RUMIVEC

The main goal for RumiVec was to develop a composite part that could allow for production of RNA bacteriophages in E. coli cells in order to validate viral nucleic acid biosensors without using higher risk pathogens. For this, the sequence coding for the phage genome must contain a modular region to allow for insertion of any control sequence to be used for testing. This system will be used in future steps to validate the Rumino biosensor by inserting the Influenza target sequence into RumiVec and then using its products for our test samples.

CYCLE 1

Design

The first step was deciding what type of RNA phage we'll produce. The most well-studied single-stranded RNA phages are from the Leviviridae family, more specifically the MS2 and QB species [1]. Given that the MS2 phage has a smaller, simpler genome than QB and can also produce higher yields when produced by bacteria [1], our group decided to assemble a genetic device for the production of MS2 phages.

In this initial iteration, the idea was to build a construct containing the genome of the wild-type MS2 phage, followed by a cassette containing a reporter which could be swapped out using Type IIS restriction enzymes. Replacement of the reporter with an insert sequence would allow for easy screening of colonies that were successfully transformed with the construct containing the insert. This construct would require two promoters: an inducible promoter for expression of an RNA molecule corresponding to the MS2 genome and the cassette; and a different promoter for expression of only the reporter. This way, the reporter protein can be produced without expression of viral proteins, which would result in death of the host cell.

We also realized that the wild-type MS2 genome is not codon-optimized for E. coli cells. However, some of the genes in the genome overlap, which means that codon optimizing might disrupt some proteins, specifically the Lysis and Replicase proteins. Due to this, we decided to test two constructs in parallel: a codon-optimized construct and a wild-type construct.

For the reporter, the team decided to use the LacZα peptide which would allow for blue-white screening to find bacterial colonies successfully transformed with either the construct with the reporter (blue colonies) or an insert sequence (white colonies) using X-Gal plates. More specifically, we combined a constitutive promoter (BBa_J23100) with a strong RBS + CDS (BBa_K329005) flanked by SapI restriction sites to generate our cassette.

To ensure that the construct could be used as a standard part, the sequences were screened for any restriction sites used for Type IIS or BioBricks assembly and removed to meet compatibility requirements.

After the construct was ordered from Twist Bioscience within a pET Kan vector, we analyzed it using the NEBridge Golden Gate Assembly tool to verify the sequence for use in Type IIS assembly. This showed us an extra SapI site in the vector backbone that was overlooked, as well as a wrong orientation on the cut sites flanking the reporter.

To fix this error, we intended on using NEB HiFi assembly to correct the orientation and swap the SapI sites for BsaI sites. For this, we designed two pairs of primers using Primer3Plus to correct the sequences. The first pair was meant to amplify the regions outside of the reporter, while the second pair was meant to amplify the reporter. The reporter primers contained overhangs that replaced the SapI sites and also added compatible extensions between it and the backbone.

Backbone Forward Primer

Backbone Reverse Primer

Reporter Forward Primer

Reporter Reverse Primer

For the insert sequence, a short 35bp sequence termed Shuffle Sequence 1 (SS1) was designed with a GC% similar to that of HPAI. Spacers were then added to flank the sequence to make it long enough for PCR and detection with Agarose Gel Electrophoresis. BsaI sites were also added to the ends of this sequence for Type IIS assembly with the construct. Similarly, primers were also designed to amplify this sequence using the Benchling primer design tool.

Build

The constructs were assembled using Benchling and then ordered through Twist Bioscience. The idea was to obtain the parts necessary for HiFi assembly to build the correct construct using the sequences we had.

Figure 1. Proposed scheme for assembly of fixed construct containing BsaI sites instead of SapI, and oriented to be inside of the reporter. PCR products would be assembled using NEB HiFi Assembly.

Test

During this first cycle, only the PCR of the reporter seemed to work. Amplification of the vector backbone seemed to result in streaking, and SS1 didn't generate a product. This roadblock prevented us from achieving our goals, and at this point we decided to close this cycle and design new constructs.

Figure 2. Agarose Gel Electrophoresis (1%) of different PCR products. Only a faint band was observed for amplification of the reporter (LacZα), while SS1 had none to low yield, and the backbone had non-specific amplification as observed by the streaking.

Learn

Based on the data collected from the PCR procedures, we concluded that these initial constructs had too many problems to be optimally used for future steps. The desired sequences for HiFi assembly couldn't be obtained, meaning we could not use these constructs to change the reporter for an insert sequence using Type IIS assembly, which is one of the main goals for our system. From this cycle we identified the main problems with our initial constructs, which were taken into consideration for the next cycles; mainly wrong orientation of restriction sites, and use of restriction sites that were also found in the plasmid backbone.

CYCLE 2

Design

The same approach for designing the new constructs as in the first cycle was taken here, but with extra care to ensure that the BsaI sites were in the proper orientation and none of the other parts used contained a restriction site of this type. The overhangs for these sites were chosen based on the prefixes and suffixes for a transcriptional unit outlined by the iGEM registry to increase standardization of these constructs.

New primers were also designed for the constructs. These primers bind to the regions outside the BsaI sites, meaning that they'll amplify the sequence inside the cassette, whether it be a reporter or an insert. This provides an alternative screening method which can be used to confirm for correct Type IIS assembly as long as the insert is a different length as the reporter (>365bp). Unfortunately, for this cycle this condition was not considered and the insert used was the exact same length as the reporter, which wouldn't allow us to differentiate the products in an agarose gel.

Build

Just as last time, the constructs were built on Benchling and then ordered as a clonal gene from Twist Bioscience. The constructs with an insert instead of a reporter were obtained using Type IIS Assembly with the NEB Golden Gate Cloning Kit directly on the constructs and the SS1 sequence. These reactions were then used to transform NEB5α cells, a strain suitable for Blue-White screening.

Test

The transformed cells were plated on X-gal plates for blue-white screening. While the construct using the WT MS2 sequence showed the expected results for the screening, the codon-optimized construct showed white colonies from cells transformed with the reporter construct.

Figure 3. X-Gal LB Agar plate of NEB5α cells transformed with the MS2 construct containing the LacZα reporter. Plate on the left was the construct with the wild-type MS2 sequences, plate on the right is the construct with codon-optimized MS2 sequences. Both constructs were inserted into a pET Kan plasmid.

Based on this screening, only the WT construct was used for the next steps. Following blue-white screening, the plasmid was purified from blue colonies and then the reporter was swapped with SS1 using Type IIS assembly as previously described. The modified plasmid underwent a second round of blue-white screening from which white colonies were selected and then used to transform BL21(DE3) cells. These cells were then induced with IPTG for expression of the MS2 phages and then the cultures were analyzed under a Bleach Agarose Gel [n] to assess size and integrity of the genome of the phages produced.

Figure 4. Bleach Gel of BL21(DE3) cells transformed with the WT MS2 construct and induced with IPTG. The left lane represents standards, the middle and right lanes are each a different colony of transformed cells.

Learn

The streaking in the Bleach Gel indicates either a degraded sample, or a highly heterogenous sample (as in many different RNA molecules were packaged into the phages instead of the single transcript). When we did literature research into possible issues, we realized that including all the proteins of the MS2 phages into a single system leads to poor results due to nonspecific RNA packaging, consistent with what was observed in our data [2]. Based on previously published work, to develop an MS2-producing system, only the Coat Protein is strictly required for phage assembly [2].

CYCLE 3: RUMIVEC

Design

Based on the results from the bleach gel of the last iteration, we changed our constructs to include only the Coat Protein (Cp) and the Maturation Protease (Mat), as these are the only ones required for structural assembly of the MS2 phage [2]. Although the Mat protein is not strictly required, including it will drastically increase stability of the phages and has a protective effect for the RNA genome [2]. Similarly, we changed the insert sequence we were testing to ensure that it was of different length as the reporter to make sure that the assembly could be screened using PCR and Agarose Gel Electrophoresis.

We also learned that the ratio between Cp and Mat proteins is crucial for accurate assembly of the MS2 phage with the correct RNA molecule [2]. Due to this, during the design of the new construct we needed to make sure that the RBS chosen for each protein had similar translation rates as the wild-type versions. The RBS were chosen based on the translation initiation rates (TIRs) predicted by the SalisLab Operon/RBS Calculator [3]. The wild-type (WT) untranslated region upstream of the Mat CDS was used as the RBS as it had similar TIRs as the WT transcript. However, the WT RBS for the Cp CDS in RumiVec had a TIR significantly lower than expected. Although the Operon Calculator has an option to design RBS, it always gives a sequence that maximizes TIR. Regardless, this option was chosen and then the sequence obtained was randomly mutated through different iterations to obtain an RBS with similar TIR as the WT transcript.

Figure 5. A.) Predicted Translation Initiation Rates of: the wild-type (WT) RBS of either the Maturation Protease (Mat) or Coat Protein (CP) with the wild-type sequence (WT RBS); the WT RBS of the two proteins within our construct (WT RBS in RumiVec); and the Custom RBS generated by SalisLab Operon Calculator inside our construct (Custom RBS). B.) TIRs of the different RBS sequences tested for the Cp CDS. Iteration 6 was chosen as the sequence for the construct.

To ensure that the phages produced by our system were mostly homogenous with the transcript produced by RumiVec, we designed the construct to include two modified pac sequences, which are recognized by the Cp for packaging the transcript into the phage capsid. This sequence was obtained from a study published by Wei, et al., which has a much higher affinity and packaging efficiency for RNA molecules by the Cp compared to the wild-type sequence. To ensure that the transcript produced from RumiVec is efficiently encapsulated into phage particles, the predicted secondary structures of the sequence were analyzed using Vienna's RNAFold to ensure that the necessary stem loop forms on the respective pac sites. The resulting model predicts with high-confidence the formation of the secondary structure in the pac sites required for adequate interactions with the Coat protein, meaning that RumiVec will be packaged into the phage progeny at high efficiency.

Figure 6. RNA Fold of the two pac sites (red arrows) in the transcript expressed by RumiVec.

The CDS for the proteins were designed by inputting the corresponding amino acid sequences obtained from the phage entry in NCBI (NC_001417.2) [4] into the IDT codon-optimization tool.

A new insert sequence, named Shuffle Sequence 2 (SS2), was also designed which has a greater length than the reporter cassette. This change allows for screening cloned plasmids using colony PCR, meaning we weren't restricted to blue-white screening for confirmation of our reactions for this last cycle.

For colony PCR, new primers needed to be designed for the construct. Given that the region downstream of the reporter remained unchanged, the same reverse primer as the one used in the last iteration could be used. The forward primer was designed using Benchling and then the pair was verified with Primer3Plus for the constructs including either the reporter or insert sequences.

Figure 7. Primer3Plus using the newly designed primers on the construct with the reporter. The predicted PCR product has a length of 362 bp and no issues were identified with the primers.

Figure 8. Primer3Plus using the newly designed primers on the construct with the SS2 insert sequence. The predicted PCR product has a length of 362 bp and no issues were identified with the primers.

Build

Following the same procedures as the previous cycles, the constructs were built on Benchling and then ordered as a clonal gene from Twist Bioscience. The constructs with an insert instead of a reporter were obtained using Type IIS Assembly with the NEB Golden Gate Cloning Kit directly on the construct and the SS2 sequence. These reactions were then used to transform NEB5α cells, a strain suitable for Blue-White screening.

Test

The transformed NEB5α cells were plated on LB agar plates and incubated. Two colonies from the transformation were then streaked onto fresh X-Gal plates and incubated again for Blue-White screening.

Figure 9. X-Gal LB Agar plate of NEB5α cells transformed with RumiVec containing either the reporter (a) or the SS2 insert sequence (b, c). No blue colonies were observed in any of the plates.

The same colonies used to streak the plates were also used for Colony PCR. The reporter-construct was included as a template control to compare amplification results and confirm successful insert integration.

Figure 10. Agarose Gel of different samples: Golden-Gate assembly reaction mixture (GG Assembly); Colony PCR from Colony 1 and 2 of NEB5α cells transformed using the Golden Gate reaction (GG Colony 1/2); Colony PCR from a colony transformed with the reporter-construct plasmid (pMS2_500 Transformant); and PCR on the reporter-construct plasmid itself (pMS2_500).

Learn

Our results from Colony PCR show that this is a valuable technique that can serve as a quick screening method for any insert/reporter combination using the same primer pair. This is a simple procedure which can quickly give results without needing any reagent other than the ones for PCR. The data from our experiments shows us that the reporter/cassette can be easily swapped out with any desired sequence as long as it has compatible ends for Type IIS assembly, indicating that we met the goal of generating a modular system.

Based on the results obtained from blue-white screening from this iteration, as well as our experience overall with this technique, we believe that a different reporter other than LacZα could serve a better role for our purposes. Blue-white screening seems to be unreliable and poorly replicable, which is why a new reporter is likely needed.

REFERENCES

[1] Bauman, V. R., Markushevich, V. N., Kish, A. Z., & Gren, E. I.a (1976). Sravnitel'naia fiziologiia replikatsii RNK-soderzhashchikh bakteriofagov MS2 i Q beta [Comparative physiology of the replication of RNA-containing bacteriophages MS2 and Q beta]. Mikrobiologiia, 45(4), 685-689. https://pubmed.ncbi.nlm.nih.gov/979687/

[2] Mikel, P., Vasickova, P., & Kralik, P. (2015). Methods for Preparation of MS2 Phage-Like Particles and Their Utilization as Process Control Viruses in RT-PCR and qRT-PCR Detection of RNA Viruses From Food Matrices and Clinical Specimens. Food and Environmental Virology, 7(2), 96-111. Advance online publication. https://doi.org/10.1007/s12560-015-9188-2

[3] Salis, H. M., Mirsky, E. A., & Voigt, C. A. (2009). Automated design of synthetic ribosome binding sites to control protein expression. Nature Biotechnology, 27(10), 946-950. https://doi.org/10.1038/nbt.1568

[4] GenBank. National Center for Biotechnology Information. [cited 2025 Sep 28]. Available from: https://www.ncbi.nlm.nih.gov/nuccore/NC001417