Abstract Chromatin immunoprecipitation (ChIP) is a commonly used technique to investigate which parts of a genome are bound by a particular protein. The result of ChIP is often interpreted in a binary manner: bound or not bound. Due to this focus, ChIP protocols frequently lack the ability to quantitatively compare samples with each other, for example in a time series or under different growth conditions. Here, using the yeast S. cerevisiae transcription factors Cbf1, Abf1, Reb1, Mcm1 and Sum1, we optimized the five major steps of a commonly used ChIP protocol: cross-linking, quenching, cell lysis, fragmentation and immunoprecipitation. Quenching with glycine is inefficient and can lead to large degrees of variability, an issue that is resolved by using tris(hydroxymethyl)aminomethane (Tris). Another source of variability is degradation of the protein of interest during the procedure. Enzymatic cell lysis with zymolyase can lead to extensive protein degradation, which is greatly reduced by mechanical lysis through bead beating. Degradation also occurs during sonication of chromatin, affecting large proteins in particular. An optimal mix of protease inhibitors and cross-linking with a higher percentage of formaldehyde reduces the extent of this degradation. Finally we also show that the immunoprecipitation step itself can be greatly improved with magnetic beads and optimized incubation/washing steps. The study results in a highly optimized protocol, which is shorter, easier to perform and has a stronger, more reproducible signal with less background. This protocol is presented in detail. In addition, the results highlight the greatest sources of variability in many other protocols, showing which steps are important to focus on for reproducible and quantitatively comparable ChIP experiments.

An extensively optimized chromatin immunoprecipitation protocol for quantitatively comparable and robust results