J. Nig. Soc. Phys. Sci. 5 (2023) 1154

Journal of the
Nigerian Society

of Physical
Sciences

Degradation of PET Nanoplastic Oligomers at the Novel PHL7
Target:Insights from Molecular Docking and Machine Learning

C. E. Durua,∗, C. E. Enyohb, I. A. Duruc, M. C. Enedohd

a Surface Chemistry and Environmental Technology (SCENT) Research Group, Department of Chemistry, Imo State University, Owerri, PMB 2000, Imo State,
Nigeria.

bGraduate School of Science and Engineering, Saitama University, 255 Shimo-Okubo, Sakura-ku, Saitama City, Saitama 338-8570, Japan.
c Department of Chemistry, Federal University of Technology Owerri, PMB 1526, Imo State Nigeria.

d Department of Chemistry, Imo State University, Owerri, PMB 2000, Imo State, Nigeria.

Abstract

The versatility of Polyethylene terephthalate (PET) as a material with numerous applications in the food industry and its recalcitrance to chemical
and microbial degradation has recently made it an environmental nuisance. In this study, we applied computational methods to ascertain the
dependence of PET nanoplastic (NP) degradation on the chain length of the oligomer. The binding affinities of the NPs on the novel enzyme
Polyester Hydrolase Leipzig 7 (PHL7) were used to relate their ease of degradation at the enzyme active site. The results revealed that the binding
affinity of PET NPs at the enzyme target decreased from -5.2 kcal/mol to -0.8 kcal/mol, with an increase in PET chain length from 2.18 nm to
5.45 nm (2-5 PET chains). The binding affinities became positive at chain lengths 6.54 nm (6 PET chains) and above. These findings indicated
that PET NP degradation at this enzyme’s active site is most efficient as chain length decreases from 5-2 units and is not likely to occur at longer
PET chains. A feedforward Artificial Neutral Network (ANN) analysis predicted that the energy of the PET NPs is a very important factor in its
degradation.

DOI:10.46481/jnsps.2023.1154

Keywords: Polyethylene terephthalate; Nanoplastic; Polyester Hydrolase Leipzig 7; binding affinity; Artificial Neutral Network.

Article History :
Received: 28 October 2022
Received in revised form: 05 January 2023
Accepted for publication: 10 January 2023
Published: 27 January 2023

c© 2023 The Author(s). Published by the Nigerian Society of Physical Sciences under the terms of the Creative Commons Attribution 4.0 International license

(https://creativecommons.org/licenses/by/4.0). Further distribution of this work must maintain attribution to the author(s) and the published article’s title, journal citation, and DOI.

Communicated by: K. Sakthipandi

1. Introduction

Plastics are synthetic materials of long carbon chains that
can be formed into different shapes when still molten and
then transform into a slightly elastic solid form. Nanoplastics
(NPs) are particles ranging from 1 to 1000 nm, unintentionally
produced from the degradation and manufacturing of plastic

∗Corresponding author tel. no: +2348037131739
Email address: chidiedbertduru@gmail.com (C. E. Duru )

objects [1]. Because they may pass through biological mem-
branes, nanoplastics have a higher potential for danger than
microplastics [2, 3]. NPs can be ingested by various creatures
[4, 5], thereby raising concerns about possible bioaccumulation
and biomagnification. There is increasing evidence that marine
creatures consume NPs, some evidence of translocation outside
the stomach, and much less evidence of transfer between
trophic levels. Many studies have also demonstrated the
exposure and toxicity of NPs to human health [6, 7].

1



C. E. Duru et al. / J. Nig. Soc. Phys. Sci. 5 (2023) 1154 2

Plastic degradation is the change in the polymer’s shape,
colour, tensile strength, and molecular weight under the influ-
ence of chemicals, light, heat, or applied force. Though plastic
degradation represents the failure of the polymer to perform
a required service, the process can be useful as it concerns
the recycling of polymer wastes to reduce the environmental
pollution they can cause [3]. These days, the most popular
methods for getting rid of plastic garbage in the environment
include landfilling, burning, and mechanical and chemical
recycling [8]. Landfilling is the main method of getting rid
of plastic trash in most countries, especially impoverished
ones because it is straightforward and inexpensive. However,
accumulating plastic trash has consumed a substantial amount
of space. While burning plastic garbage may assist reduce
the need for landfill space and provide thermal energy, we
also need to consider the environmental effects of secondary
pollutants, including dioxins, carbon monoxide, and nitrogen
oxides formed during the incinerating process. Even though
mechanical recycling has grown to be the most popular method
for recovering thermoplastic wastes, the bulk of recovered
materials significantly degrade after a few processing cycles,
limiting their economic value. The success of chemical
recycling on the other hand depends on the accessibility of the
processes and the effectiveness of the catalysts [9].

Currently, there are reports that natural enzymes can
catalyze the hydrolysis of micro and nanoplastics as an
alternative to chemical processes [10, 11]. The ester bond
linkages in PET can be hydrolyzed by various esterases like
PETase [12], cutinase [13], and lipase [14]. The extent of
hydrolysis of PET by these enzymes is quite low and yields the
monomeric units mono-2-hydroxyethylterephthalate (MHET)
and bis(2-hydroxyethyl)terephthalate (BHET) [15]. These
monomers can further be degraded into ethylene glycol and
terephthalic acid, the initial reactants used for their formation.
The high recalcitrant nature of PET waste is a major bottleneck
in its hydrolysis by enzymes. Factors such as hydrophobicity,
the crystallinity of PET, low accessibility, and structure usually
limit enzyme function, thus making depolymerization very
difficult [16]. Therefore, finding more effective and sustainable
methods for breaking down PET plastic and other polymers
could have significant benefits in terms of reducing pollution
and increasing the recyclability of plastic waste.

In this study, we applied computational techniques [17] for
the first time in studying the ease of hydrolysis of different PET
nanoplastic chains by a microbial enzyme. Molecular dock-
ing of some PET nanoplastic oligomers was performed at the
active site of the novel enzyme Polyester Hydrolase Leipzig 7
(PHL7). The binding affinities of the oligomers at the active
site of the enzyme were used to estimate the chain lengths at
which PET hydrolysis is most effective. Also, artificial neural
network (ANN) was used to identify ligand-dependent factors
responsible for the degradation process.

2. Computational Methods

2.1. Preparation of PET nanoplastic oligomers

Polyethylene terephthalate (PET) nanoplastic (NP) chains
ranging from 1-10 polymer units were designed in Chemdraw
and saved in MDL SDfile fileformat [18]. They were opti-
mized using Open Babel in Python Prescription (version 0.8),
which converted them to their most stable structures using
Merk Molecular Force Field 94 (MMFF94). The optimized
structures (Table 1) were used as small molecules in the study.

2.2. Identification and preparation of enzyme target

The 3D X-ray crystallographic structure of the novel en-
zyme Polyester Hydrolase Leipzig 7 (PHL7) with identity 7NEI
[19] was retrieved from the protein data bank (PDB). The Chain
A of the enzyme was used as target to study the effect of chain
length on PET NP hydrolysis. Removal of the interfering crys-
tallographic water molecules and minimization of the protein
was done using UCSF Chimera 1.14 [20, 21, 22].

2.3. Molecular docking studies

Site-directed docking of the PET NP chains was performed
on the active site of the enzyme with Autodock Vina in PyRx
software version 0.8 [23]. The amino acids at the active site
were selected and toggled on the enzyme surfaces in the Pyrx
software. The specific site on the receptor was set using the grid
box with dimensions:center x: 22.249, center y: – 1.869, cen-
terz: – 22.211, and size x: 23.795, size y: 14.112, sizez: 15.463.
At the end of the molecular docking, the binding poses of the
enzyme-ligand complex were generated, and their scoring re-
sults were also created. Theinteractions between the enzyme-
NP complex were visualized using BioviaDiscovery Studio 4.5
[24].

2.4. Artificial Neural Network (ANN) Analysis

Computational models with numerous processing layers
may learn data representations at various abstraction levels
[25, 26]. The current study employed a feedforward artificial
neural network (ANN) made up of many perceptron layers
(with threshold activation).

Backpropagation is a supervised learning method that the
ANN uses during training. A minimum of three layers of nodes
make up this ANN: the input layer, the hidden layer, and the
output layer. Each node, except the input nodes, is a neuron
that employs a nonlinear activation function. The network in-
formation for the ANN is summarized in Table 2.

The input layers include the determined variables from the
PET NPs oligomers, including energy, molecular mass (MM),
and chain length (CL). The ANNs used 70 % of the input data
to train the model, while 30 % was used for testing the model.
The ANN had two hidden layers based on a hyperbolic tangent
activation function. The dependent variable or binary classifica-
tions are contained in the output layer. The ANN will examine
the output layer’s dependent and the input layer’s independent

2



C. E. Duru et al. / J. Nig. Soc. Phys. Sci. 5 (2023) 1154 3

Table 1: Optimized NP chains used for the study

NP Chain length Structure Chemical formula Molecular mass Minimized energy (Ha)

NP1 C10H10O5 210.18 76.01

NP2 C20H18O9 402.35 195.80

NP3 C30 H26 O13 594.52 418.82

NP4 C40 H34 O17 786.69 577.84

NP5 C50 H42 O21 978.86 633.86

NP6 C60 H50 O25 1171.02 995.98

NP7 C70 H58 O29 1363.19 1427.46

NP8 C80 H66 O33 1555.36 1159.54

NP9 C90 H74 O37 1747.53 2118.52

NP10 C100 H82 O41 1939.70 2415.53

3



C. E. Duru et al. / J. Nig. Soc. Phys. Sci. 5 (2023) 1154 4

Table 2: Network Information for the ANNs

Input layer
Covariates

1 Energy
2 Molecular mass (MM)
3 Chain length (CL)

Number of Unitsa 3
Rescaling Method for Covariates Standardized

Hidden Layer(s)

Number of Hidden Layers 2
Number of Units in Hidden Layer 1a 1
Number of Units in Hidden Layer 2a 1
Activation Function Hyperbolic tangent

Output Layer

Dependent Variables 1 Binding affinity
Number of Units 1
Rescaling Method for Scale Dependents Adjusted normalized
Activation Function Hyperbolic tangent
Error Function Relative error Sum of

Squares
a Excluding the bias unit

variables during training to see how they relate to one another.
The hidden layer’s nodes include mathematical functions that
define the relationships. Once the connections have been es-
tablished, the testing data will be used to validate them. Error
functions, such as the relative error (RE) and the sum of squares
error (SSE) shown in equations (1) and (2), would be used to
verify and evaluate how well an ANN model predicts the out-
put.

Figure 1: Structure of PET

RE =
∣∣∣∣∣ BAA − BAPBAP

∣∣∣∣∣ × 100 (1)
S S E =

∑
(BAP − BAA)

2 , (2)

where (BA)P is the estimated value of the binding affinity in
kcal/mol by ANN model, (BA)A is the experimental value of
the binding affinity in kcal/mol.

Table 3: Chain lengths and binding affinity of NPs chains at the PHL7 active
site

NP oligomers Chain length
(nm)

Binding affinity
(kcal/mol)

NP1 1.09 - 5.4
NP2 2.18 - 5.2
NP3 3.27 - 5.1
NP4 4.36 - 3.8
NP5 5.45 - 0.8
NP6 6.54 23.2
NP7 7.63 38.5
NP8 8.72 51.4
NP9 9.81 137.8
NP10 10.90 139.8

Figure 2: Plot of PET chain length against binding affinity at PHL7 active site

4



C. E. Duru et al. / J. Nig. Soc. Phys. Sci. 5 (2023) 1154 5

Figure 3: Fitting of NP oligomers at the enzyme binding pocket (A) NP1 (B) NP2 (C) NP3 (D) NP4 (E) NP5 (F) NP6

Figure 4: ANNs for predicting PET NPs degradation based on its properties-
Energy, Molecular mass (MM), and Chain length (CL)

Figure 5: Linear correlation of predicted binding affinity and actual values from
molecular docking

5



C. E. Duru et al. / J. Nig. Soc. Phys. Sci. 5 (2023) 1154 6

Figure 6: The most important property for NPs degradation by PHL7 from the
ANN prediction

3. Results and discussion

Polyethylene terephthalate is a semi-crystalline polymer
produced from the reaction of terephthalic acid and ethylene
glycol. Each monomer unit (Figure 1) has a physical length of
about 1.09 nm and a molecular weight of ≈ 200 [27].

It is a thermoplastic material with excellent chemical re-
sistance, melt mobility, and spinnability. The bacterial strain
Ideonellasakaiensis 201-F6 was recently found to exhibit a
rare ability to grow on PET as a major carbon and energy
source [28]. A novel enzyme, Polyester Hydrolase Leipzig
7 (PHL7), isolated from a compost metagenome, can com-
pletely hydrolyze amorphous PET films within hours has been
freshly reported by Sonnendecker and coworkers [19]. We have
shown in this study the binding affinities of modeled PET NP
oligomers at the PHL7 active site, and the results are given in
Table 3.

The PET monomer, which served as control in this study,
had a binding affinity of - 5.4 kcal/mol at the enzyme target and
was closely followed by NP2 and NP3 with binding affinities
of -5.2 kcal/mol and - 5.1 kcal/mol respectively. These results
suggested that the degradation of PET at this enzyme target
would be most efficient at 2.18 nm and 3.27 nm chain lengths.
The drop in the binding affinity from -3.8 kcal/mol for NP4 to
– 0.8 kcal/mol for NP5 indicated a reduction in the hydrolyzing
ability of the enzyme as the chain length increased from 4.36
nm to 5.45 nm.

Positive binding affinity values, which increased steadily,
were obtained with NP6-NP10 oligomer units. This implied
that the hydrolysis of PET within the range of 6.54 nm chain
length and above was not feasible with this enzyme. Therefore,
PET polymer materials should be within 5.45 nm and below
for efficient binding and hydrolysis at the PHL7 active site. A
plot of the binding affinity of PET NP oligomers as a function
of NP chain length is shown in Figure 2. As chain length

increased, three notable degradation characteristics of PET at
the enzyme target were observed. There is a reactive stage
at NP2-NP5 where hydrolysis was feasible, an intermediate
nonreactive stage at NP6-NP8 where enzyme action is highly
limited, and a recalcitrant stage at NP8 and above where total
recalcitrance of the NP was manifest.

The 3D views of enzyme-NP interactions showed that
the PET trimer (NP3) is the maximum chain unit that can fit
perfectly in the enzyme binding cavity (Figure 3). This fitting
becomes increasingly difficult from NP5 and above. The recal-
citrant nature of PET polymer and the difficulty in its hydrolysis
could therefore be reduced by subdividing it to less than 5.45
nm chain lengths which have good fitting in the enzyme pocket.

The ANN for the degradation of PET NPs oligomers by
PHL7 based on the intrinsic properties of the oligomers is
shown in Figure 4. The network involved two hidden layers,
which processed the input variables. The output results from
this network were compared with the actual binding affinity
values from the molecular docking by linear regression (Figure
5). The results showed that the ANNs could predict the binding
affinity with high accuracy with R2 of 0.985. The ANN was
further checked using error models, which gave small errors
for SSE (0.045) and RE (0.028). The results indicated that the
energy of the oligomers was the most important property for
its degradation, with 100 % normalized importance (Figure
6). The energy of the oligomers increases with an increase in
their sizes. The high energy of the longer oligomers reduced
their binding affinity at the active site, which makes their
degradation difficult at the enzyme target. Efficient degradation
would therefore occur at a lower energy of the NPs, which is
obtainable at chain lengths between NP2-NP4.

4. Conclusion

The molecular docking of PET NP oligomers ranging from
1.09 nm - 10.90 nm at the PHL7 active site was performed
and their binding affinities were used to determine the effect
of chain length on the degradation of PET. Binding affinities
of the NPs decreased from -5.2 kcal/mol to -0.8 kcal/mol, as
PET chain length increased from 2.18 nm to 5.45 nm (2-5
PET chains). At chain lengths of 6.54 nm (6 PET chains) and
above, the binding of the PET NPs on the enzyme was non-
spontaneous, as was seen in the resulting positive binding affin-
ity values obtained with these chains. Artificial Neural Network
analysis revealed that structural energy is the major determinant
factor in the PET degradation process.

References

[1] J. Gigault, A. T. Halle, M. Baudrimont, P. Y. Pascal, F. Gauffre, T. L.
Phi, H. El Hadri, B. Grassl & S. Reynaud, “Current opinion: What is a
nanoplastic?”, Environmental Pollution 235 (2018) 235.

[2] E. L. Ng, L. E. Huerta, S. M. Eldridge, P. Johnston, H. W. Hu, V. Geissen
& D. Chen, “An overview of microplastic and nanoplastic pollution in
agroecosystems”, Science of the Total Environment 627 (2018) 1377.

6



C. E. Duru et al. / J. Nig. Soc. Phys. Sci. 5 (2023) 1154 7

[3] C. E. Enyoh, Q. Wang, T. Chowdhury, W. Wang, S. Lu, K. Xiao & M. A.
H. Chowdhury, “New analytical approaches for effective quantification
and identification of nanoplastics in environmental samples”, Processes 9
(2021) 2086.

[4] Q. Wang, C. E. Enyoh, T. Chowdhury & A. H. Chowdhury, “Analytical
techniques, occurrence and health effects of micro and nano plastics de-
posited in street dust”, International Journal of Environmental Analytical
Chemistry 102 (2022) 6435.

[5] M. R. M. Zaki & A. Z. A. Aris, “An overview of the effects of nanoplas-
tics on marine organisms”, Science of the Total Environment 831 (2022)
154757.

[6] L. M. Hernandez, N. Yousefi & N. Tufenkji, “Are there nanoplastics in
your personal care products?”, Environmental Science and Technology
Letters 4 (2017) 280.

[7] A. Banerjee, L. O. Billey & W. L. Shelver, “Uptake and toxicity of
polystyrene micro/nanoplastics in gastric cells: Effects of particle size
and surface functionalization”, PLOS ONE 16 (2021) e0260803.

[8] J. Ru, Y. Huo & Y. Yang, “Microbial degradation and valorization of plas-
tic wastes”, Frontiers in Microbiology 11 (2020) 442.

[9] A. Rahimi & J. M. Garcı́a, “Chemical recycling of waste plastics for new
materials production”, Nature Reviews Chemistry 1 (2017) 0046.

[10] C. E. Duru, I. A. Duru & C. E. Enyoh, “In silico binding affinity analysis
of microplastic compounds on PET hydrolase enzyme target of Ideonella
sakaiensis”, Bulletin of the National Research Centre 45 (2021) 104.

[11] R. Wei & W. Zimmermann, “Microbial enzymes for the recycling of re-
calcitrant petroleum-based plastics: how far are we?”, Microbial Biotech-
nology 10 (2017) 1308.

[12] H. P. Austin, M. D. Allen, B. S. Donohoe , N. A. Rorrer, F. L. Kearns,
R. L. Silveira, B. C. Pollard, G. Dominick, R. Duman, K. El Omari, V.
Mykhaylyk, A. Wagner, W. E. Michener, A. Amore, M. S. Skaf, M. F.
Crowley, A. W. Thorne, C. W. Johnson, H. L. Woodcock, J. E. McGee-
han & G. T. Beckham, “Characterization and engineering of a plastic-
degrading aromatic polyesterase”, Proceedings of the National Academy
of Sciences 115 (2018) 4350.

[13] S. Sulaiman, S. Yamato, E. Kanaya, J. J. Kim, Y. Koga, K. Takano & S.
Kanaya, “Isolation of a novel cutinase homolog with polyethylene tereph-
thalate degrading activity from leaf-branch compost by using a metage-
nomic approach”, Applied and Environmental Microbiology 78 (2012)
1556.

[14] G. A. Macedo & T. F. Pio, “A rapid screening method for cutinase pro-
ducing microorganisms”, Brazilian Journal of Microbiology 36 (2005)
388.

[15] A. Bollinger, S. Thies, E. Knieps-Grünhagen, C. Gertzen, S. Kobus, A.
Höppner, M. Ferrer, H. Gohlke, S. H. J. Smits & K. E. Jaeger, “A novel
polyester hydrolase from the marine bacterium Pseudomonas aestusnigri-

structural and functional insights”, Frontiers in Microbiology 11 (2020)
114.

[16] R. Koshti, L. Mehta & N. Samarth, “Biological recycling of polyethylene
terephthalate: a mini-review”, Journal of Polymers and the Environment
26 (2018) 3520.

[17] C. W. Chidiebere, C. E. Duru & J. P. C. Mbagwu, “Application of com-
putational chemistry in chemical reactivity: A review”, Journal of the
Nigerian Society of Physical Sciences 3 (2021) 292.

[18] A. K. Nelapati & K. Meena, “An approach to increase the efficiency of
uricase by computational mutagenesis”, Physical Chemistry Research 11
(2023) 481.

[19] C. Sonnendecker, J. Oeser, P. K. Richter, P. Hille, Z. Zhao, C. Fischer, H.
Lippold, P. Blázquez-Sánchez, F. Engelberger, C. A. Ramı́rez-Sarmiento,
T. Oeser, Y. Lihanova, R. Frank, H. G. Jahnke, S. Billig, B. Abel, N.
Sträter, J. Matysik & W. Zimmermann, “Low carbon footprint recycling
of post-consumer PET plastic with a metagenomic polyester hydrolase”,
ChemSusChem 15 (2022) e202101062.

[20] E. F. Pettersen, T. D. Goddard, C. C. Huang, G. S. Couch, D. M. Green-
blatt, E. C. Meng & T. E. Ferrin, “UCSF Chimera- a visualization system
for exploratory research and analysis”, Journal of Computational Chem-
istry 25 (2004) 1605.

[21] C. E. Duru, I. A. Duru & A. Bilar, “Computational investigation of sugar
fermentation inhibition by bergenin at the pyruvate decarboxylate isoen-
zyme 1 target of Scharomyces cervisiae”, Journal of Medicinal Plants
Studies 8 (2020) 21.

[22] C. E. Duru, I. A. Duru & C. W. Chidiebere, “Virtual screening of selected
natural products as humantyrosinase-related protein 1 blocker” Journal of
the Nigerian Society of Physical Sciences 3 (2021) 154.

[23] O. Trott & A. J. Olson, “AutoDock Vina: improving the speed and accu-
racy of docking with a new scoring function, efficient optimization and
multithreading”, Journal of Computational Chemistry 31 (2010) 455.

[24] BIOVIA, Dassault Systemes, San Diego, Discovery studio modeling en-
vironment, (2020).

[25] Y. LeCun, Y. Bengio & G. Hinton, “Deep learning”, Nature 521 (2015)
436.

[26] C. L. Udeze, I. E. Eteng & A. E. Ibor, “Application of Machine Learning
and Resampling Techniques to Credit Card Fraud Detection” Journal of
the Nigerian Society of Physical Sciences 4 (2022) 769.

[27] S. Venkatachalam, S. G. Nayak, J. V. Labde, P. R. Gharal, K. Rao & A. K.
Kelkar, “Degradation and recyclability of poly (ethylene terephthalate)”,
In Polyester , IntechOpen, 2012, 75-98.

[28] G. J. Palm, L. Reisky, D. Böttcher, H. Muller, E. A. P. Michels, M. C.
Walczak, L. Berndt M. S. Weiss, U. T. Bornscheuer & G. Weder, “Struc-
ture of the plastic-degrading Ideonella sakaiensis MHETase bound to a
substrate”, Nature Communications 10 (2019) 1717.

7