Team:Bielefeld-CeBiTec/Model

Modeling

„Any modeling project should be tempered by the morality of laziness.“ (Barnes et. al., 2010)

Short Summary

This phrase illustrates that lab results can be predicted before actually working in the lab, saving precious time.

Our project revolves around new discovered or theoretical biological systems like for example RNAi in prokaryotes or newly discovered copper transporters. Our project required four different modeling approaches:

toxicity modeling
reactor modeling
siRNA modeling
ferritin structure modeling

Toxicity modeling

The toxicity of copper ions on the cell is well characterized (Ning et al., 2015) but for important parts of this project like the crossflow reactor we needed to know the exact point of time when our cells will die to achieve highest possible yields of copper.

The residence time should not exceed the rate of dying and cell lysis in the system. If cell lysis kicks in copper gets released back again into the substrate media and the yield minimizes. The modeling started with the copper uptake in our cells containing the BioBrick BBa_K2638204, which expresses oprC under pAraBAD (BBa_I0500)) control and induction at 1.0 % arabinose in. The toxicity is calculated for a single cell.

The first step was to calculate the rate of expression of oprC. Therefore the characterization of BBa_I0500) of Groningen was used to calculate the expression speed. At 1.0 % arabinose induction a raise of fluorescence of approx. ΔF = 82,000 within of t = 36,000 s was measurable. The conversion from fluorescence units to concentration in mol/L was calculated as k = 2.5 ∙ 10-6 mol/L (Furtado and Henry, 2002) and the volume of the used capillaries was V = 3.14 * 10^-9 L The rate of protein expression with 1.0 % arabinose is:

$$\frac{3.14 * 10^{-9} L * 2.5 * 10^{-6} mol L^{-1} * 87,500}{36,000 s} = 7.85 * 10^{-14} mol*s^{-1} (1)$$

According to (1) the increase of OprC with t₀ = 0
can be described with (2) and (3): $$\frac{dOprC}{dt} = 7.85 * 10-14 mol*s^{-1}$$ (2) $$OprC_t = OprC_0 + t ∙ 7.85 * 10^{-14} mol*s^{-1}$$ (3)

The next step is to determine the increase of the copper concentration inside the cell. OprC transports copper ions with a rate of v_max = 0.144 mol/s into the cell and has a substrate affinity of K_M = 1.66 mM (Parmar et al., 2018). The volume per cell in this modell contains V_C = 1∙10^-15 L^-1. The starting concentration was taken from the first accumulation test with [Cu²⁺₀] = 0.003 M. Due to the missing catalytic interchange number k_cat the other definition of the Michaelis-Menten equation had to be used:

$$\frac{C_inside}{dt} = \frac{-d(C_outside)}{dt} = v_max * \frac{Cu^{2+}_outside}{Cu2+outside + K_M} (4)$$

After solving the differential equation (4) with the conditions t₀ = 0:

$$[C_t]- K_M * ln{[C_t]} = v_max * t + [C_0] - K_M * ln {[C_0]} (5)$$

Due to the fact that this only describes the kinetic of a single molecule of OprC the (3) and the Avogadro number N_A = 6.022 * 10²³ mol^-1 . That way the addition was possible of the both terms on the right side of (5) and the end equation results in:

$$[C_t]- K_M * ln{[C_t]} = v_max*t^2*N_A*7.85*10-14 mol*s^{-1} + [C_0]-KM*ln{[C0]} (6)$$ At this point the equation is not further to simplify. Because of this a short python script trying out multiple different number constellations had to be tried out.
kurve = []

for each in listCt[:end+1]:
for t in ts:
nOPRC = t*5*(10**(-29))
#print each
#print t

gleichung1 = vmax*t*nOPRC*6.022*(10**(23)) - C0 - math.log(C0)*KM
gleichung2 = -each - KM * math.log(each)
gleichung2small = gleichung2 - gleichung2*0.00001
gleichung2big = gleichung2 + gleichung2*0.00001
if gleichung2small <= gleichung1 and gleichung2big >= gleichung1:
kurve.append((each,t))
print each
print t

print kurve

with open ("Tupel_fuer_Kurve3.txt", "w") as out:

This way the toxic copper concentration c_tox = 1 * 10^-5 mol/L (Ning, 2015) inside the cell is approx. reached after the induction with 1.0 % arabinose.

siRNA promoter model

First the most fitting siRNAs for our project has to be calculated. A program will be written which returns a selection of compatible siRNAs for a gene you want to silence. Possible secondary structures should also be considered (Mathews, 1999) because they could prevent the silencing mechanism of siRNAs.

**Figure 1:** The siRNA with the compatible secondary structure binds to the mRNA and enables its digestion and blocks the translation of the sequence (A). The wrong secondary structure can’t bind and that way the mRNA can be translated and won’t be digested (B).

We assume that the rate of silencing by siRNAs depends in the first place on the promoter region in front of the RNAi gene. By gathering expression rates of the most popular promoters in the iGEM parts registry valuable information for modelling will be acquired. Furthermore additional factors like secondary structures, diffusion rates and degradation rates of RNA fragments. With that a model of the actual expression rate of proteins can be modeled with a given sequence that shall be silenced. The other way round this model should tell which is the best fitting promoter for a specific protein concentration. Combining both applications in one will result in a useful tool for the iGEM community.

Discussion

The modeling for the point of time where the toxicity kicks in showed, that due to the troublesome way to find the right promoter strength a universal promoter library would be an important addition to the synthetic biology community. That way the promoter library in the parts collection represents an universal standard for the iGEM community to characterize their promoters.
The crossflow reactor design and modeling was strongly influenced by the toxicity modeling.Another way to approach this problem were the BioBricks BBa_K2638100, BBa_K2638101, BBa_K2638103 BBa_K2638105, BBa_K2638106, BBa_K2638120 and BBa_K2638121 enrich now the parts registry.
The siRNA/RNAi modeling went over into our software project because it showed way more potential like this. The ferritin modeling was a important measure to make sure the characterization of the ferritin molecules didn’t encountered any issues.

References

Barnes, D. J., & Chu, D. (2010). Introduction to modeling for biosciences. Springer Science & Business Media.
Delihas, N., & Forst, S. (2001). MicF: an antisense RNA gene involved in response of Escherichia coli to global stress factors. Journal of molecular biology, 313(1), 1-12.
Furtado, A., & Henry, R. (2002). Measurement of green fluorescent protein concentration in single cells by image analysis. Analytical biochemistry, 310(1), 84-92.
Ning, C., Wang, X., Li, L., Zhu, Y., Li, M., Yu, P., ... & Zhang, Y. (2015). Concentration ranges of antibacterial cations for showing the highest antibacterial efficacy but the least cytotoxicity against mammalian cells: implications for a new antibacterial mechanism. Chemical research in toxicology, 28(9), 1815-1822.
Mathews, D. H., Sabina, J., Zuker, M., & Turner, D. H. (1999). Expanded sequence dependence of thermodynamic parameters improves prediction of RNA secondary structure. Journal of molecular biology, 288(5), 911-940.
Parmar, J. H., Quintana, J., Ramírez, D., Laubenbacher, R., Argüello, J. M., & Mendes, P. (2018). An important role for periplasmic storage in Pseudomonas aeruginosa copper homeostasis revealed by a combined experimental and computational modeling study. Molecular microbiology.
Phan, C. M., & Nguyen, H. M. (2017). Role of capping agent in wet synthesis of nanoparticles. The Journal of Physical Chemistry A, 121(17), 3213-3219.
Stach, J. E., & Good, L. (2011). Synthetic RNA silencing in bacteria–antimicrobial discovery and resistance breaking. Frontiers in microbiology, 2, 185.