Team:Hong Kong HKUST/Design
This page explains the core parts of our switch:
Our sgRNA and asRNA parts were designed with many references to this great study on “Programmable control of bacterial gene expression with the combined CRISPR and antisense RNA system” by authors Y. J. Lee, A. Hoynes-Oconnor, M. C. Leong, and T. S. Moon. The link to the study can be found here:
https://www.ncbi.nlm.nih.gov/pubmed/26837577
dCas9-6XHis
BBa_K3017011, BBa_K3017029
The CRISPR associated protein, in our case, dCas9, is a non-specific deactivated endonuclease. dCas9 is catalytically dead by mutations D10A and H840A. It binds to the target DNA sequence when coupled with single-guide RNA, but does not induce any DNA breakage as in Cas9.
In our project, Combined CRISPRi and antisense RNA Toggle Switch, dCas9 plays a central role in CRISPR interference. dCas9 acts as an agent of steric hindrance to block mRNA polymerization to repress expression levels of target genes.
We sought help from Professor Ho Yi Mak from our university, who is also working with dCas9. The plasmid Professor Mak offered contains a Streptococcus pyogenes dCas9 gene optimized for mammalian cells but also is functional in E.coli. We designed multiple PCR overhang primers to make the dCas9 gene compatible with the standard assembly so that we could clone it into RFC10 plasmids in further steps.
The dCas9 protein is tagged with 6XHis at the C-terminus (BBa_K3017011) for easy characterization. In BBa_K3017069, parts BBa_K608002 and BBa_B0015 are added before and after dCas9-6XHis respectively by PCR overhangs. As the dCas9 protein coding region contains intrinsic PstI cut sites, we could not add an iGEM suffix after the part. Instead, we added iGEM prefix and SpeI and SalI cut sites as suffix for further digesting-ligation steps.
Single-guide RNA (sgRNA)
BBa_K3017001, BBa_K3017002
dCas9 is directed to the specific DNA locus by a sgRNA, where it binds to suppress downstream gene expression. sgRNA, a common abbreviation of short guide RNA or single guide RNA is a combination of crisprRNA (crRNA) and tracrRNA. Their respective functions are to guide the Cas9 protein to the target sequence, and scaffold the Cas9 protein. With reference to the research on reversible CRISPRi switch[1], we redesigned the traditional sgRNA by adding an artificial linker behind crRNA and tracrRNA as well as modified the 3-component-sgRNA to suit our suppression purposes. The dCas9 protein used in our project originated from Streptococcus pyogenes (sp), therefore the design of sgRNA is compatible with spCas9.
In our circuit, two sgRNAs target the protein coding sequences of gfp and mrfp respectively to suppress these genes under different conditions. sgRNA transcripted under arabinose induction represses mrfp, while sgRNA transcribed under IPTG induction represses gfp.
 |
 |
| sgRNA for gfp repression (BBa_K3017001) | sgRNA for mrfp repression (BBa_K3017002) |
|
Secondary structure of sgRNA is predicted by a software, NUPACK. |
|
Spacer - crRNA
crisprRNA(crRNA) is also commonly referred to as the spacer. A spacer encodes the sequence complementary to the target binding site, and is usually 19-20bp in length. When choosing the target binding region, we considered mainly 2 factors: the location of the PAM sequence, and the suppression effect upon binding.
The PAM sequence for spCas9 is 5′-NGG-3′ (where “N” can be any nucleotide base), which is located downstream of the binding region on the non-target strand. Recognition of the PAM[3] by the Cas9 nuclease is thought to destabilize the adjacent sequence, allowing interrogation of the sequence by the sgRNA, and thus resulting in RNA-DNA pairing in the presence of a matching sequence. Therefore the presence of a PAM sequence is essential for the Cas9 protein to function[3].
The research shows that CRISPRi suppression effect is the strongest at 35 nucleotides upstream from the start codon of the coding region[1]. However, upstream area of our coding region is a generic constitutive promoter. To avoid non-specific binding, we compromised the suppression efficiency and chose a region shortly after the start codon where suppression is only a few percent weaker than the ideal region. We found a PAM sequence (TGG) 27 nucleotides in GFP part BBa_E0040, which lead to an sgRNA binding region spanning 20bp, 7 nucleotides into the protein coding sequence. To accommodate the PAM sequence in BBa_E1010 mRFP, the spacer is arranged on the opposite DNA strand 14 nucleotides into the gene. When sgRNA with spacer that is complementary to mRFP sequence is transcribed, it binds with dCas9 to suppress the expression of mRFP. Vice versa, when sgRNA specific to GFP is transcripted, GFP is suppressed.
 |
 |
| sgRNA bound with target gfp DNA | sgRNA bound with target mrfp DNA |
Handle - tracrRNA
tracrRNA is an RNA loop that acts as a handle for dCas9 to hold on to, such that the dCas9 protein is delivered to the target site together with the sgRNA. Earlier research[4] has proved that tracrRNA is strictly required for Cas9-mediated DNA interference both in vitro and in vivo. The tracrRNA forms a loop on the sgRNA after transcription to provide a scaffolding site for the dCas9 to form a duplex with the spacer.
Loop - artificial linker
Destroying the secondary structure of the handle in sgRNA could theoretically cause separation of the dCas9 protein from the sgRNA, thus removing the suppression effect. The study mentioned above had proved this hypothesis. The team then tried to design an artificial linker, which also forms a loop as a secondary structure, after the handle. After several trials and modifications, the research team[1] discovered that extending the artificial loop, i.e. destroying the secondary structure, could further increase the derepression.
In our sgRNA designs, the loop is designed mostly following the study. However, our spacer is different from what the study uses. RNA secondary structure prediction by NUPACK shows that the probability of artificial linkers forming secondary structures with the spacer is higher than forming a loop with itself. Therefore, we changed some nucleotides to promote this predicted event.
 |
 |
| before modification | after modification |
|
Secondary structure of sgRNA for gfp DNA binding |
|
Antisense RNA (asRNA)
BBa_K3017003, BBa_K3017004
asRNA is the key to the reversibility of our CRISPRi switch. The asRNA binds to the artificial linker loop of the sgRNA, and extends the loop to become a linear structure. The study shows that this linearization of the loop causes derepression.
The synthetic asRNA is composed of 2 functional domains: the extensor, for linearizing the loop; and Spot 42, which aids the stability of the asRNA.
asRNA BBa_K3017003 reverses the suppression effect on GFP induced by sgRNA BBa_K3017001, while asRNA BBa_K3017004 reverses the suppression effect on RFP induced by sgRNA BBa_K3017002.
 |
 |
Extensor
The extensor sequence is complementary to the artificial linker loop. By forming a duplex, the loop is extended. It was hypothesized in the study that an asRNA complementary to the spacer would displace the DNA target and yield a higher depression effect. But after optimization, the study found out that changing the sgRNA’s secondary structure by extending a loop formed by the artificial linker would maximize the derepression effect to at least 95%, compared to only 15-55% of complementing the spacer.
Since the derepression has to be specific to the gene, the asRNA extensor and sgRNA artificial linker are designed in pairs; one for each of the toggling states.
asRNA-sgRNA-target GFP DNA complex
asRNA-sgRNA-target RFP DNA complex
“Programmable control of bacterial gene expression with the combined CRISPR and antisense RNA system,”[1]The study found out that changing the sgRNA’s secondary structure by extending a loop formed by the artificial linker would maximize the derepression effect to at least 95%, compared to only 15-55% of complementing the spacer.
Spot 42
Bacterial Hfq protein is known to modulate the stability or the translation of mRNAs and interact with small regulatory RNAs. A study provides evidence that Hfq strongly cooperates in intermolecular base pairing between the asRNA and its target RNA. Spot 42 is an RNA antisense regulator that has several A/U-rich regions that provide specificity to Hfq protein. Hfq protein also provide protection to the asRNA at these A/U-rich regions.[5]
An identical Spot 42 sequence is present in every asRNA we have designed as it is highly specific to Hfq protein and the duplex is necessary for the asRNA to work. Hfq is endogenous to E.coli[6], therefore the protein is not encoded in any part of the circuit.
References:
[1]Y. J. Lee, A. Hoynes-Oconnor, M. C. Leong, and T. S. Moon, “Programmable control of bacterial gene expression with the combined CRISPR and antisense RNA system,” Nucleic Acids Research, vol. 44, no. 5, pp. 2462–2473, Feb. 2016.
[2]C. Anders, O. Niewoehner, A. Duerst, and M. Jinek, “Structural basis of PAM-dependent target DNA recognition by the Cas9 endonuclease,” Nature, vol. 513, no. 7519, pp. 569–573, 2014.
[3]S. H. Sternberg, S. Redding, M. Jinek, E. C. Greene, and J. A. Doudna, “DNA Interrogation by the CRISPR RNA-Guided Endonuclease Cas9,” Biophysical Journal, vol. 106, no. 2, 2014.
[4]T. Karvelis, G. Gasiunas, A. Miksys, R. Barrangou, P. Horvath, and V. Siksnys, “crRNA and tracrRNA guide Cas9-mediated DNA interference inStreptococcus thermophilus,” RNA Biology, vol. 10, no. 5, pp. 841–851, 2013.
[5]T. Møller, T. Franch, P. Højrup, D. R. Keene, H. P. Bächinger, R. G. Brennan, and P. Valentin-Hansen, “Hfq,” Molecular Cell, vol. 9, no. 1, pp. 23–30, 2002.
[6]G. M. Cech, A. Szalewska-Pałasz, K. Kubiak, A. Malabirade, W. Grange, V. Arluison, and G. Węgrzyn, “The Escherichia Coli Hfq Protein: An Unattended DNA-Transactions Regulator,” Frontiers in Molecular Biosciences, vol. 3, 2016.