inefficient at forming stars. Low star-formation efficiency is a critical parameter in galaxy evolution: it is why stars are still forming nearly 14 billion years after the Big Bang and why star clusters generally do not survive their births, instead dispersing to form galactic disks or bulges. Yet the existence of ancient massive bound star clusters (globular clusters) in the Milky Way suggests that efficiencies were higher when they formed ten billion years ago. A local dwarf galaxy, NGC 5253, has a young star cluster that provides an example of highly efficient star formation. Here we report the detection of the J = 3→2 rotational transition of CO at the location of the massive cluster. The gas cloud is hot, dense, quiescent and extremely dusty. Its gas-to-dust ratio is lower than the Galactic value, which we attribute to dust enrichment by the embedded star cluster. Its star-formation efficiency exceeds 50 per cent, tenfold that of clouds in the Milky Way. We suggest that high efficiency results from the force-feeding of star formation by a streamer of gas falling into the galaxy.

A dust-obscured massive maximum-starburst galaxy at a redshift of 6.34.

Massive present-day early-type (elliptical and lenticular) galaxies probably gained the bulk of their stellar mass and heavy elements through intense, dust-enshrouded starbursts--that is, increased rates of star formation--in the most massive dark-matter haloes at early epochs. However, it remains unknown how soon after the Big Bang massive starburst progenitors exist. The measured redshift (z) distribution of dusty, massive starbursts has long been suspected to be biased low in z owing to selection effects, as confirmed by recent findings of systems with redshifts as high as ~5 (refs 2-4). Here we report the identification of a massive starburst galaxy at z = 6.34 through a submillimetre colour-selection technique. We unambiguously determined the redshift from a suite of molecular and atomic fine-structure cooling lines. These measurements reveal a hundred billion solar masses of highly excited, chemically evolved interstellar medium in this galaxy, which constitutes at least 40 per cent of the baryonic mass. A 'maximum starburst' converts the gas into stars at a rate more than 2,000 times that of the Milky Way, a rate among the highest observed at any epoch. Despite the overall downturn in cosmic star formation towards the highest redshifts, it seems that environments mature enough to form the most massive, intense starbursts existed at least as early as 880 million years after the Big Bang.

Risk assessment of dietary exposure to methylmercury in fish in the UK.

Risk assessment of chemicals in food is generally based upon the results of toxicological studies in laboratory animals, allowing for uncertainties relating to interspecies differences, human variability, and gaps in the database. Use of quantitative human data is preferable if available, as in the example of methylmercury. Methylmercury is a neurotoxic environmental contaminant, for which fish is the main source of dietary exposure. Human data from poisoning incidents and epidemiological studies have been used by expert committees to derive a guideline intake level for methylmercury, based on the susceptibility of the most sensitive lifestage, the developing fetus. In the UK, an expert group of nutritionists and toxicologists was formed to review the benefits and risks associated with fish consumption. A formal risk-benefit analysis was not possible because the nutritional data were not sufficiently quantitative. The Committee on Toxicity of Chemicals in Food, Consumer Products and the Environment (COT), therefore, modified the risk assessment approach to derive different guideline intake levels for different subgroups of the population. The COT opinion was used to provide targeted advice on how much fish can be consumed without undue risk from the contaminants. Consumption by adults of one weekly portion (140 g) of shark, swordfish or marlin, would lead to an exceedance of the guideline intake for methylmercury of 40-90%, set to protect the developing fetus, without considering intake from the rest of the diet. Pregnant women and women who may become pregnant within 1 year were, therefore, advised to avoid consumption of these species. Intakes in other adults would be within a higher guideline intake, set to protect groups of the population other than the developing fetus. However, consumption by children of one weekly portion of these species could lead to an exceedance of this guideline intake by up to 60%, without considering intake from the rest of the diet. It was, therefore, advised that consumption of these species by children should be avoided.

Mechanistic studies on the activation of biphenyl 2-hydroxylation by glucocorticoids.

In order to establish the mechanism by which the selective activation of biphenyl 2-hydroxylation by betamethasone occurs the effect of modifying possible critical factors in the hydroxylation process has been examined. Activation of biphenyl 2-hydroxylation by betamethasone was found in detergent-solubilized rat liver microsomes indicating that intact microsomal membranes are probably not necessary for the activation. Betamethasone had no effect on the spectrally apparent binding of biphenyl or of other type I, type II or reverse type I model substrates. The activation process did not appear to be greatly influenced by changing the ratio of cytochrome P-450 reductase to cytochrome P-450 nor by changing the amount of NADPH. Addition of NADH increased the extent of activation suggesting that betamethasone facilitates transference of the second electron to cytochrome P-450. However, betamethasone also stimulated cumene hydroperoxide supported biphenyl 2-hydroxylation; therefore a step subsequent to cytochrome P-450 reduction is also involved in the activation. Activation did not correlate with increased uncoupling of an active oxygen-cytochrome P-450 complex to form hydrogen peroxide.

Tissue and sex differences in the activation of aromatic hydrocarbon hydroxylases in rats.

Betamethasone and alpha-naphthoflavone produced similar activation of biphenyl 2-hydroxylase and benzo[a]pyrene 3-hydroxylase in control male rat liver microsomes. In small intestinal epithelial microsomes, betamethasone had no effect whereas alpha-naphthoflavone caused a pronounced activation of benzo[a]pyrene hydroxylation and a lesser activation of biphenyl 2-hydroxylation. In lung microsomes, betamethasone had no effect on either enzyme activity whereas alpha-naphthoflavone had no effect on biphenyl 2-hydroxylase but inhibited benzo[a]pyrene hydroxylase. In kidney cortex microsomes from male rats both compounds caused inhibition or had no effect whereas in kidney cortex microsomes female rats betamethasone activated whereas alpha-naphthoflavone had no effect. Activation also occurred in isolated viable hepatocytes from male rats. The response of biphenyl 2-hydroxylase was very similar to that found in male rat liver microsomes but benzo[a]pyrene hydroxylase was more sensitive to activation and less sensitive to inhibition than in microsomes. The findings are interpreted as demonstrating the presence of more than one 'latent' aromatic hydrocarbon hydroxylase in rodents.

Factors influencing the carcinogenicity of food chemicals.

The relationship between food and cancer is extremely complex. It is generally accepted that diet is a contributory factor in the aetiology of a large proportion of cancers, but with very few exceptions, we are unable to identify specific causal agents. Many food components have genotoxic potential and more are produced endogenously during digestion. Conversely, there is increasing evidence that consumption of some foods may decrease the risk of cancer, and a number of plant constituents have been shown to have the potential to inhibit various stages of the carcinogenic process. Yet we have little understanding of the interactions between the different food-related genotoxic and protective factors. A further complication is the variation in individual susceptibility and vulnerability. As a result we are still not able to determine the optimal diet for minimising cancer risk. In recognition of these issues, the UK Ministry of Agriculture, Fisheries and Food (MAFF) is funding a number of projects aimed at providing greater mechanistic understanding of the links between food and cancer, in order to offer detailed advice to the public. This report summarises the proceedings of a workshop entitled 'Factors influencing the carcinogenicity of food chemicals', held in London on 1 June 1998, providing overviews of some of the key issues, and demonstrating how the MAFF-funded research is contributing to advances in these areas. It includes discussion of genetic polymorphisms and how they may contribute to individual susceptibility and help to identify causal links between food components and colorectal cancer. Biomarkers of DNA damage in human studies and of inhibition of carcinogen activation and endogenous formation of genotoxic reactive nitrogen species are examined. Also considered are the potential uses of physiologically based pharmacokinetic modelling techniques for providing more accurate estimates of risk and reducing the uncertainty in extrapolation between species and doses. Research now in progress will help to establish the critical risk and protective factors involved in diet-related colorectal cancers, in order to provide a sound scientific basis for formulation of dietary advice to the public.

Investigations of the genotoxicity and cell proliferative activity of dichlorvos in mouse forestomach.

This study investigated the possible mechanism by which dichlorvos may have caused forestomach tumours in mice in a chronic corn oil gavage cancer bioassay [NTP (1989) Toxicology and carcinogenesis studies of dichlorvos in F344/N rats and B6C3F1 mice (gavage studies). National Toxicology Program Technical Report 342, NIH Publ. No 89-2598]. For this purpose, a method has been developed to assess the genotoxicity of irritant substances on mouse forestomach epithelium. Groups of five B6C3F1 mice were given a single oral dose of dichlorvos, the genotoxic forestomach carcinogen 1-methyl-3-nitro-1-nitrosoguanidine (MNNG) or the irritant, non-genotoxic forestomach carcinogen butylated hydroxyanisole (BHA). After periods of 2-48 h, three parameters were assessed: unscheduled DNA synthesis (UDS) by autoradiography of tissue sections, replicative DNA synthesis (RDS) also by autoradiography of incorporated [3H]thymidine, and histopathological changes, including hyperplasia. MNNG induced UDS but not RDS or hyperplasia in forestomach epithelium, consistent with its genotoxic mode of action. BHA and dichlorvos did not induce UDS, consistent with absence of genotoxic activity in the forestomach after in vivo exposure. In contrast, BHA and dichlorvos induced RDS and subsequent hyperplasia, which is likely to result from irritant damage. These data suggest that the chronic effects of dichlorvos on mouse forestomach epithelium in the oral gavage bioassay were mediated via enforced cell proliferation, rather than by a genotoxic mechanism.

Chemically-induced unscheduled DNA synthesis in cultures of adult human epidermal keratinocytes.

Skin is a major target organ for many experimental carcinogens that exist in our environment and the majority of previous carcinogenicity studies have utilised animal derived models. In view of the fact, that many of these environmental chemicals exhibit species- and tissue-specific metabolism, a human skin tissue derived model would be a distinct advantage. Squamous epithelial carcinoma is a predominant form of skin cancer in man and, in theory, human epidermal keratinocytes present an appropriate target cell to employ as an in vitro system to study epidermal carcinogenesis. This report demonstrates the valuable potential of human keratinocyte cultures as a suitable model for mechanistic studies on factors which may influence DNA damage and, hence, the subsequent development of cancer in human epidermis. Keratinocytes were serially cultivated from adult human skin samples and maintained in culture for at least 3 passages. Tertiary cultures, isolated from 3 separate individuals, were exposed to the direct-acting experimental carcinogen, methyl methanesulphonate (CAS No. 66-27-3), and benzo[a]pyrene (CAS No. 50-32-8), which requires metabolic activation. DNA repair was assessed by a quantitative autoradiographic technique. Methyl methanesulphonate and benzo[a]pyrene both elicited a dose-related increase in unscheduled DNA synthesis in cultures prepared from each individual. Inter-individual variation in the response was observed for each chemical, but this was greater in the case of benzo[a]pyrene, which indicates inter-individual variation in both xenobiotic metabolism activity and DNA repair capacity.

The application of a wedge perfusion technique to the in vivo-in vitro rat hepatocyte DNA-repair assay.

The in vivo-in vitro rat hepatocyte DNA-repair assay is regarded as labour-intensive and time-consuming to perform. This has tended to impose limitations on its use as a routine procedure for assessing the potential genotoxicity of chemicals. We have developed a simple wedge-perfusion technique which enables hepatocytes to be isolated from several different rats simultaneously. Hepatocyte yield and metabolic capacity are comparable to those isolated by conventional whole-liver perfusion. Hepatocyte viability was generally superior to that obtained when performing multiple in situ perfusions for the rat hepatocyte UDS assay. The median lobe is routinely used but no difference was observed in the UDS response to the positive control genotoxic agents, methyl methanesulphonate (MMS, CAS No. 66-27-3) and 2-acetylaminofluorene (AAF, CAS No. 53-96-3), in hepatocytes isolated from the median or either lateral lobe. The use of Williams medium E or Leibovitz L15 culture medium did not influence the response. This perfusion technique greatly reduces the time, equipment and personnel required and therefore the cost for hepatocyte isolation. It also facilitates the inclusion of concurrent control groups at each time point of assay.

Prostaglandin E(2) release in keratinocyte cultures following exposure to various tumour promoters.

The interaction of tumour promoters with the target cell type (keratinocyte) may be an essential feature of their promoting activity and their ability to initiate an inflammatory response. The role of prostaglandin E(2) (PGE(2)), particularly in the keratinocyte, remains largely unknown, but it is closely associated with inflammation and regenerative epidermal hyperplasia, which appear critical for tumour promotion. Rat keratinocytes derived from sublingual mucosa represent a suitable model to investigate the ability of several irritants, of varying tumour-promoting potency, to stimulate PGE(2) release. Cytotoxicity was evaluated by the neutral red uptake assay and a concentration that reduced cell viability to 50% of control was selected as a maximum concentration for subsequent measurement of PGE(2) release. Phorbol-12-myristate-13-acetate, ionophore A23187 and mezerein stimulated PGE(2) release at non-toxic concentrations. Anthralin, benzoyl peroxide, sodium dodecyl sulfate and acetic acid did not stimulate PGE(2) at non-toxic concentrations, but release was associated with a toxic response. Epidermal growth factor and phospholipase C, which are closely associated with intracellular signalling systems that modulate keratinocyte proliferation and differentiation, also stimulated PGE(2) release. Epidermal growth factor elicited PGE(2) release at a concentration reported to be mitogenic in keratinocyte cultures. The stimulation of PGE(2) release in the absence of a toxic response by PMA, mezerein and ionophore A23187 may be indicative of a direct interaction of the chemicals with intracellular pathways involved in regulation of keratinocyte differentiation and proliferation. This interaction may also reflect the ability of such chemicals to initiate an inflammatory response. Measurement of PGE(2) release may be useful to investigate further the mechanism of action of tumour promoters in the target cell type. In contrast, other tumour promoters did not stimulate release at non-toxic concentrations, which implies that their ability to initiate an inflammatory response and possibly their promoting activity is associated with the induction of a toxic response in the target cell population.

Development of an optimal method for the cryopreservation of hepatocytes and their subsequent monolayer culture.

Adult rat hepatocytes, of good viability, were routinely isolated using a versatile biopsy perfusion technique. In order to develop an optimal cryopreservation regimen we investigated a variety of factors including the cooling rate, the constituents of the medium, the cryoprotective agents and the thawing conditions. The preferred freezing medium consisted of Leibovitz L15 culture medium supplemented with 10% (v/v) foetal calf serum, 10% (v/v) tryptose phosphate broth, 20% (v/v) dimethylsulphoxide, 100 mug/ml streptomycin and 100 U/ml penicillin. A freezing rate faster than the generally used 1-2 degrees C/min appeared to be optimal for subsequent hepatocyte attachment and survival in culture. Handling conditions after thawing were important for maximal recovery of cells in culture. To assess the functional integrity of hepatocytes in monolayer culture after cryogenic storage a number of parameters were monitored. Biochemical assays included lactate dehydrogenase leakage, protein synthesis, cellular ATP and ADP content, glutathione (reduced form), cytochrome P-450 content, ethoxycoumarin-O-deethylase activity, glucuronide and sulphate conjugation capacity. Cytochrome P-450 content was maintained at levels comparable with those in freshly prepared cultures for up to 24 hr in culture; ethoxycoumarin-O-deethylase activity was also similar to that of freshly prepared cultures. Changes in various biochemical parameters indicated that cryopreservation resulted in subtle damage to hepatocytes and that even cells that excluded dye did not survive as well as non-frozen cells. When human hepatocytes were subjected to the method developed for rat hepatocytes, they survived the trauma of cryopreservation but fewer cells subsequently attached and spread out to form a monolayer culture. These results indicate that human hepatocytes are more sensitive to freezing under the conditions developed for rat hepatocytes. This may reflect a greater sensitivity of the cells to cryopreservation and/or that the cryogenic conditions were not optimal for human hepatocytes.

Detection of chemical-induced unscheduled DNA synthesis in cultures of normal adult human keratinocytes.

Human keratinocytes are the most appropriate target cells to employ in an in vitro model to study the potential ability of a chemical to react with the genome of human skin cells. The resulting damage to DNA may be a key initiating factor in the subsequent development of squamous epithelial carcinoma, which is a common form of cancer in humans. Keratinocytes were routinely established from adult human skin samples and maintained in culture for at least three passages. Genotoxic damage was assessed by a quantitative autoradiographic technique which enables detection of unscheduled DNA synthesis without prior treatment of the cultures with agents, such as hydroxyurea, to suppress S-phase synthesis. Tertiary cultures from two individuals were exposed to the experimental carcinogens, methyl methane sulphonate (direct-acting genotoxin) and benzo[a]pyrene (requiring metabolic activation). Both chemicals induced unscheduled DNA synthesis in human keratinocytes. This study indicates that human keratinocyte cultures are a suitable model for the detection of genotoxic potential in human epidermis.

Unscheduled DNA synthesis in tracheal epithelial cell cultures.

An in vitro method has been established for the isolation and culture of tracheal epithelial cells for the evaluation of chemically induced genotoxicity using an unscheduled DNA synthesis assay. Cell cultures were derived from the Wistar albino rat and the golden Syrian hamster. Epithelial cells were isolated with protease type XIV for 16 hr and allowed to attach for 24 hr on collagen-coated coverslips in multi-well plates. Cells were exposed to the experimental carcinogens benzo[a]pyrene (metabolism-dependent) and methylmethanesulphonate (direct acting) for 24 hr. Benzo[a]pyrene and methylmethanesulphonate induced DNA repair in cultures isolated from hamsters, whereas only methylmethanesulphonate induced unscheduled DNA synthesis in rat tracheal epithelial cells, thus indicating the lack of metabolic activation in the rat cultures. These results could indicate the suitability of this culture system for the evaluation of airborne carcinogens.

Comparison of cultured keratinocytes and fibroblasts as models for irritancy testing in vitro.

Surfactants are widely used and often cause irritation to human skin. Three groups of surfactants, the trimethylammonium bromides (cationic), sodium dodecyl sulphate (anionic) and the polyoxyethylene sorbitans (Tweens, nonionic) were tested on a rat keratinocyte line (RTE) and an established fibroblast line (3T3-L1) to assess their potential as models for skin irritancy testing. Acid phosphatase (AP) release seems to parallel the development of signs of irritation in vivo and therefore AP activity was assayed after 4 hours' treatment to give an early indication of toxicity. AP activity in RTE keratinocytes rose to a peak and fell sharply, whereas in 3T3-L1 it did not change with treatment. Therefore AP may be a specific indicator of toxicity in keratinocytes. Neutral red (NR) uptake and kenacid blue (KB) staining were both assayed after 3 days' treatment as an indicator of cell proliferation. RTE and 3T3-L1 were equally sensitive in terms of NR and KB-ID(50) values for the anionic and nonionic compounds; however, 3T3 was more sensitive to the cationic compounds.

Toxicity of paracetamol and cyclophosphamide in monolayer cultures of rat and human hepatocytes.

Hepatocyte structural and functional integrity was characterized in short-term primary monolayer cultures. Ethoxycoumarin-O-deethylase activity, cellular glutathione, ATP and ADP content, protein synthesis and lactate dehydrogenase (LDH) leakage were monitored. Cytotoxicity and perturbation of hepatocyte function involving measurement of these biochemical parameters was investigated following exposure to paracetamol and cyclophosphamide. Evaluation of several biochemical parameters appears to be a more appropriate assay of a perturbation in cellular integrity than the use of a single parameter. In the case of paracetamol the response of human hepatocyte cultures was broadly similar to that observed in rat hepatocyte cultures. An increase in LDH leakage occurred following exposure to high dose levels and was associated with depletion of cellular glutathione levels. At both toxic and non-toxic dose levels protein synthesis was impaired and a decrease in the cellular ATP/ADP ratio was evident. A decrease in protein synthesis and cellular glutathione was observed in both rat and human hepatocytes following exposure to cyclophosphamide. The data highlight the potential use of hepatocyte cultures for investigation of specific cytotoxic events and also emphasize the incorporation of human tissue in such systems.

Carcinogen-induced nuclear enlargement in cultures of rat tracheal primary epithelial cells.

Rat tracheal epithelial cells were cultured for 72 hr and then exposed for 3 hr to a range of test chemicals, including non-carcinogens, and direct-acting and activation-dependent carcinogens. The cells were then washed and cultured in fresh medium without the test chemicals for a further 24, 72 or 120 hr. Nuclear size measurements were then made. At 72 and 120 hr after exposure, those cultures treated with the non-carcinogens pyrene and anthracene had distributions of nuclear area similar to those of the solvent controls. All cultures treated with activation dependent carcinogens (N-nitrosopyrrolidine, N-nitrosodimethylamine, benzo[a]pyrene, dimethylbenz[a]anthracene or 4-nitroquinoline-N-oxide) or direct-acting carcinogens (nitrogen mustard or methylmethanesulphonate) showed a shift in distributions to the right, indicating enlarged nuclei. These results indicate that carcinogen-induced nuclear enlargement can occur in cultures of rat primary tracheal epithelial cells, and that this may be a useful indicator system for respiratory carcinogens.

Effects of some non-genotoxic chemicals on DNA synthesis in rat hepatocytes.

The effect of the antihistaminic drug methapyrilene (MP) on DNA synthesis in primary cultures of rat hepatocytes has been compared with the effects of sodium phenobarbitone (PB), clofibric acid (CA), 2-acetylaminofluorene (AAF) and dimethyl sulfoxide (DMSO). The response for all chemicals was dependent on the concentration of epidermal growth factor (EGF) in the culture medium. MP at concentrations between 0.1 and 1 mum stimulated DNA synthesis. PB had a stimulatory effect on DNA synthesis at 0.5-1 mm. A greater increase in DNA synthesis was observed in the absence of EGF in the culture medium, for both chemicals. CA (0.1 mm) increased DNA synthesis in the absence of EGF and inhibited DNA synthesis at concentrations of 10 ng/ml or more. A dose-related increase in DNA synthesis with DMSO was observed in the presence of 10 ng EGF/ml. AAF did not stimulate DNA synthesis and inhibited it in the presence of 10 ng EGF/ml.

Variations in the response of cell lines to metabolism-mediated toxicity.

One major criticism of cytotoxicity tests is the negligible capacity for metabolism of foreign compounds, which exists in most cell lines. We have investigated the use of an exogenous metabolizing system, comprising 9000 g supernatant of liver from Aroclor 1254-pretreated rats (S-9) and an NADPH regenerating system, in the kenacid blue test. BCL-D1 cells (a human embryonic lung finite cell line) appeared to be sensitive to the toxicity of active metabolites of several model toxins. Other cell lines appeared to be less sensitive but the effects varied with the cell line and the test compound. These differences may, in part, be due to differences in the intrinsic ability of different cell lines to detoxify the active metabolites. It is therefore possible to use this type of test system for mechanistic studies but results with unknown compounds should be treated with caution.

Measurement of eicosanoid release in keratinocyte cultures to investigate skin irritation and tumour promoting activity.

Skin irritation, inflammation and hyperplasia appear to be intimately associated with the phenomenon of tumour promotion, but the mechanism of action remains elusive. Prostaglandins and leukotrienes play an important role in skin inflammation and prostaglandin E(2) (PGE(2)) modulates several events associated with phorbol ester-induced tumour promotion. This study investigated the release of eicosanoids (PGE(2) and leukotriene B(4)) and markers of cytotoxicity [neutral red (NR) uptake and intracellular acid phosphatase (AP) activity], after exposure of rat tongue epithelial (RTE) keratinocyte cultures to chemicals of different irritating and tumour promoting activity. The potent phorbol ester tumour promoter phorbol-12-myristate-13-acetate (PMA), and the less potent, structurally related diterpene ester, mezerein (MEZ) were compared with the known skin irritant sodium dodecyl sulfate (SDS). Cytotoxicity data reflected the in vivo skin irritation potential of the test chemicals and intracellular AP activity was increased after exposure to SDS (160 mug/ml), but did not appear to be increased by the more cytotoxic chemicals PMA and MEZ. Extracellular levels of PGE(2) were increased (200 to > 1000% of control levels) after an 18-hr exposure to PMA or MEZ over a concentration range of 0.01 to 20 mug/ml (NR(50) values 8.0 +/- 6.6 and 15.5 +/- 4.8 mug/ml, respectively). These data indicated that PGE(2) release occurred in the absence of cytotoxicity. In contrast, SDS only elicited PGE(2) release after exposure to 80 mug/ml (a cytotoxic dose level, NR(50) 82.5 +/- 9.9 mug/ml). The potency of a chemical to elicit PGE(2) release in keratinocytes in the absence of a cytotoxic response may reflect intracellular pathways intimately associated with the initiation of an inflammatory response and possibly with tumour promoting activity.

Release of inflammatory mediators in human keratinocyte cultures following exposure to a skin irritant.

The most common pathological response of the skin, associated with chemical-induced primary irritation, is inflammation. Release of inflammatory mediators occurs in primary irritant dermatitis in both laboratory animals and humans. Inflammatory mediators (prostaglandin E(2), leukotrienes C(4), D(4), E(4) and 15-hydroxyeicosatetranoic acid) were determined by radioimmunoassay, and leukotriene B(4) and the cytokine, interleukin 1alpha, were measured by enzyme-linked immunosorbent assay, in cultures of normal adult human keratinocytes. Release of inflammatory mediators into the culture medium was assessed at various time points following treatment of the cultures with the anionic surfactant sodium dodecyl sulphate (SDS), a known human skin irritant. A cytotoxic response was confirmed by the neutral red uptake assay after a 24-hr exposure to SDS. A dose-related release of inflammatory mediators was observed. The magnitude of the response varied between different mediators and as a function of time. The results show that release of inflammatory mediators occurs in human epidermal keratinocyte cultures, following chemical insult. Large inter-experimental variations in release of the various mediators probably preclude their use in a routine assay to determine irritation potential. In the case of SDS, release of inflammatory mediators is not a sensitive indicator of cytotoxicity. However, such techniques may provide important mechanistic data on the role of inflammatory mediators in the irritative response to certain chemicals, or on the possible role and interaction of certain mediators in the irritative process.