Engineering Success

1. DESIGN

Synthesizing the Mitochondrial GCL

    The synthesis of the mitochondrial Durio zibethinus Glutamate-Cysteine Ligase (mDzGCL) protein used in this project was done through replacing the chloroplast transit peptide found in D. zibethinus Glutamate-Cysteine Ligase (DzGCL) with a mitochondrial transit peptide found in Arabidopsis thaliana succinate dehydrogenase (SDH).


Plant and Gene Selection

    DzGCL was selected as the ideal candidate GCL protein for relocalization due to its remarkable attributes observed in D. zibethinus . This unique plant species displays significantly elevated levels of glutathione (GSH)[1] in its ripe fruit pulp much higher than that in other plants.[2] This distinct characteristic suggests the DzGCL as a candidate enzyme with possibly high activity. GCL serves as the rate limiting enzyme in the production of GSH, the primary antioxidant.


    A. thaliana SDH, on the other hand, served as our model for obtaining the mitochondrial transit peptide. A. thaliana was chosen as our model plant due to its well-established status in plant biology research, with its genetic characteristics widely studied and understood.[3] Furthermore, given that SDH is a fundamental component within the mitochondrial electron transport chain, the team hypothesized that it must contain a mitochondrial transit peptide facilitating its localization to the mitochondria. Through literature review, this hypothesis was found to be valid as Figueroa et al[4], successfully identified the transit peptide in A. thaliana SDH and validated its role in localizing SDH to the mitochondria.


Gene Modification

    The first step was to identify the amino acid sequence of DzGCL from the NCBI database (accession number: XP_022738907.1). After obtaining the amino acid sequence, the natural chloroplast transit peptide of DzGCL must be predicted in order to be deleted and replaced accordingly. The chloroplast transit peptide within DzGCL was predicted through the TargetP-2.0 software[5](Fig. 1), utilizing its deep learning algorithm trained on a diverse dataset of biological sequences, including known chloroplast transit peptides, to recognize indicative features of sorting signals.[6]


Figure 1. Prediction of chloroplast transit peptide in DzGCL by TargetP-2.0, where all of the amino acids from the left peak and onwards to the left boxed in gray were predicted to be part of the chloroplast transit peptide.

    After deleting the chloroplast transit peptide from DzGCL, the next step was to identify the amino acid sequence for A. thaliana SDH from NCBI database (accession number: CAC19855.1), in which the mitochondrial transit peptide has already been identified as amino acids positions 1 to 29 by Figueroa et al.[4] The identified mitochondrial transit peptide was subsequently copied and inserted into the DzGCL in the stead of the deleted chloroplast transit peptide, yielding mDzGCL as shown in Figure 2. The amino acid sequences were finally translated back to nucleotides to obtain mDzGCL gene utilized in the plasmid.

Figure 2. Diagram of the replacement of the transit peptide in the GCL amino acid sequence

Evaluation of the Modified Protein through In Silico Analysis

Localization

    After acquiring mDzGCL, In silico subcellular localization prediction of the modified GCL was performed through deep learning using TargetP-2.0 to confirm the success of the modification in producing a mitochondrial GCL. As shown in the table in Figure 3, it was predicted that mDzGCL will localize in the mitochondria with a 92.25% probability that its transit peptide is mitochondrial, confirming the success of the modification in relocalizing the DzGCL to the mitochondria.


Figure 3. Subcellular localization prediction of the modified mDzGCL by TargetP-2.0

Function

    After confirming the localization of the mDzGCL to be mitochondria, the next step was to confirm that the enzymatic function of the modified GCL (mDzGCL) was not disturbed by the modification of the transit peptide and was still the same as that of the unmodified GCL (DzGCL).


    Having the FASTA sequences of the chloroplastic DzGCL and the mitochondrial mDzGCL, the first thing done was to acquire predictions of their maps through the HMMER program,[7] HMMScan, which utilizes the hidden Markov model and statistical analyses to find matches in the sequence with known functional domain sequences in a database. In this case, the Pfam protein database[8] was used. The maps outputted by HMMScan of DzGCL and mDzGCL are displayed in Figure 4 and Figure 5, respectively. Both figures show that HMMScan was able to identify the signature glutamate-cysteine ligase family 2 functional domain of a GCL represented by the brown section in the maps. Furthermore, the length of the functional domains calculated by finding the difference in the end and the start amino acid positions were found to be the same. Thus, the modifications in the mDzGCL were shown to have not disturbed the functional domain of GCL.


Figure 4. Map of DzGCL (unmodified) predicted through HMMScan (HMMer) showing the signature glutamate-cysteine ligase family 2 functional domain of a GCL.

Figure 5. Map of mDzGCL (modified) predicted through HMMScan (HMMer), which also shows the signature glutamate-cysteine ligase family 2 functional domain of a GCL.

    To further verify the integrity of the enzymatic function of GCL in mDzGCL, 3D structures of DzGCL and mDzGCL were predicted using AlphaFold2, a software that combines empirical knowledge on protein structure with a deep-learning algorithm, leveraging evolutionary data through multiple-sequence alignment. This approach often yields protein models of comparable accuracy to experimentally determined structures.[9,10] The predictions obtained from AlphaFold2 were then visualized in PyMOL v2.5.5 (by Schrödinger) for analysis. Shown in Figure 6, the 3D structures of DzGCL and mDzGCL, respectively, are observably similar except for the tail, in which the difference is due to modifications in the N-terminal when replacing the chloroplast transit peptide in mDzGCL with mitochondrial one. The similarity between the two 3D structures is better displayed in Figure 7, where the DzGCL (green) and mDzGCL (red) are overlaid, and a complete overlap is observable. Therefore, since the structure of the modified GCL (mDzGCL) closely resembles that of the unmodified GCL (DzGCL), the enzymatic function should also remain approximately the same.


Figure 6. The 3D structures of DzGCL (unmodified) mDzGCL (the modified DzGCL) predicted through AlphaFold2. The functional domain of DzGCL and mDzGCL are highlighted in green and red, respectively.

Figure 7. 3D structures of DzGCL (unmodified) and mDzGCL predicted through AlphaFold2, overlaid, showing complete overlap.

Vector Construction


Figure 8. mDzGCL in pCAMBIA1301 Plasmid Construct

Figure 9. GFP in pCAMBIA1301 Plasmid Construct

    In total, two plasmids were used in the project: one is the mDzGCL in pCAMBIA1301 to infiltrate plants with the mDzGCL gene designed by the team and the other is the GFP in pCAMBIA1301 as a positive control


    The pCAMBIA1301 mDzGCL construct (Fig. 8) contains the the DzGCL gene acquired from D. zibethinus and the gene for the mitochondrial transit peptide from A. thaliana.. This transit peptide allows for localization of DzGCL into the mitochondria instead of the chloroplast. Additionally the plasmid was constructed using the pCAMBIA1301 (AF234297.1) template, which includes antibiotic resistance genes. Specifically, it features a kanamycin resistance gene to select bacteria, as only transformed bacteria will survive in environments with this antibiotic. Additionally, a hygromycin marker enables the selection of transformed plants. The 35s gene promoter in this construct enhances the expression level of the mDzGCL within the infiltrated plants, aligning with the project's aim to make the GSH pathway more efficient.


    The GFP construct (Fig. 9) contains Green Fluorescent Protein which makes the infiltrated plant grow when exposed to blue to UV range light. This construct does not interfere with the GSH Pathway so it can serve as a positive control for comparing mDzGCL-infiltrated plants under different environmental conditions. Similarly to mDzGCL, it also incorporates the 35s promoter. Other plasmid components also mirror those of the mDzGCL plasmid, ensuring a precise comparison of the impact of the mitochondrial transit peptide and GCL on plant stress response.


    All the components mentioned above were acquired from the NCBI database (https://www.ncbi.nlm.nih.gov/) and put into one sequence. This sequence was then sent to the Biomatik Company (USA) for gene synthesis and its insertion into the vector.

2. WET LAB EXPERIMENTS

Approach

     Agroinfiltration approach was chosen in order to localize GCL into mitochondria in plants. Agrobacterium tumefaciens is a phytopathogenic bacterium that infects plant cells, leading to crown gall disease by transferring part of its genetic material into the plant genome. This type of bacterium is now modified to be harmless for use as a key tool in biotechnology in order to transfer genes of interest into plant cells for a variety of biological uses. For localizing mitochondria, A. tumefaciens was transformed by the construction of mitochondrial transit peptide and Durio zibethinus GCL gene and subsequently agroinfiltrated into Nicotiana benthamiana leaves. As N. benthamiana is one of the plants with the most mapped genomes, so it would be easier to interpret results. Moreover, many studies have shown that N. benthamiana is a good host system for studying the functional characterization of many genes, making it an efficient system for infiltration.


Procedure

1. Plant and Material Preparation

    The seeds of N. benthamiana were scattered in pots containing peat moss and were grown under controlled conditions. The plants were cultivated in an environment maintained at 25°C, with a light-dark cycle of 16 hours of illumination followed by 8 hours of darkness (provided by artificial lights at an intensity of 4,500 lux). Two-week-old plants were successfully grown and then planted individually into new pots and were grown under similar conditions.

N. benthamiana plants growing under controlled conditions at 25°C, with a light-dark cycle of 16 hours of illumination followed by 8 hours of darkness (provided by artificial lights at an intensity of 4,500 lux)

2. Bacteria Transformation

    The construct and control (GFP in pCAMBIA1301) vector were transformed into A. tumefaciens strain GV3101 and cloning was confirmed with Colony PCR. GFP was chosen as it has no effect towards the GSH pathway.

    After obtaining the colonies of A. tumefaciens, certain colonies were selected (Fig. 10a). These colonies were swirled into a mixture of PCR solutions (buffer, MgCl2, dNTP, Taq DNA Polymerase (Thermo Scientific, UK), Nuclease Free H2O, primers) and put into a thermal cycler.


    To verify the presence of the mDzGCL in plasmid, specific primers were designed to anneal to both ends of the target DNA sequence. Figure 10b depicts a successful gel electrophoresis result as the primers are specific to the band lengths of the mDzGCL with mitochondria in the pCAMBIA1301 plasmid.


Figure 10. (A) Bacteria colony growth on an agar medium with antibiotic, only transformed A. tumefaciens should be resistant to. and (B) Gel electrophoresis of the colony PCR for mDzGCL-pCAMBIA1301 plasmid.

3. Plant Transformation

    A. tumefaciens containing the construct or control was cultured at 30°C for 48 hours, then pelleted and resuspended in an OD600 = 0.5 in MM buffer (10 mM MES pH 5.6, 10 mM MgCl2). For agroinfiltration, the A. tumefaciens solution containing either gene or a control vector was mixed with the A. tumefaciens solution harboring the gene encoding the silencing inhibitor protein p19 at a ratio of 1:1. Thereafter, 100 mg/l acetosyringone was added, and the mixed culture solution was incubated at room temperature for 3 hours under dark conditions. The solution was then used to infiltrate the surface of three individual leaves per N. benthamiana plant using a needleless 1-mL syringe. For each construct, 6-week-old plants were used.11 At least 4 leaves, considered as 4 replicates, were infiltrated for each construct.

Incubation of culture solution and infiltration of N. benthamiana

4. Conditions/Treatments

    The transiently overexpressed mDzGCL and control in N. benthamiana were treated under normal, drought, heat, and salt stress conditions.


  1. In normal conditions, the plants were grown under 25°C receiving 16 hours of light and 8 hours of darkness while being watered daily.

  2. Drought conditions were primarily simulated by not watering the N. benthamiana plants, starting from the first day. This was done immediately due to the previously wet soil taking a set amount of time to fully dry. The plants were still grown in the same temperatures and light/dark conditions as the normal conditions.

  3. For the heat stress conditions, plants were grown under 25°C receiving 16 hours of light and 8 hours of darkness while being watered every day. After 5 days, the N. benthamiana were placed in conditions under 30°C with 12/12 hours of light/dark. mDzGCL will not be expressed until after 4 days, so variables are not introduced until after that amount of time.

  4. Salt stress conditions consisted of the N. benthamiana plants growing under 25°C with 16/8 hours of light/dark and daily watering. After 5 days, 30mL of 150 mM of NaCl was added to the soil, and an additional 10mL every 2 days for a total of 11 days.

5. γ-EC Content Analysis

    The construct and control were agroinfiltrated on N. benthamiana leaves. After 5 days, the infiltrated leaves were collected, immediately frozen in liquid nitrogen, ground into a fine powder, and freeze dried, which was then used for the metabolite analysis.

The γ-EC and GSH were extracted and analyzed by HPLC.12

3. RESULTS

Data Collection & Analysis

    To understand the role of mDzGCL on γ-EC production, the transient overexpression of this gene on N. benthamiana leaves was done. γ-EC content significantly increased by 1.6 times compared to the control group (Fig. 11). This result indicates that mDzGCL plays a role in γ-EC synthesis. Even though DzGCL was modified to be localized and overexpressed in the mitochondria, its function is still the same, which is to regulate the amount of γ-EC in the cell. The pathway of mDzGCL is only shortened in this project but still adheres to the same function.

Figure 11. γ-EC contents of N. benthamiana leaves that have been agro-infiltrated with mDzGCL in comparison to the control group (GFP). An asterisk above the bar indicates a significant difference between samples (Student's t-test, *: p < 0.05).

Raw data can be accessed here.

    Figure 12 demonstrates that the overexpression of mDzGCL did not affect the growth of any plants, as both mDzGCL and GFP groups can be seen growing at similar rates under consistent conditions.

Figure 12. N. benthamiana leaves before and after agroinfiltration with mDzGCL under normal conditions. The leaves with tags represent the infiltrated leaves. All the N. benthamiana pictures can be accessed through the google sheets above.

    Figure 12 demonstrated that the overexpression of mDzGCL did not affect the growth of any plants, as both mDzGCL and GFP groups can be seen growing at similar rates under consistent conditions.

Figure 13. The phenotype N. benthamiana leaves are agro-infiltrated with mDzGCL compared to the control (GFP) under drought stress conditions. The leaves with tags represent the infiltrated leaves. pictures can be accessed through the google sheets above.

    As shown in Figure 13, both N. benthamiana in the control and mDzGCL transiently overexpressed line were exposed to drought stress for the same amount of time. The mDzGCL transiently overexpressed line visually seems to be in better condition in terms of stress reduction when compared to the control. The wilting of leaves starts at around day 4 after treatment of the control whereas the mDzGCL transiently overexpressing leaves do not show signs of wilting until around day 6 after treatment. These results indicate that mDzGCL is able to help plants resist drought stress conditions for around 48 additional hours.

Figure 14. The phenotype N. benthamiana leaves are agro-infiltrated with mDzGCL compared to the control (GFP) under heat stress conditions. The leaves with tags represent the infiltrated leaves. The green highlight indicates the day stress is applied. pictures can be accessed through the google sheets above.
Figure 15. The phenotype N. benthamiana leaves are agro-infiltrated with mDzGCL compared to the control (GFP) under salt stress conditions. The leaves with tags represent the infiltrated leaves. The green highlight indicates the day stress is applied. Pictures can be accessed through the google sheets above.

    On the other hand, there was no observable difference in phenotypes between the controls and transiently overexpressed plants under heat and salt stress conditions (Fig. 14 and 15). However, by comparing just physical characteristics, it is not enough to prove the function of mDzGCL on stress response. Further investigation would be needed to reliably confirm the effect of mDzGCL on heat and salinity.

Conclusion

    The results on the agroinfiltrated N. benthamiana leaves reflect the effectiveness of modifying GCL to be localized to the mitochondria resulting in an improved stress response. The data shows that the overexpression of mDzGCL did not affect plant growth but can support the reduction of ROS (Fig.12 and 13). The enabling of a faster stress response for N. benthamiana with transiently overexpressed mDzGCL was effective in delaying the negative effects of drought. This is especially impactful within the context of Thai agriculture, where low crop yields frequently follow intensive droughts. This affects both farm owners and the farmers themselves. As Thai farm owners and farmers rely on their crop stability for money, unpredictable weather can have a significant negative impact on their means of living.


    Thus, Team Thailand-RIS has successfully modified DzGCL to shorten the GSH pathway by turning the protein locality to mitochondrial in hopes of improving plant tolerance to external stressors. The team has listed the gene (mDzGCL) in the part registry as part BBa_K4629001.

Figure 16. Phenotype changes in N. benthamiana leaves under drought stress from days 0-6 after infiltration. The mDZGCL transiently overexpressing line wilts considerably slower than GFP transiently overexpressing line.

    In early September, the Thailand-RIS Human Practices Team surveyed Thai farmers and farm owners who cultivate durians in Chanthaburi, a major agricultural province in Thailand. The majority of those surveyed were most concerned about drought out of all the conditions tested within the project. In Figure 13 and 16, data shows that the modified GCL pathway transit is able to delay plant wilt by 48 hours after 4 days. An additional 2 days before plant wilt occurs is hugely beneficial for Thai farmers. It allows them more leeway to react by acquiring a sufficient water supply and implementing necessary measures to protect their crops from wilting further. This extra time can make a substantial difference in preserving their livelihoods and ensuring a more stable agricultural yield. Due to the nature of this project and the reliance of Thailand's economy and community on agriculture, Thailand-RIS believes that the results are noteworthy and can create positive impacts for farmers and consumers alike.

Human practices team interviewing farmers and Chanthaburi governor.

Evaluation

Strengths

    The stress tests featured 4 total trials for each type of stress (2 trials for GFP (control) and 2 trials for mDzGCL), with each construct including 6 infiltrated leaves for observation. This allowed for the results to have greater reliability.


    Throughout the lab duration, the hypothesis of mDzGCL improving ROS response was shown to be valid on 2 different plant species including N. benthamiana and Solanum lycopersicum (more details below). This allowed for better understanding of the modification and its potential drawbacks. Having an accurate understanding of the limitations of this project allows for a clearer prediction of the possible implementations of GCL localized to the mitochondria.

Agroinfiltration of Solanum lycopersicum.

Limitations

    When first propagating the N. benthamiana plants for stress treatments, two-week-old sprouts were transferred to new pots. These younger N. benthamiana sprouts are weaker and more sensitive, and a large portion of the plant roots were damaged due to packing in the soil with too much pressure. Our lab advisor was able to identify the N. benthamiana with abnormal leaves, a sign of poor health and unsuitability, and then removed them from the sample. This small mistake cost roughly half of the original plant stock due to broken roots. Thankfully, a sufficient number of plants were retained for the stress treatments. However, it limited the number of stress responses that could be tested, and there was not enough time to regrow more N. benthamiana for additional stress treatments.


    The observation of N. benthamiana phenotypes was not sufficient to conclude the effects of mDzGCL under high salinity and heat stress. If there was more time available, metabolite and gene expression analysis could have been done for better confirmation of the phenotypic results. More experiments are needed to understand the exact mechanisms which caused the result of this project.


    While waiting for the N. benthamiana plants to mature, the Thailand-RIS team additionally experimented on a minor subproject with the same A. tumefaciens in S. lycopersicum, a dwarf cultivar of tomatoes. This functionally performed as training for wet lab team members before testing on N. benthamiana. This was done in hopes of observing the effects of mDzGCL on the nutrient levels and ripening of S. lycopersicum fruits. During the agroinfiltration process for S. lycopersicum, uncertainties arose as to when exactly to inject the tomatoes as visually observing the ripeness of S. lycopersicum was difficult. Regrettably, the effect of different stages of tomato ripeness on the fruit's pH levels and, subsequently, the expression of proteins within it was overlooked. pH levels can impact enzyme activity, cellular environment, protein structure and amino acid charges, all of which can change the expression of proteins. This resulted in inconclusive metabolite analysis results.

Phenotype data collection and preparation for metabolite analysis in S. lycopersicum.

    As the modification of S. lycopersicum with mDzGCL was a subproject and results were unclear, hence it was not pursued further. The ripeness of the S. lycopersicum fruit requires another level of complexity that is difficult to analyze with the currently available resources and time of the team. Additionally, fruits may not be a good model to study environmental stresses: fruits have shorter life spans, go through separate physiological processes in comparison to the plant itself, and have nutrient reserves which affect stress responses. As the main focus of this project was studying a period of time sufficient enough to create implications for long term effects on plant stress responses, S. lycopersicum would not constitute compelling evidence for this project.

N. benthamiana and S. lycopersicum kept in their respective trays saved for agroinfiltration.

References

  1. Singcha P, Khaksar G, Sirijan M, Sirikantaramas S. Durian (Durio zibethinus L.) fruit: A superior dietary source of natural glutathione and γ-glutamylcysteine. SSRN. 2023; 4566631. doi:10.2139/ssrn.4566631
  2. Pinsorn P, Oikawa A, Watanabe M, Sasaki R, Ngamchuachit P, Hoefgen R, Saito K, Sirikantaramas S. Metabolic variation in the pulps of two durian cultivars: Unraveling the metabolites that contribute to the flavor. Food Chemistry. 2018; 268:118-125. doi:10.1016/j.foodchem.2018.06.066
  3. Meinke DW, Cherry JM, Dean C, Rounsley SD, Koornneef M. Arabidopsis thaliana: A model plant for genome analysis. Science. 1998 Oct 23;282(5389):662-682. doi:10.1126/science.282.5389.662
  4. Figueroa P, Leon G, Elorsa E, Holuigue L, Jordana X. Three different genes encode the iron-sulfur subunit of succinate dehydrogenase in Arabidopsis thaliana. Plant Molecular Biology. 2021;46:241-250. doi:10.1023/A:1010612506070
  5. Emanuelsson O, Brunak S, von Heijne G, Nielsen H. Locating proteins in the cell using targetp, signalp and related tools. Nature Protocols. 2007;2(4):953-971. doi:10.1038/nprot.2007.131
  6. Armenteros JJ, Salvatore M, Emanuelsson O, Winther O, Von Heijne G, Elofsson A, Nielsen H. Detecting sequence signals in targeting peptides using deep learning. Life Science Alliance. 2019;2(5). doi:10.26508/lsa.201900429
  7. Finn RD, Clements J, Eddy SR. HMMER web server: interactive sequence similarity searching. Nucleic Acids Research. 2011;39(suppl_2):W29-37. doi:10.1093/nar/gkr367
  8. Bateman A, Coin L, Durbin R, Finn RD, Hollich V, Griffiths-Jones S, Khanna A, Marshall M, Moxon S, Sonnhammer EL, Studholme DJ. The Pfam protein families database. Nucleic Acids Research. 2004;32(suppl_1):D138-141. doi:10.1093/nar/gkh121
  9. Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, Tunyasuvunakool K, Bates R, Žídek A, Potapenko A, Bridgland A. Highly accurate protein structure prediction with AlphaFold. Nature. 2021;596(7873):583-589. doi:10.1038/s41586-021-03819-2
  10. Evans R, O&aposs;Neill M, Pritzel A, Antropova N, Senior A, Green T, Žídek A, Bates R, Blackwell S, Yim J, Ronneberger O. Protein complex prediction with AlphaFold-Multimer. bioRxiv. 2021:2021-10. doi:10.1101/2021.10.04.463034
  11. Khaksar G, Sangchay W, Pinsorn P, Sangpong L, Sirikantaramas S. Genome-wide analysis of the DOF gene family in durian reveals fruit ripening-associated and cultivar-dependent DOF transcription factors. Scientific Reports. 2019;9(1):12109. doi:10.1038/s41598-019-48601-7
  12. Pinsorn P, Oikawa A, Watanabe M, Sasaki R, Ngamchuachit P, Hoefgen R, Saito K, Sirikantaramas S. Metabolic variation in the pulps of two durian cultivars: Unraveling the metabolites that contribute to the flavor. Food Chemistry. 2018; 268:118-125. doi:10.1016/j.foodchem.2018.06.066
© 2023 - Content on this site is licensed under a Creative Commons Attribution 4.0 International license.
The repository used to create this website is available at gitlab.igem.org/2023/thailand-ris