Aim
In the course of evolution, plants obtained defence mechanisms against pathogen infections. Upon infestation, the activation of specific promoters (i.e. the FAD7 promoter) leads to the expression of antimicrobial chemicals, proteins and enzymes protecting the plant against pathogen infection. But often the plant’s own defence mechanisms are not strong enough to completely protect against infections, especially if the plant is stressed (drought, heat, etc.). Our project is about boosting the plant’s immune system by introducing additional defence genes into the plant. A key objective for our system is that the normal gene regulation is also used for the regulation of our synthetically introduced defence and is only activated during plant infection.
How to tackle this problem
First, it is crucial to understand which molecular mechanisms are activated during infection and how the respective gene regulation works. Plants recognise infection leading to the activation of inducible promoters and expression of defensive genes starting a cascade to release proteins. We have to find a way to interfere with this control system, without disturbing it.
With genetic engineering, it is theoretically possible to compute outputs depending on specific inputs or design multiple outputs from a single input. For our project, we want to couple a synthetic pathogen resistance output with the normal reaction of the cell to an infection. The result is two outputs with only one given input. The easiest way to do so would be by introducing a new resistance gene controlled by the same inducible promoter to the plant.
Unfortunately, the same promoters often work differently in specific gene environments. Due to the structural complexity of the DNA far-away parts can interact with the promoter and operate as an enhancer or silencer. Therefore, it is not certain that the same promoters at different places inside the genome lead to the same amount of gene expression. In some applications, it is absolutely crucial that genes are expressed in the same amount. Since we imagine this synchronisation to be used in multiple applications, where those identical gene expressions are more important, we had to come up with another system to accomplish it.
Nature Can Be a Great Inspiration
In prokaryotes gene regulation is less complex, but a unique feature is the expression of functionally connected genes in operons. All parts/enzymes needed for a specific cellular process are expressed from the same mRNA, their amount is very similar to each other. For eukaryotes this system has not been used to a great extent. If we want to transfer this model in the eukaryotic cell, how do we achieve that?
Transcriptional Synchronisation
When a gene is transcribed mRNA is produced and after processing is translated by the ribosome: Initiation of translation is different in eukaryotes than in prokaryotes. There is no Shine-Dalgarno sequence, but the ribosome scans the mRNA for a start codon after interaction of initiation factors with the 5’cap and poly-A tail. It is also possible to have an internal ribosome entry site (IRES), where the ribosome can directly bind to the mRNA. But if an IRES is used less protein is produced compared to initiation using initiation factors. Our idea to solve this problem is to split the unprocessed mRNA by ribozymes and add processing signals for 5’ cap and poly-A tail to allow for translation. This process results either in a protein or RNA therapeutic output. A disadvantage of proteins is that they need to be processed and have a high metabolic burden to the cell.
Plant Design
Our initial idea for an application of transcriptional synchronisation is a defence mechanism against pathogen infection in plants. siRNA is a widely used immune mechanism in plants.
Small interfering RNA (siRNA) is a class of non-coding RNA of 20-30 nucleotides length that is cleaved from a longer, double-stranded RNA. This is the first step of the process called RNA-interference and performed by the enzyme dicer. The resulting siRNAs are then bound to the multiprotein RNA-induced silencing complex (RISC). Here, the strands are separated and the more stable one is integrated into the RISC complex. This antisense siRNA strand is used to guide the RISC to complementary mRNAs and more or less stable alignment to this mRNA is performed. The catalytic component of the RISC, a protein from the argonaute family, cleaves the target mRNA and prevents its translation.[1] This mechanism has first been discovered in plants, where it is part of the defence response against viruses, bacteria and fungi [2].
We are going to synchronise the expression of a pathogen targeting siRNA to a by the infection upregulated gene. By using self-cleaving ribozymes the mRNA is split into 2 functional parts, one of which is normally processed by the cell and an additional therapeutic RNA against the pathogen infection. On DNA level the transcriptional synchronisation device is inserted into an endogenous gene between its ORF and transcription termination sequence. It consists of a synthetic poly-T DNA-Sequence, followed by a 3’ self-cleaving ribozyme, a spacer, a 5’ self-cleaving ribozyme, a siRNA output and another 3’ self-cleaving ribozyme. Upon inducible activation of a promoter the whole DNA sequence, including the poly-T sequence (poly-A tail) is transcribed into a pre-mRNA. After transcription, the ribozymes on the pre-mRNA undergo self-cleavage and a 5’ cap is added to the ORF. The part containing the ORF is now a functional mRNA and can be exported and translated into a protein due to the 5’ cap and the synthetic poly-A tail, which enables nuclear export and cytoplasmic translation of the mRNA. The cut out self-cleaving ribozymes will be degraded by nucleases. The spacer between the ribozymes is necessary to allow them to assume their active secondary structure without interrupting each other. The DNA sequence of the spacer was chosen to be poly T to minimize secondary structure formation and interaction with neighbouring sequences, especially the poly-A tail. After cleavage by the ribozymes, the output siRNA can execute its function independently. The last 3’ self-cleaving ribozyme and the attached part of the transcribed terminating sequence are degraded by nucleases as well. The stability of the synchronised part of the 3’ design can be additionally influenced by the length of the synthetic poly-T tail: The shorter the tail, the earlier the synchronised part is degraded by the cell.
Figure 1: Depicted is the design for the eukaryotic device for transcriptional synchronisation. The design has saiRNA output that is flanked by a 5’ and a 3’ self-cleaving ribozyme. After transcription, the output is cut from the pre-mRNA by the ribozymes and can fulfil its intended function. The pre-mRNA is also cut by another 3’ self-cleaving ribozyme, which releases the ORF together with the synthetic poly-A tail. This part can be further modified by the cellular machinery and act as normal mRNA. Abbreviations: open reading frame (ORF), hammerhead ribozyme (HHRz) hepatitis delta virus (HDVRz), terminating sequence (Terminator).
Proof of Concept in E. coli
Because of corona restrictions and long generation times, we couldn’t directly test transcriptional synchronisation in plants. For that reason, we had to come up with a proof of concept to show that the system works in a different context. Since we are limited on which model organism we can use in our lab, we had to develop a proof of concept for E. coli. Since RNAi is a eukaryote specific regulatory mechanism we needed an RNA-based workaround. We decided to implement a regulatory system based on short transcription activating RNAs (STARs) that can function as RNA-logic regulators. The STAR mRNA expression is synchronised to an RFP reporter gene, which is under control of the arabinose inducible pBAD promoter. The self-cleaving Hammerhead Ribozyme (3’-Ribozyme) and Hepatitis Delta Virus (5’-Ribozyme) will cut the mRNA and let RFP and STAR work independently. The STAR mRNA will then interact with the complementary T500 terminator and activate the expression of an amilGFP by resolving a preexisting hairpin structure. By measuring both amilGFP and RFP fluorescence in the cell, we can determine whether synchronisation works.
More Information on Proof of ConceptBuild & Test
After designing the mechanism, we came up with an assembly scheme to build the E. coli part and test some function of it. Due to the Covid-19 pandemic, access and activity in the lab was restricted and we couldn’t completely test our system. We are planning on realizing this in a second year project.
Future Plans for E. coli (Outlook 2021)
To optimise transcriptional synchronisation in E. coli we planned different experiment to ascertain the function of our system and to find out if adjustments are necessary:
- Cleavage kinetics of the ribozyme: To prove that the STAR mRNA is cut out of the vector, we will characterise the cleavage kinetics of the ribozyme. Additionally, we plan to investigate the effect of permutation in the sequence on the cleavage efficiency.
- Functionality of STAR: We will show the necessity of the cleavage of STAR by the flanking ribozymes by creating a plasmid without functional ribozymes. In this plasmid, the coding sequence for both HDVRz and the HHRz is replaced by a non-coding base sequence that was created by base randomization of the original ribozyme sequences. This will ensure the loss of the characteristic stem-loops that are essential for the cleavage function of ribozymes [1][2].
- Inducible promoter for attacker plasmid: The attacker plasmid will receive an inducible pBAD promoter that will replace the constitutive promoter currently present in the plasmid to make the system responsive to external inputs.
- Characterisation of the complete system: We plan to transform E. coli with both, the attacker and the target plasmid and show that the expression of mRFP1 and the subsequent release of cleaved STAR-RNA induces the translation of the reporter amilGFP.
Future plans for Plants
The implementation of transcriptional synchronisation in plants is more difficult, since there are different plant models available.
Plants can basically be divided into two groups, mono- and dicotyledons. The appearance of the plant already reveals to which group it belongs. Monocotyledons only have one cotyledon, dicotyledons have, as the name suggests, two. In monocot plants, the leaves have an elongated structure, whereas dicotyledonous plants have broad leaves and the arrangement of the veins differs from a parallel course to a network of veins. Based on the shape of the flower, it can also be determined whether it is a mono- or dicotyledon plant. When planning our project, we decided to use corn, which belongs to the monocotyledons, and Arabidopsis (dicotyledons) as model organisms. In this way, we can show that our mechanism works in both groups of plants.
We plan to establish transiently transformed Zea mays as a monocotyledon and stably transformed Arabidopsis thaliana as a dicotyledon plant model. Our design of the mechanism allows an uncomplicated transfer between different model organisms through adding specific cutting sites for each vector flanking the insert. For Zea mays we will use the vector E040pHSP-mGFP-o and for Arabidopsis thaliana pABindmCherry. In both vectors, there is an inducible promoter, which we will use for the activation of our mechanism. The inducible promoter will be subject to changes depending on which context we want to use our system.
The constructs contains following parts:
- Flanking cutting sites: For inserting the construct in Zea mays or Arabidopsis thaliana
- Cutting sites inside the insert: For exchanging parts (e.g. Poly-A-tail)
- mCherry: synchronised reporter protein, simulating a naturally occurring protein in the plant
- Poly-A-tail (48 bp): Ribozymes prevent natural polyadenylation → Need of synthetic Poly-A-tail
- Hammerhead Ribozyme (HHRz): Cuts at 3‘-end of ribozyme with complementary sequence to 5‘-end of first YFP sequence (important for secondary structure of HHRz)
- YFP (two sequences separated by spacer): Complementary to each other
The sequences form a double strand (300 bp) which will be cut by a Dicer into shRNA for RNAi - Spacer: Separates the two YFP sequences
- Hepatitis Delta Virus Ribozyme (HDVRz) : Cuts at 5‘-end of the ribozyme
- Terminator 35S: Not included in the vector
The ultimate goal is to use the system against pathogen infection. But before we can tackle this, we have to show that the system works in plants as expected. So we first plan to test our mechanism in Zea mays as a proof of concept: Zea mays mutants that were stably transfected with YFP will be transiently transfected with a plasmid via gene gun. The plasmid contains an inducible mCherry cassette that is transcriptionally synchronised with an siRNA output against the YFP of the plants. Upon biolistic transformation, the inducible heat shock promoter of the RFP cassette will be activated by exposure to higher temperatures - triggering the production of RFP and the siRNA against the plants YFP. The siRNA will inhibit the production of YFP proteins, reducing the yellow fluorescence. Thus, if the fluorescence of the plant cells changes from yellow to red, we will know that our mechanism works as intended and that the synchronised protein (RFP) as well as the added output (siRNA) are functional.
Improvement of the Different Parts of the System
If the possibility should arise that the synchronised gene is not stably expressed, we have some ideas how to improve the system.
- Use of a terminator sequence as an alternative for the Poly-A-tail: to see, if necessary for proper nucleus export and translation
- Different length of the Poly-A-tail to test different stabilities of potential mRNA outputs
- Use of a Group-I-like Ribozyme (GIR) instead of the Poly-A-tail or terminator for 5’-capping as an alternative for 3’-polyadenylation In this case the order of synchronised gene and siRNA is changed.
- Use of a transcription factor instead of siRNA. to show that synchronisation also works on a protein level
Fine tuning of siRNA production
In another experiment we will determine the produced amount of RNAi. Therefore, we will use different external triggers and compare the amounts of synthesized RNAi. The RNAi will later defend the plant against the pathogen, so it is important that it is produced in sufficient quantities. We need to find the best conditions for RNAi synthesis before we can test the system with real pathogens. Examples of possible external influences could be temperature or light intensity. In addition, it is important to find out exactly what happens during a pathogen infection. We are planning to monitor the effect of different pathogens on the cell and hope that we will find out by RNA sequencing what specific effects these pathogens have on the cell. Then we will be able to select the right inducible promoter with which to synchronize our system. This for example could be an abscisic acid promoter. We are also gonna test whether we can detect differences in siRNA expression after infection with different types of pathogens once we have found the right inducible system. Through all these experiments we hope to find the optimal conditions for the system for specific pathogen infection. In the last step, we want to infect plant cells with pathogens and compare transcriptional synchronisation with other pathogen resistance mechanisms using RNAi. It is possible to apply RNAi directly to the leaves from the outside to protect the plant from pathogen infection [3]. We assume that it is more effective to use the plants for the production of RNAi than to apply RNAi directly to the leaves and are going to test how our system works in direct competition.
References
[1] Dana H, Chalbatani GM, Mahmoodzadeh H, et al. Molecular Mechanisms and Biological Functions of siRNA. Int J Biomed Sci. 2017;13(2):48-57.
[2] Muhammad T, Zhang F, Zhang Y, Liang Y. RNA Interference: A Natural Immune System of Plants to Counteract Biotic Stressors. Cells. 2019;8(1):38. Published 2019 Jan 10. doi:10.3390/cells8010038
[3] Dubrovina AS, Kiselev KV. Exogenous RNAs for Gene Regulation and Plant Resistance. Int J Mol Sci. 2019;20(9):2282. Published 2019 May 8. doi:10.3390/ijms20092282
