Team:Wageningen UR/Results/Phage Engineering


subproject hexbadge



To eliminate Xylella fastidiosa from plants without collateral damage Xylencer must be very specific. Phages are very specific and have been found to be effective against X. fastidiosa [1]. However, phage therapy is not yet effective for practical use in plants, as it is difficult to apply and bacteria develop resistance [2]. Phage engineering can be the solution to the current flaws of phage therapy. To solve the problems concerning application we can equip the phage with chitin-binding proteins that allow phage to bind to insects that feed on plants and use them to spread. This will allow the phages to follow X. fastidiosa, which uses the same mechanism to spread. Read more on the chitin-binding proteins here. To solve the problem of X. fastidiosa acquiring resistance against phages and escaping the plant immune system, the phage will induce plant immunity through Microbe-Associated Molecular Patterns (MAMPs) [3]. Apart from lysing X. fastidiosa, the phage will produce proteins that the plant recognizes as pathogens, activating the plant immune system against X. fastidiosa. Read all about the plant immunity here.

Figure 1: An overview of the combined phage engineering methods used and the steps used to connect them

To create a phage that can do this, we need to engineer the phage genome. Unfortunately, we cannot work with X. fastidiosa because of quarantine measures and slow growth [4]. Instead we work with model organisms E. coli and phage Lambda. Phage Lambda is a good model phage as it is well studied and has capsid structures that are similar to those of X. fastidiosa phages, that can be used to fuse proteins to. To edit phage Lambda, we use two novel methods of genome engineering and phage production. Namely yeast based genome engineering and cell free production of phages using a Transcription Translation (TXTL) mixture, which are shown in figure 1. Together, these methods allow genetic engineering and production of phages without use of the host. This makes our work applicable to future work with X. fastidiosa phages.

We were able to assemble genomes in yeast and produce phages using TXTL. While application to engineered phage Lambda was not reached yet, many important steps have been made towards it.


Increasing interest in phage therapy requires new methods for phage engineering [5]. The current methods are in vitro engineering of phages or engineering through homologous recombination after infection of bacterial cells with wild type phage. The in vitro approach consists of digesting the phage genome using native restriction sites, subcloning the fragments that should be modified and, finally, ligate them. This technique is often inefficient due to lytic phage genomes being too large. The second method consists of transforming bacteria with a fragment of modified phage DNA, with arms that are homologous to the phage genome. Cells are then infected with the phage and the engineered DNA sequence is inserted into its genome through homologous recombination. Consequently, a fraction of the progeny will be genetically modified. This method requires the insertion or deletion of DNA sequences in a time frame that is too long for most phage lytic cycles. The percentage of engineered phages is directly linked to the homologous recombination efficiency of the host cells, which is generally very low. Therefore, the screening times are very time-consuming. However, using CRISPR-Cas counterselection screening workload could be reduced, making the method more efficient. A method using recombination and CRISP-Cas was used as an alternative phage engineering method. You can read about that here. As the CRISPR-Cas method for phage Lambda might not be applicable to X. fastidiosa, we prefer the in vitro methods.

Novel methods, such as yeast-based engineering of phages offers some advantages, like high-efficiency recombineering of DNA sequences, allowing for a significant reduction of screening time [6,7]. Moreover, it enables the simultaneous engineering of multiple non-contiguous loci in the phage genomes, as the native homologous recombination efficiency in Saccharomyces cerevisiae is higher than in bacteria [6]. This method also avoids the inclusion of unwanted selection markers in the genome. Most importantly though, yeast-based engineering of phage genomes does not require the genetic modification of the host bacterial strains with complicated engineering and selection mechanisms. In our case, this is very beneficial, since we could engineer X. fastidiosa specific bacteriophages without working with the pathogenic bacteria.

However, there is a major drawback to this technique, which is that it relies completely on the efficient transformability of the bacterial strains that will be used to reboot the modified phages. This is why the engineering of phages in a completely synthetic way is the final goal for phage engineers. By removing the requirement of genetically accessible cells, phage genome editing techniques become available for many bacterial species. Phages with pathogenic hosts could be genetically edited without the safety precautions that are normally attached to these procedures.

In 2012, Shin et al demonstrated that it is possible to produce bacteriophage T7 in absence of host cells, using a cell free Transcription and Translation (TXTL) system [8]. This publication was followed by others that show production of E. coli phages M2 (RNA phage), phi X174 (ssDNA phage) and phage T4 (dsDNA phage) [9][10]. This means that it is possible to produce active bacteriophages by adding the phage genome to a cell extract. So far only E. coli phages have been produced using this technique. Research on extending TXTL systems to other organisms is currently ongoing [11]. To our knowledge, using TXTL for the production of engineered phages has not been shown yet.

Yeast Engineering

We aim to engineer phage Lambda using a yeast-based approach, which has been proven to be highly efficient in previous research [6]. With this method, phage genomes are amplified in different fragments, which are assembled in S. cerevisiae together with a backbone sequence. This backbone sequence has an origin of replication for S. cerevisiae and for E. coli, turning the genome into a yeast-bacterium shuttle vector. Engineered sequences can be transformed into the same yeast cells and, once there, genome modification can take place via homologous recombination. The plasmid is finally transformed into bacteria, where the modified phages are rebooted.

  • How does this method work? arrow_downward

    DNA fragments of a size ranging from 5.5 – 11.5 kbp were PCR amplified, with 50 bp homologous flanking regions. A gap-repair recombination system recognizes the homologous ends and assembles the fragments together to form a full vector, in combination with a bacterial artificial chromosome cloned next to a yeast artificial chromosome (BAC-YAC) backbone. Therefore, this BAC-YAC sequence needs to have homology arms that are complementary to the ends of the end fragments of the phage genome. The resulting plasmid, a transformation-associated recombination plasmid (TAR), can be engineered by transforming yeast with the engineered DNA sequence of interest. This sequence must also include homology arms.

    The engineered TAR plasmid is extracted from yeast by enzymatically disrupting the cells. Finally, the modified phages can be rebooted by transforming the TAR plasmid into bacteria. There, the host is able to express all the phage proteins that will lead to the assembly of new, modified phage particles. Plaques can then be picked, used to make a phage stock and sequenced to verify that the insertion or deletion was correctly engineered. The use of S. cerevisiae permits an easy genetic manipulation of the phage genomes and, additionally, the simultaneous modification of several loci.



We diverged from the work of previous researchers that used the yeast-based platform to engineer T7 phages, due to the difference in the replication cycles of the T7 and phage Lambda [6,7]. In this project, we decided to include BsaI restriction sites in the TAR plasmid exactly between the phage genome and the backbone sequence. We also designed two novel approaches to reduce the number of fusion proteins that are displayed on the phage Lambda capsid.

  • Experimental design of the TAR plasmid arrow_downward

    The experiments performed with the T7 phage worked in the past due to the design-favorable replication cycle of the phage [6,7]. Interestingly, the T7 phage has a bidirectional replication cycle, which would exclude the BAC-YAC part of the TAR plasmid from being replicated. In the case of phage Lambda, however, this replication cycle is more complicated, as it forms long chains of phage genomes connected to each other. These long DNA chains, called concatemers, contain cos sites. Once the cos sites are recognized by a capsid protein, the concatemers are excised at the cosn site, leaving the phage Lambda genome ready for packaging in the capsid. Furthermore, the recognition of cos sites relies on their proximity in the concatemer sequence.

    When the concatemer is cut, the cos sites get split along both ends of the linear phage Lambda genome. In our strategy, once all fragments are PCR amplified from the linear genome and assembled to the BAC-YAC sequence in yeast, the cos sites are on both sides of the BAC-YAC backbone, too far apart to be recognized. For this reason, we introduced BsaI sites flanking the backbone. These BsaI sites can be used to digest the TAR plasmid after isolation. The overhangs then allow the genome to ligate into the naturally occurring circular form.

  • Experimental design of the fusion protein arrow_downward

    For the fourth pillar of Xylencer, we want the phages to spread to other infected plants. To achieve this the Xylencer phage is engineered to display chitin-binding domains on the capsid. First, we wanted to prove its feasibility in phage Lambda, using it as a proof of principle. To do that, the phage Lambda capsid decorator protein gpD was used. Fusion proteins have been reported to be displayed on the phage Lambda capsid when fused to this decorator protein [12-14]. This protein is present on the phage capsid in 405 copies and is encoded by the gpd gene. However, it is not possible to substitute all wild-type copies of the decorator protein for the fusion protein, as it will affect the assembly of the phage capsid [12]. In fact, phage lambda is known to tolerate only a fraction of engineered fusion proteins on its capsid. For that purpose, new strategies were used in this project to decrease the number of engineered capsid proteins.

    One of the strategies is based on ribosomal frameshifting sequences. The exact function of these sequences, present in bacteria and phages, is not completely understood. However, the purpose of a few of them has been recently described [15,16]. Ribosomal frameshifting sequences are used by bacteria to translate two different proteins with different functions from the same DNA sequence. As can be seen in figure 2, the sequence of the truncated protein is followed by a slippery sequence that contains a stop codon at the –1 reading frame. When the ribosome goes over this slippery sequence, it sometimes slips back one nucleotide, thus provoking a frameshift that unveils the stop codon and ends the translation. Right after the slippery sequence, a pseudo-knot contributes to the de-assembly of the ribosomes. The sequence of the C-terminal part of the untruncated protein is located downstream.

    Figure 2:The mechanism behind ribosomal frameshifting, leading to a co-expression of truncated and untruncated proteins from the same DNA sequence. Below, part BBa_K3286283 is shown, which codes for the ribosomal frameshifting fusion protein of Chitinase A1 to gpD

    In our project, we cloned this ribosomal frameshifting sequence, amplified from the genome of E. coli DH10B, to fuse the gpd gene from phage Lambda with the chitin-binding domain Chitinase A1. To assess the ratio between truncated and untruncated protein, we included a 6x-His tag at the N-terminus of gpD for western blotting purposes. We cloned this construct into a plasmid under expression of a strong constitutive promoter (BB_J23100) and transformed it into E. coli. As a control, a second plasmid was cloned harboring a fusion of gpd with Chitinase A1 using a linker in between, as used in a previous research article [12]. Full cell lysate was used to observe protein expression by Western blotting, using an anti-His tag antibody as the primary antibody.

    As a backup strategy we also worked on engineering phage Lambda to contain a second copy of the gpd gene, directly fused to the chitin-binding protein. With this second copy directly downstream of the first copy in the same operon we expect to have a similar level of expression, leading to a 50/50 mix of gpd and gpd fusion on the capsid. Testing this strategy also has value in extrapolating to X. fastidiosa, as the ribosomal frameshifting might not be as efficient in every bacterial species.

  • Why use BAC and YAC sequences? arrow_downward

    In this project, the YAC sequence of the TAR plasmid was used to replicate all the TAR plasmids in yeast once they were correctly assembled. The HIS3 gene was used as an auxotrophic marker to select for yeast colonies that had assembled a plasmid containing, at least, the BAC-YAC sequence.

    With a similar purpose, the BAC part of the TAR plasmid was included in the construct so that the plasmid could be replicated in bacteria. This would lead to the acquisition of a higher concentration of the TAR plasmid, required for its efficient digestion with BsaI and subsequent ligation.

  • Materials and methods arrow_downward

    The phage Lambda genome was bought from Sigma Aldrich. Sold as methylated genome from E. coli host strain W3110

    S. cerevisiae CEN.PK-1D strain was used for the assembly of the DNA fragments amplified from the phage Lambda genome. For the rebooting of phages Lambda , the E. coli K-12 DH10B strain was utilized.

    The BAC vector used for the construction of the TAR plasmid was pBelobac_LO01, a derivative from pBelobac11. The YAC plasmid pHLUM was used to amplify the CEN/ARS region plus the HIS3 gene, which were used as the origin of replication in yeast and the auxotrophic marker to select for positive yeast colonies, respectively.

    The genes encoding for the fusion proteins were cloned into the BB_J23100 part, derived from the plasmid J61002. The gpd gene was amplified from isolated phage Lambda genome. The ribosomal frameshifting sequence was amplified from the E. coli K-12 DH10B strain. The Chitinase A1 chitin-binding domain was ordered as a gene block from IDT.

    E. coli strain DH10B was used for cloning of constructs.

    Culture Conditions

    S. cerevisiae CEN.PK-1D cells were grown in YPD medium at 30 °C. After the TAR plasmid was transformed into yeast, S. cerevisiae was cultured in SC medium Leu+, Trp+, Ura+, His- at 30 °C. E. coli K-12 DH10B cells were cultured in LB medium at 37 °C.

    PCR amplification of the phage genome

    All DNA fragments used in the assembly of the TAR plasmid were amplified using Q5 Polymerase from New England Biolabs, supplemented with DMSO. Homology arms of approximately 50 bp were designed to be flanking every fragment by ordering overlapping primers.

    General protocols

    Protocols for yeast competent cells, yeast transformation, plasmid digestion and Gibson assembly can be found here.

    Yeast extraction of TAR plasmid

    For extraction of the TAR plasmid from S. cerevisiae CEN.PK-1D cells, MiniJET plasmid isolation kit was used following manufacturer instructions. After resuspension in resuspension buffer, 3 μL of 1000 U*mL-1 zymolase was added to the samples. The samples were then incubated at 37 °C for 30 minutes. The next steps were performed according to the protocol.

    Rebooting of phage genomes

    After assembly, the TAR plasmids, digested with BsaI and ligated overnight, were transformed into E. coli DH10B cells. For that, 20 μL of competent cells were mixed with 2 μL of the purified ligation mix. From this solution, 20 μL was used for electroporation in a 1 mm gap electroporation cuvette (Gene Pulser Cuvette from Bio-Rad) at 1800 V, 25 μF and 200 ω using a ECM 630 from BTX. Cells were recovered in 2 mL Eppendorf tubes 1 mL LB medium for 1.5 hour at 37 °C. Then, they were mixed with 3 mL LB soft agar (LB + 0.7% agar) warmed at 52 °C and 200 μL of overnight LB culture with E. coli DH10B, and finally poured onto a LB plate. These plates were incubated at 37 °C overnight.



As a first attempt at the yeast-based platform to engineer phage genomes, we tried to obtain wild-type phage Lambda. To do that, we first amplified the phage Lambda genome in six different fragments, at the same time introducing two silent point mutations to eliminate two BsaI restriction sites within the wild-type genome of the phage. The BAC-YAC sequence was previously obtained by cloning the YAC sequence into the corresponding BAC plasmid, in S. cerevisiae CEN.PK-1D cells. Once the BAC-YAC plasmid was constructed, it was PCR-amplified with the appropriate homology arms to the phage Lambda genome. The six phage Lambda genome fragments and the linearized BAC-YAC sequence were co-transformed into yeast and after overnight incubation individual colonies were screened. Different junctions between genome fragments were PCR amplified to verify the correct assembly of the TAR plasmids (Figure 3).

The plasmids of the positive candidates were extracted and digested with the BsaI enzyme. After this digestion and subsequent ligation of the genome, we transformed it into E. coli DH10B overnight incubation observed to visualize plaques. However, no plaques were obtained for any of the three successfully assembled TAR plasmids.

Figure 3: Screening of 15 different yeast colonies transformed with the TAR plasmid. the junctions between assembled fragments were amplified to check for the correct assembly of the TAR plasmid. Colored arrows show the positive candidates from which the plasmid was extracted, as they show positive results for all the amplifications. the amplification of the region around the junction between the first phage Lambda fragment and the BAC-YAC sequence never rendered the correct band for any of the › 35 screened yeast colonies

As we were not able to reboot wild-type phage Lambda , we did not attempt to engineer it with the fusion proteins. However, we tested the differential expression of the fusion protein (gpD + Chitinase A1) compared to the wild-type capsid decorator protein when introducing the ribosomal frameshifting sequence in the construct. We lysed the cells that had been transformed with either the plasmid containing the fusion gene with the linker or the plasmid containing the fusion gene with the ribosomal frameshifting sequence. The lysates were run on a SDS-PAGE gel, next to a negative control consisting of the lysate from cells transformed with an empty plasmid. The gel was then Western blotted and the resulting membrane was treated with an anti-6x His Tag primary antibody and a HRP conjugated secondary antibody (Figure 4). The image shows how the ribosomal frameshifting sequence worked as expected and produced both the truncated and the untruncated versions of the fusion protein, which means that phages could display both the wild-type and the fusion versions of the decorator protein on their capsid. Moreover, it is observable that the truncated protein is less expressed than the untruncated protein, indicating that the frequency of the ribosomal frameshift is below 50 %.

When revising the sequencing results, we discovered that there was an insertion of a nucleotide before the start of the chitinase a1. This implies that there was a change in the reading frame that generated a new stop codon close to the beginning of said protein. Consequently, the expected mass of the fusion protein with the linker ended up being 14.65 kDa and, the one of the truncated protein, 13.4 kDa. The mass of the untruncated protein, considering the early stop codon, is expected to be 16.8 kDa. The last band of the PageRulerTM Prestained Protein Ladder was not visualized correctly on the SDS-page gel.

Figure 4: Western blot of the lysates of the bacterial cells transformed with: 1) the plasmid containing the gpd + chitinase a1 fusion gene with a linker (BBa_K3286282), 2) the plasmid containing the gpd + chitinase a1 fusion gene with a ribosomal frameshifting (RFS) sequence (BBa_K3286283). The samples were run next to the PageRulerTM Prestained Protein Ladder

Protein model

We constructed a 3D model of the untruncated version of the fusion protein with the ribosomal frameshifting sequence (figure 5). In this model, it is observable that the gpD protein can correctly fold and is not sterically disturbed by the fused domain. Moreover, there is a clear spatial separation between the Chitinase A1 and the gpD protein, generated by the presence of the linking peptide. This model, together with previous studies [12], gives us an indication that Chitinase A1 could be correctly displayed on the phage capsid, giving the phage the ability to bind to Chitin.

Figure 5: 3D model of the gpD to Chitinase A1 domain fusion protein with a ribosomal frameshifting sequence between the fused proteins. In blue, the gpD protein, in yellow the linker derived from the ribosomal frameshifting sequence and in red the Chitinase A1 domain.


The TXTL system is based on E. coli cell free extract to which energy, in the form of ATP and maltose, and building blocks, in the form of amino acids, are added. The TXTL system allows for in vitro production of proteins.

  • What is the TXTL mixture? arrow_downward

    TXTL mix is made by adding energy carrying molecules, e.g. ATP, energy generating molecules, e.g. maltose and building blocks, e.g. amino acids. Cell free extract is made by lysing cells and using centrifugation to separate cell contents [17]. The additional compounds, which are shown in table 1, are a diverse mix, of which not all compounds might be strictly necessary.

    Contents TXTL mix
    Table 1: contents of the TXTL mixture used for phage production in literature
    Ingredients Concentration Shin 2012 [8] Concentration Rustad 2018 [9] Unit
    ATP 1.5 0.75 mM
    GTP 1.5 0.75 mM
    CTP 0.9 0.45 mM
    UTP 0.9 0.45 mM
    tRNA 0.2 0.2 mg/mL
    cAMP 0.75 0.75 mM
    Folinic acid 0.068 0.068 mM
    Mg-glutamate 6.5 5 mM
    K-glutamate 100 60 mM
    Amino acids 1.5 1.5 to 3 mM
    GamS 0.0033 0 mM
    Hepes pH8 50 50 mM
    Coenzyme A 0.26 0.26 mM
    NAD 0.33 0.33 mM
    Spermidine 1 1 mM
    3-PGA 30 30 mM
    DTT 2 0 mM
    Maltose or 0 10 to 15 mM
    Maltodextrin 0 20 to 40 mM
    PEG 8000 2 2 %
    dNTPs 0.5 ? mM

    The contents of the mixture used to produce phages in literature are known. The contents of the mixture that Arbor Biosciences distributes are considered proprietary and were not shared with us. It seems likely that the contents of the Arbor Biosciences match the contents of the mix used in literature, as Arbor cites the phage production papers on their website and is working together closely with the Noireaux Lab, which developed the TXTL mixture and published the papers on phage production using TXTL[8,9].

    Amino acids

    The amino acid mixture contains the amino acids necessary for protein production in cell free extract. Usually the amino acids are added in equal quantities, except for glutamate. Glutamate is added at a higher quantity as it is also partially used as an energy source and can be converted into other amino acids. Furthermore, it is added as potassium-glutamate and as magnesium-glutamate. In this purpose it acts as a carrier for these salts.

    Energy mixture

    The energy carrying molecules, ATP, GTP, CTP and UTP are crucial to the TXTL system. The ratio between these compounds differs between papers [8,9]. cAMP is closely related to energy production and fulfills a key signaling role. In the most recent paper on phage production by TXTL, maltose/maltodextrin is used. These compounds serve as energy source for the TXTL system. Coenzyme A and NAD are important in the energy metabolism of TXTL, but might be present in large enough quantities in the cell free extract [18]. Addition of tRNAs, which supplement the tRNAs that are already present in cell extract, and folinic acid, which functions as a precursor for formylmethione synthesis (required for prokaryotic translation initiation), have also been shown to be unnecessary and even reduce yields [18]. Spermidine is primarily associated with mRNA stabilization.

    Other ingredients

    GamS inhibits the E. coli RecBCD complex, which is involved in the degradation of bacteriophage DNA. GamS is encoded by phage Lambda to circumvent this degradation. Addition of GamS could help stabilize the phage DNA [19]. Hepes is used to buffer the pH at 8. DTT is used to stabilize proteins and prevents dimerization of thiolated DNA [20]. Magnesium and potassium are important for the stabilization of RNA [21]. PEG8000 is used to increase the density of the reaction mixture, which helps emulate the molecular crowding in a living cell. This can lead to lower expression of certain proteins, but significant increases in yield of phages were reported [9]. The addition of each of the DNA nucleotide triphosphates (dNTPs) is essential for genome replication of phages [8].



Design of the TXTL experiments is focused on reaching the perfect conditions for phage Lambda production in a TXTL reaction. This includes testing of different temperatures, modifications to the genome and addition of other compounds. Production of phages using TXTL has been shown before for phages T7 and T4 [9]. This serves as the inspiration to produce phage Lambda using TXTL.

  • Experimental design TXTL arrow_downward

    Phage Lambda differs from phage T7 in its genome replication, transcription and capsid assembly. Phage Lambda is more dependent on its host than phage T7, as it chooses between lysogeny and lysis in an extensive regulatory circuit. T7, on the other hand, is always lytic. The key in optimizing the TXTL reaction to be suitable for phage Lambda, is finding reaction conditions that can take care of these differences.

    TXTL reactions typically run at 30 °C. However, for phage Lambda lytic development it seems that 37 °C is the optimal temperature [22]. Phage Lambda has a regulatory circuit that makes the choice between lytic and lysogenic development. If the choice for lysogenic development is made, transcription of capsid proteins is repressed [23]. This means that lytic development is preferred for producing phages. Performing the reaction at 37 °C did, by itself, not yield active phages.

    As we hypothesized that phage Lambda might be going ’lysogenic’ in the TXTL we decided to try to stimulate Lambda to lytic reproduction by adding the cro protein to the TXTL mix. This protein is one of the key regulators of phage Lambda. By repressing Cl, which in turn represses the lytic cycle, it allows the phage to develop lytically [23]. The cro gene was amplified from the phage Lambda genome and cloned into the J23119-pBR322 vector under the control of the BB_J23105 promoter. This is Anderson promoter (relative strength 24%) was chosen because of trouble during cloning the cro gene under the BB_J23119 Anderson promoter (relative strength 100%). Cloning of the cro gene resulted in mutations of the promoter and ribosome binding site, indicating toxicity. A likely reason for toxicity is the DNA binding capability of the cro protein. To avoid the toxicity, we decided to put the cro gene under a weaker Anderson promoter and under the T7 promoter (figure 6). These were successful in cloning. As using the T7 promoter in TXTL requires addition of T7 polymerase, we chose to use the J23105 cro plasmid for TXTL. Due to time constraints, it was not possible to verify the production of the cro protein. The addition of the cro plasmid to the TXTL reaction was, by itself, not enough to produce phage Lambda. Further experiments are necessary to determine whether cro addition can be useful.

    Figure 6: A genetic map of the cro expression construct

    During phage reproduction concatemers (figure 7) are formed that consist of many copies of the phage Lambda genome. During assembly of the capsid the concatemers are recognized by cos sites. These cos sites are subsequently cut, and single copies of the phage Lambda packaged into the capsid. As the single linear copies of phage Lambda genome do not have this full cos site, they are not efficiently packaged into the phage capsid. If genome replication takes place in the TXTL mixture this is not a problem. However, if there is no genome replication, it is necessary to simulate the concatemers. The packaged phage Lambda genome has 12 base overhangs at the cos site. These overhangs are used to circularize the Lambda genome upon entry in the host [24]. In order to create phage Lambda concatemers in vitro, the phage genome was incubated with T4 ligase. When the resulting ligation product was used in a TXTL reaction, it did not yield active phage Lambda. Whether the problem lays with the concatemerization or with other requirements for phage Lambda is unclear.

    Figure 7: A schematic showing a phage lambda concatemer



We were able to produce T7 using TXTL, reaching a concentration of ~3*108 Plaque Forming Units (PFU)/mL. The production of T7 was verified by plaque assays, using the TXTL mixture and the genome as controls. The resulting plates can be seen in figures 8-10. These controls showed no plaques, confirming that the plaques were caused by phages produced in the TXTL mixture. Using a genome concentration of 0.04 nM we reached a phage-to-genome ratio of 0.5.

In literature a yield of 3.35*1011 PFU/mL was reached [9]. The difference in phage production can be explained mainly in the amount and quality of genome used for the reaction. A phage genome concentration of 0.25 nM is considered optimal. We used a lower genome concentration, due to low concentration of isolated phage genome [9]. The highest reported phage-to-genome ratio is 2.2. To reach this ratio, improvements can be made in the quality of the inserted genomes and other reaction conditions.

  • TXTL reaction conditions arrow_downward

    TXTL reactions were performed with a total reaction volume of 12 μL in 96 wells plates (Greiner v-bottom). Reactions were run overnight at 30 °C or 37 °C. P70A(2)-deGFP plasmid supplied by Arbor Biosciences was used as a positive control. Phage Lambda genome was acquired from Sigma-Aldrich (Methylated from E. coli W3110). T7 genome was isolated from T7 phage stock using the phenol-chloroform method. The TXTL reaction was supplemented with 3% W/W PEG-8000, 0.5 mM dNTPs and 3.3 μM GamS. PEG-8000 helps create the correct thermodynamic for phage capsid assembly by increasing molecular crowding [9]. dNTPs are added to allow for phage genome multiplication. GamS is a protein that protects linear DNA fragments from degradation [19].

Figure 8: Plaque assay of 10-6 diluted TXTL produced T7
Figure 9: Plaque assay of T7 genome
Figure 10: Plaque assay of the TXTL mixture

We also visualized the phages using Transmission Electron Microscopy (TEM). PEG8000 is used during phage production and this forms a coat over a TEM sample, making visualization difficult. To circumvent this, we produced phage T7 without PEG8000. The detected phage yield for TXTL T7 without PEG8000 was ~2*10 3 PFU/mL. This yield of phages is 105 lower than that of the reaction with PEG8000.

Using the TEM image the phage stock was verified to contain phage T7. The diameter of phage T7 is 55 nM, which is matched perfectly by the diameter measured on the TEM image. The phage clearly shows the expected sharp edges and the tail structure (figure 11). In the TXTL sample structures were found with the correct length and the correct shape. As can be seen in figure 12, they do not clearly show the tail, but it matches the other characteristics of phage T7. In the TXTL mixture some other structures were found, which can be seen in figure 13 and are similar in shape to phage T7. However, the diameter of the structures is significantly smaller at 42 nM. These structures might be unfinished T7 phages. Packaging of DNA into the capsid increases internal pressure and can increase the size, as well as increase the sharpness of the corners [25].

  • Phage detection arrow_downward
    Transmission Electron Microscopy (TEM)

    The electron microscope used for the TEM was a JEOL 1400plus. Pictures were made at different levels of magnification using a voltage of 120 kV and a 2k x 2k Matataki camera. Samples were applied to 200 and 400 meshes with carbon film. This film has been treated with ion glow discharge to make it hydrophilic. 10 μL of sample was applied to every mesh and incubated for 2 minutes. After incubation liquid was removed using a filter paper and the samples that contained PEG8000 were washed with two drops of distilled water. The samples were then stained using 2% phosphotungstic acid (pH 6,8). The liquid was once again removed using filter paper. Samples were then airdried and visualized in the electron microscope.

    Plaque assay

    Production of phages was quantified using plaque assays. Plaque assays were carried out using either the full plaque assay protocol or the spot assay protocol, for quick quantification. These protocols can be found on the protocols page.

Figure 11: Phage T7 from the phagestock
Figure 12: Phage T7 from the TXTL reaction

Production of phage Lambda was not successful. Our efforts on testing the correct condition using different temperature, genome concatemerization and changing regulation using the Cro protein did not result in plaques. To investigate what was happening within the TXTL reaction we visualized the TXTL mixture using TEM. The hypothesis was that the capsid might be produced but genome packaging was not taking place. We were not able to find complete phage Lambda, but we did find structures which appear to be unfinished capsids (figure 14). This suggests that some transcription and translation of phage Lambda took place within the TXTL reaction. These structures have a diameter of around 50 nM, while the mature phage Lambda capsid has a diameter of 60 nM. This difference can be explained by the expansion that takes place upon packaging of DNA [26].

Figure 13: Structures found in the T7 TXTL reaction
Figure 14: Phage lambda capsid like structures found in the TXTL reaction


Phage engineering using yeast-based genome engineering and TXTL phage production remains an interesting option for future research. Rebooting of yeast assembled genomes in E. coli and production of phage Lambda in TXTL were not yet successful. More research into the complicated reproduction and replication system of phage Lambda will be necessary to solve this. The successful assembly of phage Lambda genome and the production of phage T7 are promising steps for phage engineering.

As the engineered phage is essential to Xylencer, we also attempted phage engineering using alternative methods. Read about them here.

  • References arrow_downward
    1. Das, M., Bhowmick, T. S., Ahern, S. J., Young, R., & Gonzalez, C. F. (2015). Control of Pierce's disease by phage. PLoS One, 10(6), e0128902.
    2. Jones, J. B., Jackson, L. E., Balogh, B., Obradovic, A., Iriarte, F. B., & Momol, M. T. (2007). Bacteriophages for plant disease control. Annu. Rev. Phytopathol., 45, 245-262.
    3. Jeannette N Rapicavoli, Barbara Blanco-Ulate, Artur Muszynski, Rosa Figueroa- Balderas, Abraham Morales-Cruz, Parastoo Azadi, Justyna M Dobruchowska, Claudia Castro, Dario Cantu, and M Caroline Roper. Lipopolysaccharide o-antigen delays plant innate immune recognition of Xylella fastidiosa. Nature communications, 9(1):390, 2018.
    4. John M Wells, Boligala C Raju, Hsueh-Yun Hung, William G Weisburg, Linda Mandelco-Paul, and Don J Brenner. Xylella fastidiosa gen. nov., sp. nov: gram-negative, xylem-limited, fastidious plant bacteria related to xanthomonas spp. International Journal of Systematic and Evolutionary Microbiology, 37(2):136–143, 1
    5. Pires, D. P., Cleto, S., Sillankorva, S., Azeredo, J., Lu, T. K. (2016). Genetically Engineered Phages: a Review of Advances over the Last Decade. Microbiology and Molecular Biology Reviews, 80(3), 523-543.
    6. Ando, H., Lemire, S., Pires, D. P., Lu, T. K. (2015). Engineering Modular Viral Scaffolds for Targeted Bacterial Population Editing. Cell Systems, 1(3), 187-196.
    7. Lu, T. K. T., Koeris, M. S., Chevalier, B. S., Holder, J. W., McKenzie G. J., Brownell D. R. (2013). Recombinant phage and methods. Patent no. US20130122549 A1.
    8. Shin, J., Jardine, P., & Noireaux, V. (2012). Genome replication, synthesis, and assembly of the bacteriophage T7 in a single cell-free reaction. ACS synthetic biology, 1(9), 408-413.
    9. Rustad, M., Eastlund, A., Jardine, P., & Noireaux, V. (2018). Cell-free TXTL synthesis of infectious bacteriophage T4 in a single test tube reaction. Synthetic Biology, 3(1), ysy002.
    10. Garamella, J., Marshall, R., Rustad, M., & Noireaux, V. (2016). The all E. coli TX-TL Toolbox 2.0: A platform for cell-free synthetic biology. ACS synthetic biology, 5(4), 344-355.
    11. Moore, S. J., MacDonald, J. T., Wienecke, S., Ishwarbhai, A., Tsipa, A., Aw, R., ... & Polizzi, K. M. (2018). Rapid acquisition and model-based analysis of cell-free transcription–translation reactions from nonmodel bacteria. Proceedings of the National Academy of Sciences, 115(19), E4340-E4349.
    12. Pavoni, E., Vaccaro, P., D’Alessio, V., De Santis, R., Minenkova, O. (2013). Simultaneous display of two large proteins on the head and tail of bacteriophage Lambda. BMC Biotechnology, 13, 79.
    13. Gupta, A., Onda, M., Pastan, I., Adhya, S., Chaudhary, V. K. (2003). High-density functional display of proteins on bacteriophage Lambda. Journal of Molecular Biology, 334(2), 241-254.
    14. Hayes, S. (2019). Bacterial Virus Lambda Gpd-Fusions to Cathelicidins, α- and β-Defensins, and Disease-Specific Epitopes Evaluated for Antimicrobial Toxicity and Ability to Support Phage Display. Viruses, 11, 869.
    15. Meydan, S., Klepacki, D., Karthikeyan, S., Margus, T., Thomas, P., Jones, J. E., Khan, Y., … & Mankin, A. S. (2018). Programmed ribosomal frameshifting generates a copper transporter and a copper chaperone from the same gene. Molecular Cell, 65(2), 207-219.
    16. Napthine, S., Ling, R., Finch, L. K., Jones, J. D., Bell, S., Brierley, I., Firth, A. E. (2017). Protein-directed ribosomal frameshifting temporally regulates gene expression. Nature Communications, 8, 15582.
    17. Sun, Z. Z., Hayes, C. A., Shin, J., Caschera, F., Murray, R. M., & Noireaux, V. (2013). Protocols for implementing an Escherichia coli based TX-TL cell-free expression system for synthetic biology. JoVE (Journal of Visualized Experiments), (79), e50762.
    18. Cai, Q., Hanson, J. A., Steiner, A. R., Tran, C., Masikat, M. R., Chen, R., ... & Yin, G. (2015). A simplified and robust protocol for immunoglobulin expression in Escherichia coli cell‐free protein synthesis systems. Biotechnology progress, 31(3), 823-831.
    19. Wilkinson, M., Troman, L., Ismah, W. A. W. N., Chaban, Y., Avison, M. B., Dillingham, M. S., & Wigley, D. B. (2016). Structural basis for the inhibition of RecBCD by Gam and its synergistic antibacterial effect with quinolones. Elife, 5, e22963.
    20. Fjelstrup S., Andersen, M., Thomsen, J., Wang, J., Stougaard, M., Pedersen, F., & Knudsen, B. (2017). The Effects of Dithiothreitol on DNA. Sensors, 17(6), 1201.
    21. Pyle, A. (2002). Metal ions in the structure and function of RNA. JBIC Journal of Biological Inorganic Chemistry, 7(7-8), 679-690.
    22. Groman, N. B., & Suzuki, G. (1962). Temperature and Lambda phage reproduction. Journal of bacteriology, 84(3), 431-437.
    23. Dodd, I. B., Shearwin, K. E., & Egan, J. B. (2005). Revisited gene regulation in bacteriophage λ. Current opinion in genetics & development, 15(2), 145-152.
    24. Catalano, C. E., Cue, D., & Feiss, M. (1995). Virus DNA packaging: the strategy used by phage λ. Molecular microbiology, 16(6), 1075-1086.
    25. Guo, F., Liu, Z., Fang, P. A., Zhang, Q., Wright, E. T., Wu, W., ... & Chiu, W. (2014). Capsid expansion mechanism of bacteriophage T7 revealed by multistate atomic models derived from cryo-EM reconstructions. Proceedings of the National Academy of Sciences, 111(43), E4606-E4614.
    26. Dokland, T., & Murialdo, H. (1993). Structural transitions during maturation of bacteriophage Lambda capsids. Journal of molecular biology, 233(4), 682-694.