How Mechanistic Enzymology Helps Industrial Biocatalysis: The Case for Kinetic Solvent Viscosity Effects

Abstract

Biocatalysis is one of the oldest fields that has been used in industrial applications, with one of the earliest purposeful examples being the mass production of acetic acid from an immobilized Acinetobacter strain in the year 1815. Efficiency, specificity, reduced reaction times, lower overall costs, and environmental friendliness are some advantages biocatalysis has over conventional chemical synthesis, which has made biocatalysis increasingly used in industry. We highlight three necessary fields that are fundamental to advancing industrial biocatalysis, including biocatalyst engineering, solvent engineering, and mechanistic engineering. However, the fundamental mechanism of enzyme function is often overlooked or given less attention, which can limit the engineering process. In this review, we describe how mechanistic enzymology benefits industrial biocatalysis by elucidating key fundamental principles, including the k_cat and k_cat/K_m parameters. Mechanistic enzymology presents a unique field that provides in-depth insights into the molecular mechanisms of enzyme activity and includes areas such as reaction kinetics, catalytic mechanisms, structural analysis, substrate specificity, and protein dynamics. In line with the objective of protein engineering to optimize enzyme activity, we summarize a range of strategies reported in the literature aimed at improving the product release rate, the chemical step of catalysis, and the overall catalytic efficiency of enzymes. Further into this review, we delineate kinetic solvent viscosity effects (KSVEs) as a very efficient, cost-effective, and easy-to-perform method to probe different aspects of enzyme reaction mechanisms, including diffusion-dependent kinetic steps and rate-limiting steps. KSVEs are cost-effective because simple kinetic enzyme assays, such as the Michaelis–Menten kinetic approach, can be combined with them without the need for specialized and costly equipment. Other techniques in protein engineering and genetic engineering are also covered in this review. Additionally, we provide information on solvent systems in enzymatic reactions, details on immobilized biocatalysts, and common misconceptions that misguide enzyme design and optimization processes.

1. Introduction

Hudges and Lewis defined biocatalysis as the application of enzymes to transform molecules into desired products [1] (Figure 1). Biocatalysis has been identified as an alternative to conventional chemical synthesis because it uses natural or engineered enzymes and microorganisms to accelerate chemical reactions [2,3]. Several thousand years of history can be traced in industrial biocatalysis, with the earliest applications around 6000 BC or earlier [4,5]. One of the oldest purposeful applications of biocatalysis is the large-scale production of acetic acid from ethanol by an Acinetobacter strain in 1815 [6]. Biocatalysis has since been a rapidly developing and expanding field in industry because of its benefits, which include improved efficiency and specificity, shorter reaction times, lower overall costs, and environmental friendliness [7]. The production of several pharmaceutical compounds, including those used to treat heart disease, cancer, and bacterial and viral infections, has been achieved through biocatalysis [8,9].

The primary focus of industrial protein engineering is to develop enzymes with optimized catalytic activity, substrate affinity and specificity, binding affinity, and stability by adopting a rational design or directed evolution approach [10,11] (Figure 2). We noticed a trend in some biocatalytic studies, especially in industrial research, where, in the quest to modify enzymes for optimal turnover rates, the underlying mechanisms of enzyme function were overlooked or given less priority. When there is a limited understanding of the mechanistic changes occurring in a modified enzyme, the likelihood of encountering challenges in predicting the stability and scalability of the enzyme is high [10,12]. Additionally, the potential for further optimization may be reduced by a lack of mechanistic information. To illustrate, by identifying the rate-limiting steps in a reaction, approaches that target modifications of such steps in the catalytic pathway can be prioritized. Another disadvantage of a results-oriented approach is the lack of control over tradeoffs during enzyme modifications since the absence of thorough knowledge of the mechanistic pathways makes it challenging to monitor and determine the effect of such changes. A prime example is a study that sought to modify the activity of formate dehydrogenase (FDH) from Candida methylica (CmFDH) [13]. Following site-directed mutagenesis, the mutants showed a 6.5-fold increase in the k_cat value but also an increased K_m. The authors described the observed higher K_m value as a necessary compensatory mechanism in the enzyme’s active site to make room for the substrate when the bulky residues, Tyr and His, were substituted at Asn119 [13]. Therefore, the precise engineering of enzymes requires the integration of mechanistic studies.

Protein engineering has made significant progress, with a number of noteworthy advancements in rational design and directed evolution techniques. The rational design of enzymes requires an in-depth understanding of enzyme structures and mechanisms and utilizes a moderately sized library [14] (Figure 3). On the other hand, directed evolution only needs a basic understanding of the structure and function of enzymes and requires a larger library size [14] (Figure 3). Regardless of which approach is used in enzyme design, the desired functionality can be achieved by addressing tradeoffs between activity and stability. In protein engineering, the two common techniques of rational design include site-directed mutagenesis and domain swapping. Site-directed mutagenesis is a technique that is used to introduce specific, targeted changes to the DNA sequence of a protein, whereas domain swapping involves the exchange of structural elements between two or more identical protein monomers to form dimers or multimers with desired novel protein properties [15,16]. Merz et al. used a rational approach to enhance the catalytic activity of the thermophilic enzyme indoleglycerol phosphate synthase (IGPS) at low temperatures by mutating the trpC gene, which encodes IGPS [17]. Further investigation of the IGPS mechanistic pathway showed that the low activity of the enzyme was caused by its rigid structure. Specific amino acid substitutions were conducted to increase flexibility, which significantly increased catalytic turnover at low temperatures. The Nobel Prize-winning technique, directed evolution, aims to accelerate the process of natural evolution in biological molecules and systems in a test tube through repetitive rounds of gene variation and screening [18,19,20,21]. The development of continuous directed evolution systems, like yeast OrthoRep, is a recent advancement in directed evolution. Such systems allow for simultaneous hypermutation and selection in vivo at rates that are roughly 100,000 times faster than the host genome [22]. Protein engineering processes have been integrated with computational design and machine learning techniques to improve the prediction of protein properties and to facilitate the rapid exploration of vast sequence spaces [23]. ProteinMPNN has demonstrated high success rates in de novo enzyme design [24], while deep learning models, such as AlphaFold and RoseTTAFold, have greatly enhanced protein structure prediction [25]. Ancestral sequence reconstruction (ASR) is another trend in enzyme engineering design and optimization, which involves inferring and resurrecting ancestral protein sequences to create enzymes with enhanced properties [26]. Thermophilic enzymes that maintain high catalytic activity at lower temperatures have been engineered using ASR [26]. Historical sequence changes in the structure and function of enzymes can also be identified via ASR, which makes it possible to reconstruct a protein family’s evolutionary history and provide causal insights into functional diversities [27].

Biocatalysis has benefited from genetic engineering, which is a closely related field to protein engineering, differing in focus and techniques used. Genetic engineering is a process that involves the direct manipulation of an organism’s genes to enhance, modify, or remove specific properties pertaining to the organism [28]. In the context of biocatalysis, genetic engineering utilizes techniques such as Clustered Regularly Interspaced Short Palindromic Repeats (CRISPRs), cloning, and transformation to create new DNA constructs, commonly referred to as recombinant DNA [29]. Genetic engineering was used to construct a novel chimeric lysin (Ply187N-V12C) by fusing the catalytic domain of the bacteriophage lysin Ply187 with the cell-binding domain of the lysin PlyV12, which resulted in a hybrid that showed a broader lytic spectrum or host range [30]. Similarly, a soluble recombinant chimeric protein (CmAP/MBL-AJ) was created by fusing the genes encoding the marine bacterium Cobetia marina alkaline phosphatase (CmAP) and Far Eastern holothurian Apostichopus japonicus mannan-binding C-type lectin (MBL-AJ) [31]. The bifunctional hybrid demonstrated higher alkaline phosphatase activity and specificity relative to the wild-type.

Table 1. Genetically engineered proteins with their modified structural properties and industrial applications.

Protein	Major Structural Characteristic	Application	Ref.
Bacteriophage lysine, Ply187N-V12C	Catalytic domain of Ply187 lysin; bacterial cell-binding domain of PlyV12 lysin	Enhanced antimicrobial therapy due to an increase in lytic activity against human pathogens	[30]
Tumor-associated antigens MUC1 and HER2	MUC1subunit; HER2 subunit	Enhanced breast cancer detection in ELISA against MUC1 or HER2	[32]
α-Glucanohydrolase, mut-dexA	N-terminal mutanase domain; C-terminal dextranase domain	Enhanced α-glucan hydrolysis in fighting dental biofilm formation	[33]
Bacillus subtilis phytase	N-terminal β-1,4 endoglucanase domain; C-terminal phytase domain	Enhanced nutrition uptake as a potential feed additive for monogastric animals	[34]
Glucose 1-dehydrogenase IV, BmGlcDH-IV	Mutation G259A	Altered substrate specificity to exclusively D-glucose for clinical testing of blood glucose levels	[35]
Bacillus subtilis aminopeptidase, BSAP	Mutations I387A, I387C and I387S	Modified substrate specificity to enable hydrolysis of phenylalanine derivatives with bulky side chains	[36]
Transmembrane zinc metallopeptidase Neprilysin, NEP	Mutations G399V/G714K	Altered protein cleavage site specificity to target the Phe20-Ala21 bond in amyloid β 1-40 for Alzheimer’s disease treatment	[37]
Mucor hiemalis endo-b-N-acetylglucosaminidase, Endo-M	Mutation N175Q	Enhanced glycosynthase activity with oxazoline and natural N glycan for the synthesis of glycoproteins	[38]
Candida antarctica lipase B, CalB	Mutations G39A, T103G, W104F, A141Q, I189Y, L278A	Increasing amidase activity to three times that of the wild type for the production of optically active amines	[39]
Guanylate kinase (GK)	Mutation S68P	Altering GK function from enzymatic activity to protein-binding to regulate mitotic spindle orientation and cell adhesion	[40]

below summarizes some genetically engineered proteins with their corresponding modified structural properties and industrial applications.

While enzyme engineering has been significantly improved by developments such as directed evolution and computational design, mechanistic enzymology holds a unique and crucial role by focusing on the molecular basis of enzyme-catalyzed reactions. Mechanistic enzymology examines detailed aspects of enzymes, including reaction kinetics, catalytic mechanisms, structural analysis, substrate specificity, protein dynamics, and inhibition studies [41,42]. Knowledge obtained from a detailed mechanistic study is relevant in the rational design and optimization of enzymes for specific industrial purposes. Mechanistic studies enable the successful identification and alteration of key residues and motifs that are essential to catalytic pathways [14,43]. Mechanistic studies also promote sustainable chemical synthesis by expanding the repertoire of available biocatalytic transformations [44,45]. Furthermore, through mechanistic enzymology, conserved residue networks have been discovered to connect distal regions of enzymes to their active sites [46,47]. This discovery implies that the conventional paradigm of “structure encodes function” should be expanded to incorporate the role of protein dynamics and energy flow in enzyme activity [3].

In Figure 4, we present a conceptual framework that highlights three important areas of industrial biocatalysis: biocatalyst engineering, solvent engineering, and mechanistic enzymology. Biocatalyst engineering includes those processes that directly modify the intrinsic properties of the enzyme, such as protein engineering and genetic engineering. Processes that involve changes to the enzyme’s surrounding environment (extrinsic), including the choice of solvent system, are classified under solvent engineering. Mechanistic enzymology, although sometimes overlooked, represents a core area that provides insights into enzyme behavior and will be the focus of this review.

2. The Significance of Mechanistic Studies in Enzyme Engineering

Different techniques are implemented by mechanistic enzymology to obtain in-depth details of the structural and functional properties of enzymes [48]. One of the most widely used techniques in enzymology, site-directed mutagenesis, is used to probe the potential roles of specific amino acid residues by comparing the function and structure of wild-type and mutant enzymes [49,50,51]. Specific mutagenesis experiments not only complement structural and kinetic data but also confirm the catalytic role of active-site side residues of an enzyme [52]. For instance, research on alkaline phosphatase has shown that enzyme activity is reduced by 88,000 times when the metal-coordinating catalytic residue R166 is changed to a serine, suggesting that R166 plays a crucial role in Mg²⁺ coordination and transition-state stabilization [53]. In addition to single-site mutations, combinatorial mutagenesis of multiple residues can be used to assess cooperative interactions between spatially distant residues within a protein’s structure. In the same alkaline phosphatase study, mutants of five structurally linked residues, namely, D101, D153, R166, E322, and K328, were individually restored, and the results showed that the enzyme’s catalytic efficiency depended on the combined presence of all five residues, hence indicating synergistic interactions among them [53].

In addition to understanding the unique roles of amino acid residues in enzyme dynamics, information about enzyme function can be provided by solvent-sensitive steps in catalysis. Kinetic solvent viscosity effects (KSVEs) are a powerful technique that analyzes how changes in solvent viscosity affect the kinetic parameters of an enzymatic reaction to provide unique perspectives on the factors influencing enzyme function [48]. The theoretical basis of KSVEs, Kramers’ theory, lies in the principle that increased solvent viscosity creates friction against proteins in solution, decreasing motion and potentially inhibiting catalysis in motile enzymes [54]. The KSVE technique is considered a highly efficient and cost-effective technique because simple kinetic enzyme assays, such as the Michaelis–Menten kinetic approach, can be combined with it without requiring specialized and costly equipment. Techniques such as site-directed mutagenesis, pH effects, and kinetic isotope effects have been combined with KSVEs in recent research to provide detailed insights into the mechanisms of action of several enzymes, including D-2-hydroxyglutarate dehydrogenase, NADH:quinone oxidoreductase, and nitronate monooxygenase [55,56,57]. In a study on Pseudomonas aeruginosa NADH:quinone oxidoreductase, KSVEs with glucose and sucrose were used to probe the presence of a solvent-sensitive isomerization within the enzyme–substrate complex [56]. Further analysis revealed that flavin reduction was the rate-limiting step for the substrate benzoquinone, whereas hydride transfer from NADH to the FMN and quinol product release were partially rate-limiting for the substrate juglone [56]. Further details on the methods used to study KSVEs are covered in a review by Gadda and Sobrado [48]. Different aspects of enzyme reaction mechanisms can be probed using KSVEs, including diffusion-controlled substrate binding and rate-limiting steps, which could be one of or a combination of isomerization steps, the product release step, or the chemical step of catalysis [48]. According to the Stokes–Einstein equation, a molecule’s diffusion coefficient (D) is inversely proportional to solvent viscosity (η) [58]. An increase in solvent viscosity raises the frictional resistance experienced by enzymes and substrates in solution [59]. Consequently, any kinetic step that is diffusion-dependent, including substrate association, substrate dissociation, or product release, will slow down in direct proportion to the viscosity change [59]. In KSVE experiments, this is reflected as a linear decrease in the k_cat/K_m and k_cat values, with slopes approaching 1 for fully diffusion-controlled steps, intermediate slopes (0 < m < 1) for mixed control, and a slope nearing zero when the chemical step of catalysis is rate-limiting [48]. Diffusion controls how rapidly enzyme and substrate molecules encounter each other and how quickly products exit the active site, establishing an upper limit to catalytic efficiency (k_cat/K_m ≈ 10⁸–10⁹ M⁻¹·s⁻¹ for large enzymes) [48]. The KSVE methodology essentially involves setting up protocols to measure initial velocity data in the presence of varying concentrations of an additive, usually polyhydroxylated compounds. When two substrates are required for complete enzyme activity, the concentration of one substrate may be fixed, while the concentration of the other is varied [11]. The same initial velocity measurement is repeated with the addition of different concentrations of a viscosigen, i.e., 0%, 9%, 15%, 30%, and 40%. Solvent microviscosigens, such as glucose, glycerol, and sucrose, create resistance to the motion of molecules in solution and are associated with the internal diffusion of a substrate into the active site of an enzyme [48] (Figure 5). Solvent macroviscosigens, on the other hand, include polymeric additives, such as Ficoll-400 and polyethylene glycol (PEG), and determine the free volume available to molecules moving in solution [48]. Increasing solvent macroviscosity can be used as a control for KSVEs to distinguish whether the observed viscosity effect arises from diffusion limitations rather than macromolecular crowding [48] (Figure 5).

Table 2. Microviscosigens and macroviscosigens, with their corresponding molecular weight (MW) ranges.

Microviscosigen		Macroviscosigen
	MW (Da)		Typical MW Range (kDa)
Glucose	180.2	Polyethylene glycol	200–20,000
Sucrose	342.3	Polyacrylamide	Varies, reaching several millions
Glycerol	92.1	Ficoll	300–700
Ethylene glycol	62.1	Dextran	40–500
Trehalose	342.3	Polyethylene oxide	100–1000
Erythritol	122.1	Polyvinylpyrrolidone	10–360
Fructose	180.2	Methylcellulose	30–150
Maltose	342.3	Carboxymethylcellulose	90–700
Mannitol	182.2	Agarose	~120,000
Sorbitol	182.2	Xanthan gum	~2000

below lists some microviscosigens and macroviscosigens, along with their respective typical molecular weights.

The data are plotted using the Michaelis–Menten equation to establish a hyperbolic dependence of the rate on substrate concentration. Further analysis may be followed by making a plot of normalized kinetic parameters, such as the k_cat or k_cat/K_m parameters (thus, “(k_cat)_o/(k_cat)_η or (k_cat/K_m)_o/(k_cat/K_m)_η)” against relative viscosity (η_rel) using Equation (1). The kinetic parameters k_cat/K_m and k_cat are often preferred because they can easily be measured, and significant mechanistic information can be obtained [48,60]. In-depth details and references on KSVEs were summarized in a review by Gadda and Sobrado [48].

$${\displaystyle \frac{{\left(k\right)}_{\mathrm{o}}}{{\left(k\right)}_{\mathsf{\eta }}}}=\mathrm{m}\left({\eta }_{rel}-1\right)+1$$

(1)

In Equation (1), the slope of the line, m, indicates how much the normalized kinetic parameter depends on viscosity [61]. A linear correlation between the relative k_cat ((k_cat)_o/(k_cat)_η) and the relative viscosity (η/η₀) is indicative of the presence of viscosity-dependent or solvent-sensitive steps in the enzyme catalytic cycle [59]. Analysis of the resulting trends can provide insights into the reaction mechanism, for instance, the slope of the line that fits the data (Equation (1)) for linear dependence of the data in a plot of (k_cat/K_m)_o/(k_cat/K_m)_η versus η_rel (Figure 6) is consistent with the capture of the substrate in enzyme−substrate complexes that is fully (m = 1) or partially (0 < m < 1) committed to catalysis in the forward direction [48]. Similarly, the slope of the line that fits the data (Equation (1)) for linear dependence of the data in a plot of (k_cat)_o/(k_cat)_η versus η_rel (Figure 7) is consistent with the overall turnover of the enzyme being fully (m = 1) or partially (0 < m < 1) limited by the product release rate [48,56]. A zero slope, for both plots, indicates a reaction that is consistent with the absence of forward and reverse commitments to catalysis on the k_cat or k_cat/K_m values [48]. Additionally, a solvent-sensitive internal isomerization of the enzyme−substrate or enzyme−product complex is consistent with a hyperbolic trend in a plot of (k_cat)_o/(k_cat)_η or (k_cat/K_m)_o/(k_cat/K_m)_η versus η_rel during enzyme catalysis [48].

3. Experimental Methods in Enzymology

The broader implications of using the technique of KSVEs were discussed in the previous section; however, the knowledge obtained from KSVEs can also be complemented by combining several other important techniques in mechanistic enzymology. Nonetheless, these techniques may require expertise and resources that are not readily available to most laboratories [11]. The X-ray crystallography technique can be used to determine the 3D structure of a protein by producing an electron density map of the atoms within the protein [62]. The detailed architecture of the active site of an enzyme and the structures of intermediates occurring in the reaction pathway can also be elucidated using X-ray crystallography [63,64]. Hydrogen/deuterium exchange mass spectrometry (HDX-MS) is a solution-state structural technique that can be used for the direct monitoring of enzyme behavior by linking structure, conformational dynamics, and function under native conditions [65,66,67]. HDX-MS provides information on protein conformational dynamics [68], protein–membrane interactions [69], and folding dynamics [70] by analyzing the mass differences resulting from isotopic exchange between the protein backbone amide hydrogens and the surrounding solvent. Nuclear magnetic resonance (NMR) is another technique that is used to elucidate the structure and dynamics of an enzyme. NMR determines the 3D structure, monitors conformational changes during catalysis, and identifies important catalytic residues in the reaction pathway by using the magnetic properties of the enzyme’s atomic nuclei [71,72,73]. NMR can also monitor the interaction between an enzyme and its substrate throughout a reaction process, and the ionization states of catalytic residues can be determined using NMR [74,75]. However, interpreting data from structural techniques, such as NMR, without kinetic data can lead to errors, as enzymes typically complete their catalytic cycles in milliseconds, whereas NMR data collection usually takes hours [76,77,78]. Reaction by-products produced at rates slower than the enzyme turnover, for example, could be mistakenly identified as important intermediates in the reaction pathway when relying only on NMR data [76,79,80].

Reaction pathways are monitored using enzyme kinetic techniques, from substrate binding at the active site of an enzyme to product release. The transient-state kinetic technique provides valuable insights into the pathways of reactions catalyzed by enzymes, including the direct observation of intermediates and products [81]. The individual, rapid steps occurring in a reaction can also be directly observed and quantified in real time via transient kinetic analysis [76,82]. Another technique, kinetic isotope effects (KIEs), has been used for decades to investigate mechanisms of reactions by exploiting the differences in reaction rates when atoms of reactant molecules are replaced by their respective isotopes [83]. In addition to predicting whether two or more pathways of reactions share a similar transition state [83,84,85], KIEs can reveal which bond is broken or formed at each stage of a catalytic reaction [85,86]. The KIE method can also be used to determine protein dynamics, hybridization states, rate-limiting steps, and binding effects [48,87]. The ionization states of substrates or active site amino acids involved in enzyme activity can also be deduced by the method of pH-dependent enzyme kinetics [48,88].

A critical component to the stability and efficiency of enzymes is the reaction medium used for the enzymatic reaction. Solvent engineering is a tool that presents numerous opportunities for optimizing enzyme properties and involves the strategic design and manipulation of solvent systems to improve the solubility of reactants and products, enzyme stability, activity, and selectivity [89]. In biocatalysis, water is the most frequently used solvent system in biocatalytic transformations; however, the high polarity of water creates a drawback for hydrophobic reactants, which are commonly used in synthetic reactions [90]. The use of organic solvent systems for enzymatic reactions has become a trend because of the increased solubility of hydrophobic substances, especially when water-dependent side reactions need to be suppressed [91]. Ionic liquids have also been used as reaction media for the past two decades because of their wide thermal stability, low flammability, and low vapor pressure [92,93,94,95]. Butyl imidazolium chloride has demonstrated superior catalytic performance compared to other ionic solvents in the synthesis of the bio-based bis-epoxide compound, bis(oxiran-2-ylmethyl) furan-2,5-dicarboxylate, offering a simpler, more scalable, and cost-effective solution suitable for large-scale applications [96]. Other ionic solvents that were tested were summarized in a study by Chícharo et al. [96]. However, the usage of ionic solvents is often limited due to low sustainability, high cost, poor biodegradability, and poor biocompatibility. Deep eutectic solvents (DESs), as alternatives to ionic liquids, were first introduced by Abbott et al., 2003 [97] when they observed that mixtures of choline chloride and urea possessed significantly lower melting points relative to their pure components [98,99]. DESs are eutectics formed by mixing a hydrogen bond acceptor (HBA) and a hydrogen bond donor (HBD) with simultaneous heating and stirring until a liquid is formed [99]. This results in a self-association to create a new eutectic phase that is characterized by a significantly reduced melting point compared to the individual components [100]. An example of DESs is choline chloride–ethylene glycol (1:2) with a T_m of 66 °C. DESs are increasingly used as solvent systems in biocatalysis because they are nontoxic, nonvolatile, nonflammable, biodegradable, and considered “green” solvents [90]. Other advantages include their ease of preparation and lower cost [90]. The efficacy of different organic solvent systems, such as isopropylamine, oxalic acid, formic acid, and ethyl acetate, was studied and reported by He et al. [101]. Other approaches in solvent engineering have utilized surfactants as various reaction media due to their unique amphiphilic nature to enhance the solubility of hydrophobic substances. An advanced approach known as Pickering emulsion uses solid particles, called Pickering particles, which possess unique properties that enable them to adsorb onto and stabilize the interface between two immiscible components, thereby preventing separation [102]. A study by Wang et al. utilized certain bacteria as Pickering particles, taking advantage of their size and surface hydrophobicity to stabilize Pickering emulsions [103]. A repertoire of diverse Pickering particles was provided in a review by Omar and Sadeghi [104].

Table 3. DESs with their respective HBAs, HBDs, mole ratios, and melting temperatures.

Hydrogen Bond Acceptor (HBA)	Hydrogen Bond Donor (HBD)	Mole Ratio	Melting Temperature (°C)	Ref.
Choline chloride	1-Methyl urea	1:2	29	[105]
Choline chloride	Benzamide	1:2	92	[105]
Choline chloride	1,1-Dimethyl urea	1:2	149	[105]
Choline chloride	4-Hydroxybenzoic acid	1:2	97	[106]
Choline chloride	1,3-Dimethyl urea	1:2	70	[105]
Choline tetrafluoroborate	Urea	1:2	67	[105]
Diethylethanolammonium chloride	Glycerol	1:2	−1.2	[107]
Diethylethanolammonium chloride	Trifluoroacetamide	1:2	0.05	[108]
N,N-Diethylenethanolammonium chloride	Glycerol	1:2	−1.33	[109]
Tetrabutylammonium chloride	Ethylene glycol	1:4	−16.84	[110]
Tetrabutylammonium chloride	Urea	4:1	27.14	[111]
Tetrabutylammonium chloride	Decanoic acid	1:2	−11.95	[112]
Tetrapropylammonium bromide	Glycerol	1:3	−16.1	[109]
Tetrapropylammonium bromide	Ethylene glycol	1:3	−13.3	[113]
Propylammonium bromide	Glycerol	1:2	−4	[114]

below contains a list of DESs with their corresponding HBAs and HBDs, mole ratios, and melting temperatures. An extensive database of DESs was summarized in a review by Omar and Sadeghi [97].

While solvent systems can be considered for enzyme stability reasons, the state of the enzyme is also vital to its function. Immobilizing enzymes onto fixed support systems can overcome certain drawbacks of enzymes in the free state, including limited reusability, especially for industrial applications like waste removal, and susceptibility to denaturation [115]. Enzymes can be anchored via adsorption, covalent binding, encapsulation, or cross-linking in order to maintain high catalytic activity over extended periods, operate across a broader pH and temperature range, and withstand other harsh conditions [116]. In order to retain high catalytic activity, laccase from Pleurotus florida NCIM 1243 was immobilized by adsorption onto cellulose nanofibers, and the study results showed that the enzyme was capable of removing micropollutants, over 90% for Remazol Brilliant Blue R and 50% for Remazol Black 5 [117]. In a different study, laccase was adsorbed onto polyacrylonitrile–biochar composite nanofibrous membrane, and it was observed that the immobilized enzyme showed over 60% of chlortetracycline removal efficiency from simulated wastewater [118]. Both studies on laccase demonstrated that immobilizing enzymes maintain significant activity rates and that the enzymes can be reused repeatedly.

4. The k_cat Parameter—Enzyme Turnover Number

For an enzyme that catalyzes a reversible reaction with a single substrate yielding a single product, Scheme 1 below illustrates the simplest, but realistic, kinetic model. Note that, unlike in most biochemistry textbooks, the chemical step of catalysis and the physical step of product release are treated independently here.

We begin by explaining the kinetic parameter k_cat, first by stating the concept of steady-state kinetics introduced by Briggs and Haldane in 1925 [119]. Steady-state kinetics is described by the condition of an enzyme-catalyzed reaction, where the concentration of the enzyme–substrate (ES) complex remains constant during the catalytic process [119,120]. Under steady-state conditions, the reaction rate for an enzyme (E) with a single substrate (S) can be given by the Michaelis–Menten equation (Equation (2)):

$$v={\displaystyle \frac{k_{cat}\left[{\mathrm{E}]}_{\mathrm{T}}\right[\mathrm{S}]}{K_m+\left[\mathrm{S}\right]}}$$

(2)

When the substrate concentration is much greater than K_m, Equation (2) is simplified as

$$v=V_{max}=k_{cat}[{E]}_T$$

(3)

where [E]_T is the total concentration of enzyme, v is the velocity of reaction, and V_max is the maximum reaction velocity [121]. k_cat represents the maximum velocity at which substrate molecules are converted by an enzyme into a product per unit time when the substrate fully saturates the enzyme [122]. Thus, k_cat sets a lower limit on the rate of any step following substrate binding and is a function of the time it takes for the enzyme to bind the substrate through to the release of the product into solution by the enzyme [123]. Under saturating substrate conditions, k_cat reflects the rate-limiting step(s) in the reaction pathway [121].

The conventional definition of the rate-limiting step of an enzyme-catalyzed reaction, i.e., the slowest step in the catalytic pathway that determines the overall rate of reaction, misrepresents the operation of a sequence of enzyme-catalyzed reactions [124]. Several studies have shown that reactions may have two or more steps, influencing the total reaction rate to varying extents [3]. Thus, the overall reaction rate can proceed only as fast as the combination of all rate-limiting steps, implying that the identification of the rate-limiting step(s) in enzyme-catalyzed reactions through mechanistic enzymology is important. Depending on the enzyme system, an isomerization step, the chemical step of catalysis (k₃), and/or the product release step (k₅) can be the rate-limiting step(s) [125]. In some enzymes, the rates of bond formation and cleavage occur faster than the rate of conformational change, suggesting that the conformational change is the rate-limiting step in the catalytic process [126,127,128]. It is essential to note that substrate binding is not considered a rate-limiting step, as the enzyme turnover number (k_cat) is defined under conditions of substrate saturation [11]. A further explanation will be provided later in this section.

In Scheme 1, k₂, k₃, k₄, and k₅ are first-order rate constants, whereas k₁ is a second-order rate constant for substrate association [129]. The rate equation for Scheme 1 can be derived using the steady-state approximation as [60,130]

$$v={\displaystyle \frac{k_1{k}_3{k}_5\left[S\right]}{k_2{k}_4+k_2{k}_5+k_3{k}_5+k_1(k_3+k_4+k_5)\left[S\right]}}[{\mathrm{E}]}_{\mathrm{T}}$$

(4)

When the substrate concentration is much greater than K_m, the rate is determined by Equation (3), and Equation (4) is simplified as

$$V_{max}=k_{cat}[{\mathrm{E}]}_{\mathrm{T}}={\displaystyle \frac{k_3{k}_5}{k_3+k_4+k_5}}[{\mathrm{E}]}_{\mathrm{T}}$$

(5)

From Equation (5),

$$k_{cat}={\displaystyle \frac{k_3{k}_5}{k_3+k_4+k_5}}$$

(6)

It is clear from Equation (6) that k_cat may be dependent on the catalytic step and product release, but it is not affected by substrate association and dissociation (k₁ and k₂, respectively). Equation (6) simplifies to Equation (7) when the chemical step of catalysis is irreversible (k₄ ≈ 0) or when the enzyme–product complex tends to partition forward and release the product rather than reverting to the enzyme–substrate complex (k₅ >> k₄) [48].

$$k_{cat}={\displaystyle \frac{k_3{k}_5}{k_3+k_5}}$$

(7)

Three scenarios will be described to illustrate how the rate-limiting step(s) of a reaction determine the overall turnover number (k_cat) and how identifying such steps can increase the likelihood of successful targeted attempts to improve enzyme activity. For an enzyme-catalyzed reaction with a single substrate and product and an irreversible catalytic step, suppose k₃ and k₅ are 10 s⁻¹ and 10 s⁻¹, respectively; the estimated k_cat from Equation (7) would be 5 s⁻¹.

Scenario 1.

If the chemical step of catalysis is the only rate-limiting step, i.e., k₅ >> k₃, with k₅ = 5 s⁻¹ and k₃ = 0.20 s⁻¹, then k_cat = 0.19 ≈ k₃, implying that any efforts to improve the product release step (k₅) will not increase the overall turnover number of the enzyme.

Scenario 2.

If the product dissociation step is the only rate-limiting step, i.e., k₃ >> k₅, with k₃ = 5 s⁻¹ and k₅ = 0.20 s⁻¹, then k_cat = 0.19 ≈ k₅, implying that any efforts to improve the chemical step of catalysis (k₃) will not increase the overall turnover number of the enzyme.

Scenario 3.

If both the chemical step of catalysis and the product dissociation step are rate-limiting, i.e., k₃ ≈ k₅, with k₅ = 0.20 s⁻¹ and k₃ = 0.20 s⁻¹, then k_cat = 0.1 s⁻¹, implying that two rate-limiting steps generally give lower turnover numbers relative to a single rate-limiting step; however, any efforts to improve either the product release step (k₅) or chemical step of catalysis (k₃) will increase the overall turnover number of the enzyme.

A clear example of scenario 1 is found in a study by Krooshof et al. [130], where the role of Asp260 in the catalytic triad of haloalkane dehalogenase (DhlaA) was investigated by the method of site-directed mutagenesis [131]. A 12-fold increase in bromide release (the product), a 220-fold reduction in carbon–bromine bond cleavage (the chemical step of catalysis), and a 10-fold decrease in k_cat were observed. Further analysis revealed that the chemical step of catalysis was rate-limiting, explaining why the increased product release did not enhance the overall turnover rate [131]. From the three scenarios above, performing mechanistic studies first to determine the rate-limiting step(s) of a reaction will guide approaches aimed at increasing the turnover number of an enzyme by targeting such steps that are relevant to k_cat.

We mentioned earlier in this section that substrate binding is not considered a rate-limiting step because enzyme turnover is defined under conditions of substrate saturation. Thus, when the substrate concentration is much less than K_m, the rate is determined by k_cat/K_m, and Equation (2) is simplified as

$$v={\displaystyle \frac{k_{cat}}{K_m}}[\mathrm{S}][{\mathrm{E}]}_{\mathrm{T}}={\displaystyle \frac{k_1{k}_3{k}_5}{k_2{k}_4+k_2{k}_5+k_3{k}_5}}[{\mathrm{E}]}_{\mathrm{T}}$$

(8)

From Equation (8),

$${\displaystyle \frac{k_{cat}}{K_m}}={\displaystyle \frac{k_1{k}_3{k}_5}{k_2{k}_4+k_2{k}_5+k_3{k}_5}}$$

(9)

Despite the simplicity of the kinetic model in Scheme 1, it remains challenging to determine from Equation (9) how each step in the kinetic model influences the k_cat/K_m parameter as compared to Equation (7) [129]. As a result, enzyme turnover is better defined under conditions of substrate saturation, where k_cat is dependent on kinetic steps beyond the substrate binding step, i.e., the isomerization step, the chemical step of catalysis, or the product release step.

5. Strategies for Enhancing the Product Release Step

For the past few decades, approaches such as mutagenesis, nanobiocatalysis, modulation of loop dynamics, allosteric modulation, and engineering of transport tunnels have been used to enhance the product release step in enzyme catalysis.

A widely used technique, nanobiocatalysis, involves immobilizing enzymes on nanoparticles to improve enzyme catalytic efficiency, durability, and recyclability [132]. Nanoparticles have high surface areas and unique properties, which will enhance the turnover rate of enzymes via mechanisms including metal ion activation [133], electron transfer [134], morphology effects [134], reducing mass transfer limitations [135], conformation changes [136], increasing enzyme stability [137], and increasing substrate affinity [135].

Product release tunnels have been directly targeted by some studies through the modification of specific residues involved in product delivery or by the de novo synthesis of transport tunnels to improve product release [138,139,140]. Residue modification involves widening tunnels to influence product movement by altering the sizes of tunnel-lining residues or by adjusting the hydrophobicity/hydrophilicity content of such tunnels. For example, in a study on toluene-ortho-xylene monooxygenase, it was found that the size of the residue at position 314 was inversely proportional to enzyme activity [141]. The glutamine residue at this position was mutated to smaller residues, which resulted in an increased product efflux rate and a higher enzyme activity [141]. Conversely, de novo tunnel synthesis involves identifying potential locations for new tunnels through computational modeling and introducing mutations to create openings or channels in the protein structure. A notable example is a study on haloalkane dehalogenase, in which a transport tunnel was designed and tested computationally to show an increased product release rate and enhanced activity of the enzyme [138].

The effects of enzyme motion, such as loop dynamics, on catalysis have been a focus in protein engineering. Research on enhancing or reducing the flexibility of specific enzyme loop regions has led to increased product release and enzyme turnover rates [142,143]. Allosteric modulation is another approach used to enhance product release rates by using specific molecules that influence conformational dynamics, thereby increasing enzyme turnover rates [144,145,146].

In

Table 4. Strategies for improving the product release step of an enzyme-catalyzed reaction.

Category	Methodology	Key Findings	Ref.
Nanoparticle technology	β-Galactosidase was coupled with CdSe/ZnS core/shell quantum dots of different sizes and wavelengths. Quantum dots were stabilized by DHLA-CL4 and DHLA-PEG-CL4 ligands.	β-Galactosidase with quantum dots exhibited a 3-fold increase in the kcat value, resulting from an increased product dissociation rate. The unique microenvironment surrounding the hydration layer of the enzyme–QD complex was attributed to the observed effect.	[147]
	CdSe/ZnS core/shell quantum dots (QDs) of different sizes were conjugated with phosphotriesterase (PTE), followed by extensive characterization.	When conjugated to QDs, a 4-fold increase in the kcat value and a 2-fold increase in the kcat/Km value were shown by PTE. The enhanced activity was attributed to an increase in the enzyme–product dissociation rate, likely due to the varying microenvironment of the hydration layer of the PTE-QD complex.	[148]
	A model system was used to test a proposed detailed reaction scheme involving various possible interactions between a substrate and an enzyme–nanoparticle bioconjugate.	The enzyme attached to nanoparticles showed a 3.2-fold increase in the kcat value and a 1.7-fold increase in the kcat/Km value. The proposed detailed reaction scheme suggested that product release becomes rate-limiting when the enzyme is attached to the nanoparticles.	[149]
Mutagenesis	Site saturation and iterative mutagenesis were performed on the identified N-terminal residues 5-YWQN-8 in GH11 xylanase XynLC9 from Bacillus subtilis to produce mutants W6F/Q7H and N8Y.	The W6F/Q7H and N8Y mutants exhibited increased kcat values by 2.6-fold and 1.8-fold, respectively, and kcat/Km values by 1.5-fold and 1.2-fold, respectively. Both mutants were more thermostable and had enlarged active site clefts.	[150]
	DNA shuffling was employed to mutagenize the trpC gene, which encodes indole glycerol phosphate synthase, from the hyperthermophilic archaeon Sulfolobus solfataricus.	Single and double amino acid mutants had improved activity and increased product release rates. The rate-determining step changed from product dissociation to the chemical step of catalysis.	[17]
	Site-directed mutagenesis was performed on specific residues of orotidine 5′-monophosphate decarboxylase from Saccharomyces cerevisiae to give the mutants Q215A and Y217F.	The Q215A and Y217F mutants exhibited increased kcat and decreased kcat/Km, resulting from increased product release rate and a weakening of protein–phosphodianion interactions.	[151]
	The structural and functional effects of the evolutionarily changing active site in the biliverdin reductase B (BLVRB) family were investigated by creating multiple human BLVRB mutants through site-directed mutagenesis.	A novel mechanism was proposed where coenzyme “clamps” formed by arginine side chains at positions 14 and 78 slowed coenzyme release and enzyme turnover in wild-type enzymes. However, an inverse association was observed between coenzyme release and enzyme turnover for different BLVRB mutants.	[152]
	Enantioselectivity was modified in alcohol dehydrogenase A (ADH-A) from Rhodococcus ruber DSM 44541 by performing saturation mutagenesis on three targeted sites in the catalytic cavity of ADH-A.	The stereoselectivity of the enzyme was altered from (S)- to (R)-1-phenylethanol by the W295A mutant, resulting in a 4-fold increase in the kcat value. W295A showed an enhanced NADH release rate, shifting the rate-determining step from coenzyme release to hydride transfer. Other mutants also showed an increased kcat value relative to the wild type.	[153]
	Directed evolution was applied to Escherichia coli S-1,2-propanediol oxidoreductase (FucO) in the study, followed by site-directed mutagenesis of specific residues, including F254 and N151.	A dual role in enzyme function was found to be played by the residue F254, contributing to increased coenzyme dissociation from the active site and a 4-fold increase in the kcat value with S-1,2-propanediol.	[154]
	Both natural and non-canonical amino acids were incorporated into bacterial phosphotriesterase (PTE) at position T309 to create different mutants.	Replacing Tyr309 with unnatural amino acids improved the phosphotriesterase turnover rate up to 11-fold and the kcat/Km value up to 4-fold. The rate-limiting product release rate was observed to be increased by the deprotonated 7-hydroxyl group of L-(7-hydroxycoumarin-4-yl)ethylglycine through the electrostatic repulsion of 4-nitrophenolate.	[155]
	A novel approach to in vitro evolution of RNA using droplet-based microfluidics was presented, and cycles of random mutagenesis and selection were conducted to evolve the X-motif.	Evolved X-motif ribozyme variants showed an ~28-fold higher kcat value relative to the wild type. This increase in enzyme turnover was attributed to an improved product release rate, which was identified as the rate-limiting step.	[156]
	The catalytic activity of Agrobacterium radiobacter phosphotriesterase was enhanced by the genetic incorporation of the unnatural amino acid selenotyrosine (SeHF) at Tyr309.	The SeHF309 mutant resulted in a 12-fold increase in the kcat value and a 3.2-fold increase in the kcat/Km value. The product release pocket was observed to be opened by the SeHF309 mutant, resulting in an increased rate of product release.	[157]
Loop dynamics	Site-directed mutagenesis was performed on E246 in loop L2 located at the entrance of the active site of Pseudomonas aeruginosa d-arginine dehydrogenase to produce the mutants E246Q, E246G, and E246L.	The E246G mutant exhibited a 4-fold increase in the kcat value, while the E246G and E246L variants also showed increased kcat values. The increase in the turnover rate was attributed to an increase in the product release rate resulting from altered loop dynamics.	[158]
	Site-directed mutagenesis was performed on residues at the interface of the mobile loop of the NS2B-NS3pro complex, a two-component viral protease.	Product release was primarily affected by mutations in the NS2B mobile loop, which triggered a conformational change that activated the catalytic center. A 3.4-fold increase in the kcat value and a 1.5-fold increase in the kcat/Km value were observed in the glycine-linked S83W mutant.	[143]
	The thermal stability and catalytic activity of GH11 xylanase XynA from Streptomyces rameus L2001 were improved by rational design when the combined mutations 11YHDGYF16, 23AP24/23SP24, and 32GP33 were introduced to give Mut1 and Mut2 variants.	A 4-fold increase in the kcat value was observed for both Mut1 and Mut2, resulting from a broader catalytic cleft and increased “thumb” flexibility, directly affecting product release. Improved thermal stability was observed in both mutants, with over 85% residual activity after 12 h at 80–90 °C.	[159]
	Product inhibition was relieved in chorismate–pyruvate lyase (CPL) by the modulation of its conformational dynamics. Residues at the enzyme’s product-binding site were mutated to increase flap dynamics and facilitate product release.	Mutants exhibited enhanced flap dynamics in chorismate–pyruvate lyase (CPL), characterized by an 8-fold reduction in product inhibition and a 3-fold increase in the kcat value. The product release rate was increased through flap opening without significantly changing the kcat/Km value.	[160]
Modifying transport tunnels	The role of tunnels in DhaA haloalkane dehalogenase was investigated by creating eight mutants via site-directed mutagenesis.	Five pathways were discovered for product release and water exchange, with chloride ions released solely through p1 and alcohol released through all five pathways. All mutants exhibited increased kcat values, ranging from ~3- to 30-fold.	[140]
	A new transport tunnel was designed in haloalkane dehalogenase via computational design and directed evolution. Mutants were designed to block the native tunnel, allowing for the observation of the properties of the newly opened auxiliary tunnel.	The LinBW mutant, with a blocked native tunnel, showed an increase in both product release rate and a 4-fold increase in the kcat value when the new auxiliary tunnel was opened. The LinBCC mutant, on the other hand, had a 2-fold decrease in enzyme turnover.	[138]
	The activity of Rhodococcus rhodochrous haloalkane dehalogenase was enhanced by creating mutants using a novel engineering strategy involving rational design and directed evolution to identify key residues in access tunnels.	Higher enzyme activities were observed by all mutants, with up to a 32-fold increase in the kcat value and a 26-fold increase in the kcat/Km value. There was an increased rate of carbon–halogen breakage and a change in the rate-determining step to product release.	[161]
	The conformational dynamics of lysine cyclodeaminase from Streptomyces pristinaespiralis were modulated to address substrate and product inhibitions. The Val61-Val94-SpLCD variant was designed by mutating Ile61 and Ile94 residues associated with substrate and product delivery processes.	The Val61-Val94-SpLCD variant showed improvements in the kcat, kcat/Km, Ki-LYS, and Ki-LPA values by 20, 4, 19, and 9 times, respectively. The expanded substrate and product delivery tunnels of the mutant reduced substrate and product inhibition.	[162]
Using allosteric effectors	VHH domain antibodies as allosteric effectors on the nucleoside hydrolase from Trypanosoma vivax (TvNH) were studied.	VHH 1589 inhibited N-glycosidic bond cleavage while increasing the rates of product release. The data indicated a substrate-dependent effect on kcat and kcat/Km, with a net increase in the kcat value observed for the natural nucleoside 7-methylguanosine, where product release was rate-limiting.	[163]
	The allosteric effects of heparan sulfate (HS) on cruzipain-mediated kinin release from kininogen (HK) were studied in Trypanosoma cruzi.	HS reduced the inhibitory activity of HK on cruzipain by 10-fold and enhanced the kcat and kcat/Km values by 2.5-fold and 6-fold, respectively. The rate of cruzipain-mediated kinin release was also increased by HS up to 35-fold.	[164]
	The structural basis of allosteric product release in Escherichia coli dihydrofolate reductase (DHFR) was investigated, with a focus on the rate-determining step of tetrahydrofolate (THF) product release.	The transient entry of the NADPH nicotinamide ring into the active site was observed to involve allosteric product release in DHFR, which resulted in the displacement of the THF’s pterin ring. A remodeling of the enzyme structure and conformation of the THF was attributed to the entry of NADPH, which facilitated a rapid product release.	[145]
	The mechanism of substrate binding and product release in the geranylgeranylation of Rab proteins by geranylgeranyltransferase type II (GGTase-II) was studied.	GGpp was observed to be an allosteric activator of GGTase-II by causing an increase in enzyme affinity for the Rab7:REP-1 complex. GGpp was also found to facilitate product release by increasing the dissociation rate of prenylated Rab7:REP-1 complex.	[146]

below, we present a summary of diverse studies, briefly describing the experimental approach that the researchers used to increase the enzyme turnover, mainly by improving the rate of product release from the active site of enzymes. A significant increase in k_cat, often by several folds, can be observed across all studies, indicating that accelerating product release rates is not only a theoretical approach but a practical way of significantly increasing enzyme turnover rates.

6. Strategies for Enhancing the Chemical Step of Catalysis

Enhancing the rate of the chemical step of catalysis is mostly the main focus in engineering enzymes, especially when it is rate-limiting in the enzyme-catalyzed reaction. This step records events from the enzyme–substrate complex to the enzyme–product complex, except for any isomerization steps following the enzyme–substrate complex. Computational approaches have been used in diverse studies to optimize enzymes by targeting certain active site properties, including catalytic motifs [165], as well as virtually simulating newly designed active sites for improved catalytic rates [166].

The unique roles of certain non-canonical amino acids in the active sites of enzymes have been investigated in some studies, where residues such as Nδ-methylhistidine have been shown to act as catalytic nucleophiles in certain hydrolytic enzymes, resulting in catalytic rate enhancements exceeding 9000-fold [167]. Similarly, canonical amino acids of specific properties have been substituted into enzyme active sites via site-directed mutagenesis to increase the k_cat value by stabilizing the transition state, enhancing nucleophilicity, altering steric repulsion, modulating electronic effects, or changing protein dynamics [168].

Several studies that have used various strategies to improve the chemical step of catalysis (k₃) and enzyme turnover rate are examined and recorded in

Table 5. Strategies for improving the chemical step of catalysis of an enzyme-catalyzed reaction.

Category	Methodology	Key Findings	Ref.
Computational design and directed evolution	Computational design was used to create eight enzymes with two distinct catalytic motifs for the Kemp elimination reaction. Directed in vitro evolution was subsequently applied to refine the computational designs.	The designed enzymes achieved rate enhancements of up to 105 and demonstrated multiple turnovers. Catalytic efficiency was further improved by directed evolution, yielding over 200-fold increase in the kcat/Km value and 106-fold increase in the kcat value relative to the uncatalyzed reaction (kuncat).	[165]
	Active sites in retro-aldolases were constructed using computational algorithms, which consisted of four distinct catalytic motifs for catalyzing carbon–carbon bond breakage in a nonnatural substrate. The use of explicit water molecules for proton shuffling and charged side-chain networks was included in the design.	Of the 72 experimentally characterized designs, 32 exhibited detectable retro-aldolase activity across various protein folds. Designs that incorporated explicit water molecules were more successful, achieving significant increments in reaction rates.	[166]
	Computational design was used to create enzymes capable of catalyzing the bimolecular Diels–Alder reaction. A transaminase enzyme was modified through directed evolution to recognize complex ketone substrates.	High stereoselectivity and substrate specificity were exhibited by the computationally designed Diels–Alder enzymes. The evolved transaminase successfully catalyzed the reaction with a complex ketone substrate, with an 87-fold increase in the kcat value.	[169]
	A droplet-based microfluidic screening tool was used to enhance an artificial aldolase by combining computational design and directed evolution to achieve properties comparable to those of the natural enzyme.	The improved enzyme achieved a 109-fold increase in the enzyme turnover rate, with high stereoselectivity and broad substrate tolerance. It was revealed by further analysis that a Lys-Tyr-Asn-Tyr tetrad was important for enzyme activity.	[170]
Incorporating non-canonical amino acids	The active site of an Escherichia coli transketolase variant (S385Y/D469T/R520Q) was incorporated with non-canonical amino acids via site-specific saturation mutagenesis to produce different variants.	The mutants showed a significant increase in the kcat and kcat/Km values. The residue 385 variants exhibited a 43-fold increase in specific activity, a 100% increase in kcat value, a 290% increase in the Km value, and a 240% increase in the kcat/Km value.	[171]
	Nδ-methylhistidine, a non-canonical amino acid, was incorporated as a catalytic nucleophile into the active site of a novel hydrolytic enzyme, followed by optimization procedures.	Over a 9000-fold increase in the ester hydrolysis rate was observed in the engineered enzyme. Nδ-methylhistidine was revealed as a genetically encodable surrogate for dimethylaminopyridine through crystallographic analysis. Histidine methylation was noted to be important for enzyme activity by blocking the production of nonreactive acyl-enzyme intermediates.	[167]
Mutagenesis	Rational design was used to reconstruct the catalytic pocket and enzyme–substrate interactions of dehydrogenase/reductase EbSDR8 through targeted mutations at specific residues, namely, G94 and S153.	Over a 15-fold improvement in the kcat/Km value was shown by the mutants compared to the wild-type. An increase in the kcat value was also observed for the mutants, and steric repulsion and C−H…π interactions were attributed to this enhanced activity.	[168]
	Disulfide bonds were eliminated via site-directed mutagenesis in Endoglucanase II (Cel5A) from Trichoderma reesei to create two mutants: C99V and C323H.	A 2-fold increase in both the kcat/Km and kcat values was shown by the C99V mutant, whereas a 1.3-fold increase in the kcat/Km value was shown by the C323H mutant. Both mutants had increased flexibility in the substrate-binding cleft.	[172]
	Rational design involving active-site-targeted, site-specific mutagenesis to construct a broad substrate spectrum version of the phenylalanine dehydrogenase (PheDH) from Bacillus halodurans.	Mutant PheDH (E113D-N276L) showed a 6-fold increase in the kcat/Km value for oxidative deamination and a 1.6-fold increase for reductive amination.	[173]
	Site-directed mutagenesis of Tobacco etch virus (TEV) protease, with mutations at T17S/N68D/I77V, was used to produce two different mutants: S219N and S219V.	S219N and S219V showed ~100-fold and ~50-fold increases in their kcat values, respectively. S219N was twofold faster than S219V without a change in its Km value.	[174]
	Saturation mutagenesis was performed at two positions (165 and 264) near the catalytic Trp171 residue of lignin peroxidase from Phanerochaete chrysosporium.	A significant increase in the kcat value (2-fold) and the kcat/Km value (13-fold) was shown by the mutants. A 10-fold increase in affinity for azo dye substrates was also observed.	[175]
	Site-directed mutagenesis was used to create a lac2-9 mutant from recombinant laccase rlac1338, followed by screening to identify those with improved enzyme activity.	Mutant laccase lac2-9 showed a 3.5-fold increase in specific activity, a ~4-fold increase in the kcat/Km value, and a ~2-fold increase in the kcat value.	[176]
	In the study, various mutants were created from arginine deiminase from Pseudomonas plecoglossicida (PpADI) through directed evolution and site-directed mutagenesis, and improved enzyme activity and thermal stability were tested.	The PpADI M9 mutant showed a 64.7-fold increase in the kcat value, a 1.75-fold increase in half-life, and an increase in thermal stability from 47 °C to 54 °C.	[177]
	Mutagenesis and strong promoter replacement strategies were used to create mutants of the PAE-degrading enzyme EstJ6, with the aim of enhancing its catalytic activity and expression level.	The EstJ6M1.1, EstJ6M2, and EstJ6M3.1 mutants exhibited an increased kcat value by up to 2.9-fold, 3.1-fold, and 4.3-fold, respectively, and an increased kcat/Km value by up to 3.7-fold, 4.6-fold, and 6-fold, respectively.	[178]
	Candida antarctica lipase B (CALB) was modified using proline ionic liquids with varying lengths of hydrophobic alkyl chains on the side.	CALB modified with [ProC12][H2PO4] showed a 3.0-fold increase in hydrolysis activity and a 2.8-fold improvement in the kcat/Km value. Thermal stability was increased as well.	[179]
	Random mutagenesis, site-saturation mutagenesis, and combinatorial mutagenesis were used to enhance the activity of a synthesized stereoselective short-chain carbonyl reductase (SCR), an enzyme that produces tert-butyl(3R,5S)-6-chloro-3,5-dihydroxyhexanoate ((3R,5S)-CDHH) for rosuvastatin synthesis.	Mut-Phe145Met/Thr152Ser showed a 1.60-fold increase in the kcat value and a 5.11-fold improvement in the kcat/Km value. A 1.91-fold increase in the kcat value and an 8.07-fold increase in the kcat/Km value were shown by mut-Phe145Tyr/Thr152Ser. A yield of more than 99% in (3R,5S)-CDHH production was also achieved by both mutants.	[180]
	Site-directed mutagenesis and N-glycosylation analysis were employed to modify the active site residues of the xylanase XynCDBFV from Neocallimastix patriciarum, resulting in the creation of various mutants.	The W125F/F163W double mutant showed a ~20% increase in the kcat value, and N-glycosylation at Asn-37 was found to play a dominant role in both catalytic activity and thermal stability.	[181]

. Table 5 below summarizes these studies, briefly highlighting the methodology and key findings obtained. These methods center on improving the bond-breaking and bond-forming reactions within the catalytic pathway, and in most cases, the overall turnover rate of the enzymes involved was either fully or partially rate-limiting by the chemical step of catalysis. A significant fold increase in k_cat was observed across these studies when specific alterations affecting the catalytic step were performed.

7. The k_cat/K_m Parameter—Enzyme Catalytic Efficiency

k_cat/K_m measures how efficiently an enzyme converts substrates into products and is one of the parameters used to compare the level of preference an enzyme has for different substrates [182]. This steady-state kinetic parameter is often used to measure and compare the catalytic usefulness of engineered or mutant enzymes relative to wild-type or other mutants of the enzyme [11]. The parameter combines two important kinetic constants: k_cat, which represents the turnover number, and K_m, known as the Michaelis constant. From Equation (2) in the k_cat section, when the substrate concentration is much less than the K_m value, Equation (2) is simplified to Equation (10) below.

$$v={\displaystyle \frac{k_{cat}}{K_m}}\left[\mathrm{S}\right][{\mathrm{E}]}_{\mathrm{T}}$$

(10)

Thus, under the condition where [S] << K_m, k_cat/K_m is a second-order rate constant with units of M⁻¹s⁻¹. Under this condition, can the physical meaning of k_cat/K_m as a rate constant be interpreted? Cleland demonstrated that k_cat/K_m represents the “net” rate constant for the substrate association step (E + S → ES) and that this parameter is determined by all the elementary rate constants up to and including the first irreversible step [60,129]. In order to report the kinetics of the production of the enzyme–substrate complex (ES) that “are destined to yield the product,” Northrop suggested using “k_cap” rather than k_cat/K_m, as the latter is the apparent rate constant for the “capture” of the substrate [129]. The parameter k_cat/K_m should not be mistaken for merely the result of dividing k_cat by K_m, as students often incorrectly use it to evaluate catalytic efficiency and substrate affinity simultaneously, which is inaccurate in many cases [183]. k_cat/K_m is indicative of how effectively the free state of an enzyme captures the substrate for catalytic turnover. Thus, k_cat/K_m defines the rate of productive substrate binding, which governs specificity [123]. The Michaelis constant, K_m, is the ratio of k_cat to k_cat/K_m and corresponds to the concentration of the substrate at which the enzyme converts the substrate into product at half the maximum reaction rate. K_m is a pseudo-dissociation constant and does not report on the chemical affinity of enzymes unless rigorously determined through advanced methods, which are beyond the scope of most enzyme kinetic analyses. K_m may be considered a measure of substrate affinity only under specific conditions, which will be discussed later in this paper.

8. Strategies for Enhancing the Catalytic Efficiency of an Enzyme

The k_cat/K_m parameter is another focus of enzyme engineering and optimization, and it is particularly relevant under low substrate conditions where most of the enzyme is in its free form, which makes the reaction rate dependent on how efficiently the enzyme binds the substrate and initiates the catalytic process. Nanoparticle technology has been applied in enzyme engineering, where enzymes are immobilized on nanoparticle surfaces to influence substrate binding and catalysis. The catalytic efficiency of phosphotriesterase has been increased by nanoparticles, such as cadmium selenide (CdSe) and zinc sulfide (ZnS), through the alteration of the microenvironment surrounding the hydration layer of the enzyme–nanoparticle bioconjugate [148].

Understanding loop dynamics can be useful when optimizing enzymes because the conformationally dynamic nature of enzymes has been shown to be essential in regulating specificity and biological function. For instance, residues at the interface of the mobile loop of the NS2B-NS3pro complex, a two-component viral protease, were mutated, resulting in a 1.5-fold increase in the k_cat/K_m value [143]. Studies have incorporated non-canonical amino acids into the active site of enzymes to increase the k_cat/K_m value as a result of their unique properties, including transition state stabilization and enhanced nucleophilicity [171].

Selected studies, including their methodologies and key findings, that have used various strategies to enhance the catalytic efficiency (k_cat/K_m) of enzymes are summarized in

Table 6. Strategies for improving the catalytic efficiency of an enzyme-catalyzed reaction.

Category	Methodology	Key Findings	Ref.
Mutagenesis	Corner engineering was used to target the transition region in Bacillus subtilis lipase A to create multiple DES-resistant variants with improved activity and thermal stability.	The K88E/N89K mutant showed a 10-fold increase in the kcat/Km value in 30% ChCl–acetamide and a 4.1-fold increase in 95% ChCl–ethylene glycol. A 6.7-fold improvement in thermal resistance was observed for the mutants at 50 °C. The approach was validated with two additional industrial enzymes.	[184]
	Site-directed mutagenesis was used to modify two main N-glycosylation residues (N45, N64) in cellobiohydrolase I (CBH I) from Trichoderma reesei.	Increased affinity for carboxymethylcellulose, an improved kcat/Km value, and stability up to 80 °C were shown by both mutants.	[185]
	The catalytic pocket of dehydrogenase/reductase EbSDR8 was reconstructed via targeted mutations at specific residues: G94 and S153.	The mutants showed over 15-fold increase in the kcat/Km value relative to the wild-type. The increased catalytic efficiency was due to steric repulsion and C−H…π interactions.	[168]
	Two rounds of random mutagenesis were performed on a novel esterase (EstKa) from marine Klebsiella aerogenes to obtain the mutants I6E9 and L7B11.	Mutants I6E9 and L7B11 showed 1.56- and 1.65-fold higher kcat/Km values relative to the wild-type EstKa. The Pro96Thr and Gln156Arg mutations increased catalytic efficiency by 1.29- and 1.48-fold, respectively. In contrast, the compound mutant Pro96Thr/Gln156Arg exhibited a 68.9% decrease in the Km value and a 1.41-fold increase in the kcat/Km value.	[186]
	The N- and C-terminal domains of keratinase KerSMD were replaced with those from a homologous protease, KerSMF, to generate three different mutants.	N-terminal domain replacement increased the kcat/Km value by >2-fold, whereas C-terminal domain replacement increased the kcat/Km value by 1.5-fold. The kcat/Km value was enhanced by 1.75-fold when both the N- and C-terminal domains were replaced.	[187]
	The conformational dynamics of lysine cyclodeaminase from Streptomyces pristinaespiralis were modulated to address substrate and product inhibitions. The Val61-Val94-SpLCD variant was designed by mutating Ile61 and Ile94 residues associated with substrate and product delivery processes.	The Val61-Val94-SpLCD variant showed improvements in the kcat, kcat/Km, Ki-LYS, and Ki-LPA values by 20, 4, 19, and 9 times, respectively. The expanded substrate and product delivery tunnels of the mutant reduced substrate and product inhibition.	[162]
	Saturation mutagenesis was performed on position 233 of loop 6 of an acidic, mesophilic GH5 cellulase from Gloeophyllum trabeum to create variants N233A and N233G.	The N233A and N233G mutants showed increased specific activities of 27% and 70%, respectively, and catalytic efficiencies of 45% and 52%, respectively.	[142]
	Focused site-directed iterative saturation mutagenesis (FSISM) was applied to rationally design mutations at a non-catalytic cavity in 3-ketosteroid-Δ1-dehydrogenase (Δ1-KstD) from Mycobacterium smegmatis.	All four mutants, H132M, L113F, V419W, and M51L, exhibited a ~30-fold increase in the kcat/Km value and a nearly 10-fold increase in specific activity.	[188]
	Site-directed mutagenesis was performed to create four variants based on multiple sequence and 3D structure alignments of endo-β-1,4-xylanase TmxB from Thermotoga maritima.	The T74Y mutant showed a 1.3-fold increase in specific activity and a 1.6-fold increase in the kcat/Km value. The DY mutants (two amino acid insertions) showed a 1.2-fold increase in the kcat/Km value.	[189]
	Rational design was used to engineer a substrate-binding cavity and entrance tunnel in L-amino acid deaminase from Proteus mirabilis.	The Q92A mutant exhibited an increased binding cavity volume (994.2 Å3), enzyme activity (191.36 U mg1), and kcat/Km value (1.23 mM1min1). The T436/W438A mutant also showed an increased binding cavity volume and kcat/Km value.	[190]
Nanoparticle technology	CdSe/ZnS core/shell quantum dots (QDs) of different sizes were conjugated onto phosphotriesterase (PTE), followed by extensive characterization.	There was a 2-fold increase in the kcat/Km value when the enzyme was conjugated with QDs. The results indicated that the increase in catalytic efficiency was due to the different microenvironment of the hydration layer of the PTE-QD bioconjugate.	[148]
	A proposed model system consisting of a detailed reaction scheme was used to test different possible interactions between a substrate and an enzyme conjugated with nanoparticles.	There was a 1.7-fold increase in the kcat/Km value when the enzyme was conjugated to nanoparticles. The data revealed that the product release step becomes rate-limiting when the enzyme is attached to the nanoparticles.	[149]
Loop dynamics	The residues at the interface of the mobile loop of the NS2B-NS3pro complex, a two-component viral protease, were mutated via site-directed mutagenesis.	The NS2B mobile loop mutants altered the rate of product release by causing a conformational change that activated the catalytic center. The glycine-linked S83W mutant had a 1.5-fold increase in the kcat/Km value relative to the wild-type.	[143]
Incorporating non-canonical amino acids	Non-canonical amino acids were incorporated into the active site of an Escherichia coli transketolase variant (S385Y/D469T/R520Q) through site-specific saturation mutagenesis to produce different variants.	A 43-fold increase in specific activity, a 290% increase in the Km value, and a 240% increase in the kcat/Km value were exhibited by the residue 385 variants.	[171]
	Tyr309 was replaced with two non-canonical amino acids, L-(7-hydroxycoumarin-4-yl)ethylglycine (Hco) and L-(7-methylcoumarin-4-yl)ethylglycine, in bacterial phosphotriesterase.	The mutants with unnatural amino acids had increased phosphotriesterase turnover rates, up to 11-fold, and a kcat/Km value up to 4-fold. The deprotonated 7-hydroxyl group of Hco was observed to increase the rate-limiting product release rate via electrostatic repulsion of 4-nitrophenolate.	[155]
	The non-canonical amino acid, Nδ-methylhistidine, was used as a catalytic nucleophile to design a novel hydrolytic enzyme.	The novel enzyme exhibited an over 9000-fold increase in the ester hydrolysis rate and a 15-fold increase in the kcat/Km value. Crystallographic analysis proved Nδ-methylhistidine as a genetically encodable surrogate for dimethylaminopyridine.	[167]

. In these studies, the approaches used affected either the chemical step of catalysis, the product release step, or both, which resulted in a significant fold increase in the k_cat/K_m value. It is worth noting that some approaches increased the k_cat/K_m value despite a decrease in k_cat, due to a more substantial reduction in K_m relative to k_cat.

9. Revisiting Misconceptions in Enzyme Design and Optimization

We define “misconception” as an incorrect or false theory, concept, principle, or claim resulting from inaccurate information, oversimplification, and/or incorrect inferences or interpretations of data. In this section, we highlight common fallacies made in enzyme kinetic studies and provide accurate information and formulas to refute them.

Misconception 1: K_m is a measure of binding affinity and is equivalent to the dissociation constant K_d.

Refuting misconception: The relationship between K_m and K_d is often misinterpreted. As K_m is represented by the substrate concentration at half the maximum reaction velocity, substrate affinity is sometimes interpreted as K_m. K_m, however, is not a direct measure of affinity but rather is considered an indicator of how efficiently the substrate is bound and converted into product by the enzyme. In contrast, the equilibrium between the free and bound states of a substrate is quantified by K_d, which makes it the proper kinetic parameter for describing substrate affinity. Assuming K_m is equivalent to K_d is incorrect. Although related to each other, K_m is only equal to K_d under the following three conditions for an irreversible reaction: (1) when the substrate is non-sticky—the substrate quickly dissociates after binding the active site, (2) when the product is non-sticky—the product quickly dissociates after its formation, and (3) when the product release step is not rate-limiting. Consider Scheme 1, described earlier, which involves a reaction with a single substrate and product. The k_cat and k_cat/K_m given in Equations (7) and (9) can be used to derive an expression for the K_m parameter as follows:

$$K_m={\displaystyle \frac{k_{cat}}{(\frac{k_{cat}}{K_m})}}={\displaystyle \frac{k_3{k}_5}{k_3+ k_4+ k_5}}.{\displaystyle \frac{k_2{k}_4+ k_2{k}_5+ k_3{k}_5 }{k_1{k}_3{k}_5}}$$

(11)

Equation (11) is simplified to Equation (12) below.

$$K_m={\displaystyle \frac{k_2{k}_4+k_2{k}_5+k_3{k}_5 }{k_1(k_3+k_4+k_5)}}$$

(12)

Condition 1: For a non-sticky substrate, where k₂ >> k₃, Equation (12) further simplifies as:

$$K_m={\displaystyle \frac{k_2{k}_4+ k_5(k_2+ k_3) }{k_1{k}_3+ {k_{1}k}_4+ k_1{k}_5}}= {\displaystyle \frac{k_2{k}_4+ k_2{k}_5 }{k_1{k}_3+ {k_{1}k}_4+ k_1{k}_5}}$$

(13)

Condition 2: For a non-sticky product, where k₅ >> k₄, Equation (13) further simplifies as:

$$K_m={\displaystyle \frac{k_2{(k}_4+ k_5) }{k_1{k}_3+ {k_1(k}_4+ k_5)}}= {\displaystyle \frac{k_2{k}_5 }{k_1{k}_3+ {k_{1}k}_5}}$$

(14)

Condition 3: Given that the product release step is not rate-limiting where k₅ >> k₃, K_m from Equation (14) then simplifies as:

$$K_m={\displaystyle \frac{k_2{k}_5 }{k_1(k_3+ k_5)}}= {\displaystyle \frac{k_2{k}_5 }{{k_{1}k}_5}}$$

(15)

Therefore, under the three conditions mentioned above, K_m is equivalent to K_d, as shown in Equation (16) below.

$$K_m={\displaystyle \frac{k_2 }{k_1}}=K_d$$

(16)

The conditions become even more complex when considering an enzyme-catalyzed reaction with two or more substrates, for which the analytical formulations for the k_cat and k_cat/K_m parameters are considerably more complex than Equations (7) and (9).

Misconception 2: Increasing the rate of substrate binding increases enzyme turnover (k_cat).

Refuting misconception: This claim may arise from studies where alterations or modifications that enhance substrate binding in the active site also led to increased enzyme turnover. The improved substrate binding might be correlated with enzyme turnover in such experiments. Nonetheless, it is important to underline that k_cat is defined under saturating substrate conditions, more especially following the binding of the substrate to the enzyme [120]. When substrate levels are relatively low ([S] << K_m), the rate of the reaction is determined by k_cat/K_m (Equation (8)), which presents a complex model that makes it challenging to determine how each step influences the reaction rate [128]. In contrast, under substrate-saturating conditions, Equation (5) provides a mechanistically feasible understanding of the impact of each kinetic step on enzyme turnover. Therefore, if k_cat increases in the modified enzyme, the factor causing this increase likely affects the rate-limiting step(s), such as an isomerization step, the chemical step of catalysis, or the product release step, rather than substrate binding itself (Equation (7) derived from Equation (4) under saturating substrate conditions). The enzyme turnover number is dependent on all reaction steps, excluding the substrate association and dissociation steps.

Misconception 3: An alteration that affects the k_cat value implies that the chemical step of catalysis is affected.

Refuting misconception: This statement is often made in studies that seek to optimize enzyme activity by causing an alteration, which may be in the form of a mutation of specific residues in the enzyme active site, and then measuring kinetic parameters, such as k_cat, K_m, and k_cat/K_m. An alteration that affects enzyme turnover does not necessarily imply that catalysis is affected. We established earlier from Equation (7) that k_cat is a composite rate constant that can be influenced by either the isomerization step, the chemical step of catalysis, or the product release step, whichever is rate-limiting or a combination thereof. Without detailed mechanistic studies, it is impossible to attribute changes in enzyme turnover solely to the chemical step of catalysis. An alteration that speeds up the chemical step of catalysis may not affect overall turnover if product release is already the rate-limiting step. Conversely, an alteration that affects product release could change turnover without altering the chemical catalytic step if product release is rate-limiting.

10. Conclusions

Mechanistic enzymology is a strong foundation for understanding the molecular mechanisms of the catalytic activity of enzymes, which is essential for engineering and optimizing enzymes in industrial biocatalysis (Figure 8). Enzyme mechanisms can be elucidated by combining techniques such as X-ray crystallography, NMR spectroscopy, transient-state kinetics, kinetic isotope effects, site-directed mutagenesis, and kinetic solvent viscosity effects [48,62,81]. The information obtained from enzyme mechanisms is considered useful for improving industrial yield and productivity, as well as for other applications, including the design of new enzymes for food processing, drug development, and the production of biofuels, detergents, and pulp and paper [190]. The solvent systems of enzymatic reactions are critical to the stability, activity, and selectivity of enzymes. Deep eutectic solvents (DESs) have emerged as alternatives to most solvent systems because they are nontoxic, nonvolatile, nonflammable, biodegradable, and considered “green” solvents [90]. The state of the enzyme is also vital, as studies have shown that immobilized enzymes tend to exhibit greater stability, retain much of their activity, and are especially useful for various industrial applications, such as waste disposal, where enzymes can be reused.

The technique of kinetic solvent viscosity effects (KSVEs) is regarded as a highly efficient and cost-effective method that can be applied to simple enzyme kinetic assays, such as the Michaelis–Menten kinetic approach, without the need for advanced instruments and expertise. In KSVEs, microviscosigens, such as glucose and sucrose, can be used to increase the frictional resistance experienced by enzymes and substrates in solution, whereas macroviscogens, such as Ficoll-400 and polyethylene glycol, can be used to set up controls to distinguish whether the observed viscosity effect arises from diffusion limitations rather than macromolecular crowding. KSVEs can be used to probe different aspects of enzyme reaction mechanisms, including diffusion-controlled substrate binding and rate-limiting steps, which may involve one or a combination of isomerization steps, the product release step, or the chemical step of catalysis [48] Determining the effect of diffusion-dependent steps is important in biocatalysis, as it can guide targeted approaches in enzyme optimization.

We have demonstrated in this review that, beyond optimizing enzymes through approaches such as rational design and directed evolution, the integration of mechanistic enzymology techniques, particularly those involving KSVEs, enables the investigation of diffusion-dependent kinetic steps and the precise identification of rate-limiting steps. Knowledge of rate-limiting steps in an enzyme-catalyzed reaction can be used to guide targeted approaches in enzyme design and optimization by prioritizing such steps in the pathway. Moreover, challenges such as limited control over tradeoffs, difficulty in predicting enzyme behavior under different conditions, and issues with industrial scalability can all be addressed through mechanistic enzymology.