A Cell-free Infection System to Study Translation, Replication and Phage-particle Production during Infection of E. coli

Background: Viruses infect all kingdoms of life, and new species are continuously being discovered. The single-stranded (+)-RNA viruses comprise the largest group of viruses, which includes pathogens such as Dengue virus, Corona virus and West Nile virus. Also, the simple bacteriophage Qβ belongs to this group of viruses. Studies of the mechanism of Qβ infection can increase our general understanding of the singlestranded (+)-RNA viruses, which can be exploited in the pursuit to treat and prevent diseases caused by pathogenic (+)-RNA viruses. Methods: In this study, we have analysed the production of infectious Qβ phage particles in three different cell-free infection systems upon addition of the Qβ genome as a template. The cell-free infection systems were based on cell-free protein expression systems: two commercial systems and one custom-made system. We studied the course of infection by analysing the production of viral RNA, proteins and phage particles produced in the cell-free reactions. The replication of the viral RNA was determined by RTPCR, while the translation of the viral proteins was J Biotechnol Biomed 2022; 5 (2): 94-116 DOI: 10.26502/jbb.26421280051 Journal of Biotechnology and Biomedicine 95 examined by radiolabelling, and the production of infectious phage particles was evaluated by doublelayered plaque assays. Results: Bacteriophage Qβ was found to replicate in two of the three tested cell-free infection systems. Specifically, the viral RNA was replicated, the viral proteins were translated, and infectious phage particles were produced in the cell-free infection systems. The pattern of translation regulation of the viral proteins appeared similar to in vivo infection. Infectious Qβ phage particles were produced at yields of 2.5 × 10 PFU/μL reaction and 2.5 × 10 3 PFU/μL reaction in the commercial and custom-made system, respectively. Importantly, intact Qβ phage particles were shown not to replicate in the cell-free infection systems under the tested conditions. Conclusion: Cell-free infection systems can support replication of viral RNA, translation of viral proteins and self-assembly of infectious Qβ phage particles. We provide opportunities for further optimisation of the phage particle yield. Cell-free infection systems can be used in the future to study newly discovered viruses, the development of antiviral and antibacterial drugs, and in biotechnology.

complex pathogenic (+)-RNA viruses [7]. Bacteriophage Qβ infects the gram-negative bacterium Escherichia coli (E. coli), but it only targets so-called male bacteria with F-pili expressed on the surface. The Qβ genome is only 4217 nucleotides long and encodes a maturation protein, a coat protein, a read-through protein and a RNAdependent RNA polymerase (RdRP; known as the βsubunit) [7,8]. The read-through protein is occasionally produced as a result of leaky termination during translation of the coat protein [9]. The icosahedral Qβ capsid enclosing the (+)-RNA genome is composed of 176 copies of the coat protein, 3-5 read-through proteins and one molecule of the maturation protein [10,11].
During the first step of the Qβ life cycle, the phage particle adsorbs to the surface of E. coli via contact between the maturation protein on the viral surface and the bacterial F-pilus [12,13]. This interaction promotes the release of the Qβ (+)-RNA genome into the host cell via a poorly understood mechanism, which may rely on the retraction of the F-pilus to pull the complex of the maturation protein and the Qβ (+)-RNA into the host cell [13,14]. Next, the Qβ (+)-RNA is translated by host ribosomes in a tightly controlled process [15]. Initially, the coat protein is translated, which causes a rearrangement of the secondary structure of the (+)-RNA to enable the subsequent translation of the replicase gene and the maturation gene [16]. The coat proteins form dimers that bind to the operator region of the replicase gene, thereby blocking the binding of ribosomes and consequently abolishing translation [17]. only approximately 450 molecules of the β-subunit are produced [10]. The Qβ (+)-RNA is replicated via formation of a complementary (-)-strand intermediate as soon as the β-subunit is produced. The active Qβ replicase complex is composed of the viral β-subunit and the host proteins EF-Tu, EF-Ts and ribosomal protein S1 [18]. Also host factor Q (Hfq) assists the Qβ replicase complex by recognising the 3' end of the Qβ (+)-RNA [19], and ensuring a higher affinity for Qβ (+)-RNA than for host mRNAs [20]. In the uninfected cell, Hfq acts as a global regulatory protein that controls the expression of many mRNAs by facilitating interactions between small regulatory RNAs and mRNA molecules [21]. Late in the infection cycle, the (+)-RNA virus particles are assembled. The maturation proteins accumulate in the cell wall of the host cell, which results in cell lysis and release of the progeny viruses [22].
The study of the kinetics of viral infections in vivo relies on the ability to synchronise the infection [23,24], which depends on the infectivity and cellular conditions of the individual cells. Common synchronisation methods include inoculation at 4°C and starvation of which the first approach lowers the infectivity of the cells, and both approaches may affect other cellular processes [24,25]. Magnetofection is an alternative method of synchronisation of in vivo infections, which however, is technically demanding [26]. Cell Importantly, CFI systems allow the study of the role of essential host proteins in viral replication, without considering the survival of host cells [27]. Furthermore, viral infections can be studied under different environmental conditions e.g. the cell-extract can be prepared from cells grown under stressed or other potentially interesting conditions, since a cell-extract made from stressed cells can still support protein synthesis [28]. In this study, we test the ability of three different cell-free protein synthesis systems to support the production of infectious phage particles by directly adding the Qβ RNA genome or a cDNA copy hereof to the cell extract, thereby generating a CFI system. We show that the Qβ phage can replicate in two out of three tested CFI systems. Specifically, the Qβ (+)-RNA is translated and replicated, and infectious progeny phage particles are successfully assembled. Furthermore, we show that intact Qβ phage particles cannot replicate in a CFI system.

Preparation of viral templates
The pQβ7 plasmid is an infectious plasmid encoding a full-length cDNA copy of the Qβ genome [29]. The pQβ7 plasmid holds an ampicillin resistance gene. The plasmid was purified using a Midiprep kit (Nucleobond Xtra Midiprep, Macherey-Nagel) followed by phenol/chloroform extraction and ethanol precipitation.
The purified pQβ7 plasmids were used as a template in two different ways: directly as a template in CFI systems or as a template for in vitro RNA transcription of Qβ (+)-RNA, which subsequently served as the template in CFI systems. In the latter case, the pQβ7 plasmid was linearized using SmaI (Thermo Scientific) to ensure well-defined 3'-ends of the resulting transcripts, while the 5'-ends are dictated by the upstream T7 promoter. The linearized DNA was purified by phenol/chloroform extraction and ethanol precipitation. Qβ (+)-RNA was produced by the "TranscriptAid T7 High Yield Transcription Kit" (Thermo Scientific) using the linearized pQβ7 as a template. The RNA products were purified by phenol/chloroform extraction followed by ethanol precipitation and the yield and homogeneity of the RNA was determined by 1% agarose gel electrophoresis.
The cells were harvested and resuspended in column buffer (20 mM Tris/HCl pH 8, 500 mM NaCl, 1 mM EDTA and 1 mM PMSF pr. gram cells) and subsequently lysed with lysozyme (0.5 mg/g cells) and deoxycholate (0.1 mL/g cells). The lysed cells were DNase treated (0.5 mg/mL, Roche) and centrifuged (4,500 rpm, 40 minutes). The supernatant was applied to a chitin column (New England Biolabs) equilibrated with column buffer followed by washing overnight with the same buffer at 4°C. On the following day, the column was flushed with flush buffer (20 mM Tris/HCl pH 8, 500 mM NaCl, 1 mM EDTA and 50 mM DTT), the flow was stopped, and the column was incubated for 24 hours at 4°C and for 24 hours at room temperature.
Hfq liberated by intein splicing was eluted with column buffer and dialysed overnight at 4°C against dialysis buffer (20 mM Tris-HCl, 200 mM NaCl). The protein concentration was determined by the Bradford procedure (Bio-Rad).

Cell-free infection systems
The ability of the Qβ virus to replicate in a bacterial cell-free system was tested in three different cell-free protein expression systems, which were either of commercial origin or homemade. The Qβ genome was used as a template either as a plasmid-borne cDNA copy or as an in vitro transcribed (+)-RNA, which were directly added to the systems. The "PURExpress ® In Vitro Protein Synthesis" (PURE) kit (New England Biolabs) contains all the purified E. coli components required for cell-free translation. Additionally, it contains T7 RNA polymerase to enable transcription from a DNA template driven by a T7 promoter. 25-μL reactions were set up following the instructions from the manufacturer with intact pQβ7 plasmid (0.8 μg) or Qβ (+)-RNA (6 μg) as templates. Since Hfq was not expected to be part of the system, this host protein was either added (4.4 μg) in 10-1,000 times molar excess over added genomes or left out to study its importance. as templates. The protocol from the manufacturer was followed and the reactions were incubated at 25°C for 8-16 hours or at 37°C for 3 hours at 160 rpm. The ability of intact Qβ phage particles to replicate in a CFI system was tested using the S30 T7 kit. Reactions containing ~50 or ~200 Qβ phage particles instead of the DNA or RNA template were prepared. A separate reaction with 1.6 μg pQβ7 was included as a positive control. The reactions were incubated at 25°C for 16 hours at 160 rpm. A CFPS system was made by growing and processing BL12(DE3) E. coli into a functional S30 T7 extract following the protocol established by Krinsky et al. [31]. The extract contains T7 RNA polymerase, since the strain contains the λDE3 lysogen that carries the gene for T7 RNA polymerase. The buffers and solutions were prepared as described by Krinsky et al. [31]. The aromatic amino acids were dissolved by adding a small amount of NaOH, and undissolved amino acids were removed by filtration (0.45 μm). E. coli BL21(DE3) cells were grown as described by Krinsky et al. [31].
The cells were harvested, and the pellet was resuspended

Analysis of phage production
The production of infectious Qβ phages in the CFI systems was evaluated with a double-layered plaque assay [32]. Purified Qβ phages were included as a positive control, and top agar, uninoculated overnight culture, LB medium and H 2 O were included as negative controls to assess potential sources of contamination.
An area of the gel containing a high concentration of free [ 14 C]-serine caught in the front was omitted to achieve higher sensitivity, and the gel was exposed for an additional 6 days and scanned again. The brightness and contrast were optimised equally on the entire blot.

Analysis of viral RNA replication by RT-PCR
The viral RNA replication in the CFPS system was analysed by reverse transcription polymerase chain reaction (RT-PCR). Total RNA was purified from CFPS reactions added either pQβ7, Qβ (+)-RNA or pET vector as a template and from purified, intact Qβ phages using the "Nucleospin RNA" kit (Macherey Nagel).
Additional DNase treatment of the RNA was performed using the "RNase-free DNase I" kit (Thermo Scientific).
The purified RNA was analysed on a 1% agarose gel, and the RNA concentration was determined on a "Nanodrop" instrument measuring the absorbance at 260 nm. First-strand synthesis was carried out using the "RevertAid First Strand cDNA Synthesis" kit (Thermo Scientific) using the Qβ_genome_3'UTR primer (Sigma) ( Table 1)

Results
In this study, three different CFPS systems were tested for their ability to sustain the production of infectious phage particles and thereby work as CFI systems by adding the viral genome as a template. Two of the tested systems are commercial (the "PURExpress ® In Vitro Protein Synthesis" system from New England Biolabs (PURE) and the "S30 T7 High-Yield Protein Expression System" Promega (S30 T7)), while the third system is based on a homemade CFPS system [31]. Table 2 provides an overview of the three systems.   (29), or a RNA template (Qβ (+)-RNA). "Reactions, pQβ7" indicates the processes taking place in the reaction, when pQβ7 is added in the CFI system. Translation and replication are normal steps in the viral life cycle (italics), while T7 transcription of DNA is an additional step that is independent of the viral infection cycle (non-italics) and happens initially, when pQβ7 is used as a template in the CFI systems. "Reactions, Qβ (+)-RNA" indicates the processes involving the Qβ genome during infection, when Qβ (+)-RNA is added in the CFI system. "Hfq" indicates if Hfq is added externally or is expected to be intrinsically present in the CFI system.

Phage production in commercial proteinexpression systems
The two commercial protein-expression systems, PURE and S30 T7, were tested for their ability to support viral infection and cause phage particle production upon addition of the Qβ genome either as an infectious plasmid (pQβ7) [29] or as Qβ (+)-RNA.  dilution" indicate the observed number of plaque-forming units (PFU) on the plates containing the 10 -2 , 10 -4 and 10 -6 dilutions, respectively. The number of plaques was counted, where density allowed a sound determination of the number of PFU. "Yield PFU/μL" indicates the number of PFU related to the volume of the CFI reaction. "nd": not determined due to a high plaque density making the precise counting impossible.

Intact Qβ phage particles do not replicate in a CFI system
The ability of intact Qβ phage particles to replicate in a CFI system was tested using the S30 T7 system to verify that the observed number of PFU in the CFI systems are products directly derived from the added Qβ genomes.
Purified Qβ phage particles were added to the   μg) were added to the S30 T7 system to investigate their ability to sustain the production of infectious phage particles upon incubation in the system. A negative control without template was made in parallel. The PFU yield of the reactions was analysed with a double-layered plaque assay. "PFU count undiluted", "PFU count 10 -2 dilution", "PFU count 10 -3 dilution" and "PFU count 10 -4 dilution" indicate the observed number of plaque-forming units (PFU) on the plates containing the undiluted sample or the 10 -2 , 10 -3 and 10 -4 dilutions, respectively. The number of plaques was counted, where density allowed a sound determination of the number of PFU. *1 PFU is not considered statistically sound. "Yield PFU/μL" indicates the number of PFU related to the volume of the CFI system. "nd": Not determined due to a high plaque density making the precise counting impossible. "na": Not analysed.
The expected number of phage particles in the S30 T7 reactions with added Qβ phages were 1 and 4 PFU/μL, respectively, based on the assumption that intact phage particles cannot serve as templates in a CFI system.
These numbers are comparable to the observed PFU counts of 0.8 and 3.8 PFU/μL, respectively (Table 4).
This confirms that intact Qβ phage particles in the CFI system do not support the production of Qβ phages and justifies that the observed PFU counts reported in Table   3 entirely results from the replication of the added DNA template. The included positive control (Table S1) confirms the functionality of the system with a yield equivalent to the replication of Qβ reported in Table 3.

Production of infectious Qβ phage particles in the CFPS system
The homemade CFPS system is based on the protocol described by Krinsky et al. [31]. The reaction involves an E. coli cell extract, which has been prepared by S30 centrifugation after cell lysis and removal of debris.

Analysis of the production of Qβ-encoded proteins in the CFPS system
The Qβ genome encodes the maturation protein (48.5 kDa), the coat protein (14.3 kDa), the read-through protein (36 kDa) and the catalytic RdRP subunit called the β-subunit (65.5 kDa) [8]. The production of viral proteins in the CFPS system was evaluated by including

CFPS system
The production of Qβ virus particles requires translation of viral proteins, replication of the viral genome and assembly of phage particles [10]. The production of Qβ indicating the presence of Qβ RNA, is 1036 bp. Figure 4 shows that a PCR product of ~1036 bp appears in the reaction based on RNA extracted from the CFPS system with pQβ7 as a template (Figure 4, 1 and Table 1). RNA extracted directly from purified Qβ phages was included as a positive control. Negative controls without primers, template or reverse transcriptase were also included. Next, PCR reactions were run with primers annealing to the Qβ genome, and the products were analysed by 1.

Production of infectious phage particles in CFI systems
In the present study, E. coli-based CFI systems were tested for their ability to support Qβ infection. The goal was to develop a model CFI system allowing viral translation and replication, and the production of infectious phage particles to be studied. The commercial protein expression system PURE is based on purified components from E. coli essential for T7 transcription and translation. The potential production of infectious Qβ phage particles in the PURE system was evaluated with a double-layered plaque assay (Table 3). In general, viruses are well-known to exploit a variety of host proteins during infection, and EF-Tu, EF-Ts, S1 and Hfq have been found to be important for replication of Qβ (+)-RNA [18]. Unlike the other mentioned host proteins, Hfq is not directly associated with E. coli transcription or translation, and it was therefore not expected to be included in the PURE system. Thus, PURE reactions were set up with or without added Hfq to evaluate its importance. Hfq was added in 10-1,000 times molar excess relative to added genomes. Surprisingly, the PURE system did not support the production of infectious phage particles irrespective of the presence of Hfq (Table 3). The replication of (+)-RNA viruses is dependent on the interaction between a wide range of host factors with a few viral components during cell entry, gene expression, virus assembly and release. (+)-RNA viruses, especially, rely on the interaction with RNA-binding host proteins [1], and new host proteins employed by viruses are continuously discovered [34].
Thus, it is possible that Qβ is unable to replicate in the PURE system due to the absence of one or more host proteins essential for phage particle production. The host proteins involved in the replication of the Qβ (+)-RNA appear to be well established [35][36][37][38], while other steps during the infection cycle (e.g. the assembly of phage particles) may rely on unknown interactions with host proteins. The commercial S30 T7 system was initially tested for its ability to replicate Qβ from an added genome template. The double-layered plaque assay showed that this system successfully produced infectious Qβ phage particles at a yield of 3.5 × 10 4 to 2.5 × 10 5 PFU/μL reaction (Tables 3 and 4). The production of Qβ phage particles is solely originating from the added viral genomes, since intact Qβ phage particles were found unable to replicate in the S30 T7 system (Table 4). This observation seems reasonable, since the Qβ (+)-RNA is expected to be inaccessible, when it is encapsulated in the particle. Specifically, the Qβ (+)-RNA is highly structured with stem-loop structures binding to coat-protein dimers and the maturation protein composing the viral capsid [22].
Most likely, the release of Qβ (+)-RNA relies on the interaction between the maturation protein and the Fpilus on the surface of live E. coli cells [39], which is unlikely to occur in a CFI system, where cell walls have been removed by S30 centrifugation.

Production of viral proteins during Qβ infection in the CFPS system
In addition to the detection of infectious phage particles, the production of viral proteins in the homemade CFPS system was analysed by radiolabelling ( Figure 3)

Detection of Qβ RNA in the CFPS system
The replication of viral RNA in the CFPS system was analysed by RT-PCR using primers specific for the Qβ genome ( Figure 4).

Cell-free infection systems
A few viruses have previously been synthesised in cellfree transcription/translation systems under different conditions and with varying yields.  Table 6: Overview of cell-free virus synthesis. "Host extract" lists the host cell from which the extract in the cell-free system is originating. "Virus" indicates the virus, which is produced in the CFI system. "Genome" shows the type and size of the viral genome, while "RNA or DNA copy" indicates that either an RNA or DNA copy has been added as a template in the CFI system, and "Genome concentration" gives the concentration of the added genomic template. "PFU/μL reaction" indicates the yield of infectious viruses produced per μL cell-free reaction. "Genome replication" and "Protein synthesis" show if viral genome replication and protein synthesis has been analysed. "Ref." lists the references of the included studies, where "*" indicates the unoptimized CFI systems. This table is not meant to provide a complete overview of all cell-free virus productions. na: Not analysed.
Notably, the present study monitors both replication and translation of the viral genome in the CFI system, in addition to the production of infectious phage particles.
The yield of the CFPS system used in this work is 2.4-2.5 × 10 3 PFU/μL reaction, which is comparable to or higher than the yields obtained in other unoptimised cell-free systems (marked with * in Table 6). It has previously been shown that optimisation of various components of the CFI systems can improve the phage particle yield dramatically. Amongst the possible steps of optimisation is the addition of molecular crowders (e.g. PEG 8000). PEG 8000 mimics the molecular crowding in the cell, which at an optimal concentration can stimulate self-assembly of macromolecular complexes, while an excessive amount of PEG can result in aggregation and misfolding [43]. In the CFPS system presented here, 3% PEG 6000 was included based on Krinsky et al. [31], but an optimisation of the concentration and degree of polymerisation of PEG might increase the yield significantly. Ideally, the concentration of proteins and other essential molecules in the cell extract should be adjusted to mimic the composition of the E. coli cytoplasm [44]. Specifically, an increased concentration of translation factors and an optimisation of NTP concentrations have been shown to increase the yield [45,46], together with the adjustment of the magnesium ion concentration, since the magnesium concentration affect ribosome functionality [47]. In the present CFPS reaction, the cytoplasmic extract is diluted to contain 24.3 mg/mL protein, compared to ~10 mg/mL protein in other cell-free systems [44,48], which indicates a potential for further adjustment of concentrations and ratios of individual components.

Future perspectives
In this study, infectious Qβ phage particles have been produced in a cell-free system. Qβ has previously been used in the development of anti-viral drugs and vaccines [49,50]. Additionally, Qβ has been used in phage therapy targeting specific pathogenic bacteria, which are eliminated by lysis. This method enables specific targeting of pathogenic bacteria [51]. In the future, CFI systems can potentially be employed in similar biomedical applications. Furthermore, Qβ-based CFI systems may be suitable in evolution studies, since the high mutation rate of the Qβ RdRP complex causes the accumulation of point mutations in the produced RNA genomes [52]. CFI systems are advantageous in allowing the study of interactions between viral and host cell components, which can pose a challenge in vivo e.g.
if the host protein of interest is crucial to the cell.
Ribosomal protein S1 and Hfq are examples of essential, RNA-binding host proteins that are exploited by the Qβ virus. However, the functional relevance of Hfq and S1 during viral infection cannot be studied by mutagenesis in vivo, since mutations are prone to cause cell death or other unwanted reactions in the cells [19]. This, however, would not be a problem in a cell-free extract depleted of the protein of interest e.g. via degron tagging [53]. Additionally, the in vitro infection systems may facilitate the study of newly discovered bacteriophages [5,6], and knowledge of their life cycle can be employed for specific medical and biotechnological  [5,6]. In general, the expansion of the phage diversity contributes to the study of the impact of viruses on their hosts.

Conclusion
Here, we have shown that CFI systems can be established for the bacteriophage Qβ based on a commercial transcription/translation S30 T7 system as well as a custom-made CFPS system [31]. In the latter system, we show that viral proteins are produced with an overall pattern of translation regulation comparable to in vivo conditions. In addition, replication of the viral RNA is demonstrated for this CFI system. Both CFI systems support the self-assembly of infectious Qβ phage particles with yields that are comparable to other unoptimized CFI systems, and the home-made CFI system provides ample opportunity for further improvements via careful titration of crucial components. For future studies, the CFI systems can be used to study known and newly discovered viruses.
Understanding the molecular details of viral infections can lead to the development of anti-viral drugs and vaccines, and it can be used in phage therapy.

Ethics approval and consent to participate
Not applicable.

Consent to publication
Not applicable.

Availability of data and materials
All data generated or analysed during this study are included in this published article and its supplementary information files.