j. Soc. Cosmet. Chem., 37, 199-210 (July/August 1986) Predicting surfactant irritation from the swelling response of a collagen film j. c. BLAKE-HASKINS, D. SCALA, L. D. RHEIN, and C. R. ROBBINS, Colgate-Palmolive Research and Development Center, 909 River Road, Piscataway, NJ 08854. Received November 8, 1985. Presented at the Annual Meeting of the Society of Cosmetic Chemists, New York, December 5-6, 1985. Synopsis The objective of this research was to develop an in vitro method to predict the irritancy of surfactant .systems without the use of human or animal tissues. The swelling (tritiated water uptake) of a collagen film substrate has been found to correlate with the irritation potential of anionic surfactants and products based on these ingredients. Swelling response is concentration-dependent, and higher substrate swelling indicates greater irritation potential. The method was used to study the relationship of surfactant structure to swelling activity for alkyl sulfates and alkyl ether sulfates, in which the carbon chain length and ethylene oxide content were varied systematically. Results showed that the C12 homolog produced the most swelling and that membrane swelling is inversely proportional to ethylene oxide content. The assay was also used to investigate interactions between anionic and amphoteric surfactants. Addition of an amphoteric surfactant to an anionic surfactant red'uced the amount of swelling produced by the anionic surfactant. Results of this in vitro test correlate with findings from established in vitro and in vivo laboratory and clinical assessments. INTRODUCTION Frequent exposure to some surfactants in cleaning products can result in skin irritation, characterized by redness, flaking, and swelling. A relationship between irritation in- duced by surfactants or surfactant-based products and the swelling of isolated skin and epidermal membranes has been reported several times (1-3). Choman (4) reported that the exposure of dermal collagen and skin to various anionic surfactants resulted in expansion and contraction of these substrates. In particular, the swelling of isolated epidermis has been established (3-6) as a parameter relating to irritancy of anionic surfactants. This work prompted us to develop analytical methods for quantitatively determining swelling and to study alternative membranes as substrates for swelling assays. The swelling of surfactant-treated isolated epidermis, gelatin, collagen, and related membranes was observed in the course of these studies, and the amount of water absorbed appeared to vary in proportion to the known irritancy of the test surfactants. A commercially prepared collagen film was found which had excellent swelling character- istics, and through a cooperative project with the manufacturer the film was optimized 199

200 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS for this assay. We developed a convenient method to quantitatively determine swelling and to predict the irritancy of anionic surfactants by measuring the uptake by the collagen film of tritiated water from a treatment solution. The swelling response of collagen film was used to study the relationship between surfactant structure and activity, and how surfactants swell and disrupt the three di- mensional structure of protein substrates. The method has been used routinely by our laboratory to evaluate the irritation potential of liquid detergent and soap products. MATERIALS AND METHODS CHEMICALS 3H-H20 was obtained from New England Nuclear (Boston, MA) and was used to pre- pare working stock solutions of 5 X 107 dpm/ml by dilution. Radioactivity was mea- sured with a Beckman LS-6800 liquid scintillation spectrometer using Aquasol-2 (New England Nuclear) scintillation cocktail. Linear dodecylbenzenesulfonate (LAS), am- monium lauryl sulfate (ALS), ammonium laureth-3 sulfate (AEOS-3EO), ammonium laureth-6 sulfate (AEOS-6EO), ammonium laureth-9 sulfate (AEOS-9EO), and sodium hexadecyl sulfate were obtained from Vista (Conoco, Saddlebrook, NJ). Sodium octyl, decyl, dodecyl, tetradecyl sulfates, and dodecyltrimethyl ammonium bromide (DTAB) were obtained from Eastman Chemicals (Rochester, NY). Cocamidopropyl betaine (CAPB) was obtained from Miranol (Dayton, NJ). Tween 20 (Polysorbate 20, a mixture of laurate esters of sorbitol and sorbitol anhydrides) was obtained from ICI Americas (Wilmington, DE). COLLAGEN FILM Collagen film supplied by Helitrex Inc. (Princeton, NJ) was prepared from bovine deep flexor (Achilles) tendon. While the precise treatment is proprietary to Helitrex, key elements are as follows. The tendon collagen was purified by enzymatic treatment and salt extraction to remove contaminating non-collagenous proteins and lipids. The film was cast from an acid dispersion, neutralized with ammonia, and then subjected to a cobalt-60 radiation dose of 1.0 Megarad. When collagen film is produced as a medium for tissue culture purposes, it is sterilized with radiation doses of 2.5 Megarad. Con- trolled radiation treatment produces crosslinking in collagen, resulting in changes in physical properties. We studied collagen samples subjected to varying radiation doses and found the 1.0 Megarad treatment to produce the optimum swelling response and membrane integrity. PROCEDURE Samples of collagen film were cut into 1.27 • 1.27 cm (0.5 • 0.5 inch) squares and each square was placed in a 20-ml screw cap vial and weighed (ca 10 mg). Each piece of film was treated with 10 ml solution containing the test surfactant in 0.1 mM phos- phate buffer, pH 7.0, and enough tritiated water to give 1 • 105 dpm/ml. The solu- tions containing the film pieces were incubated at 50øC for 24 hours. The collagen film was then removed from solution with tweezers, rinsed in a liter of deionized water