Numerical Study on Microbial Depolymerization Process of Xenobiotic Polymer ()

Masaji Watanabe^{}, Fusako Kawai^{}

Okayama University, Okayama, Japan.

**DOI: **10.4236/msce.2021.91004
PDF HTML XML
67
Downloads
163
Views
Citations

Okayama University, Okayama, Japan.

This study demonstrates mathematical analysis of biodegradation processes of xenobiotic polymers. A model for microbial population is based on the fact that growth rate of microorganisms is proportional to the microbial population and consumption rate of parts of carbon sources. The model is paired with a model for weight distribution. Those models lead to inverse problems for a molecular factor and a time factor of degradation rate. Solution of the inverse problems allows us to simulate the biodegra-dation process.

Share and Cite:

Watanabe, M. and Kawai, F. (2021) Numerical Study on Microbial Depolymerization Process of Xenobiotic Polymer. *Journal of Materials Science and Chemical Engineering*, **9**, 43-50. doi: 10.4236/msce.2021.91004.

1. Introduction

Microbial depolymerization processes are classified into exogenous type and endogenous type. Molecules liberate their parts to reduce in an exogenous type depolymerization process. Polyethylene (PE) and polyethylene glycol (PEG) are polymers subject to exogenous type depolymerization processes. Studies showed utilization of PEG of average molecular weight 20,000 by Psedomonas aeruginosa [1], degradation of PEG 20,000 by anaerobic bacteria isolated from sludge of a municipal anaerobic digester [2], and efficient biodegradation of PEG by Pseudomonas stutzeri was documented [3]. A mathematical model was proposed and numerical techniques were developed for PE biodegradation [4]. The mathematical techniques were reapplied to a biodegradation process of PEG [5].

Random breakdown of molecules is the primary factor of endogenous type depolymerization processes. Polyvinyl alcohol (PVA) and polylactic acid (PLA) are polymers subject to endogenous depolymerization processes. A mathematical model was proposed and numerical techniques were developed for an enzymatic degradation of PVA [6]. Those mathematical techniques were reapplied to an enzymatic hydrolysis of polylactic acid (PLA) [7]. Techniques originally developed for endogenous type processes were replied to exogenous type depolymerization processes of PE and PEG [8].

This study revisits an exogenous type depolymerization process of PEG to demonstrate mathematical techniques. Experimental outcomes before and after cultivation of microbial consortium E-1 in culture media were incorporated into a computational analysis based on a mathematical model. Inverse problems for a molecular factor and a time factor of a degradation rate were formulated. Solutions of those inverse problems allow us to simulate a biodegradation process of PEG.

2. Formulation of Exogenous Type Depolymerization Process

Consider an exogenous type depolymerization process in a culture medium. Let w(t, M) [mg] be the weight distribution of a polymer with respect to the molecular weight M at time t, and v(t) [mg] be the total weight of polymer molecules with molecular weight between A and B at time t. The total weight v(t) over the interval [A, B] is expressed in terms of the integral

$v\left(t\right)={\displaystyle {\int}_{A}^{B}w\left(t,M\right)dM}$ (1)

The total weight v(t) of the residual polymer at time t is expressed in terms of the integral

$v\left(t\right)={\displaystyle {\int}_{0}^{\infty}w\left(t,M\right)dM}$ (2)

Integral (1) well approximates the integral (2) with suitable choices of A and B. In this study, integrals with the lower limit were replaced with integrals with the lower limit A = 10^{3.0}, and integrals with the upper limit ∞ were replaced with integrals with upper limit B = 10^{4.2}.

Let σ(t) be the population of viable cells at time t. Equations (3) for the weight distribution w(t, M) and the microbial population σ(t) was proposed in previous studies [8] [9] [10].

$\frac{\partial w}{\partial t}=\sigma \left(t\right)\left[-\lambda \left(M\right)w+c\left(M\right){\displaystyle {\int}_{M}^{\infty}\lambda \left(K\right)d\left(K\right)w\left(t,K\right)dK}\right]$ (3)

$c\left(M\right)=M{e}^{\rho M},d\left(K\right)=\frac{\rho {e}^{-\rho K}}{K\left(1-{e}^{-\rho K}\right)},\rho =\frac{\mathrm{log}2}{L},$

Parameter L is the molecular weight of a monomer unit, e.g. PE: 28 (CH_{2}CH_{2}), PEG: 44 (CH_{2}CH_{2}O), and k and h are positive parameters. The function
$\lambda \left(M\right)$ corresponds to a molecular factor of degradation rate.

This study proposes the equation

$\frac{d\sigma}{dt}=ku\left(t\right)\sigma -h\sigma ,$ (4)

where $u\left(t\right)=-{v}^{\prime}\left(t\right)$ . Note that $u\left(t\right)$ is the rate of consumption of total weight. System of Equations (3), (4) is associated with the initial conditions

$w\left(0,M\right)={f}_{0}\left(M\right)$ , (5)

$\sigma \left(0\right)={\sigma}_{\text{0}}$ , (6)

where ${f}_{0}\left(M\right)$ and ${\sigma}_{\text{0}}$ are the initial weight distribution and the initial microbial population, respectively.

3. Numerical Solutions of Inverse Problems for Molecular Factor and Time Factor

The initial value problem (3), (4), (5) (6) is solvable for $w\left(t,M\right)$ and $\sigma \left(t\right)$ provided the molecular factor $\lambda \left(M\right)$ and values of parameters ${\sigma}_{\text{0}}$ , k, and h are given beforehand. In order to obtain the function and values of parameters, consider change of variables from t to $\tau $ defined by the equation

$\tau ={\displaystyle {\int}_{0}^{t}\sigma \left(s\right)ds}.$ (7)

Denote by by $w\left(\tau ,M\right)$ , $S\left(\tau \right)$ , and $V\left(\tau \right)$ functions that correspond to w(t, M), $\sigma \left(t\right)$ , and v(t) according to the change of variables (7), respectively.

Note that

$\frac{\partial W}{\partial \tau}=\frac{dt}{d\tau}\frac{\partial w}{\partial t}=\frac{1}{\sigma \left(t\right)}\frac{\partial w}{\partial t},\text{\hspace{0.17em}}\text{\hspace{0.17em}}\frac{dS}{d\tau}=\frac{dt}{d\tau}\frac{d\sigma}{dt}=\frac{1}{\sigma \left(t\right)}\frac{d\sigma}{dt},\text{\hspace{0.17em}}\text{\hspace{0.17em}}\frac{dV}{d\tau}=\frac{dt}{d\tau}\frac{dv}{dt}=\frac{1}{\sigma \left(t\right)}\frac{dv}{dt}$

and that Equations (3) and (4) lead to

$\frac{\partial W}{\partial \tau}=-\lambda \left(M\right)W+c\left(M\right){\displaystyle {\int}_{M}^{\infty}\lambda \left(K\right)d\left(K\right)W\left(\tau ,K\right)dK}\text{\hspace{0.17em}},$ (8)

$\frac{dS}{d\tau}=kU\left(\tau \right)S-h,$ (9)

respectively, where $U\left(\tau \right)=-{V}^{\prime}\left(\tau \right)$ .

Suppose that F_{1}(M) is the weight distribution for
$\tau ={T}_{1}$ , that is,

$W\left({T}_{1},M\right)={F}_{1}\left(M\right),$ (10)

and that F_{2}(M) is the weight distribution for
$\tau ={T}_{2}$ , that is,

$W\left({T}_{2},M\right)={F}_{2}\left(M\right).$ (11)

Equation (8), the initial condition (10), and the final condition (11) form an inverse problem for the molecular factor $\lambda \left(M\right)$ , for which the solution of the initial value problem (8), (10) also satisfies the final condition (11).

Numerical techniques to solve the inverse problem for
$\lambda \left(M\right)$ were developed in previous studies. Weight distributions of PEG 6000 before and after cultivation of microbial consortium E-1 for one day, three days, five days, seven days, and nine days f_{0}(M), f_{1}(M), f_{2}(M), f_{3}(M), f_{4}(M), and f_{5}(M) were introduced into analysis. Note that w(t_{i},M) = f_{i}(M) for i = 0, 1, 2, 3, 4, and 5, where t_{0} = 0, t_{1} = 1, t_{2} = 3, t_{3} = 5, t_{4} = 7, and t_{5} = 9. Weight distributions after cultivation of the microbial consortium E-1 for one day and three days were the functions F_{1}(M) and F_{2}(M), respectively, and the inverse problem (8), (10), (11) was solved numerically for T_{1} = 0 and T_{2} = 1.0 [10]. Note a positive scalar multiple of a molecular factor is also a molecular factor.

Once the molecular factor $\lambda \left(M\right)$ was obtained, Equation (10) was solved for $W\left(\tau ,M\right)$ with the initial condition

$W\left(0,M\right)={f}_{0}\left(M\right).$ (12)

A previous study shows the exponential function

$V\left(\tau \right)={v}_{0}{e}^{-\mu \tau}\text{\hspace{1em}}\left({v}_{0}={\displaystyle {\int}_{0}^{\infty}{f}_{0}\left(M\right)dM}\right),$ (13)

well approximates function $V\left(\tau \right)$ so that ${V}^{\prime}\left(\tau \right)=-\mu {v}_{0}{e}^{-\mu \tau}$ . In this study, function $V\left(\tau \right)$ was approximated by the exponential function (13) with $\mu \approx 0.878$ .

Once the function $V\left(\tau \right)$ was obtained, Equation (9) was solved with the initial condition

$S\left(0\right)={\sigma}_{0}.$ (14)

Denote by $S\left(\tau ,{\sigma}_{0},k,h\right)$ the solution of the initial value problem (9), (14). In view of the Formula (7), $t=q\left(\tau ,{\sigma}_{0},k,h\right)$ , where

$q\left(\tau ,{\sigma}_{0},k,h\right)={\displaystyle {\int}_{0}^{\tau}\frac{dr}{S\left(r,{\sigma}_{0},k,h\right)}}.$

Given m pairs of values of t and τ, $\left({t}_{i},{\tau}_{i}\right)\text{\hspace{0.17em}}\left(i=1,2,3\right)$ , consider the system of equations

${g}_{i}\left({\sigma}_{0},k,h\right)=0\text{\hspace{1em}}\left(i=1,2,3\right)$ (15)

where ${g}_{i}\left({\sigma}_{0},k,h\right)=q\left({\tau}_{i},{\sigma}_{0},k,h\right)-{t}_{i}$ . Here ${\tau}_{i}$ is the value of τ that satisfy $v\left({t}_{i}\right)=V\left({\tau}_{i}\right)$ . In particular,

${t}_{\text{1}}=\text{1},\text{\hspace{0.17em}}{\tau}_{i}\approx 0.0\text{24275},\text{\hspace{0.17em}}{t}_{\text{2}}=\text{3},\text{\hspace{0.17em}}{\tau}_{2}\approx 0.\text{288317},\text{\hspace{0.17em}}{t}_{\text{3}}=\text{5},\text{\hspace{0.17em}}{\tau}_{3}\approx \text{2}.\text{62864}.$

Intervals $\left[{\sigma}_{0,0},{\sigma}_{0,1}\right]$ , $\left[{k}_{0},{k}_{1}\right]$ , $\left[{h}_{0},{h}_{1}\right]$ were divided into ${n}_{1},{n}_{2}$ and ${n}_{3}$ intervals, respectively, and the initial value problem (9), (14) was solved for

${\sigma}_{0}={\sigma}_{0,0}+i\Delta {\sigma}_{0},\text{\hspace{0.17em}}\text{\hspace{0.17em}}\Delta {\sigma}_{0}=\frac{{\sigma}_{0,1}-{\sigma}_{0,0}}{{n}_{1}},\text{\hspace{0.17em}}\text{\hspace{0.17em}}i=0,1,\mathrm{...},{n}_{1},$

$k={k}_{0}+j\Delta k,\text{\hspace{0.17em}}\text{\hspace{0.17em}}\Delta k=\frac{{k}_{1}-{k}_{0}}{{n}_{2}},\text{\hspace{0.17em}}\text{\hspace{0.17em}}j=0,1,\mathrm{...},{n}_{2},$

$h={h}_{0}+l\Delta h,\text{\hspace{0.17em}}\text{\hspace{0.17em}}\Delta h=\frac{{h}_{1}-{h}_{0}}{{n}_{3}},\text{\hspace{0.17em}}\text{\hspace{0.17em}}l=0,1,\mathrm{...},{n}_{3},$

${\sigma}_{0,0}=0.0015,\text{\hspace{0.17em}}\text{\hspace{0.17em}}{\sigma}_{0,1}=0.0016,$

${k}_{0}=0.0045,\text{\hspace{0.17em}}\text{\hspace{0.17em}}{k}_{1}=0.0055,$

${h}_{0}=0.35,\text{\hspace{0.17em}}\text{\hspace{0.17em}}{h}_{1}=0.45,$

${n}_{1}=100,{n}_{2}=100,{n}_{3}=100$ .

The value of ${\sigma}_{0}$ , k, and h that minimize the error

$\sqrt{\underset{i=1}{\overset{5}{{\displaystyle \sum}}}{\left[{g}_{i}\left({\sigma}_{0},k,h\right)\right]}^{2}}$

were approximately 0.00156, 0.00518, and 0.404, respectively.

Given $\lambda \left(M\right)$ and $\sigma \left(t\right)$ , the initial value problem (3), (5) was solved numerically. Figures 1-5 show some numerical results. Figures 1-4 show comparison between experimental results and numerical results for weight distribution after cultivation of microorganisms for one day, three days, five days, and seven days, respectively. Figure 5 shows the graph of the function $\sigma \left(t\right)$ .

Figure 1. Weight distribution after cultivation of microbial consortium E-1 for one day. An experimental result and numerical results are shown.

Figure 2. Weight distribution after cultivation of microbial consortium E-1 for three days. An experimental result and numerical results are shown.

Figure 3. Weight distribution after cultivation of microbial consortium E-1 for five days. An experimental result and numerical results are shown.

Figure 4. Weight distribution after cultivation of microbial consortium E-1 for seven days. An experimental result and numerical results are shown.

Figure 5. Transition of microbial population over seven days.

4. Discussion

Equation (4) is based on the fact that the increasing rate of the microbial population is proportional to both the consumption rate and the microbial population. Parameter k is the conversion rate from the carbon source to the microbial population per unit population, It is also based on the fact that the decreasing rate of the microbial population is proportional to the microbial population. Parameter h is the rate of loss of active microorganisms. Our analysis shows that the microorganisms convert approximately 0.5% of carbon source consumed per day to their growth, and that approximately 40% of microorganism loses degradability per day. The numerical results show that our model is appropriate.

Acknowledgements

The authors thank Ms. Y. Shimizu for her technical support.

Conflicts of Interest

The authors declare no conflicts of interest regarding the publication of this paper.

[1] | Haines, J.R. and Alexander, M. (1975) Microbial Degradation of Polyethylene Glycol. Applied Microbiology, 621-625. https://doi.org/10.1128/AM.29.5.621-625.1975 |

[2] | Dwyer, D.F. and Tiedje, J.M. (1983) Degradaton of Ethylene Glycol and Polyethylene Glycols by Methanogenic Consortia. Applied and Environmental Microbiology, 185-190. https://doi.org/10.1128/AEM.46.1.185-190.1983 |

[3] | Obradors, N. and Aguilar, J. (1991) Efficient Biodgradation of High-Molecular-Weight Polyethylene Glycols by Pure Cultures of Pseudomonas Stutzeri. Applied and Environmental Microbiology, 2383-2388. https://doi.org/10.1128/AEM.57.8.2383-2388.1991 |

[4] | Watanabe, M., Kawai, F., Shibata, M., Yokoyama, S., Sudate, Y. and Hayashi, S. (2004) Analytical and Computational Techniques for Exogenous Depolymerization of Xenobiotic Polymers. Mathematical Biosciences, 192, 19-37. https://doi.org/10.1016/j.mbs.2004.06.006 |

[5] | Watanabe, M. and Kawai, F. (2010) Effects of Microbial Population in Degradation Process of Xenobiotic Polymers. In Howlett, P., Nelson, M. and Roberts, A.J., Eds., Proceedings of the 9th Biennial Engineering Mathematics and Applications Conference (EMAC-2009), Vol. 51, c682-c96. http://journal.austms.org.au/ojs/index.php/anziamj/article/view/2433 https://doi.org/10.21914/anziamj.v51i0.2433 |

[6] | Watanabe, M. and Kawai, F. (2006) Mathematical Modelling and Computational Analysis of Enzymatic Degradation of Xenobiotic Polymers. Applied Mathematical Modeling, 30, 1497-1514. https://doi.org/10.1016/j.apm.2005.12.011 |

[7] | Watanabe, M., Kawai, F., Tsuboi, S., Nakatsu, S. and Ohara, H. (2007) Study on Enzymatic Hydrolysis of Polylactic Acid by Endogenous Depolymerization Model. Macromolecular Theory and Simulations, 16, 619-626. https://doi.org/10.1002/mats.200700015 |

[8] | Watanabe, M. and Kawai, F. (2019) Exponential Approximation of Consumption Rate of Carbon Source in Numerical Simulation of Microbial Depolymerization Process. International Journal of Advances in Science Engineering and Technology, 7, 2321-8991. http://www.ijaseat.iraj.in/volume.php?Volume_id=553 |

[9] | Watanabe, M. and Kawai, F. (2019) Mathematical Modeling and Inverse Analysis for Microbial Depolymerization Processes of Xenobiotic Polymers. Journal of Materials Science and Chemical Engineering, 7, 39-46. https://doi.org/10.4236/msce.2019.712005 |

[10] | Watanabe, M. and Kawai, F. (2018) Numerical Study of Microbial Depolymerization Process with Weight Distributions before and after Microbial Cultivation for One Week. Proceedings of the IRES International Conference, August 2018, Kuala Lumpur, 1-5. |

Journals Menu

Copyright © 2021 by authors and Scientific Research Publishing Inc.

This work and the related PDF file are licensed under a Creative Commons Attribution 4.0 International License.