Team:XJTLU-CHINA/Model

Modeling

Bioinformatics methods play irreplaceable roles in biology science explorations nowadays. Various theoretical models can be used to simulate, construct, and predict the evolutions of microworld in a visualized way, thus assisting scientists with better understanding of biological processes.

In our project, the design of ExoCar is an extensive of last year's project and we made a considerable progress. We found an ideal protein EAAT2 that may contribute to effective treatment of neurodegenerative diseases. Additionally, a 3 plasmids regulation system is designed to facilitate and improve the exosome yield in the primary procedure of drug production. Among these designs, model is supposed to be the core and linkage of the whole story.

Model & project

**You may ask: “Why do you so emphasize the importance of modeling? "**

Modeling is one of the axes of the project of XJTLU-CHINA iGEM, 2019. This year, our models not only anticipate the results of our core experiments, and simplify the task of exploring experimental conditions, but also confirm the feasibility and rationality of our project design. The project design now we present on our wiki was adopted and implemented after our modeling results shows enough consistency with our expectations.

EAAT2 Functional Verification model (EFV model)

To be Specific, when we found EAAT2 protein as a potential candidate to used in our project, we had no idea whether this protein can be expressed in our cell lines and intake excessive L-glutamate in culture medium with high efficiency. After building our EAAT2 Function Verification (EFV) model with the help of Matlab and online literatures, we predicted that by utilizing EAAT2 protein as the tool, we could achieve our purpose since our results provided a positive feedback for our idea, which suggested the EAAT2-induced L-glutamate intake rate can reach up to our expectation. Furthermore, from our model, we also predicted that the rate of L-glutamate rate was in the unit of μmol/min. This detail greatly reduced the work of exploration of experiments condition, shortened the experiment duration and ensured the progress of the project.

Riboswitch Optimization model (RO model)

Besides the attribution to project evaluation and implementation, our models also facilitate us to design the part which we expected to improve and provide to the whole iGEM community. We use thermodynamic model to predict the translation strength of different Ribosomal Binding Site (RBS) regions when they are connected with our theophylline riboswitch sequence. By this way, we perfectly improved our reporting part, enabled it with higher sensitivity to its ligand, made it more detectable and stabilized its reporting function.

3 Plasmids Regulation System model (PRS model)

Not only for this year’s project, our models also achieved a long-term impact on the design of our project in the future. Specifically, our 3 plasmids regulation system (PRS) model is built up to mimic the process of therapeutic exosome production. By building up this model, we successfully predict the producing condition when cells are used to generate exosome drugs. More importantly, we added regulation parts to control gene expression intracellularly. This model approves our project design about cell production and can be extracted as a method to build up engineered cell lines. Our model provides guidance and some predicted details for cellular precisely-control production and it may have an impact on pharmaceutical industries.

Figure 1. Interactions of project design with modeling & experiment with modeling

Model Blocks

The main models we make for this project can be divided into such 3 parts

EAAT2 Functional Verification model (EFV model)

3 Plasmids Regulation System model (PRS model)

Riboswitch Optimization model (RO model)

EAAT2 Functional Verification model
(EFV model)

Why this model?

When EAAT2 was suggested to be our treatment RNA drug, which was not involved in last year’s project. There was coming a problem that how the function of EAAT2 improves and qualifies glutamate absorption into the neural system. Therefore, we design experimentsto measure the glutamate absorbing ability of EAAT2 under certain condition.

Before operating the wet experiment, EAAT2 functional verification model (EFV model) was built to preverify the function of EAAT2. At the same time, details of wet experiment was modified to adapt to the real condition under the guidance of this model.

Figure 2. EAAT2 proteins absorb L-glutamate in N2A extracellular environment.

Methods

This model is a multi-step mathematical model based on Central Dogma Principle, and will predict the overall glutamate absorbing ability of N2A cell under our experimental condition. Following the gene expression process, we divide this model into 5 subparts, including plasmids transfection, gene transcription, mRNA translation, EAAT2 protein translocation and glutamate absorption. Each of the steps is described by the specific mathematic and statistic method.

However, Due to high level complexity of cell physiology. We make following assumptions to optimize and simplify the process.

Assumption

Refer to literatures and experimental design, we assume that the cell physiology has reached a relatively steady state, thus the number of EAAT2 proteins will keep in a fixed level when the absorption of glutamate begins.
Once plasmids are transfectedd into cells, they will be transcribed soon and then mRNAs will be generated. Simultaneously, mRNAs will start decay when they enter into cytoplasm. The generation and degradation of mRNA occur at the same time, with their own speed. Finally, the copy number of intracellular mRNA would approach to a constant at equilibrium.
Like mRNAs, protein translation and degradation also occur at the same time. However, based on our results, we predict that the time for protein to approach to a stable level is comparatively negligible. Therefore, we build our model mainly based on the number of proteins at the equilibrium stage.

Figure 3. Main procedure of EFV model.

1. Transfection

The first part of our model is transfection. To mimic cell transfection process in vitro experiments, the transfection efficiency can be expressed with a proportion constant F.

2. Transcription

Then in the process of transcription, mRNA of EAAT2 will be transcribed in cells once plasmids are transfected. Meanwhile, generated mRNAs are degrading simultaneously. Taken together the process of mRNA generation and degradation, according to assumption 1, overall number of mRNAs can be represented with a constant Nr.

3. Translation

The next step is translation, like the transcription, proteins will be produced and degrade at the same time in cells. Here, Law of Mass Action kinetics (LMA) is applied to describe this process.

**Parameter Table**
Parameter	Value	Unit	Description	Source
Nr	600	no unit	the mean number of mRNA per cell	[5]
R	122.3	s^-1	translation rate	[8]
D	6*10^-3	s^-1	protein degradation rate	[1]

**Variable Table**
Variable	Unit	Description
P(t)	protein	the number of mRNA
t	second	time

With proper parameter, the model is solved by the ODE45 MATLAB function in file EFV_Trans_main.m and file EFV_Trans_function.m. The simulation results are indicated in the following figure.

Figure 4. The number of EAAT2 in N2A cell versus time. When the simulation exceeds 1000s, the number of EAAT2 reminds in a stable level Ne (~10⁵)

4.Translocation

After translation, EAAT2 proteins will translocate to cell membrane. For this part, we read from online literature that most EAAT2 proteins will be engulfed into cell membrane immediately after translation. And only few proteins will be sent into proteasome and be degraded. Since protein translation and degradation have been taken into consideration in previous equation. We make a deduction that the number of membrane-valid EAAT2 equals to Ne (from Translation part figure 4).

Figure 5. The distribution of EAAT2 in astrocyte cell. This figure shows that most of the EAAT2 proteins are engulfed into cell membrane

5.Absorption

Finally, EAAT2 proteins anchored on cell membrane will absorb L-glutamate acid from extracellular medium. This process followes first-order Michaelis-Menten kinetics and Eadie-Hofstee transformation. Combined with the data from literatures, V_max the maximum absorption rate of glutamate in our experimental system, can be calculated.

**Parameter Table**
Parameter	Value	Unit	Description	Source
V_max	1.41	μmol min^-1	maximum reaction rate	[2]
K_glum	52.7	μmol	k_m of EAAT2	[2]

**Variable Table**
Variable	Unit	Description
[Glu]	μmol	L-glutamate acid concentration in medium
V_abs	μmol min^-1	glutamate absorption rate

Result

By running the equation with Matlab (EFV_Absor_main.m) , we get Figure 6 which predicts the absorption of L-glutamate acid in different concentration of glutamate in medium. When glutamate concentration approaches to V_max, glutamate are absorbed in a comparative high speed. EAAT2-induced L-glutamate intake rate can reach up to the unit of μmol/min, which is considered to be efficiently enough for our project. In conclusion, induced up-regulated expression of EAAT2 protein can absorb glutamate significantly and we can achieve our therapeutic purpose by utilizing EAAT2 proteins.

Figure 6. The rate of EAAT2 absorbing extracellular glutamate versus extracellular glutamate concentration.

Finally, EAAT2 wet experiments can be performed with the insurance from modeling. Here, what should be mention is that an important adjustment is supported by EFV model. the initial concentration of glutamate used to treat N2A cell is adjust to 80μmol/L since this concentration mimic the real condition of early neurodegenerative diseases and also is a comparatively reasonable concentration for absorption test.

The result of EAAT2 wet experiment, as we expected, shows a high feasibility in absorbing glutamate. The absorption rate, consistent with our predictions, reaches up to the unit of μmol/min. More experiment details can be find by clicking this.

(Back to Model Blocks)