Organic Contaminant Biodegradation by Oxidoreductase Enzymes in Wastewater Treatment

Organic contaminants (OCs), such as pharmaceuticals, personal care products, flame retardants, and plasticisers, are societally ubiquitous, environmentally hazardous, and structurally diverse chemical compounds whose recalcitrance to conventional wastewater treatment necessitates the development of more effective remedial alternatives. The engineered application of ligninolytic oxidoreductase fungal enzymes, principally white-rot laccase, lignin peroxidase, and manganese peroxidase, has been identified as a particularly promising approach for OC remediation due to their strong oxidative power, broad substrate specificity, low energy consumption, environmental benignity, and cultivability from lignocellulosic waste. By applying an understanding of the mechanisms by which substrate properties influence enzyme activity, a set of semi-quantitative physicochemical criteria (redox potential, hydrophobicity, steric bulk and pKa) was formulated, against which the oxidoreductase degradation susceptibility of twenty-five representative OCs was assessed. Ionisable, compact, and electron donating group (EDG) rich pharmaceuticals and antibiotics were judged the most susceptible, whilst hydrophilic, bulky, and electron withdrawing group (EWG) rich polyhalogenated compounds were judged the least susceptible. OC susceptibility scores were in general agreement with the removal rates reported for experimental oxidoreductase treatments (R2 = 0.60). Based on this fundamental knowledge, and recent developments in enzyme immobilisation techniques, microbiological enzymic treatment strategies are proposed to formulate a new generation of biological wastewater treatment processes for the biodegradation of environmentally challenging OC compounds.


Introduction
Wastewater treatment plant (WWTP) effluents and biosolids are regularly found to contain a wide variety of heterogeneously structured novel micropollutants as chemically engineered production of anthropogenic compounds continues to diversify. Environmental organic contaminants (OCs) can be broadly characterised by their prevalence as ubiquitous necessities (e.g., antibiotics, pharmaceuticals, personal care products, hormones, plasticisers, flame retardants, surfactants, biocides, and preservatives), their recalcitrance to conventional wastewater treatment and natural degradation processes, their environmental persistence and bioaccumulation, their ubiquity at analytically challenging yet ecologically harmful ppb/ppt environmental concentrations, and their general lack of current source control legislation [1]. Importantly, whilst the environmental impacts of OC emission include endocrine disruption, food-web accumulation, ecological toxicity/mutagenicity/teratogenicity, and antibiotic resistance dissemination, the continuous discharge of OCs in wastewater effluent allows for even the least persistent OCs to potentially impart some degree of ecological harm upon receiving

Criterion Selection and Its Rationale
Four physicochemical properties of contaminants that affect their suitability as fungal enzyme substrates are considered as criteria for the assessment process and are described, as follows.
• Criterion 1 (C1): Net electron donating group density As the one-electron oxidation reactions oxidoreductases employ initialise substrate transformation are driven by the redox potential differential (∆E) between the enzyme and its substrate, substrate molecules with lower redox potentials, weaker electron affinities, and, thus, greater electron transfer tendencies will be more susceptible to enzymatic oxidation, and subsequent transformation/removal [16]. Unfortunately, the structural density of electron donating groups (EDG) and electron withdrawing groups (EWG) was adopted as a proxy for redox potential, as the availability of redox potential data for the shortlisted contaminants is extremely limited. Here, the term: Net Electron Donating Group Density, represents the presence, balance, and strength of donating and withdrawing groups within a given contaminant.
• Criterion 2 (C2): Hydrophobicity Hydrophobicity represents the tendency of a compound to partition and sorb from the aqueous phase into the nonaqueous phase [17]. Hydrophobicity is known to heavily impact upon the removal rates that were observed in wastewater treatment processes [18,19] and also upon the susceptibility of contaminants to purified enzyme degradation [20,21].
• Criterion 4 (C4): pKa Some studies [24,25] showed that compounds with lower pKa values display faster oxidation rates under physiological pH conditions and micropollutants can be more readily degraded by oxidoreductase at a pH that is higher than its pKa.
As such, the criteria weightings were set to 9-3-3-1 for net electron donating group density, hydrophobicity, steric hindrance, and pKa, respectively. To produce a value of C1, the following methodology was adopted: • Firstly, each donating and withdrawing substituent within the molecular structure of the shortlisted contaminants was identified and then assigned a score of 3, 2, or 1, relative to its classification as a strong, medium, or weak group, respectively (Table 2).

•
The donating and withdrawing group scores of each contaminant were separately tallied and the difference between these summations calculated.

•
This difference was then divided by the contaminants molecular weight before being multiplied by 100 to provide a final normalised density-based parameter with units of net donating group score per 100 g/mol.
Overall, the contaminants with a final net electron donating group density value >1 (strong donating group dominance), 0-1 (slight/moderate donating group dominance), or <0 (withdrawing group dominance) were assigned a susceptibility score of 9, 6, or 3, respectively. Contaminants only containing withdrawing groups were assigned a score of 0. However, it must be clarified that, as donating and withdrawing substituents influence redox potential via electron density modification of the aromatic moiety to which they are attached, only the redox potential of aromatic compounds is influenced by their presence. The functional groups that are attached to aliphatic molecules are, therefore, considered to have no positive impact on oxidoreductase susceptibility in this criterion category and, as such, aliphatic contaminants were assigned a net electron donating group density susceptibility score of 0.

Criterion 2: Hydrophobicity
Hydrophobicity of a substance is quantified by the log octanol-water partition coefficient (Log K ow ). The hydrophobic nature of a compound can be broadly classified, as follows: Log K ow < 2.5 = low hydrophobicity, Log K ow 2.5-4 = moderate hydrophobicity, and Log K ow > 4 = high hydrophobicity [29]. As such, shortlisted contaminants possessing high, moderate, and low log K ow values were assigned susceptibility scores of 3, 2, and 1, respectively.

Criterion 3: Steric Bulk
The steric bulk of a substituent is numerically represented by its A-value. Highly, moderately, and minimally bulky substituents were defined as those with A-values >3, between 3-1 and <1, respectively ( Table 2). Contaminants with high, medium, and low steric bulks were, respectively, assigned susceptibility scores of 1, 2, and 3, respectively.

Criterion 4: pKa
Shortlisted contaminants possessing pKa values that were above and below the typical pH of wastewater (pH 7.5) were assigned susceptibility scores of −1 and 1, respectively. The score of contaminants that was devoid of ionisable functional groups was left unchanged.

Data Collection and Analysis
For assessing the susceptibility of the OCs, Appendix A (Table A1) presents the detailed breakdown of EDG/EWG content, log K ow , steric bulk, and pKa data used to derive susceptibility scores for each shortlisted contaminant. Relevance analysis was carried out by statistical linear regression. Figure 1 presents the shortlisted contaminants in order of their assessed susceptibility to oxidoreductase degradation within a wastewater treatment context. The higher the total susceptibility score, the greater the predicted susceptibility to oxidoreductase degradation. From a possible total susceptibility score of 16, Gemfibrozil, a lipid regulating pharmaceutical, scored the highest (14/16), whilst CP-10, which is a short-chained halogenated paraffin/polychlorinated naphthalene, scored the lowest (2/16). The ten most susceptible contaminants comprised four non-halogenated antibiotics (trimethoprim, erythromycin, tetracycline, and sulfamethoxazole), five non-halogenated pharmaceuticals/personal care products (gemfibrozil, metoprolol, benzophenone-3, galaxolide, and oestrone), and one non-halogenated plasticiser (DEHP). Conversely, the fifteen least susceptible contaminants comprised two halogenated surfactants (PFOA, PFOS), four halogenated flame retardants (PBEB, TDCPP, BDE-209, and 2,3,6,7-PCN), one halogenated plasticiser (CP-10), one halogenated antibiotic (ofloxacin), three pharmaceuticals (carbamazepine, diclofenac and fluoxetine), three biocides (OBT, TBT, triclosan), and one combustion by-product (TCDD).
Overall, the halogenated contaminants possessed, on average, a 32% lower susceptibility score (6.6/16) than non-halogenated compounds (9.7/16). This is consistent with the fact that the halogen (-X) groups are electron withdrawing groups; if not balanced with other electron donating groups within the molecule, they tend to increase the redox potential and electrophilicity of a compound by withdrawing electron density away from the aromatic ring [16]. The resulting electron sufficiency of the aromatic moiety decreases the compounds' susceptibility to oxidative catabolism.

Comparison between the Susceptibility Scores of OCs and Their Removal by Oxidoreductase Treatment
The results of the susceptibility assessment were compared with the published studies of the removal rates that were attained by various oxidoreductase treatment strategies ( Figure 2). Linear regression analysis of the relationship showed that the degradation potential of a diverse range of OCs showed good agreement with the susceptibility scores for the compounds and gave an R 2 = 0.60 (p value = 0.002), providing confidence in the assessment framework to predict the susceptibility of OCs to oxidoreductase degradation. withdrawing electron density away from the aromatic ring [16]. The resulting electron sufficiency of the aromatic moiety decreases the compounds' susceptibility to oxidative catabolism.

Comparison between the Susceptibility Scores of OCs and Their Removal by Oxidoreductase Treatment
The results of the susceptibility assessment were compared with the published studies of the removal rates that were attained by various oxidoreductase treatment strategies ( Figure 2). Linear regression analysis of the relationship showed that the degradation potential of a diverse range of OCs showed good agreement with the susceptibility scores for the compounds and gave an R 2 = 0.60 (p value = 0.002), providing confidence in the assessment framework to predict the susceptibility of OCs to oxidoreductase degradation.  Table 1).
However, for certain shortlisted contaminants, susceptibility scores and experimental removal reports did not always fully agree. The oxidoreductase degradation of triclosan, in particular, is more efficient (70-90%) than is expected from its susceptibility score (8/16) [40]. Triclosan is, however, the  Table 1).
However, for certain shortlisted contaminants, susceptibility scores and experimental removal reports did not always fully agree. The oxidoreductase degradation of triclosan, in particular, is more efficient (70-90%) than is expected from its susceptibility score (8/16) [40]. Triclosan is, however, the only shortlisted OC to bear a combination of halogen and phenolic substituents and, as such, the poor association between susceptibility score and the removal rate could arise from the idiosyncratic suppression of halogenic electron withdrawing effects by the donating and polymerisation inducing phenolic substituent [41]. More generally, such inconsistencies might be explained by the non-incorporation of additional known (e.g., reorganisation energy [42]) and ionisation potential [3] and unknown susceptibility criteria.

Enzyme Reactors and Enzyme Stabilisation
Oxidoreductase enzymes have been applied experimentally within a wide array of engineered systems. Different types of enzymatic reactor have been reported, including: aqueous-phase orientated batch or continuous stirred tank, packed bed, fluidised bed, suspended nanoparticle, hollow fibre microfilter and hybrid membrane-nanoparticle suspension bioreactor systems, as well as solid-phase centric bioslurry and biopile technologies [33,39,46]. However, high rates of pH/thermal enzyme denaturation and low rates of enzyme retention, recovery, and reusability challenge the contaminant removal efficiency and the economic feasibility of enzymatic wastewater treatment [47]. Therefore, the stabilisation of the enzymes is necessary for practical implementation of enzymic treatment. Crucially, immobilisation facilitates the combinatory aggregation or successive application of disparate enzymes, which, in turn, provides the versatility that is required to treat complex pollutant mixtures under fluctuating or poorly defined conditions, which is often the case in wastewater matrices [46]. Conventional immobilisation of enzymes onto solid-supports/carriers (e.g., mesoporous silica and organic gels) by physisorption, entrapment, or encapsulation is regularly reported to reduce enzyme washout from continuous processes, widen enzymatic temperature and pH stability ranges through intermolecular stabilising forces, increase substrate availability by consolidative adsorption, and provide protection from inhibitory molecules [48][49][50]. Recently, an alternative, carrier-free immobilisation technique has been developed to stabilise the enzymes by moderate, covalent bonding with a cross-linker. This immobilisation approach reduces the activity losses that are associated with the inaccessibility of enzymes situated deep within the pores of conventional solid-supports, and it has shown potential for producing enzymatic microreactors to remove organic contaminants from wastewater [51]. For example, Lai et al. prepared nanotubes with laccases that were cross-linked to the surface to remove azo dyes. It was found that a decolourisation rate of 74-96% was possible and 90% of the initial decolourisation rate of the reactor was maintained after 10 sequential batch reactions [52]. Similar results were reported in [53] while using enzymatic membrane reactors that were produced by cross-linking. Lassouane et al. [54] also found that crosslinking of laccases prior to a conventional entrapment method increased the immobilisation efficiency by 30%, and that more than 99% of the contaminant, Bisphenol A, was removed from aqueous solution by cross-linked enzymes.

Optimisation of Enzymatic Treatment
The susceptibility assessment results that are presented here provide a rationale towards establishing potential process design recommendations for the enzymatic treatment of OCs: the treatment performance might be optimised by selecting and manipulating enzymes and wastewater conditions to exploit the fundamental OC susceptibility relationships, as follows.

•
Redox Potential: If the typical OC content of a given wastewater stream can be established, and the redox potential of these OCs approximated from EDG/EWG analysis, an enzyme, or redox mediator can be selected that delivers the redox potential that is required for oxidation. The treatment of polyhalogenated OCs in wastewater, for example, is more likely to require the application of lignin peroxidase (1.0-1.2 V) over laccase (0.3-0.8 V), or ABTS (0.68 V) over syringaldehyde (0.51 V) [55], when compared to a predominantly antibiotic/pharmaceutical contaminated influent. • Steric Bulk/Hindrance: Apart from stabilising the enzymes, the immobilisation techniques were also reported to minimise the steric hindrance of E-S complex formation. PEG, for example, can be employed as a flexible spacer that tethers and non-directly spatially disseminates enzymes across some solid-supports [56]; employment of these flexibly spaced enzymes in a treatment process may orientate certain subunits of OCs towards the enzymes active site in a more favourable way.

Conclusions
Twenty-five representative OCs were shortlisted for susceptibility assessment to degradation by lignin degrading fungal enzymes. The physicochemical properties that determine the ability of OCs to act as fungal oxidoreductase substrates, namely redox potential, hydrophobicity, steric bulk, and pKa, were translated into semi-quantitative assessment criteria. The susceptibility scores suggested that oxidoreductase treatment systems are the most effective at degrading EDG-rich, hydrophobic, compact, and ionisable aromatic contaminants, e.g., oestrone, and least effective at degrading EWG-rich, hydrophilic, bulky, and non-ionisable aliphatic contaminants, e.g., PFOS.
The assessment framework successfully incorporated the major criteria that control the efficacy of oxidoreductase OC degradation relative to the reported oxidoreductase treatment removals (R 2 = 0.60, p value = 0.002). Although susceptibility scores and oxidoreductase treatment removal rates largely agreed, the susceptibility of triclosan was underestimated. The accuracy of susceptibility scores may be improved by incorporating additional criteria, or by calibrating criteria weightings by experimentally determining the relative importance of each physicochemical property.
As far as we know, this study is the first to construct a theoretical framework for predicting contaminant susceptibility to oxidoreductase degradation. As such, it has demonstrated the potential for, and practical utility of, theoretically linking the fundamental characteristics of disparate molecules towards deriving a value that broadly signifies the degree to which they will interact. Reasonably, this concept could be adapted for assessing the susceptibility of the contaminants to other treatment processes e.g., advanced oxidation, etc.
Overall, as novel contaminants continue to emerge, the swift and strategic design and deployment of targeted wastewater treatments will become increasingly necessary, this will be more effective if based on theoretical, fundamental assessment frameworks that are similar to that proposed here, to provide rapid and reliable estimations of compound susceptibility and process efficacy. Ultimately, frameworks that predict the performance of oxidoreductase systems could, if developed alongside systems for improving mass enzyme producibility and immobilisation, make the application of fungal enzymes for sustainable wastewater treatment feasible.
Author Contributions: S.R.S. was the academic supervisor, proposed the original conceptualization, critically directed the research as it progressed and completed the final review and editing of the manuscript; E.A.B. was the MSc student undertaking the research, who developed the approach, completed the validation and prepared the thesis on which the paper was based; Z.L. was cosupervisor, contributing to the technical development of the research and prepared the manuscript for submission. All authors have read and agreed to the published version of the manuscript.
Funding: This research was unfunded.

Conflicts of Interest:
The authors declare no conflict of interest.