5 Collection and assembly of T4 genome in S. cerevisiae
6 Phage T4 rebooting
7 References
Engineering Success
Spiral Model of Software Development by Barry Boehm
1 Our Process
(January-February 2020)
(1) Determine goals, choices and constraints:
We hope to design a universal method that allows the main body of the phage to be adapted to assemble different tail fibers, thereby changing its host range, so as to achieve quantitative control of the intestinal flora, detection of food-borne pathogens, and treatment of intracellular parasitic infections.
(2) Risk analysis:
How to achieve what is expected in the design: the assembly process of the phage body and the different tail fibers. The process needs to meet the following requirements: easy to operate and capable of producing a large number of hybrid phages with heterologous body and tail fibers.
(3) Build a prototype:
We hope to design a universal platform to assemble different tail fibers with a shared main body of another phage. For example, through our platform, we can hybridize the main body of phage A with tail fibers of phage B, or phage C, etc. The host range is thereby modified. In order to produce a large number of hybrid phages with heterologous main body and tail fibers, we designed a method for assembling the body-tail fiber hybrid phages in vitro. The method was designed to use artificial main body of T4 phage and tail fibers of other phages. Tail fiber ligase was proved to simulate the intracellular assembly process of phage. In our expectation, we will invent a detection instrument which contains several tail fibers. Whenever mixing the main body of T4 phage, the tail fibers and tail fiber ligase in this instrument, the hybrid phages could be produced. When spraying those hybrid phages onto the sample, the pathogenic bacteria would be infected.
Then we started to design the method to obtain main body of T4 phage, tail fibers and tail fiber ligase. Tail fibers and the ligase could be produced through conventional genetic engineering. According to one of our senior sisters, the tail fibers could be separated from the phage body by rude shock (the data have not been published), so that the main bodies of T4 phages could be harvested by this way.
Figure 1. Scenario diagram of the application of our first project.
We package main body of phage T4, tail filament ligases and several tail filaments as capsules and put them into a toolkit. The main instrument of the detection of bacteria is a bacterial screening gun, in whose handle the capsules are slipped and the phages with various tail filaments would be generated. When spraying the phages onto surface of the samples, the existing bacteria would be infected and some phenomenon would be observed.
(4) Test and verify the prototype:
Through further investigation, we learned that the assembly process of phage is not just a simple enzymatic reaction. It needs the support of the intracellular environment and various complex metabolic pathways, so it is difficult to directly assemble in vitro; in addition, the phage body and tail fibers used as raw materials are difficult to store stably for a long time, and it is also difficult to activate and assemble them immediately.
(5) The next stage of the plan:
We decided to change the phage assembly process from in vitro to intracellular, and no longer use phage main body and tail fibers as hybrid materials, but directly carry out the allogeneic combination of gene sequences, and then use the natural way of phage replication to obtain hybrid phage.
(February-March 2020)
(6) Determine goals, choices and constraints:
We hope to design a production process that can replace specific genes or gene clusters of bacteriophages from phages with allogenic bacteriophage genes for intracellular expression to produce a hybrid phage with new traits and multifunctional properties.
(7) Risk analysis:
Adding exogenous genes to phages can lead to changes in their genomes, causing biosafety problems. We need to devise a scheme in which artificial hybrid phages can survive and function, but cannot reproduce in the natural environment.
(8) Prototyping:
We have designed a standard hybrid phage production process, taking the modification of the T4 phage tail fiber as an example. This process requires first removing the tail fiber gene from the T4 phage genome, and then using a plasmid vector to carry the heterologous tail fiber gene. The defective T4 genome and the plasmid vector are co-introduced into the same E. coli, and the expression and assembly are carried out in the cells respectively to produce hybrid phage with defective genome. The phage is the final product. It has a heterologous tail, and a defect in the genome preventing it from reproducing in nature.
Through literature research, we found some problems: First, the specificity of the phage to the host is related to many factors. In addition to the receptor binding protein on the tail fiber, the structure of the substrate, the short tail fiber, and the lyase will all affect The choice of phage on the host. Secondly, the capsid proteins of phages may be packaged into exogenous plasmids during genomic packaging, which means that the gene defects of hybrid phages may be remedied, causing potential biosafety risks.
(10) Next stage of the plan:
We decided to change our thinking and change our goal to reactivate the phage inside the cell using the genome imported into the cell. This method can not only serve as the technical basis for phage genome replacement, but also provide a possible way for phage genetic modification and foreign gene addition.
(March-August 2020)
(11) Determine goals, choices and constraints:
We hope to design a method that would divide a phage genome into multiple segments and build a phage library based on those segments. When modifying genes, we can manipulate specific segments to reduce the impact on other genes. Finally, we only need to connect and express these segments in order to produce modified phages.
(12) Risk analysis:
The combination and expression of the segments need to be carried out in the cell, so we need a technology to ensure that the plasmid transporting these segments can be introduced into the same cell simultaneously and equally, and can be stably present and expressed in the cell.
(13) Prototyping:
We designed a set of intracellular rebooting process of phage genome. Take the rebooting of T4 phage as an example. First, we performed segmental PCR processing on the genome of T4 phage, and obtained 26 small segments with a sequence length of 7-8kbp. According to their order in the genome, they were divided into four groups with similar numbers. Each group was connected by Overlapping Extension PCR to produce four large segments of about 42 kbp in length. Next, we use the artificial yeast chromosomes (YAC) with different antibiotic resistance genes as the vectors to transform four large fragments into the same E. coli. Then YACs were screened with four antibiotics to obtain the E. coli with all four YACs. After that, we introduced the CRISPR system into the E. coli to achieve linearization and recombination of the four segments. Finally, the complete T4 phage genome is expressed in E. coli, and the rebooting of the phage can be achieved.
Besides using four antibiotics to screen E. coli transformant, we found another way to achieve this goal. Kanamycin resistance of bacteria comes from the interaction of T7rnap gene and neo gene. T7rnap gene can express T7 RNA polymerase, which in turn can activate neo gene expression. This process provides the bacteria with kanamycin resistance, and these bacteria can survive on the medium containing kanamycin.
So, we plan to divide T7rnap gene into two segments: core fragment and σ fragment. They are regulated by the same promoter, so that they have similar expression levels and combine to produce active T7 RNA polymerase. In the same way, neo gene can be divided into two parts: neo-α and neo-β. Only when the two fragments exist at the same time, can neo gene function fully.
Therefore, we can use four genes to mediate kanamycin resistance and construct a four-plasmid-coexistence system. Only when the four plasmids exist at the same time and have the appropriate expression level, can the bacteria grow on the medium containing kanamycin to achieve the purpose of screening.
(14) Test and verify the prototype:
During the evaluation of this process, we noticed two problems: Firstly, technically, the process relies on a complex multi-plasmid coexistence system, which requires four plasmids to enter the cell at the same time and have similar copies in the cell. This is difficult to control; secondly, we learned through investigation that it is achievable to introduce a huge plasmid (about 170kbp, the size of the genome of the T4 phage) into the cell. Therefore, we can avoid the use of multi-plasmid coexistence system by connecting fragments in advance, so as to simplify the process.
(15) Next stage of the plan:
We believe that it is necessary to use large plasmid transduction instead of multi-plasmid transduction, namely firstly connect the four plasmids constructed outside of E. coli with the CRISPR system to obtain a YAC plasmid carrying the complete phage genome, and then introduce it E. coli.
(August-October 2020)
(16) Determine goals, choices and constraints:
We hope to improve the original process, abandon the complex multi-plasmid system, and use a more convenient method to completely introduce the phage genome into the host cell. This method needs to simplify the operator's quantitative control of each parameter, and make the detection of each step more accurate, macroscopic and visualized as much as possible.
(17) Risk analysis:
The ligation of fragments requires the help of the CRISPR system, which has a higher success rate in the intracellular environment. Therefore, we need to use new methods to integrate the four plasmids into the same cell and avoid the use of conventional transformation methods.
(18) Prototyping:
Based on the phage genome, the use of segmented PCR to construct four plasmids is the same as before, and the constructed plasmids will be introduced into yeast cells respectively. Then, using the mating induced fusion phenomenon between the different genders of yeast, the two yeasts were first fused to form diploid yeasts to achieve the integration of the two plasmids, and then the CRISPR system was introduced into them to connect to form two fragments of about 84kbp. Next, the fusion between the two diploid yeasts is carried out, and the complete phage genome is formed by the same method. Finally, we only need to extract this complete large plasmid and introduce it into E. coli to restart the phage.
To decrease the potential difficulties in direct manipulation of the huge coliphage T4 genome, the 168.9 kbp T4 genome is divided into four parts (~40 kbp each). Every part is then divided into six or seven fragments (7~8 kbp each), which can be amplified conveniently via PCR and using the natural T4 genome as the template.
The primers of two contiguous fragments are designed with overlap sequences to mediate the homologous combination between each couple of contiguous fragments. Each single PCR product is about 7~8 kbp. Six or seven contiguous PCR products containing homologous arms, along with the corresponding linearized yeast vector, are co-introduced into Saccharomyces cerevisiae via electroporation. The primers used for amplification of phage genomic fragments are listed in Table 1.
Table 1. All primers used in PCR amplify wide-type T4 genome. 1~4 refer to the four parts of the T4 genome. Fragment A~F, G~L, M~Q, R~W, are continuous gene fragments of part 1, 2, 3, 4, respectively. The temperatures of every fragments are calculate using our mathematic model.
Though it is technically feasible to amplify larger phage genomic fragments, we still chose to amplify the 7~8 kbp fragments because of the accuracy of PCR amplification. On the one hand, the efficiency of transformation-associated recombination (TAR) in yeast is limited by the number of linearized DNA fragments. It was regarded that if there are too many fragments used for TAR, the correct transformants with all assembled fragments would be precious few. On the other hand, if the gene fragments are too small in size, the efficiency of homologous combination could be decrease. As a result, we choose 7~8 kbp as the size of every piece of PCR products.
Agarose gel electrophoresis is applied to evaluate the results of PCRs. The unexpected results could be (I) no specific DNA bands; (II) bands with unexpected relative molecular weight; (III) poor yield of PCR products, which are possibly due to the mistakes in designing the upstream and downstream primers, and the inappropriate temperatures of PCRs. We would go over the primers of the unexpected products, while resetting the PCR parameters. If the above three unexpected results always occur in a specific PCR, the WT phage T4 genome should be sequenced, because the template could be mutated or changed. If (IV) all the DNA bands were dim or even no products were obtained, the reason might be the inactivation of DNA polymerase and the T4 genome, which would be then rebought and reextracted.
3 Preparation of plasmids and vectors
Firstly, we design two plasmids for the switch of mating type of Saccharomyces cerevisiae. The pTrp-gRNAα (Figure 2a) and pTrp-gRNAa (Figure 2b) express guide RNA targeting Yα and Ya sequences, respectively. The Ya and Yα sequences used as the repairing fragments are amplified via PCR. Briefly, when co-transforming the MATα yeast cells with pTrp-gRNAα and Ya fragment, and whenever the Cas9 protein is expressed in yeast, the MATα cells will be converted to MATa cells.
Figure 2a. Gene map of pTrp-gRNAα. The left map is pTrp-gRNAα, in which the gRNA targets sequence Yα.
Figure 2b. Gene map of pTrp-gRNAa. The right map is pTrp-gRNAa, in which the gRNA targets sequence Ya.
The Cas9 expression plasmid pMet-Cas9 (Figure 3a) and gRNA expression plasmid pTrp-gRNA-S1S2 (Figure 3b) are obtained from Prof. Zhongjun Qin at the Center for Excellence in Molecular Plant Science, Chinese Academy of Sciences. They are electroporated into S. cerevisiae and amplified apart. The harvested yeasts are broken and the plasmids are separated. To increase the universality and reduce the consumption of human and material resources, the Cas9 expression plasmid are expected existing in yeast cells stably. The gRNA is controlled by lac promotor, which is expected to trigger the dsDNA scissors at the specific time we want. Our plasmid that express Cas9 protein and gRNA was designed as Figure 3a.
Figure 3a. Gene map of pMet-Cas9. This plasmid is expected to exist stably in yeast cells.
Figure 3b. Gene map of pTrp-gRNA-S1S2. This plasmid mediates the separation of the genome fragments and linearized vectors.
We use linearized pRS42X series vectors with different marker genes, pRS422 (ADE2), pRS423 (HIS3), pRS426 (URA3) and pRS427 (LYS2). Given that our linearized plasmids are amplified via PCR, to eliminate the contamination of cyclical plasmids, we used SmaI digested vectors as templates for PCR amplification. We will use Agarose gel electrophoresis to evaluate the results of PCR. The potential unexpected results and the troubleshooting processes are similar to those mentioned above. (Figure 4)
Figure 4. The seven linearized vectors.
The grey parts on both the ends is the homologous arms which homologous with sequences of the T4 genomic fragments. The selection yeast marker genes are HIS3, URA3, and ADE2, respectively.
4 Switching the Mating type of S. cerevisiae
Figure 5 shows the schematic diagram for switching the mating type of yeast. The yeast mating type determined sequence, Yα and Ya sequence, are PCR amplified, and the templates are de novo synthesized. The Cas9 expression plasmid pMet-Cas9, pTrp-gRNAα plasmid for the expression of gRNA targeting the MATα locus, and the Ya fragment homologous with Yα sequence, are co-electroporated into S. cerevisiae VL6-48 (MATα) to convert it from MATα to MATa. The transformants are selected on SC-MET-TRP (omitting methionine and tryptophan) medium for 2 days. The Cas9 expression plasmid contains another gRNA targeting the plasmid pTrp-gRNAα, which is controlled by GAL1 promoter, so that the gRNA expression plasmid pTrp-gRNAα could be linearized and degraded in the yeast cells on SC-MET medium. As a result, the auxotrophic marker gene TRP1 is lost by the yeast cells, and the gene could be reused to select other introduced plasmids. The Cas9 expression plasmid is expected to exist in the yeast stably, for Cas9 protein would be employed to generate other double-strain breaks in further steps.
Figure 5. Diagrammatic sketch of switching the mating type of S. cerevisiae.
The green linearized sequence Yα and the blue linearized sequence Ya determine the mating type of S. cerevisiae as MATα and MATa, respectively. The red arrows represent the double-strain breaks caused by CRISPR/Cas9. The break at Yα would result in homologous recombination between Yα and Ya, so that the mating type of S. cerevisiae could be switched. The break at gRNA expression plasmids can lead to degradation.
Simultaneously, we build the MATα yeast cells harboring the same Cas9 expression plasmid by electroporating. The transformants are selected on SC-TRP (omitting tryptophan) medium for 2 days, too.
The switched mating type is tested by crossing the transformants with tester strains 17α and 14a, both carrying the HIS3 auxotrophic marker. Lawns of tester strains are prepared by spreading 500 ml overnight YPD culture of tester strains on YPD agar plates and incubating at 30 degree Celsius for 2 days. The lawns and transformant colony plates are replica plated onto a fresh YPD plate at the same time. A second replica plating onto synthetic dextrose medium is performed after incubating at 30 degree Celsius overnight, and the plates are incubated at 30 degree Celsius to reflect mating response and scored at the next day. Haploid strains and MATα/α or MATa/a diploid strains should form a patch on just one of the two tester strains if the tester strain is of the opposite mating-type. MATa/a strains are nonmaters and will not form a prototrophic patch with either tester strain.
The unexpected results could be low efficiency of mating-type switching, limited by the efficiency of co-electroporation and homologous recombination. To our knowledge, the strategies to increase the harvest of correct sex-converted yeast cells are culturing and manipulating more yeast cells and change the reagents if necessary. Of course, if the harvest cells are still few, we will redesign our gRNA expression plasmids, because we regard the low efficiency of generating double strand breaks as the main reason. Also, we will add more Ya fragment into the electroporation cuvette to increase the content of the Ya sequence in transformants. The concentration of other pre-introduced plasmids will be reset to find out the most suitable electroporation conditions.
5 Collection and assembly of T4 genome in S. cerevisiae
Figure 6 shows the schematic diagram of the following manipulation in yeasts.
Figure 6. Assembly of the phage T4 genome in yeast.
The haploid yeasts are arranged at the top of this diagram. They are transformed with group 1 to 4, respectively. The red arrows represent double strand breaks caused by CRISPR/Cas9 system. The vermeil, yellow, violet, clear green and sapphire squares represent homologous sequences mediated assembly, and the blue strips represent PCR products of T4 genome. The gene fragments, plasmids and vectors emerging in the yeast cells are all the nucleic acid elements associated with our project in corresponding cells.
As mentioned above, we obtain all plasmids and vectors used in our project, and the T4 genome is also obtained as 26 7~9 kbp fragments. The fragments and linearized YAC backbones are PCR amplified separately. Then every group, including a linearized YAC backbone and corresponding fragments of T4 genome, are co-electroporated into yeast cells and circulated to the YAC-phage plasmid. Specifically, group 1 and group 3 are introduced into MATα yeast cells with the linearized YAC backbones with auxotrophic marker HIS3; while group 2 and 4 are introduced into MATa strains with the backbones with auxotrophic marker URA3. The conditions for co-electroporation are described previously (Ando et al., 2015). Then we select the four transformants on corresponding minimal medium, SC-MET-HIS (omitting methionine and histidine) and SC-MET-URA (omitting methionine and uracil), respectively. After two days, the single colonies are then picked out and inoculated on another fresh synthetic medium, simultaneously colony PCR is operated to check whether the linearized fragments are assembled successfully via homologous recombination. Colony PCRs are carried out for all single colonies until the discovery of transformants harbored the correctly assembled cyclical YAC-phage plasmid. The primers utilized in colony PCR are listed in Table 1. The four obtained yeast strains are named yMATa-1His-Cas9Met, yMATα-2Ura-Cas9Met, yMATa-3His-Cas9Met, and yMATα-4Ura-Cas9Met.
We use mating-mediated fusion and Cas9-facilitated homologous recombination assembly (Zhou et al., 2016) for the construction of a huge plasmid containing the whole phage T4 genome. First we cross yMATa-1His-Cas9Met and yMATα-2Ura-Cas9Met mediated by yeast mating. The diploid strains are selected on the synthetic medium SC-MET-HIS-URA (omitting methionine, histidine and uracil) for 2 days. The resulting diploids were then transformed with gRNA expression plasmid pTrp-gRNA-S1S2, and another linearized YAC backbone with auxotrophic marker ADE2, followed by culturing and selecting the transformants on SC-MET-TRP-ADE (omitting methionine, tryptophan and adenine) for 2 days. The harvest single colonies are picked out and colony PCR is carried out once again to check whether the two huge fragments have been released from the corresponding plasmids and assembled with the later-introduced linearized YAC vector. The selected strains are then inoculated on minimal medium SC-MET-ADE (omitting methionine and adenine). Simultaneously, the gRNA in pMet-Cas9 is induced by galactose eliminate plasmid gTrp-gRNA-S1S2, so that the diploid yeast only contains YAC-phage plasmid pADE-phage12 and Cas9 expression plasmid pMET-Cas9. The harvest diploid yeast is named yMATa/MATα-12Ade-Cas9Met, which would be then mating switched to MATα/MATα as described above.
Table 2. Primers of colony PCR.
Colony PCR would be used to detect whether the fragments have been assembled successfully. Fragment II and III, fragment IV and I should be amplified and detected after the fusing of yMATα/MATα-12Ade-Cas9Met and yMATa/MATa-34Lys-Cas9Met, and the other fragments should be amplified and detected after the first fusions.
Simultaneously, yMATa-3His-Cas9Met and yMATα-4Ura-Cas9Met are fused and manipulated through a similar strategy. Besides, the later-introduced linearized YAC vector is marked with auxotrophic marker LYS2. As a result, after using the corresponding medium to culture the yeast cells, the harvest diploids should be yMATa/MATα-34Lys-Cas9Met, which would be then converted to MATa/MATa.
The same strategy was utilized to convert MATa/MATα to MATα/MATα or MATa/MATa and lost the gRNA expression plasmids (Xie et al., 2018). The obtained yeast strains are named yMATα/MATα-12Ade-Cas9Met and yMATa/MATa-34Lys-Cas9Met, respectively. Then we cross yMATα/MATα-12Ade-Cas9Met and yMATa/MATa-34Lys-Cas9Met through the same methodology as described above.
To be on the safe side, we keep yeasts which are able to survive on the minimal medium and contain the expected plasmids proliferation on the corresponding plates and transplanting regularly. These operations support the possibility to check out where the mistakes occur and prevent making a fresh start when meeting some other unexpected results. The proliferated yeast cells could amplify more needed plasmids so that the costs of operation mistakes could be decreased.
The unexpected results could be the off-targeting phenomenon of the CRISPR/Cas9 editing system. We should ensure the reconstructed T4 genome would not be destroyed by Cas9 protein, so that the potential off-target phenomenon will be checked out and redesign the gRNA if necessary. Restriction enzyme mapping and agarose gel electrophoresis are used to detect if the extracted YAC-T4 plasmid is complete or not.
The reconstructed T4 genome has to escape from the designed CRISPR/Cas9. A restriction endonuclease map would be carried out with RsrII, BssHII, FspAI, and SgrDI. The four restriction enzymes are the unique cutters of both the T4 genome and the linearized YAC vector. If the off-target or double-strain breaks occur, there would be more than 4 strips when the AGE is carried out. As a result, the CRISPR/Cas9 system should be redesigned.
The gel recovery would be left out, for the restriction enzymes would be cut the giant genome into several pieces. On the one hand, gel recovery might cause huge wastage and break of giant-sized plasmids. On the other hand, the concertation of YAC-T4 plasmids could be decrease, which might defeat the followed electroporation.
Table 3. The yeast strains used in our project.
The parent strain is S. cerevisiae VL6-48 (MATα, trp1-Δ1, ura3-Δ1, ade2-101, his3-Δ200, lys2, met14). All the other strains are obtained from our project.
Table 4. Plasmids used in our project.
6 Phage T4 rebooting
After assembling the whole T4 genome in S. cerevisiae, we beak the yeast cells and obtain the YAC-T4 plasmid. The huge plasmid is then purified via agarose gel electrophoresis. We expect with the bands of the correct size (about 175 kbp, including a 168.9 kbp genome and a 6.3 kbp vector). There might be several stripes of diverse size for the yeast cells also contain other plasmids, so that cutting gel was necessary.
The obtained YAC-T4 plasmids are then introduced into competent E. coli cells via electroporation. To avoid the potential results, such as low transformation efficiency or even no transformants generated, preliminary experiments are designed to determine parameters roughly and parallel experiments are operated to obtain enough transformants. Mathematical modeling is also built to predict the potential range of the parameters. For more detailed description of our model, please click here.
The E. coli transformants are incubated in LB liquid medium for 3~4 hours at 37 degrees Celsius, followed by adding drops of chloroform into the supernatants to assist lyse. The centrifugal supernatants are then mixed with an overnight culture of E. coli, and the phages are amplified overnight at 37 degrees Celsius.
7 References
[1]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.
[2]Xie, Z.-X., Mitchell, L. A., Liu, H.-M., Li, B.-Z., Liu, D., Agmon, N., Wu, Y., Li, X., Zhou, X., Li, B., Xiao, W.-H., Ding, M.-Z., Wang, Y., Yuan, Y.-J., & Boeke, J. D. (2018). Rapid and Efficient CRISPR/Cas9-Based Mating-Type Switching of Saccharomyces cerevisiae. G3: Genes|Genomes|Genetics, 8(1), 173–183.
[3]Zhou, J., Wu, R., Xue, X., & Qin, Z. (2016). CasHRA (Cas9-facilitated Homologous Recombination Assembly) method of constructing megabase-sized DNA. Nucleic Acids Research, 44(14), e124–e124.
