Effects of Epidermal Growth Factor and Clostridium di$&iZe Toxin B in a Model of Mucosal Injury By John P. Lawrence, Lucy Brevetti, R.J. Obiso, T.D. Wilkins, Ken Kimura, and R. Soper Iowa City, Iowa and Blacksburg, Virginia l Numerous factors have been advocated as being para- mount to the development of necrotizing enterocolitis (NEC) including hypoxia, abnormal bacterial flora, and by products of enteral feedings. In an effort to better understand mecha- nisms involved at the level of the intestinal mucosal barrier the authors have chosen the CACO-2 cell line to model the neonatal intestinal epithelium. By growing CACO-2 cells in transwell inserts, the authors have investigated the ability of Clostridium difficile toxin B, epidermal growth factor (EGF), and a model of mechanical injury to alter transepithelial resistance of CACO-2 monolayers. The findings show that toxin B diminishes resistance in this setting, and EGF can alter that resistance drop. Copyright o 7997 by W.B. Saunders Company INDEX WORDS: Clostridium difficile toxin B, epidermal growth factor, CACO-2 cells, necrotizing enterocolitis. I N AN EFFORT to independently investigate the dynamics of mucosal injury and repair as might occur in necrotizing enterocolitis (NEC), we have chosen to model the developing gastrointestinal mucosa using the CACO-2 cell line. The purpose of the present report is to study the dynamics of mucosal barrier function in response to the introduction of various components, which may be important in the evolution of NEC. The mechanical creation of a mucosal defect will be used to simulate mucosal injury, and the ability of epidermal growth factor (EGF) and Clostridium dij‘icile toxin B (toxin B) to alter mucosal function in this setting will be investigated. MATERIALS AND METHODS The CACO-2 cell line was used as our model of gastrointestinal mucosa. CACO-2 cells were seeded at 5 X 104/cm2 on transwell inserts coated with laminin. The cells were grown in Dulbecco’s minimum essential minimum (DMEM) with 10% fetal calf semm, penicillin- streptomycin, and 5% CO1 at 37°C. Media was changed every 48 hours. Transepithelial electrical resistance (TEER) was measured across the monolayer using an epithelial voltohmmeter (Work Precision Instru- ments, Inc, Sarasota, FL). Confluency of the cell monolayer was From the Depurtment of Surgery, University of Iowa Hospitals and Clinics, Iowa Ci,y, IA, and the Department qf Biochemistry, Virginia Polytechnic Institute, Blacksburg, VA. Presented at the 43rd Annual h~ternational Congress of the British Association of Paediatric Surgeons, St Helier, Jersey, Channel Islands. July 16-19, 1996. Address reprint requests to John Lawrence, MD, Department of Surgery, The Umiversity of Iowa Hospitals and Clinics, 200 Hawkins Ds Iowa City, IA 52242. Copyright 0 1997 by WB. Saunders Compar~y 0022.3468/97/3203-OOI2$03.00/0 430 determined with the combined use of TEER and light microscopy. A TEER of greater than 200 ohms cm2 indicated a morphologically confluent monolayer. After confluency of the cellular monolayer, identical transwells were randomly placed into control groups, for which no intervention or treatment was selected, or to one of the experimental groups. An injury group was created by using a micropipette to create a defect in the cellular monolayer measuring approximately 0.75 mm in diameter. Groups selected to receive C di#iciZe toxin B (TechLab, Blacksburg, VA) had the toxin added to the apical side of the cellular monolayer in a dose of 0.16 ng/mL. Groups receiving EGF (Sigma, St Louis, MO) had this peptide added to the apical side of the monolayer at a dose of 10 ng/mL. An additional subset of the transwells received a combination of the simultaneous creation of the injury and addition of toxin B or EGE Lastly, one additional group received simultaneous addition of both toxin B and EGF. Transepithelial electrical resistance was followed for all groups serially. Additionally, the morphologic appearance of the monolayers was recorded using phase microscopic photographic images. Statistical analysis was performed to compare the measured resistance in various groups by use of analysis of variance (ANOVA). RESULTS The resistance across the cellular monolayers consis- tently increased from the time of cell plating to achieve- ment of confluency. Typically 6 to 8 days of cell growth were necessary to achieve resistances in the 200 to 250 ohm . cm2 range; microscopically this corresponded with cellular confluency. The results of resistance measurements across the CACO-2 cell monolayer at various time points for all experimental groups are depicted in Table 1. Over the period of observations, no significant change in resis- tance was noted for the control groups. In contrast, all groups receiving the mechanical injury alone, or in combination with addition of either EGF or toxin B, showed significant drops in resistance at 8 hours after injury. Morphologically, the mechanical injury produces a Table 1. Resistance OHMS cm* Across Monolayer at Times Following Treatment 0 Hours 8 Hours 32 Hours Control 232 i- 13 216 i- 8 221 i 6 Injury 224 2 3 182 k 5* 204 -t 4” EGF 228k 10 234 i 15 207k 10 Toxin B 231 i 10 210 2 5 190 -t 2* EGF + injury 211 i5 186 k 4” 195 2 4s Toxin B + injury 205 2 5’ 176 -t- 5’ 181 f 4” EGF + toxin B 224 IT 3 222 t 5 213 2 6t NOTE. n = 4. *PC .05 versus control for given time point. tP < .05 versus toxin B. Journal ofpediatric Surgery, Vol32, No 3 (March), 1997: pp 430-433