In Vitro Pull-Down Assay without Expression Constructs

Abstract

The in vitro pull-down assay is a well-established method to confirm direct binding in protein-protein interactions that was inferred from other interaction assays, such as two-hybrid analysis (1). The assay is usually carried out using glutathione-S-transferasetagged or His-tagged fusion proteins as the pull-down drivers and in vitro-translated 35S-labeled proteins as probes to detect interactions. The fusion proteins and the radioactive proteins with which they interact are harvested using an affinity matrix. The co-precipitated radioactive proteins are analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The most laborious steps in the assay may be the construction of plasmids for the fusion proteins and their expression in Escherichia coli cells. In addition, fusion proteins often are expressed as insoluble forms, which further complicate their use in experiments. Here we report the development of a rapid in vitro pull-down assay that omits the use of particular expression constructs, thereby enabling evaluation of in vitro proteinprotein interactions within 1 day from sample preparation to results. Table 1 outlines our protocol. The advantage of the assay is that it uses in vitro biotinylated proteins instead of tagged proteins as the pull-down drivers. To test the protocol, we applied it to two well-known protein-protein interactions: p53-SV40 large T antigen (2) and cFos-cJun (3). Using SDSPAGE, we successfully confirmed that radioactive p53 is co-precipitated by biotinylated simian virus 40 (SV40) large T antigen (LT) but not by biotinylated luciferase (luc; negative control), cFos, or cJun (Figure 1). We obtained the same results using biotinylated p53 and radioactive LT, and similarly, we dually confirmed the cFos-cJun interaction. In addition, we also successfully detected the previously reported self-interactions of both LT and cJun (3,4). Further, we obtained almost the same results by the scintillation counting method (Table 2). This method is faster than SDS-PAGE, but seems sometimes less convincing because radioactively labeled high molecular weight proteins such as LT and luciferase tended to yield a higher variation of nonspecific counts than did smaller proteins. So far using our new method, we have successfully detected all six of the previously reported interactions that we tested. The use of in vitro biotinylated proteins in pull-down assays seems to have several advantages. First, particular plasmids for the pull-down drivers need not be prepared, because most cloning vectors have bacteriophage promoters (T7, T3, or SP6 promoters) for in vitro transcription of the insert DNA and therefore the biotinylated proteins are synthesized through in vitro transcription-translation using biotin-lysine transfer RNA (tRNA)