Prologue

This year, SZU-China has brought a unique high-quality sweet tomato. So, how exactly does our team produce such a sweet tomato? On this page, we will present the starting point, basic principles and general experimental methods of the experiments implemented by the SZU-China design. After discussion among our team, we decided to divide the experiment into three modules: Expression of Sweet Protein, Sweetness Detection and Sweet Protein Localization. These three experiments are progressive. First, starting from the expression of sweet protein, we will describe the experimental optimization in the process of producing sweet protein. Then, to ensure the produced sweet protein is effective, we conducted an experiment on detecting the effectiveness of sweet protein. Finally, we will describe the Localization experiment of sweet protein. The ultimate goal of these efforts is to ensure that tomatoes can produce as much sweet protein as possible.

Optimization of sweet protein expression module

A trial of prokaryotic expression

Before achieving our ultimate goal, we conducted a series of tentative experiments in the hope of increasing the production of sweet proteins as much as possible. We first chose the prokaryotic system Escherichia coli as the chassis and expressed the constructed plasmid in it. (Fig 1)

To verify the ability of the plasmid we constructed and the heterologous expression of sweet proteins, we conducted expression in a prokaryotic system. As the most common prokaryote in the research field, Escherichia coli has many advantages: clear genetic background, simple culture with low cost, fast growth rate, high protein expression level and mature operation technology. Therefore, we first chose Escherichia coli as the chassis for trial expression of sweet proteins.

However, we found that although both Thaumatin and Brazzein were expressed in Escherichia coli, the expression levels were very low (for details, please see the Results webpage). This may be due to the limitations of the prokaryotic expression system of Escherichia coli and the formation of inclusion bodies^[1]. In addition, we also found that extracting sweet proteins from Escherichia coli requires complicated processes and the production lacks stability. After consulting a large number of literatures and considering various advantages of expressing sweet proteins in plant systems such as safety and effectiveness (for details, please see the Description webpage), after discussion, we decided to try to use tomatoes as the chassis to ensure the expression effectiveness of exogenous sweet proteins and produce high-quality sweet proteins.

A trial of expression in the virus system

Subsequently, we commenced attempting to express sweet proteins in plants. Regarding the selection of plant chassis types, considering that sweet proteins will ultimately be utilized by us as food or supplementary food, we desire a plant chassis that can be more conveniently employed by us. Consequently, we opted for tomatoes as the chassis organism. Micro-Tom is a tomato variety. Its fruits bear a resemblance to cherry tomatoes in appearance and serve as a commonly employed plant chassis. Its plants are diminutive, typically around 10 to 20 centimeters in height. This renders it highly suitable for cultivation and growth in the limited space of laboratories. For instance, a large number of Micro-Tom tomato plants can be placed in spaces such as incubators and laboratory benches, facilitating large-scale experimental operations and research. It exhibits a short growth cycle. Generally, it only takes approximately 2 to 3 months from sowing to fruit maturity. This implies that our team can observe the entire growth and development process of tomatoes within a relatively short period and obtain experimental results rapidly, thereby enhancing research efficiency and expediting the progress of our experiments.

Additionally, Micro-Tom possesses a distinct genetic background. It has a relatively simple and well-defined genetic background. After years of research and cultivation, its genetic characteristics are relatively stable and there are relatively few genetic variations. This enables our team to more readily control and analyze the influence of genetic factors on the growth and development of tomatoes during experiments, reducing the uncertainty of experimental results. In terms of physiological research, Micro-Tom also boasts numerous advantages. For instance, its fruits are convenient for observation and analysis. The fruits of Micro-Tom tomatoes are small, yet their structure and physiological characteristics are similar to those of ordinary tomatoes. The fruits are easy to observe and handle, which aids us in researching aspects such as the fruit development process, ripening mechanism and nutritional components. In terms of experimental operations, Mico-Tom also showcases many advantages. For example, Micro-Tom has established a relatively mature genetic transformation system. Utilizing methods such as Agrobacterium-mediated transformation, exogenous genes can be conveniently introduced into Micro-Tom tomato plants to achieve operations such as overexpression, silencing, or editing of genes. This provides a powerful means for studying the function and mechanism of action of genes. Finally, the reproduction of Micro-Tom is extremely facile. Its reproduction method is simple. It can be reproduced by seeds or through asexual reproduction methods such as cuttings. The germination rate of seeds is high and the reproduction speed is rapid, capable of meeting the requirements for a large number of plants in experiments. Moreover, asexual reproduction can maintain the genetic characteristics of plants, facilitating control experiments and repeated experiments^[2][3][4]. Therefore, we selected Micro-Tom as the chassis for producing high-quality sweet proteins.

We remained uncertain as to whether the expression of sweet proteins in plants can truly avert protein misfolding. Following discussions, we determined to employ the transient expression method for verification in plants to ascertain whether tomatoes, as the chassis, can express sweet proteins and ensure correct folding. The TRV virus boasts a broad host range and can infect numerous important model plants and economic crops, including tobacco, tomatoes and Arabidopsis. This endows it with extensive applicability in the research and application of diverse plants. Researchers can utilize TRV to conduct research on gene functions and develop related biotechnologies in various plant species. Additionally, the operation of the TRV virus in transgenesis is extremely straightforward. It does not require complex equipment or technical conditions. Researchers can perform operations such as virus cultivation, preservation and inoculation in an ordinary laboratory environment, thereby reducing the difficulty and cost of experiments. Moreover, the TRV virus is highly modifiable. It can be modified through genetic engineering techniques, for instance, by inserting specific gene fragments or markers to better track the virus infection process and study the mechanism of gene silencing. This modifiability provides convenience for researchers to customize virus vectors according to specific experimental needs. Most significantly, the TRV virus causes relatively minor harm to plants. Compared with some viruses that are highly pathogenic to plants, TRV generally inflicts relatively mild damage on plants and does not lead to rapid plant death or severe growth and development disorders. This enables plants to maintain a relatively normal physiological state when using TRV for gene silencing research, facilitating better observation of the impact of gene silencing on plants. Therefore, we used the TRV virus as a vector to infect tomato plants via Agrobacterium, enabling tomatoes to carry exogenous sweet protein genes (Fig 2)

     We have respectively verified the expression ability of sweet proteins at the RNA at the protein detection level. Additionally, considering that the defense mechanism of tomatoes against exogenous gene expression may lead to the silencing of the CaMV 35S promoter (for more information, please see the Design webpage), we have detected the expression levels of sweet proteins in leaves and fruits at the protein detection level.

Practice of transgenic experiment

After the verification of the previous experiments, we considered that expressing sweet proteins in tomatoes is a feasible scheme. Therefore, we began to engage in the experiment of transgenic sweet proteins. Similarly, we verified the expression level of transgenic tomato plants at the RNA level. At the protein level, we also detected the expression levels of sweet proteins in leaves, flowers and fruits to confirm the successful expression of proteins.

To further enhance the quality of sweet protein expression in tomatoes, we focused on whether the designed genetic circuit can exhibit a more superior performance in tomato sweet protein production. Initially, we hoped to increase the overall production of sweet proteins by means of multi-copy vectors. However, after extensive literature review, we found that multi-copy vectors have many problems. First, excessive gene copies may impose a metabolic burden on host cells. Host cells need to allocate more resources for the replication and expression of multiple gene copies, which is very likely to affect the normal growth and metabolism of cells. Second, gene silencing may occur. A high copy number may trigger epigenetic regulatory mechanisms, resulting in transcriptional silencing of some or even all gene copies, which in turn leads to unstable expression of the target gene or even complete loss of expression. In addition, the frequency of gene recombination may be higher. Multiple gene copies increase the possibility of homologous recombination or other genetic rearrangements, which will lead to instability of the vector and changes in gene expression patterns. Considering that the chassis organism we used is tomato, using a multi-copy vector may be unfavorable to the growth of tomatoes themselves and the expression inhibition caused by DNA methylation modification of tomatoes will also increase. Based on this, after extensive literature review, we chose the E8 promoter. The E8 promoter is one of the most widely characterized fruit ripening-specific promoters and shows strong conservation in different tomato varieties^[5] (for more information, see the Design). We hope that through this design, sweet proteins can be efficiently and specifically expressed in fruits while avoiding excessive adverse effects on the growth of tomato plants. Therefore, we replaced the CaMV 35S promoter with the E8 promoter (Fig 3)

Fig 3. The two pictures above are the plasmid maps of the CaMV 35S promoter. The two below are the plasmid maps of the E8 promoter. As can be seen from the maps, we replaced CaMV 35S with the E8 promoter

And as before, the expression level of sweet proteins was verified at the RNA and protein levels respectively. In particular, we also detected the expression levels of sweet proteins in leaves, flowers and fruits to test the specificity of E8. Finally, we employed the Elisa kit from Shanghai Hengyuan Biological Technology Co., Ltd. to assess the yield of sweet proteins in tomato fruits.

Module for detecting the quality of sweet proteins

Metabolic detection of tomatoes

Following preliminary validation, we have confirmed that the protein can be expressed in tomatoes with a considerable expression level. However, to ensure the high quality of the sweet protein, further experiments are necessary. First, we must ensure that the expression of the foreign gene does not excessively affect the metabolism and growth of the tomato plants. To this end, we have tested for the common inorganic metabolic product, glucose. In addition, we have conducted continuous observation and comparison of the plant height and fruit growth conditions of the tomato fruits to ensure that the expression of the sweet protein does not significantly impact the metabolic state of the tomatoes, thereby not affecting the growth of the tomato fruits.

Detection of sweet protein efficacy

After extensive experimental validation, we have confirmed that the target protein has been successfully expressed in genetically modified tomatoes without significantly affecting the growth of the tomato plants. However, whether the sweet protein Thaumatin in the genetically modified tomatoes possesses sweetness remains unknown. Beyond focusing on the expression levels of sweet protein, it is more critical to determine whether the sweet protein produced in tomatoes can bind to the taste buds on the tongue and elicit a sweet taste. Consequently, we have conducted a series of attempts to verify the effectiveness of the proteins expressed in our genetically modified tomatoes.

Drawing from past experience, the most common method to determine whether sweet proteins possess sweetness is through human sensory evaluation, which involves recruiting and training volunteers to conduct taste assessments to obtain results. However, for safety reasons, we will strictly adhere to the regulations of the iGEM community and resolutely refrain from conducting experiments on our product with human subjects. Therefore, we have discarded this method. Instead, we have innovatively modified the Enzyme-Linked ImmunoSorbent Assay (ELISA) to determine whether the sweet protein can bind to human taste receptor proteins on the taste buds to produce sweetness. After reviewing the literature, we learned that the human tongue is primarily composed of three receptor proteins: T1R1, T1R2 and T1R3 and the perception of sweetness depends on the heterodimer complex of T1R2 and T1R3, which together form a functional sweet taste receptor, with T1R2 mainly responsible for binding to sweet molecules. After discussing with our Principal Investigator (PI) and professors, we decided to focus our research on T1R2, as it is one of the primary receptors for sweet sensation. At the same time, considering the optimization of time and resource utilization, we decided to preliminarily verify the effectiveness of our produced protein on T1R2. (For more details, please see the Model webpage) Therefore, we have initiated experimental exploration.

Enzyme-Linked ImmunoSorbent Assay (ELISA) is a quantitative analysis technique widely used in the fields of biomedicine and biotechnology. Its basic principle is based on the specific binding reaction between antigen and antibody. The known antigen or antibody is immobilized on a solid phase carrier (such as a polystyrene microplate) and then the sample to be detected is added. The target antigen or antibody in the sample binds to the antigen or antibody on the solid phase carrier. Then, an enzyme-labeled secondary antibody (or antigen) is added and this antibody (or antigen) specifically binds again to the bound antigen (or antibody). Finally, the substrate of the enzyme is added and the enzyme catalyzes the substrate to produce a color reaction. The absorbance of the reaction product is measured to determine the content of the target substance in the sample or to determine whether the target substance is present in the sample.

In order to be able to use this principle to determine whether the sweet protein in tomatoes can bind to the human receptor protein T1R2, we innovatively mixed two kits. First, T1R2 was fixed on a solid phase carrier carrying the T1R2 antibody. Then, the protein is extracted from the samples of E8-Thaumatin fruits and the extracted protein is incubated on the solid phase carrier to allow the sweet protein to specifically bind again to the bound human receptor protein. This step is the key to our innovation. Next, we used a secondary antibody labeled with HRP enzyme to specifically bind again to the bound sweet protein. Finally, we added the enzyme substrate TMB and the enzyme catalyzes the substrate to produce a color reaction to determine whether the sweet protein produced by tomatoes can bind to T1R2. In order to better reflect the binding degree of the sample, we incubated the Thaumatin standard carrying HRP enzyme as a secondary antibody on the solid phase carrier that has already been incubated with T1R2 and then used TMB for color development^[6]. (For details, please see the Protocol webpage). Then this is used as a positive control for analysis.(Fig 4)

Fig 4. Schematic diagram of the principle for the effectiveness testing of sweet protein in the experimental group

Our result can determine whether the Thaumatin protein can bind to the human receptor protein T1R2. However, the result of our method cannot yield the sweetness index of the sample. This is because we cannot ensure that all the standards in the positive control group can be successfully incubated. Therefore, we cannot obtain the relationship between absorbance and concentration and thus we cannot obtain the relationship between absorbance and sweetness index. Finally, we cannot estimate the sweetness index of the tomato sample. Therefore, further experimental exploration is needed to obtain the sweetness index of Thaumatin.

Sweet protein sweetness detection

After consulting relevant materials, we discovered that the currently commonly employed method for detecting the sweetness of sweeteners is to establish a concentration relationship between sweeteners and sucrose standard solutions and then calculate the sweetness of sweeteners based on this concentration relationship. In the work of Grant E. DuBois and D. Eric Walters utilized this method and established a sweetener concentration-response relationship curve between different sweeteners and sucrose.

First, they prepared standard sucrose solutions of various concentrations (2%-16%) and formulated sweetness grades corresponding to different concentrations of standard sucrose solutions. Subsequently, they asked volunteers to taste standard sucrose solutions of different concentrations and remember their corresponding sweetness grades. After completing these tasks, volunteers were required to taste sweeteners of different concentrations and rate and score sweeteners of different concentrations according to the sweetness grades of sucrose. Finally, Grant E. DuBois and D. Eric Walters established a relationship between the sweetness grade and dose concentration of sweeteners——the concentration-response relationship curve of sweeteners. Among them is the concentration-response relationship curve of standard Thaumatin (Fig 5)

Fig 5. Concentration-response curve of Thaumatin

Therefore, in their work, they established a relationship among sucrose concentration, sweetness grade and sweetener concentration. According to this relationship, we only need to know the sweetness of heterologously expressed Thaumatin in tomatoes is equivalent to the concentration of standard Thaumatin and then we can calculate how sweet our Thaumatin is according to this relationship. Subsequently, we found that an electronic tongue can be used to measure this concentration.

The electronic tongue is an electronic sensor device that imitates human taste. The electronic tongue is composed of four parts: sensor array, signal conditioning system, test platform and application software. The sensor array is composed of six electrodes and an auxiliary electrode to form an independent unit mechanism and together with a reference electrode, it forms a complete sensor array. The signal conditioning system is composed of a signal excitation unit, a signal conditioning unit and a data acquisition unit. The test platform provides a strict working environment for sample placement and detection. The application software is an application program running on a computer for controlling electronic tongue detection, analyzing data and outputting results. The electronic tongue is based on the electrode system principle, emits multi-frequency large-amplitude pulse excitation scanning signals and detects the characteristic response voltage signals of the measured substances through a sensor array composed of several specific precious metals. Then, it inputs into a mathematical model for data analysis to obtain the overall sample taste characteristics. As a powerful tool for distinguishing taste and flavor characteristics, it can perform rapid detection on samples, thereby avoiding experimental deviations caused by human subjective feelings.

Based on the principle of the electronic tongue and our ELISA experimental results, we prepared standard Thaumatin tomato solutions with concentrations of 0, 15, 30, 45 and 60 (ppm) referring to the concentration-response relationship curve of Thaumatin and input them into the electronic tongue for taste analysis. Subsequently, according to the taste characteristics of solutions of different concentrations, we established a SVR (Support Vector Regression) linear regression equation using the mathematical model provided by the electronic tongue. The parameters were set as 0, 15, 30, 45 and 60 respectively according to the concentrations. Among them, SVR (Support Vector Regression) is an application of SVM (Support Vector Machine) to regression problems, which can realize data classification and regression processing. Therefore, when we detect a sample, we can use the electronic tongue to analyze the taste characteristics of the sample and then input it into the SVR model to obtain a standard Thaumatin solution similar to the taste characteristics of the sample.

To sum up, we have achieved the sweetness detection of Thaumatin in transgenic tomatoes through the following three steps.

--Step 1
We use an electronic tongue to detect the taste characteristics of standard Thaumatin tomato solutions of different concentrations. Subsequently, according to the sweetness characteristics, parameters are set as 0, 15, 30, 40 and 60 respectively based on the concentrations to establish a SVR linear regression equation. According to the established model, the R^2 in the established linear regression equation curve is 0.9868, indicating that the model is good and usable. By detecting the taste characteristics of the control group's tomato Micro-Tom (i.e., uninfected wild-type Micro-Tom tomatoes) and transgenic tomatoes and inputting them into the SVR model, we can know the concentration of standard Thaumatin similar to the taste characteristics of Thaumatin contained in transgenic tomatoes (Fig 6)

Fig 6. We established an SVR model based on the taste characteristics of standard Thaumatin

--Step 2
According to the concentration-response regression curve of Thaumatin, we can calculate the sweetness grade corresponding to Thaumatin in transgenic tomatoes.

--Step 3
Subsequently, according to the relationship between sweetness grade and the concentration of standard sucrose solution, we can calculate the sucrose concentration corresponding to Thaumatin in transgenic tomatoes and thus obtain its sweetness.

To sum up, we have detected the sweetness of Thaumatin in transgenic tomatoes using this method. That is, in transgenic tomatoes containing the E8 promoter, Thaumatin with a concentration of 11.0951 mg/L is equivalent to a standard sucrose solution with a concentration of 8.65%.

For more information, please check the Results webpage.

Sweet protein localization module

Following the successful expression of a potent and high-quality sweet protein, we aspire to ensure its accumulation within tomato fruit to maintain a consistent level of the sweet protein. To this end, we have designed experiments focused on the accumulation of sweet protein. In our investigation concerning the vacuolar targeting of the sweet protein in tomatoes, the following steps were undertaken: We constructed a Thaumatin gene fused with a plant vacuolar targeting peptide SPS-NTPP and a fluorescent protein, which was subsequently cloned into the PGD vector (Fig 7)

Fig 7. Plasmid carrying the Thaumatin gene with the plant vacuole localization peptide SPS-NTPP and fluorescent protein

The vector was then successfully induced for expression in Escherichia coli to preliminarily validate the correct expression of the targeting peptide. Subsequently, the vector was transformed into the GV3101 strain of Agrobacterium tumefaciens. Given the time constraints and the lengthy process associated with the cultivation and genetic modification of tomatoes, we opted for Nicotiana benthamiana, a model organism from the same Solanaceae family as tomatoes, as our experimental subject. N. benthamiana offers several advantages: its genetic manipulation is facile and its transformation techniques are well-established, allowing for the introduction of foreign genes into cells through methods such as Agrobacterium-mediated transformation, similar to that in tomatoes, for studies on gene function and other aspects. Moreover, N. benthamiana has a shorter growth cycle compared to tomatoes, enabling a rapid transition from seed germination to plant maturity, thereby accelerating research progress and enhancing efficiency. Additionally, N. benthamiana exhibits exceptional performance in plant immunity research, making it a valuable model for studying plant immune responses. For instance, the Agrobacterium infiltration method facilitates the transient expression of proteins in tobacco leaf cells, rendering it an excellent in vivo experimental system for cell biology, biochemistry and protein-protein interaction studies, which can aid our team in thoroughly explore the relevant mechanisms and processes of plant immunity. Therefore, we utilized Agrobacterium-mediated infiltration of tobacco leaves and observed under a confocal laser scanning microscope to ascertain whether the sweet protein was effectively stored.

The preliminary validation in tobacco has confirmed the feasibility of constructing SPS-NTPP-Thaumatin-EGFP. However, the choice of chassis may also influence the function of the localization peptide. Therefore, we have continued to plan the expression scheme in tomatoes and aim to complete parts of it as much as possible according to the schedule.

For further details, please refer to the Results section on the associated webpage.

Module for directed evolution to remove bitter peptides

When consulting relevant literature on Thaumatin, we found that the Thaumatin sweet protein carrying bitter peptides may cause the sweet protein to have a bitter taste. Therefore, to avoid this phenomenon, we decided to conduct screening of bitter taste sites and directed evolution of proteins for Thaumatin. Considering the continuity of protein directed evolution, we conducted simulation experiments through visual analysis and molecular docking simulations. Regarding the optimization of this part, although we have not carried out experimental verification, we can calculate the approximate results in data modeling.

For more information, please see the Model webpage.