LPS from bovine serum albumin drives TNF-α release during ex-vivo placenta perfusion experiments, contaminates the perfusion system but can be effectively removed by oxidative cleaning

Abstract

IntroductionThe dual ex-vivo perfusion of human placental tissue is useful to study inflammatory pathways. We found significant TNF-α release in negative controls similar in concentration to lipopolysaccharide (LPS) stimulated placentas. The aim of the current study was to (i) identify sources driving TNF-α release and (ii) develop an approach to control for it. Method(i) To determine sources leading to TNF-α release, solutions frequently circulated through the perfusion system and perfusion media with different bovine serum albumin (BSA) quality were exposed to mouse macrophage cell lines (RAW264.7) and subsequently measured for TNF-α expression. (ii) To assess memory effects and validate cleaning procedures, sham perfusion experiments were conducted either in the presence or absence of exogenous LPS, in new tubing that was contaminated, cleaned and analyzed for the effectiveness of LPS removal. Oxidative and acid–base cleaning were tested for their effectiveness to reduce LPS contamination. ResultsTNF-α release, observed in negative control experiments, was attributed to the use of LPS-contaminated BSA as well as inadequate cleaning of the perfusion system. Once introduced in the perfusion system, LPS accumulated and created a memory effect. Oxidative but not acid–base depyrogenation effectively reduced LPS levels to concentrations that were in accordance with FDA guidelines (<0.5 EU/mL) for medical equipment redeemed appropriate for re-use. DiscussionLPS contamination of the placenta perfusion model could have confounding effects on experimental outcomes leading to misinterpretation of data. To circumvent LPS contamination LPS-free BSA and oxidative depyrogenation cleaning techniques should be implemented in future placental perfusion studies.