Secondary structure prediction of TLS

Research indicates that the folding structure of the TLS motifs triggers the movement of transcript in plant vascular bundle. A predicted long hairpin motif with a stem (8 to 12 nucleotides)–variable bulge(s)–stem (4 to 7 nucleotides)–variable loop is found crucial to the mobility of transcript, And the D loop deletion enhances the transport activity.[1]For early screening, we obtained 20 tRNA sequences of Arabidopsis thaliana from the ncbi database, and performed RNA secondary structure prediction (done on RNAfold web server), using tRNA containing above structure for subsequent experiments.

Figure 1: Secondary structure prediction of natural tRNA sequence and artificially modified tRNA sequence. tRNA of Met(A) and Met ΔΔT(B) are proven to be mobile, both of which share stem–variable bulge(s)–stem–variable loop structure. Ala tRNA(C) share the same structure while His tRNA(D) doesn't.

Plasmid construction

To verify the TLS-triggered mobility of fused transcript in vascular bundle of Nicotiana benthamiana，we amplified EGFP gene from plasmid pcDNA3.1-T2A-EGFP. 20 tRNA sequences of Arabidopsis thaliana and Nicotiana benthamiana are synthesized. Then 4 plasmids pGREENII-GFP(-TLS) is constructed, which contains 35S promoter and T-DNA repeat to efficiently express our mRNA in plant cell.

To visualize the evaluation of mobility, we constructed a series of plasmid pGREENII-GUS-TLS and pGREENII-RUBY-TLS, where RUBY( BBa_K3900028 ) is a effective reporter for noninvasively monitoring plant gene expression and plant transformation[2]. Kozak sequence are added to the 5' terminal of CDS to enhance translation efficiency.

Figure 2: To visualize the evaluation of mobility, we constructed a series of plasmid pGREENII-GUS-TLS and pGREENII-RUBY-TLS, where RUBY( BBa_K3900028 ) is a effective reporter for noninvasively monitoring plant gene expression and plant transformation[2]. Kozak sequence are added to the 5' terminal of CDS to enhance translation efficiency.

GFP qPCR result showing TLS-triggered transcript mobility

First, in order to verify that our mRNA can move in plants driven by TLS, Agrobacterium containing our pGREEN-GFP-TLS shuttle vector was injected into the marked area of plant leaves for transient expression. Unfortunately, no significant fluorescent signal was detected outside the injection region under excitation light. However, relative GFP mRNA level of leaf samples outside the injection area showed that the mRNA containing TLS moved further.

Figure 3: First, in order to verify that our mRNA can move in plants driven by TLS, Agrobacterium containing our pGREEN-GFP-TLS shuttle vector was injected into the marked area of plant leaves for transient expression. Unfortunately, no significant fluorescent signal was detected outside the injection region under excitation light. However, relative GFP mRNA level of leaf samples outside the injection area showed that the mRNA containing TLS moved further.

RUBY to visualize the movement and expression of mRNA

Since GFP mRNA can be detected by qPCR, we speculate that the fluorescence signal of GFP is too weak to be observed. Therefore, we seek more sensitive reporter. We constructed plasmid PGREENII-RUBY-TLS, transiently expressed RUBY-TLS at the tip of tobacco leaves, and observed after 4-5 days of injection. The results showed that TLS could trigger the transport of RUBY mRNA, and mRNA was successfully expressed in plant cell.

For quantitative measurement, qPCR experiments were also used to detect the movement of RUBY mRNA. The mobility of RUBY was weaker than that of GFP, probably because RUBY mRNA (~4kbp) was much longer than that of GFP (~800bp).

Figure 4: RUBY to visualize and expression of mRNA.
A) Movement of RUBY mRNA triggered by TLS. Colored spots were observed at positions outside the injection area and close to the vascular bundle.
B) qPCR result. All three TLS showed mobility in 1cm and 2cm away from the injection aera.

GUS staining showing mobility of other TLS

Considering the large difference in the length of the transcripts of GFP and RUBY, we used another commonly used reporter : GUS(~2kbp), and we also wanted to explore the contribution of other TLS to mRNA mobility. A series of PGREENII-GUS-TLS plasmids were constructed and transiently expressed in tobacco. The results showed that in addition to the above TLS, Ala tRNA from Arabidopsis thaliana can also drive mRNA movement.

Figure 5: GUS staining result. After ethanol decoloring, obvious blue GUS enzymatic reaction products were observed around the vascular bundle, indicating the successful transportation and expression of mRNA.

Since we synthesized tRNA sequences of 20 amino acids, we will further explore the mobility of different TLS. However, due to time constraints, we will improve this part in future work.

Future plan: multiple TLS

In RNA imaging systems for MS2/MCP [3], sequence-specific RNAs that bind to MS2 viral capsid proteins are often used to enhance binding to proteins by increasing the copy number, and such RNAs are characterized by the ability to maintain a stable secondary structure after multiple repeats of the stem-loop. pCR4-24xMS2SL-stable is a commonly used plasmid for the expression of MS2-specific binding RNA, which produces an RNA(SL24) containing 24 stem-loop repeat structures[4]. Expecting to enhance mRNA binding to RNA motility-associated proteins by increasing the copy number of the TLS sequence, we replaced some of the repeats of SL24 with the RNA sequence of TLS (Ara.Mer.ΔΔt).

Due to the quality of our ligase, we failed to complete this part of the verification. We will explore whether the superimposed TLS can further enhance the mobility of mRNA in subsequent experiments.

Figure 6: Secondary structure prediction of TLS(Ara.Met.multi). The sequence can form a stable multiple stern-loop structure.

RNA dependent RNA polymerase(RdRp)

This part attempts to use TMV replicon to extend the action time of RNA products. While expressing TMV RNA dependent RNA polymerase, we detect transcription and translation levels of target proteins expressed using two subgenomic promoters of TMV. Further, we try to predict the possibility of combining TLS and RdRp.

Plasimid Construction & Modification

Considering that TMV is a mature vector, we try to modify it directly based on TMV genome. First, we successfully cloned the complete genome of TMV from pUC18-TMV and the annotation showed that TMV had two subgenomic promoters (subPromoter MP & subPromoter CP) respectively expressing movment protein and capsid protein. On the one hand, we chose an existing mature transformation solution: we inserted GFP ORF after subPromoter CP to replace CP ORF for expression (Fig.1 A). Because part of the coding sequence of capsid protein constitutes the sequence of subPromoter CP, we mutant the start codon in the subpromoter(5,712-5,714,from ATG to AGA) in case of expressing a nonsense peptide(Fig.1 C) [5].

On the other hand, to reduce the pressure on TLS of transporting RNA, we want to minimize the size of the RNA product. Therefore, we try to delete the sequences of MP ORF and subPromoter CP and express the target protein after subPromoter MP (Fig.1 B). Although part of the movement protein coding sequence is also included in the TMV RNA-dependent RNA polymerase coding sequence, To ensure the functional and structural integrity of the RNA-dependent RNA polymerase, we chose to preserve a sequence at the beginning of the movement protein ORF (17 bp) and verify its effect later (Fig.1 D).

Figure 1 Construction of plasmid for RdRp verification. A. the plasimid of pGreenⅡ -TMV RNA dependent RNA polymerase-movement protein-subPromoter CP-EGFP; B. the plasimid of pGreenⅡ -TMV RNA dependent RNA polymerase-subPromoter MP-EGFP; C. the sequence of CP start codon; D. the sequence of MP start codon.

Figure 7: Construction of plasmid for RdRp verification.
A. the plasimid of pGreenⅡ -TMV RNA dependent RNA polymerase-movement protein-subPromoter CP-EGFP; B. the plasimid of pGreenⅡ -TMV RNA dependent RNA polymerase-subPromoter MP-EGFP; C. the sequence of CP start codon; D. the sequence of MP start codon.

Fluorescence microscopy of Transformed Leaves

After successfully constructing the plasmid, we infected tobacco with Agrobacterium that transformed the plasmid. At day 4, 9, 14, Fluorescence microscopy was used to observe GFP fluorescence at leaf infection sites to determine the successful transfection of the plasmid and the amount of target protein expression. At the same exposure intensity, it was found that the two groups expressing the RNA dependent RNA polymerase could directly observe stronger fluorescence than the group expressing the protein directly with the CaMV 35s promoter (Fig.2 A).

Further, 5 fields of view at low magnification were selected for each group, and the mean fluorescence intensity of each field was calculated and analyzed by Image processing software (Image J). The results showed that mean fluorescence of the two groups expressing RNA-dependent RNA polymerase was significantly higher than that of the control group (Fig.2 B). In addition, the fluorescence intensity of the subPromoter MP group (average 15.23) is about 2.5 times that of the control group (average 35.04), and the subPromoter CP group is about 3.5 times (average 50.412). Based on these results, phenotypically we demonstrate the effectiveness of two subgenomic promoters, as well as the role of RNA-dependent RNA polymerase.

Figure 8: Microfluorescence of transient transformed leaves at day 4.
A. microfluorescence image of transformed leaves, 4x means a quadruple objective, 10x means a quadruple objective;
B. mean fluorescent of multiple views at 4x, t test was used (GFP group as control, *** p<0.001, **** p<0.0001).

Detection of Protein expression

Next, to determine that the RNA dependent RNA polymerase enhances the expression of the target RNA at the protein level, we took tissue from the transfection of leaves site and extracted the protein. Leaf tissue (50 mg) expressing GFP was harvested and GFP accumulation in extracts of total soluble protein was determined by spectrofluorometry as described in Man and Epel [6]. GFP accumulation in tissues expressing GFP derived from control group is significant inferior to that obtained from the subPromoter MP group and subPromoter CP group (Fig.3 A). Tissue expressing GFP from subPromoter MP produced a mean value of 19.13 relative fluorescence units., 4-fold more GFP in comparison to control group that produced a mean of 4.87 units. While that from subPromoter CP produced a mean value of 30.07 relative fluorescence units., 6-fold more GFP in comparison to control group. In addition, Western Blot was used to further verify changes in protein expression and the results were consistent with the fluorescence intensity of the protein extract(Fig.3 B).

Figure 9: Function detection at protein level. fluorescent analysis of protein extract, t test was used (GFP group as control, **** p<0.0001); B. protein immunoblotting, actin as the loading control.

Detection of RNA expression

At the same time, the function of RNA-dependent RNA polymerase at the RNA level was also verified. 9 leaves from each group were taken for RT-qPCR to detect the transcribed RNA. The results showed that the two groups expressing RNA-dependent RNA polymerase had significantly higher RNA expression levels (Fig.4). After normalization with the control group, the average relative expression of the subPromoter MP group was about 4 and that of subPromoter CP was about 10. The results showed that RNA-dependent RNA polymerase can play a good role in replication and increase the expression of target RNA, suggesting the possibility of prolonging the lifespan of RNA.

Figure 10: RT-qPCR detects function of RdRp at the RNA level. t test was used (GFP group as control, **** p<0.0001), n=9

Prediction of RNA secondar1y structure

It has been confirmed that the self-replication of RNA dependent RNA polymerase requires the secondary structure of TMV 3’ UTR, especially the stem-ring structure plays an important role in TMV replication [7]. The mobility of the TLS sequence also depends on the stem ring in the secondary structure and is inserted near the TMV 3’utr in our design. To determine whether the combination of the two affects the correct formation of the secondary structure, we utilized the RNA secondary structure prediction tool (RNAFold Web Server ) to forecast the transformation sequence. The results show that the secondary structure of TLS and TMV 3’UTR is basically not affected, and the stem-ring structure is still preserved regardless of whether TLS is inserted directly after or before 3’ UTR(Fig.5). This indicates that our transformation will not affect the normal realization of the structure and function of the two.

Figure 11: Prediction of secondary structure after insertion of TLS

Future Plan

At present, we have verified the replication function of TMV RNA-dependent RNA polymerase and the feasibility of such modification. Due to time constraints, although Agrobacterium transfection of tobacco has been carried out, we have not had time to sample and detect RNA and protein expression levels. We hope to draw the time-expression curve of target RNA and protein under RdRP in the future to characterize the half-life and action time of target RNA and protein.

At present, we have verified the replication function of TMV RNA-dependent RNA polymerase and the feasibility of such modification. Due to time constraints, although Agrobacterium transfection of tobacco has been carried out, we have not had time to sample and detect RNA and protein expression levels. We hope to draw the time-expression curve of target RNA and protein under RdRP in the future to characterize the half-life and action time of target RNA and protein.

Time-Expression Curve

Finally, in order to verify that TLS and RdRp can work together, we chose the TLS (Ara.Met.Full) with the best comprehensive performance, and tried to install two sets of RdRp systems of different sizes respectively. The results show that although TLS mobility is reduced after RdRp is installed, it is still retained.

Figure 12: Half-life of the product after loading RdRp
A.Curve of RNA expression over time. RNA relative expression in leaf tissues was measured at 4,9,14 days post infection.we take 4 d.p.i as the peak expression of protein and RNA.
B.Curve of protein expression over time. Fluorescence intensity in leaf tissues was measured at 4,9,14 days post infection.we take 4 d.p.i as the peak expression of protein and RNA

Integration of RdRp and TLS

Finally, in order to verify that TLS and RdRp can work together, we chose the TLS (Ara.Met.Full) with the best comprehensive performance, and tried to install two sets of RdRp systems of different sizes respectively. The results show that although TLS mobility is reduced after RdRp is installed, it is still retained.

Figure 13: Microfluorescence of infection after RdRp and TLS integration
A.Statistical table of positive field of view under microfluorescence. We selected tobacco leaves 4 days after infection, counted 100 visual fields at different distances from the injection site under 40 microscope times, and recorded the visual fields with fluorescent cells as positive.
.Microfluorescence image. Representative fluorescence images of different groups1cm and 2cm away from the injection site were selected.

PART1 Construction of pGreen-Pikm1 (enhancer)

Acquisition of sequences

Acquisition of Pikm1 and Pikm2 sequences

At first, we did not notice that the Pikm gene is not found in all rice blast-resistant rice varieties, so our initial attempt to obtain the gene from general rice varieties was unsuccessful. Fortunately, we later obtained rice seeds with Pikm resistance genes from Prof. Pan Qinghua, South China Agricultural University, and obtained Pikm1 and Pikm2 genes by PCR.

Cultivation of rice seeds

We select full and complete seeds, then soak them in 75% ethanol for 1min, and disinfect them in 15% KC1O for 30min. After that, we rinse them repeatedly with distilled water for 5-6 times, then place them in 90mm petri dish covered with 2 layers of filter paper, wet the filter paper in each culture, and place the petri dish in a 28°C light incubator for germination experiments, with a light cycle of 14h/10h (day/night)

After the rice germinates, it is transferred to MS medium, and the culture is continued under the above conditions. After the rice grows a certain length of roots, they are cut as the experiment materials rather than leaves, because the PIKM gene is expressed more in the roots, which may be explained by that the rice blast will infect roots first in the soil.

Figure 13: Rice (Oryza Sativa) that we grew in MS

Plant RNA extraction

Consider that the expression of the Pikm gene in roots is much higher than in leaves, we only take rice root tissue. At the beginning, when we extracted it using the commonly used Trizol method, the extract was in a viscous state. We speculated that was because there were a large number of polysaccharides in it, and the Trizol method could not completely remove it. Therefore, later we used the RNAprep Pure kit (DP432, from Tiangen) and got better results.

Acquisition of Pikm1 and Pikm2 genes

Using the RNA obtained above as a template, We obtained rice cDNA by rt-pcr method. The Pikm1 and Pikm2 genes were then obtained by conventional PCR methods.

Figure 14: Nucleic acid electrophoresis results. Pikm is about 3200bp long, so we used a 15000bp ladder as a reference, and found that all the 4 lanes at the corresponding location had results. After sequencing, we can be sure that we have successfully obtained the Pikm sequence

Acquisition of pGreen plasmid skeleton

Thanks to igem2022 member Yan Zhijian for providing the pGreen plasmid, which is a Ti plasmid of Agrobacterium, which is compatible with and can be well expressed in Agrobacterium.

Acquisition of nanobody (named enhancer) sequence

We purchased the sequence of enhancer from Genscript Biotech Corporation. It was originally a protein that enhance the fluorescence intensity of GFP, but later it was found to be a nanobody that specifically binds to GFP, which fits our design

Acquisition of GFP: experimental materials from this laboratory

Acquisition of pUC18-TMV plasmid

We purchased this plasmid from the Hangzhou TOP Biotechnology Itd company.

Plasmid construction

Construction of pGreen-pikm2/GFP plasmid

We use the pcr method for pGreen linearization, and then the seamless cloning method connects the fragments with the skeleton.

Figure 15: PCR results of pGreen-pikm2, pGreen-GFP colonies

a) the results of colony PCR were from the Agrobacterium transforming pGreen-pikm2 plasmid, and the PCR detection area was about 7000 bp at both ends of pikm-2, and 15000 bp Marker was used as a reference, and all 8 lanes had bright bands.

b) The figure shows the results of colony PCR of the Agrobacterium transformed pGreen-GFP plasmid, the PCR detection area is at both ends of the GFP sequence, a total of about 700 bp, using 2000 bp Marker as a reference, and all 8 lanes have bright bands.

Modification of pGreen-pikm1 (enhancer)

The pGreen-pikm1 plasmid is constructed by the above method, the plasmid is linearized oppositely from the two sides of the ID sequence, and then the fragment and the skeleton are connected by seamless cloning.

Figure 16: PCR results of pGreen-pikm1 (Enhancer) colonies

The figure shows the results of colony PCR of Agrobacterium that transformed the pGreen-pikm1 (Enhancer) plasmid. The PCR detection area was the Enhancer region, a total length of 351 bp, and we use the 2000 bp Marker as a reference. Obvious bands are seen in the region of corresponding size.

Construction of Pgreen-TMV-GFP

Since we have not been able to obtain the corresponding nanobodies of some pathogens (such as viruses, etc.), we can only display GFP protein outside the coat protein of TMV, which is equivalent to making GFP a pathogenic effector, allowing the modified immune receptor (which targets at GFP) can specifically recognize GFP and complete immunity

Initially, our strategy was to insert GFP into the coat protein of TMV in the pUC18-TMV plasmid (before the stop codon [8]) firstly, but when we transformed the constructed plasmid to E. coli DH5α and did sequencing, we found that the 3' UTR at the end was lost, and we speculated that recombination of LTR (Long terminal repeat) may have occurred. Therefore, we had to change the strategy. We firstly construct pGreen-TMV (When transformed into and amplified in Agrobacterium, this is a plasmid skeleton that have less ability to conduct LTR recombination with Agrobacterium). In terms of transformation and amplification, we adopt a more stable E. coli stbl3 to avoid the possibility of LTR recombination. After all of that, we then insert GFP into the constructed pGreen-TMV and do the sequencing to make sure we got the right sequence of TMV-GFP.

Agrobacterium transient transient expression on tobacco

PART2 Experiments for results

GFP alone as effector to stimulate immunity

Material

- Agrobacterium containing pGreen-pikm1 (Enhancer) plasmid→ later referred to as pikm-1 (Enhancer)

- Agrobacterium containing pGreen-pikm2 plasmids → referred to as pikm-2

- Agrobacterium containing pGreen-GFP plasmids → referred to as GFP

Design

- Experimental group→ pikm-1 (enhancer) + pikm-2 + GFP

- Control group→ pikm-1 + pikm-2

- Positive control → GFP

- Negative control → Agrobacterium empty

Mark the injection area with a circle, inject according to the method mentioned above, and then placed in a greenhouse under suitable conditions for culture, and the phenotype was recorded one week later

Result detection

Phenotype observation

Tobacco was irradiated with 488 nm excitation light to observe the removal of GFP fluorescence

It can be seen that compared with the positive control, the negative control has a clear fluorescence clearance.

Figure 17: Phenotypic observations after transient tobacco

Negative control:
Injection of Agrobacterium with empty plasmid;

Positive control:
GFP injection;

Experimental group:
NLR (enhancer) + GFP

Immune clearance

Next, we will quantify the clearance of GFP (Total protein fluorescence intensity & WB) at the molecular level and confirm that it is cleared by plant’s immune system (ROS detection)

Protein Fluorescent Analysis

After one week of injection, we took leaf tissue from the injection area, extracted a fixed amount of total protein, and measured its fluorescence intensity.

Result shows that compared with NLR*+GFP group, GFP only and NLR+GFP had significantly higher fluorescence intensities, which indicates that our modified NLR* did produce effective clearance of GFP.

However, NLR+GFP also has a tiny reduction in fluorescence intensity compared to GFP only. We believe that this does not prove that unmodified NLR also has the ability to clear GFP, but can be explained by the increase in spontaneous dimerization after NLR overexpression [9].

Figure 18: Measurement of Fluorescence intensity in the leaves

Note

*P<0.05, ****P<0.0001, Positive control: GFP only

NLR represents the original unmodified immune molecules.

NLR* represents an artificially modified molecule whose ID sequence has been replaced with a nanoantibody Enhancer targeting GFP

WB

Since the Enhancer nanobody has the property of increasing fluorescence intensity upon binding GFP, we were not sure of the immune activation function of the artificially modified NLR only by fluorescence intensity.

Thus, we further designed WB experiment for GFP to directly measure the change of GFP content and thus we can measure the immune response more rigorously.

Figure 19: WB results for GFP

According to the information shown in the figure above, there was a more significant decrease in GFP content in the NLR*+GFP group compared to the GFP only group (positive group) and the empty vector group (negative group).

ROS Measurement

Reactive oxygen species (ROS) signaling plays an important role in the innate immune response of plants, can directly inhibit the growth of pathogens, and can also participate in the disease resistance process as a signaling molecule, and its rapid production is an important marker of plant defense system activation. ROS production is mainly mediated by NADPH oxidase, also known as Respiratory Burst Oxidase homologous Proteins (RBOHs) in plants, so plant ROS can be detected by elisa to determine the status and strength of the plant's immune response. We used the ROS kit (MM-43700M1) to experiment on plant tissues after 1 week of injection

Figure 20: Measurement of ROS content in the leaves

Note

*P<0.05, ***P<0.001,****P<0.0001

NLR represents the original unmodified immune molecules.

NLR* represents an artificially modified molecule whose ID sequence has been replaced with a nanoantibody Enhancer targeting GFP.

As shown in the table and figure above, it can be seen that compared with normal tissue, 3 groups respectively injected with empty plasmids (1.807±0.067 vs. 1.1.716±0.004, P<0.05), GFP (1.803±0.045vs.1.716±0.004,P<0.05),and unmodified NLR+GFP(1.835±0.057 vs. 1.716±0.004, P<0.001) all had significant ROS elevations and remained similar levels of immunity. This can be explained by the plant's own immune response against Agrobacterium

The experimental group injected with the modified immune receptor + GFP showed a higher immune response intensity than the uninjected leaves (2.014±0.043 vs. 1.716±0.004, P< 0.0001). Even compared with other experimental groups (empty plasmid, GFP, unmodified NLR), there was a significant increase in ROS, indicating that our modified NLR by Agrobacterium transiently exactly enhanced the immune response of plant

TMV-GFP as pathogen to stimulate immunity

Testing of the promoter

Selection and acquisition of promoters

Characterization of the promoter

Result

We observed colonies and bacterial fluid under a fluorescence microscope using blue light as the excitation source.

Figure 21: Photographs of colonies under a 4x objective (control on the left and Agrobacterium with GFP protein on the right).

Figure 22: A photograph of the bacterial solution on a slide under a 10x objective. (control on the left and Agrobacterium with GFP protein on the right)

Gray values representing relative fluorescence intensity were obtained using a multifunction imager at consistent exposure times. Examination revealed significant differences. The basal fluorescence of the culture was subtracted in this analysis. These disparities are discernible between different promoters or distinct proteins.

Figure 23: FIG.1 Agrobacterium tumefaciens without fluorescent protein; FIG.2 50Spro-GFPuv; FIG.3 Pvbp2-GFPuv;

Figure 24: Comparison of relative fluorescence intensity between the different promoters, 50Spro-GFPuv was very significantly different from the control group (p<0.001), and Pvbp2-GFPuv was significantly different from the control group (p<0.05).

Selection of phototoxic proteins

Our engineered strains are mainly used in plants, and plants need sunlight to grow, so we chose sunlight as the way to initiate suicide.

KillerRed

miniSOG

We incorporated the more observable protein miniSOG as a complementary measure. This phototoxic protein, derived from LOV protein, produces toxic substances under blue light excitation to kill bacteria [14]. It also emits green fluorescence for easy observation. To facilitate the use of this protein in Agrobacterium, its codon has been optimized.

Figure 25: The relative fluorescence intensity of miniSOG and the control group was very significant (p<0.001)

Directed evolution of promoters

Promoter analysis and mutation

We extracted the 40bp sequence before each Agrobacterium tumefaciens gene in NCBI as the key sequence of promoters and analyzed whether each base was conserved. Then we selected two of the sequence with a total of 12 bases to be mutated using degenerate primers to construct a mutation library. GFPuv was used as a reporter gene for the screening. In total, we obtained thousands of colonies with different fluorescence intensities and extracted four of them for amplification. Surprisingly, sequencing results showed that three of the brightest mutations in our screened strains were the same mutation sequence. We called it 50Spro-e1. This proved that we had successfully evolved stronger Agrobacterium promoter and stably expressed it.

Figure 26: a. Nonconserved sites in Agrobacterium promoter near the -10 and -35 regions; b. The sequencing results of Agrobacterium with higher brightness.

fluorescence detection

we observed the colonies under the UV light, extracted the protein, and detected their fluorescence intensities using a fluorescent microplate reader. This achievement will not only enhance the control of kill switches, but also contribute to the iGEM community by providing stronger constitutive promoters for future teams using Agrobacterium.

Figure 27: The protein fluorescence observation under ultraviolet light

Figure 28: The results of protein fluorescence quantitative detection which were standardized by OD600 of bacterial solution, and empty plasmid was used in the control group. All experimental groups showed extremely significant differences from the control group(p<0.0001).

Construction of plasmid vectors

The plasmid we chose was pGREENll. all of our vectors were accomplished using Gibson Assembly [15]. Given the unique ability of Agrobacterium species to deliver T-DNA to plants, we strategically designed the biosafety module to be located outside the LB and RB sequences. This was done to prevent the inadvertent transfer of genes to plants.

After experimenting with various sites, we ultimately selected to insert the fragment between the origin of replication and the kanamycin resistance gene. This decision was informed by a range of factors, including efficiency and stability.

To ensure the success of our transformations, we employed Sanger sequencing for confirmation. This allowed us to verify that the strains had been positively transformed, thereby validating our experimental approach.

Figure 29: The plasmid we constructed (take 50Spro-GFPuv as an example)

Illumination experiments

Due to the weak expression of inducible promoter Pvbp2 in promoter characterization experiments, and a high quantity of acetosyringone has toxic effects [16]. We finally selected 50Spro as the promoter to verify the suicide effect of phototoxic proteins. and sunlight was selected as the light source for simulating the plant planting environment.

In the experiment, we used ice packs and foam to control the plate temperature below 30 degrees. After different times of light experiments, the bacterial culture plates were incubated at 28 °C in the dark for 48 to 72h. Then count the number of colonies, and the control group treated with empty plasmid in the dark light was used as a benchmark.

Results:

Our experiments with the bacterial solution under light exposure revealed that the bacteria did not exhibit significant reduction. We hypothesize that the accumulation of toxins within the bacteria may have a buffering effect on the liquid environment, which could be beneficial for the preservation and transportation of our engineered bacteria.

In contrast, our plate light experiment demonstrated a noticeable reduction of engineered bacteria under natural light irradiation, regardless of whether the conditions were sunny or cloudy. However, the rate of bacterial reduction varies under these two conditions. Given the variability of sunlight in cloudy weather, we opted to use the cumulative light intensity (lux) over time as a variable for our analysis.

Interestingly, we observed that even in the absence of any promoter ( ‘empty vector’ strain), there was still a degree of clearance under light exposure. But this effect was less pronounced than in Agrobacterium containing phototoxic proteins. It suggests that the presence of phototoxic proteins can further reduce their viability.

Figure 30: Results of light experiments of Agrobacterium with KillerRed V.S. empty plasmid (on a sunny day)

Figure 31: Results of light experiments of Agrobacterium with empty plasmid V.S. minisog (On a cloudy day)

Furthermore, we conducted plate light experiments at a 1000x bacteria concentration. These experiments showed that when the cumulative lux over time reached 3*10^6 (30 min of light on a sunny day), the effect of bacterial clearance could still be achieved.

Figure 32: 1000x concentration scavenging effect of KillerRed (dark control, 10 min, 20 min, 30 min))

Further experiments

In our final experiment, we analyzed the promoter sequence of Agrobacterium tumefaciens and attempted to enhance it through directed evolution by introducing mutations. However, due to time constraints, this experiment was not completed. In future experiments, we aim to develop a stronger constitutive promoter to enhance the safety of our engineered bacteria.

Because of the weather, our light experiment could not be repeated enough. Future experiments are needed to validate the significant killing effect of our phototoxic protein. In addition, we could try the promoter Pvbp2, to verify whether this part is indeed inducible by acetosyringone as a kill switch.