An efficient method for the purification of proteins from four distinct toxin–antitoxin modules Yann G.-J. Sterckx a,b , Steven De Gieter a,b , Valentina Zorzini a,b , San Hadz ˇi a,b,c , Sarah Haesaerts a,b , Remy Loris a,b,⇑,1 , Abel Garcia-Pino a,b,⇑,1 a Structural Biology Brussels, Department of Biotechnology, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussels, Belgium b Structural Biology Research Center, VIB, Pleinlaan 2, B-1050 Brussels, Belgium c Department of Physical Chemistry, Faculty of Chemistry and Chemical Technology, University of Ljubljana, 1000 Ljubljana, Slovenia article info Article history: Received 12 November 2014 and in revised form 27 December 2014 Available online 9 January 2015 Keywords: Toxin–antitoxin Persisters Protein purification Refolding protocol abstract Toxin–antitoxin (TA) modules are stress response elements that are ubiquitous in the genomes of bacte- ria and archaea. Production and subsequent purification of individual TA proteins is anything but straightforward as over-expression of the toxin gene is lethal to bacterial and eukaryotic cells and over-production of the antitoxin leads to its proteolytic degradation because of its inherently unstruc- tured nature. Here we describe an effective production and purification strategy centered on an on-col- umn denaturant-induced dissociation of the toxin–antitoxin complex. The success of the method is demonstrated by its application on four different TA families, encoding proteins with distinct activities and folds. A series of biophysical and in vitro activity tests show that the purified proteins are of high quality and suitable for structural studies. Ó 2015 Published by Elsevier Inc. Introduction TA modules are ubiquitous in the genomes of prokaryotes in which they play an important role in a myriad of biological func- tions [1–4]. These include stress response, multi-drug tolerance, biofilm formation and the generation of persister cells [5–12]. Of all known TA loci, members of type II systems have been the most extensively studied. Type II TA modules are typically organized in small operons containing one antitoxin gene lying upstream of one toxin gene, although variations on this theme may occur [13–15]. The toxin may be a monomer or a dimer and derives its name from the observations that it inhibits essential cellular processes such as transcription or translation by corrupting the function of essential molecules such as DNA gyrase, tRNA or the ribosome [14,16–25]. This ‘poisoning effect’ of the toxin will lead to growth arrest and ultimately to cell death when unregulated [26]. The antitoxin typ- ically consists of two functionally distinct domains. The N-terminal domain adopts a well-defined fold and is a DNA-binding and dimerization domain. The C-terminal part is usually intrinsically disordered and neutralizes its cognate toxin via the formation of a tight complex [14,27–31]. The intrinsic unstructured nature of this neutralization domain makes the antitoxin highly susceptible for proteolytic degradation, reducing its in vivo lifetime [32–34] and also allows for a tight link between regulation of transcription and protein activity [31,35,36]. TA modules have found wide- spread applications in biotechnology and show potential in medi- cine and drug design [37–40]. Despite these interests, obtaining large quantities of TA proteins remains challenging. Obtaining recombinant wild-type toxins is difficult due to their toxicity in bacterial and eukaryotic cells [41,42]. Often their genes cannot even be cloned without introduc- ing unwanted mutations [43]. Several strategies have been suc- cessfully formulated to bypass toxicity issues during toxin production, although they each contain drawbacks. The first strat- egy to produce toxin is based on knowledge of bacterial strains resistant to the action of the toxin [44,45]. This approach has suc- cessfully been used to acquire large amounts of the CcdB and VapC toxins, but such resistant strains are not available for other toxin families [46]. A second tactic uses the neutralising activity of the antitoxin. Pure free toxin can be produced when the antitoxin is provided in trans as a protection against accidental expression of low amounts of toxin prior to induction [47]. However, the authors reported that subsequent purification of the toxin is thorny because over-expression of the toxin remains limited and a http://dx.doi.org/10.1016/j.pep.2015.01.001 1046-5928/Ó 2015 Published by Elsevier Inc. ⇑ Corresponding authors at: Structural Biology Brussels, Department of Biotech- nology, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussels, Belgium. Tel.: +32 2 629 18 53; fax: +32 2 629 19 63. E-mail addresses: remy.loris@vib-vub.be (R. Loris), agarciap@vub.ac.be (A. Garcia-Pino). 1 Contributed equally to this work and should be considered joint senior authors. Protein Expression and Purification 108 (2015) 30–40 Contents lists available at ScienceDirect Protein Expression and Purification journal homepage: www.elsevier.com/locate/yprep