Suppression of the toxicity of Bac7 (1–35), a bovine peptide antibiotic, and its production in E. coli

Bac7 (1–35) is an Arg- and Pro-rich peptide antibiotic, produced in bovine cells to protect them from microbial infection. It has been demonstrated to inhibit the protein synthesis in E. coli, leading to cell death. Because of its toxicity, no cost effective methods have been developed for Bac7 production in Escherichia coli for its potential clinical use. Here, we found a method to suppress Bac7 (1–35) toxicity in E. coli to establish its high expression system, in which Bac7 (1–35) was fused to the C-terminal end of protein S, a major spore-coat protein from Myxococcus xanthus, using a linker containing a Factor Xa cleavage site. The resulting His6-PrS2-Bac7 (1–35) (PrS2 is consisted of two N-terminal half domains of protein S connected in tandem) was well expressed using the Single-Protein Production (SPP) system at low temperature and subsequently purified in a single step by using a Ni column. The combination of protein S fusion and its expression in the SPP system at low temperature appeared to suppress Bac7 (1–35) toxicity. Both the purified His6-PrS2-Bac7 (1–35) and His6-PrS2-Bac7 (1–35) treated by Factor Xa were proven to be a potent inhibitor for cell-free protein synthesis.


Introduction
Antimicrobial peptides (AMPs) consisting of 10-50 amino acid residues have been discovered from insects to mammals, specifically targeting against either Gram-negative bacteria or Gram-positive bacteria or both (Daher et al. 1988;Xi et al. 2014;Jayamani et al. 2015). Some of them are fungicidal. Some AMPs form helical structures through the membrane to cause cell lysis in a broad range of micro-organisms (Jamasbi et al. 2014;Li et al. 2015), while the others having very high contents of proline and arginine residues inhibit the functions of essential intracellular components (Tu et al. 2011;Roy et al. 2015). As recent emergence of highly antibiotic-resistant or multidrug resistant pathogens, AMPs have become attractive alternatives for the treatment of patients. The use of AMPs is considered to be somehow less problematic than the use of conventional antibiotics as AMPs induce resistant strains in a much lower frequency (Zasloff 2002;Perron et al. 2006). Furthermore, the use of AMPs together with conventional antibiotics may have synergistic effects for therapeutic purpose. On the other hand, the disadvantages of AMPs are their cytotoxicity to the host, their instability in the cells, and the cost of their synthesis. Among AMPs, proline-rich antimicrobial peptides (PR-AMPs) have been isolated from mammalian neutrophils and from haemolymph of some invertebrate species (Anderson and Yu 2003;Treffers et al. 2005;Paulsen et al. 2013).
Bac7 (1-35) is a PR-AMP isolated from bovine and belongs to the cathelicidin family (Scocchi et al. 1997). The cathelicidins serve a critical role in mammalian innate immune defense against invasive bacterial infection (Zanetti 2004), and Bac7 (1-35) was found from bovine neutrophils together with cathelicidin as a antimicrobial peptides (Romeo et al. 1988;Gennaro et al. 1989). Later, Bac7 (1-35) was shown to inhibit DNA, RNA, and protein synthesis in E. coli (Mardirossian et al. 2014) after penetrating in the cells through SmbA, a peptide transporter (Mattiuzzo et al. 2007). Recently, was shown to bind to 70S ribosome resulting in inhibition of protein synthesis (Mardirossian et al. 2014). Despite the high toxicity to Gram-negative bacteria such as E. coli, Klebsiella pneumoniae, Salmonella typhimurium, and Enterobacter cloacae at 1-10 μM, Bac7 (1-35) has remarkably low cytotoxicity to the host mammalian cells (not toxic even at 50 μM) (Tomasinsig et al. 2006). Therefore, Bac7 (1-35) has been extensively studied because of its potential use for clinical application. A method to stabilize Bac7 (1-35) by PEGylation has been developed to reduce its renal clearance by which Bac7 (1-35) still retains its antibacterial activity as well as cell penetration activity .
For pharmaceutical applications, development of a method for an efficient AMP synthesis is important, but no successful expression system for Bac7 (1-35) has reported. So far, the production of some AMPs has been successfully carried out with a yeast system since these AMPs are not toxic to yeast (Jiménez et al. 2014;Wang et al. 2014;Mao et al. 2015). Furthermore, some AMPs have been produced using an E. coli system in combination with fusion tags such as thioredoxin (Feng et al. 2012) glutathione S-transferase (GST) , maltose-binding protein (MBP) (Velásquez et al. 2011) and small ubiquitinlike modifier like protein (SUMO) and subsequent cleavage of the AMPs form the fusion proteins by proteases such as thrombin, tobacco etch virus NIa protease, bovine coaglulation factor Xa, and enterokinase, which recognize only short, linear peptide sequences. However, if the AMP activity can be retained without cleavage from the fusion construct, it would be so much convenient for the toxicity assay of the AMPs. For this purpose, the SUMO technology has been successfully applied to many AMPs such as plectasin , cathelicidin (Luan et al. 2014) and CM4 (Li et al. 2011).
In the present paper, we attempted to express Bac7 (1-35) in E. coli cells. Since Bac7 (1-35) is highly toxic to E. coli, it is essential to suppress its toxicity for its production. For this purpose, we tested two different protein tags, SUMO and protein S from Myxococcus xanthus to examine if the fusion tags could reduce the toxicity and enhance the expression of Bac7 (1-35). Protein S is a major spore-coat protein, which has been used as an effective fusion tag (Kobayashi et al. 2009). Protein S consists of 173 amino acids, which is composed of two homologous domains, the 92-residue N-terminal and the 81-residue C-terminal domains (Bagby et al. 1994). The expression vector, pCold-PST, contains two N-terminal domains (PrS 2 ), repeated in tandem, to the C-terminal end of which a cloned protein or peptide is fused. pCold-PST vector has been shown to enhance the expression as well as the solubility of a cloned protein (Kobayashi et al. 2009). Protein S fused to a target protein has been shown not to severely affect the structure and function of the protein to be fused (Kobayashi et al. 2009(Kobayashi et al. , 2012. Since AMPs are toxic to the cells when expressed in E. coli, the suppression of toxicity to the cells possibly is essential for the production of AMPs. For this purpose, we attempted to use the Single-Protein Production (SPP) system for the production of PrS 2 -Bac7 (1-35), in which MazF, an ACAspecific endoribonuclease from E. coli is induced to eliminate almost all cellular mRNAs except for the mRNA for His 6 -PrS 2 -Bac7 (1-35) that is designed to have no ACA sequences without altering its amino acid sequence (Zhang et al. 2003;Suzuki et al. 2005Suzuki et al. , 2006. This enables us to produce only His 6 -PrS 2 -Bac7 (1-35) in E. coli cells without producing any other cellular proteins. Indeed, we were able to produce His 6 -PrS 2 -Bac7 (1-35) in a reasonable amount in E. coli cells, while with the SUMO tag, we were unable to express the protein.
Production and purification of His 6 -PrS 2 -Bac7 (1-35) using SPP system BL21(DE3) co-transformed with pACYCmazF and pColdPrS 2 -Bac7 (1-35) was inoculated into 1 l of LB medium and the culture was incubated at 37 °C. When the OD 600 reached 0.8, the culture was chilled on ice for 5 min, followed by incubation at 15 °C for another 1 h. Subsequently, 1 mM IPTG was added to induced MazF and the culture was incubated at 15 °C for overnight (Suzuki et al. 2006). The cells were collected by centrifugation and re-suspended into 20 ml of binding/washing buffer consisting of 20 mM Tris-HCl (pH 8.0), 500 mM NaCl, and 20 mM imidazole-HCl (pH 8.0). After breaking the cells by using a French press, the unbroken cells were removed by centrifugation at 14,000 rpm for 20 min. The supernatant thus obtained was subjected to further centrifugation at 50,000 rpm for 30 min to remove the membrane fraction. The supernatant fraction was mixed with 1 ml of Ni-resin equilibrated in binding/wash buffer and the mixture was incubated for 1 h at 4 °C. The Ni-resin was washed twice with 10 ml of washing buffer, and His 6 -PrS 2 -Bac7 (1-35) was eluted with 20 mM Tris-HCl (pH8), 500 mM NaCl and 300 mM imidazole-HCl (pH8.0). After collecting the eluted protein, the protein concentration was determined by the optical density at 280 nm using Nano Drop (Thermo Scientific), and the purity was examined by SDS-PAGE. After the protein fraction was dialyzed against 20 mM Tris-HCl (pH8.0) and 100 mM NaCl, it was concentrated to 3 mg/ ml and stored at −80 °C.

Cleavage by factor Xa and identification of Bac7 (1-35)
The 30 μg of His 6 -PrS 2 -Bac7 (1-35) was digested with 4 μg of factor Xa in a 50 μl mixture containing 20 mM Tris-HCl (pH 8.0). 50 mM NaCl and 2 mM CaCl 2 . The reaction mixture was then incubated for 4 h at 37 °C, and the cleavage product was analyzed by 19 % SDS-PAGE followed by Coomassie blue staining. In order to confirm Bac7 (1-35) by mass spectrometry, His 6 -PrS 2 -Bac7 (1-35) was cleaved by factor Xa protease and the reaction mixture was diluted 25 times by the matrix solution containing sinapinic acid (10 mg/ml) in 0.1 % trifloroacetic acid and 50 % acetonitrile) and spotted on to a target plate (Opti-TOF 384 well insert, ABSciex) and air dried, followed by mass spectrometric analysis by a MALDI-TOF (4800 MALDI-TOF/TOF, ABSciex) using the positive mid-mass linear mode from 2 to 30 kDa.

In vitro translation inhibition assay
PURExress In Vitro Protein Synthesis kit (New England BioLabs) was used in this study. The gene for dihydrofolate reductase (DHFR) was used as a positive control. The reaction mixture containing buffer A and buffer B supplied from NEB were mixed with 20 U of RNase inhibitor (Roche), linearlized DNA (10 ng/μl) and synthetic peptide Bac7 (1-16) (10 μM) or His 6 -PrS 2 -Bac7 (1-35) (10 μM) or water, and the reaction mixture was incubated for 2 h at 37 °C. Protein production was examined by SDS-PAGE followed by Coomassie blue staining.

Growth inhibition test when His 6 -PrS 2 -Bac7 (1-35) is induced in E. coli
The E. coli strain, BL21(DE3) harboring either pColdPrS 2 or pColdPrS 2 -Bac7 (1-35) was grown in the M9-glucose medium. When the OD 600 reached at 0.2, 1 mM IPTG was added into the medium to induce the protein. As a negative control, the culture medium in the absence of IPTG was also incubated and the OD 600 was monitored every 30 min.

The antimicrobial activity of purified Bac7 (1-35) in E. coli
The E. coli strain BL21 (DE3) was grown in the M9-glucose medium and purified Bac7 (1-35) was added at the final concentration of 2 μM into the medium when OD 600 reached 0.2. OD 600 was monitored every 30 min.

Results
The expression and purification of His 6 -PrS 2 -Bac7 (1-35) Since Bac7 (1-35) is highly toxic to E. coli cells, we attempted to express Bac7 (1-35) fused to the C-terminal end of protein S or SUMO at 15 °C. As a result, we were not able to observe the production of His 6 -SUMO-Bac7 (1-35), while His 6 -PrS 2 -Bac7 (1-35) was produced as detected at around 30-kDa position in SDS-PAGE gel (Fig. 1a). To further improve the expression of His 6 -PrS 2 -Bac7 (1-35), we attempted to apply the SPP system for its production. Since the SPP system allows one to produce only a target protein without producing any cellular proteins, it may help to produce His 6 -PrS 2 -Bac7 (1-35) to a better yield. As shown in Fig. 1a, the use of the SPP system indeed enhanced the production of His 6 -PrS 2 -Bac7 (1-35). Notable the use of the SPP system for the production His 6 -SUMO-Bac7 (1-35) was unsuccessful, probably because the SUMO tag could not suppress the toxicity of His 6 -PrS 2 -Bac7 (1-35).
After fractionation by ultracentrifuge, His 6 -PrS 2 -Bac7 (1-35) was fully recovered in the soluble fraction (Fig. 1b). The final yield after purification using Ni-NTA column chromatography was determined by a Nano Drop spectrophotometer to be 2.5 mg from 1 l LB medium. Higher than 90 % purification was achieved by one-step Ni-NTA purification (Fig. 1b).

The function of His 6 -PrS 2 -Bac7 (1-35)
Although PST is known not to interfere the function of the fusion partner (Kobayashi et al. 2009), we next examined the inhibitory activity of the purified His 6 -PrS 2 -Bac7 (1-35) using a cell-free protein synthesis system (New England BioLabs) comparing with the inhibitory ability of intact fusion protein. As a positive control, the expression of dihydrofolate reductase (DHFR; 20 kDa) was examined in the absence and presence of PrS 2 . As shown in Fig. 3a, the addition of PrS 2 did not have any effects on the protein synthesis. Next, the inhibitory effects were compared between His 6 -PrS 2 -Bac7 (1-35) and His 6 -PrS 2 -Bac7 (1-35) treated with Factor Xa. We also synthesized Bac7 (1-16), which was recently reported to inhibit protein synthesis (Seefeldt et al. 2016) and used as a positive control for the experiment. As shown in Fig. 3b, both His 6 -PrS 2 -Bac7 (1-35) and His 6 -PrS 2 -Bac7 (1-35) treated with Factor Xa inhibited the protein synthesis as well as Bac7 (1-16).

The antimicrobial activity of His 6 -PrS 2 -Bac7 (1-35) in the cells
Since His 6 -PrS 2 -Bac7 (1-35) was shown to retain the ribosome inhibition activity, we have tested the growth effect of induction of His 6 -PrS 2 -Bac7 (1-35) in the cells. As shown in Fig. 4a, the cell growth was totally arrested by His 6 -PrS 2 -Bac7 (1-35) after 30 min of induction, while the induction of His 6 -PrS 2 did not cause the cell growth arrest.

Discussion
The AMP production in E. coli is challenging because their antimicrobial activity. To suppress their toxicity, relatively large tags such as GST, MBP and SUMO may be used, however, for the most of AMP production, SUMO has been widely applied and many AMPs were successfully expressed as functional forms (Li et al. 2011;Zhang et al. 2014Zhang et al. , 2015. The SUMO tag has been shown to improve protein folding and solubility, and to be used for protein detection (Luan et al. 2014). Thus, we attempted to examine if the fusion of Bac7 (1-35) to the C-terminal end of SUMO could suppress the Bac7 toxicity, but the production of the fusion protein was not detected, indicating that the SUMO tag could not suppress the toxicity of Bac7 (1-35), which is known to inhibit the function of 70S ribosomes (Mardirossian et al. 2014). Thus, we next tried protein S as a fusion tag for Bac7 (1-35). The Protein S from Myxococcus xanthus is known to function as an intra-molecular chaperone without severely affecting the function of the protein fused to it, and has been applied for the expression of proteins which are insoluble and/or difficult to be expressed (Kobayashi et al. 2009(Kobayashi et al. , 2012. In the present study, we used two 88-residue N-terminal domains repeated in tandem to the C-terminal end of which Bac7 (1-35) was fused. The resultant His 6 -PrS 2 -Bac7 (1-35) was indeed expressed well in the SPP system. In this PST-SPP system, an ACAless gene encoding His 6 -PrS 2 was used as an N-terminal tag for Bac7 (1-35) to produce His 6 -PrS 2 -Bac7 (1-35) (Fig. 5). We also constructed the ACA-less His 6 -SUMO-Bac7 (1-35) system. However, His 6 -PrS 2 -Bac7 (1-35) was expressed (Fig. 1a) while the expression of SUMO-Bac7 (1-35) was not detected, indicating that the SUMO tag was not able to suppress the Bac7 (1-35) toxicity even with use of the SPP system. Notably, however, the expression of His 6 -PrS 2 -Bac7 (1-35) was rather low, possibly because protein S fusion to Bac7 (1-35) did not completely suppress the toxicity of Bac7 (1-35). Indeed, pColdPrS 2 -Bac7 (1-35) in M9-glucose medium was toxic in the presence of 1 mM IPTG (Fig. 4a).
Bac7 (1-35) has been shown to inhibit the function of 70S ribosomes to block protein synthesis (Mardirossian et al. 2014). Thus, the activity of His 6 -PrS 2 -Bac7 (1-35) was tested using a cell-free protein expression system with a synthetic peptide Bac7 (1-16) as a protein synthesis inhibitor (Seefeldt et al. 2016) as a control. As shown Fig. 3b, the production of DHFR by the cell-free system was indeed inhibited by both His 6 -PrS 2 -Bac7 (1-35) and Bac7 (1-35) which was generated from His 6 -PrS 2 -Bac7 (1-35) by Factor Xa treatment which resulted in a small amount of uncleaved His 6 -PrS 2 -Bac7 (1-35) (Fig. 2c). Notably, the cleavage mixture effectively inhibited the protein synthesis (Fig. 3b). Since the minimum inhibitory concentration of Bac7 (1-35) has been reported to be 0.5 μM (Benincasa et al. 2004), it is assumed that there was an excessive amount of Bac7 (1-35) in the reaction mixture to inhibit protein synthesis. The PrS 2 tag is known not to interfere with its fusion partner (Kobayashi et al. 2009); for example, PrS 2 fused at the N-terminal end of OmpR, a phosphor sensory protein, did not inhibit the OmpR function at all (Kobayashi et al. 2009). Thus, it is not surprising to see that His 6 -PrS 2 -Bac7 (1-35) possesses an antibacterial activity in spite of the fact that the N-terminal part of Bac7 (1-35) has been shown to be crucial for the antimicrobial activity (Guida et al. 2015).
Using the SPP system, MazF cleaves at all ACA in mRNA while only the codon-optimized ACA-less gene for His 6 -PrS 2 -Bac7 (1-35) remains intact. Therefore, upon induction of MazF, only His 6 -PrS 2 -Bac7 (1-35) is produced in  Schematic presentation of the PST-SPP system. PrS, a major spore-coat protein from Myxococcus xanthus, was directly repeated (PrS 2 ) and used as a tag (PST tag) for higher expression and solubilization of a target protein. Functional assay for the target protein can be done without cleaving the PrS 2 from the fusion protein. A codon optimized ACA-less gene for the target protein can be expressed together with pACYCmazF, cleaving the ACA sequences in mRNAs, allowing the cells to produce the only target gene from its ACA-less mRNA (the SPP system; Suzuki et al. 2005). Note that the gene for PrS 2 is also codon-optimized for E. coli and designed to be ACA-less without altering the amino acid sequence. Since the PrS 2 tag partially suppresses the antimicrobial activity of a cloned peptide or protein, the use of the PST-SPP system allows one to produce toxic peptides and proteins in E. coli and also to perform their functional assay using the cell lysate