Team:Toulouse/Modelling

Modeling is a tool used to simplify and study systems. We can try to predict behavior with bibliographic informations or informations obtained from experiment.
Our project focuses on the development of our engineered bacterium in tree. The bacterial growth in trees seems to be unknown, so we must infer Bacillus subtilis behavior.

Bacterial Growth

Aim

Bacillus subtilis is a tree endophyte strain. A study showed that Bacillus subtilis could develop and fully colonize a tree, reaching a concentration of 10⁵ cells per gram of fresh plant. We need to know in which conditions the growth of B. subtilis is optimum in a tree and if the weather can stop its development during winter. So we decided to work on the Bacillus subtilis growth in function of the temperature during the year. Modeling bacterial growth in a tree section generates some difficulties, we need to know the distance between two tree extremities (treetops and root) or the speed sap flow which can vary with temperatures during the day and seasons, cause of the type of sap (phloem, xylem). Furthermore a tree is not an homogeneous system, its roots, trunk and branch do not contain same amount of sap and wood. The average speed of the plane tree sap is 2.4m/h, which means that in a day the sap will flow from one end to the other of a tree 30m. Tree is reduced to a bioreactor.

We make the following hypothesis:

1. According to the publication of Xianling Ji, after 6 months of Bacillus subtilis growth in a tree, bacteria cells reach a concentration of 10⁵ cells per gram of fresh plant. We assume that 10⁵ cells / g is the maximum concentration.
2. The composition of the phloem is stable. There is no effect of depletion of the medium.
3. Only temperature impact on bacterial growth.
4. It is believed that there is no leakage of cells.

Method

An assessment of the Bacillus subtilis growth in a similar sap, the birch sap was performed in laboratory conditions with optimum growth medium for Bacillus subtilis. Thus, a growth rate μ opt. From this value we can extrapolate a growth curve as a function of temperature. For this we used to cardinal temperature model:

T: Temperature.
µ_opt: optimal growth rate.
µ: growth rate at T.
T_max = max temperature supported by bacteria.
T_min = min temperature supported by bacteria.
T_opt = optimum temperature for the growth.

Necessary parameters for this function are minimun temperature T_min and maximum temperature T_max, optimal temperature for the growth T_opt and optimal growth rate µ_opt.

T_min: 10°C
T_max: 52°C
T_opt: 37°C
µ_opt: 8.5968 cfu/d

The optimal growth rate (µ_opt) is obtained experimentally with a similar birch sap environment.
The growth rate is negative below 10°C (growth test performed at 10°C and 4°C under similar conditions for the measurement of μ_opt), survival rate after 24h was 0.3 % at 10°C and null at 4°C. Conditions apply:

If__| T<= 4°C -> µ = -1
____| 4°C µ = -0.97
____| T > 10°C -> µ = f(T) with f(T) egal to cardinal temperature model.

Fig 1: bacterial growth (µ) as a function of temperature

A logistic model developed by Hiroshi Fujikawa is used to study bacterial growth.

General logistics formulas

In our cases, µ depends on the temperature. N corresponds to the bacterial population, Nmin and Nmax are two asymptotes. Parameter "m" is a curvature parameter; larger is m, smaller is the curvature of the deceleration phase with the model. Parameter n is a parameter related to the period lag; larger is n, shorter is the period of lag. Nmin is slightly lower than N0, when N is small, close to Nmin, as the initial state (N is equal to N0), Nmin / N is almost equal to 1 so the term (1-(Nmin/N)) is less than 1, growth is very slow. If N decrease until reach Nmin the term (1-(Nmin/N)) is equal to 0 thus there can not be any growth. Similarly when N is equal to Nmax the term (1-(N/Nmax)) is equal to 0 and the growth is blocked.
To overcome this we labor under two conditions , positive growth and negative growth, so two equations.This led to the writing of this model:

with n = 1 and m = 0.5

Term (1-(Nmin/N)) is not taken into account when there is growth and (1-(N/Nmax)) is not taken into account when there is bacterial decay.
Meteorological records of the Toulouse region of 2011-2013 are used to do averages daily temperatures. Thus we can determine B.subtilis growth during a year on Toulouse. This values are obtained for each day by the average on the hightest and the lowest temperature.
The density of green wood plane is about 650kg/m³. The average diameter of the trunks of the trees in question is about 0.80m and 15m high. This represents a volume of 30 m³ . The weight of the trunk is therefore 19,604kg . Added to this weight the weight of branches, twigs, leaves about 25% and about 15% of roots (source).
The average weight of a plane tree is 27,446kg where in inoculated 10mL of bacterial culture at 10⁹cfu/mL, ie 10^10 bacterial cells. This represents 3.64x10² cfu/g of fresh plant (N0).

Fig 2: (black) Bacillus Subtilis growth curve during one year (N is cells quantity by g of fresh plant). (red) average temperature. (blue) threshold at 10 °C.

In our model, growth starts only from 10°C, which happens between March and April. This period seems to be suitable to put the strain in the tree. From December the temperature decreased below 4°C, threshold below which bacteria die.

Discussion

In practice, temperature variations are certainly lower in tree than outside, especially if roots extend very deep. Composition of the tree sap must also intervene in the growth rate, nutrient content of sap is also temperature dependent. The effects of the decrease of the temperature in winter also involves a fall of the sap and this must also be involved in the disappearance of our strain in the tree. The period of Bacillus subtilis growth is certainly affected by the change in temperature, the rise of sap, its composition variations can consequently slow the growth rate. The modeling work is done with the programming language 'R' script attached.

References

• Xianling Ji, Guobing Lu, Yingping Gai, Chengchao Zheng & Zhimei Mu (2008) Biological control against bacterial wilt and colonization of mulberry by an endophytic Bacillus subtilis strain. FEMS Microbiol Ecol 65: 565–573
• A. Garnier(1977) Transfert de sève brute dans le tronc des arbres aspects méthodologiques et physiologiques. Ann. Sci. Foresi. 34 (1): 17-45
• Heikki Kallio , Tuija Teerinen , Seija Ahtonen , Meri Suihko , Reino R. Linko (1989) Composition and properties of birch syrup (Betula pubescens). J. Agric. Food Chem 37 (1): 51–54
• L. Rosso, J. R. Lobry & J. P. Flandrois (1992) AN Unexpected Correlation between Cardinal Temperatures of Microbial Growth Highlighted by a New Model. J. theor. Biol. 162 : 447-463
• Hiroshi Fujikawa (2010), Development of a New Logistic Model for Microbial Growth in Foods. Biocontrol of Science Vol 15: 75-80

Annexe

Spreading