Home Physical Sciences Bridging experiment and theory: a computational exploration of UMG-SP3 dynamics
Article Publicly Available

Bridging experiment and theory: a computational exploration of UMG-SP3 dynamics

  • Luís M. C. Teixeira , Pedro Paiva , Pedro Ferreira , Laura Rotilio , Jens P. Morth , Daniel E. Otzen , Pedro A. Fernandes and Maria J. Ramos ORCID logo EMAIL logo
Published/Copyright: August 27, 2025
Download Article (PDF)
Pure and Applied Chemistry
From the journal Pure and Applied Chemistry Volume 97 Issue 10

Abstract

Understanding how dynamical behaviour and structural features influence protein function and stability is crucial. While extensive experimental data exist, studying real-time protein dynamics and enzyme catalysis remains challenging. Computational advances have been instrumental in overcoming experimental limitations, enabling molecular-level insights into biological macromolecules. The integration of experimental and computational approaches has proven to be very valuable in protein studies. Here, we demonstrate this synergy by investigating the conformational stability of the urethanase UMG-SP3, which exhibited a lower optimum temperature than expected and rapid loss of activity. Molecular dynamics simulations of the UMG-SP3-substrate complex at various temperatures revealed structural rearrangements outside the optimum temperature range (25–35 °C), leading to loss of the native protein fold and impaired substrate binding. Even at the optimum temperature for activity, the enzyme struggled to maintain a catalytically favourable orientation, aligning with experimental findings. Unfolding profiles were determined through differential scanning fluorimetry. Notably, the computational results provided a rationaly for the structural instability observed experimentally, emphasizing the strength of computational methods in elucidating protein behaviour at the atomic level. This study highlights the importance of combining experimental and computational approaches to deepen our understanding of protein stability and function.

Keywords: Enzyme stability; molecular dynamics; protein dynamics; quantum science and technology; UMG-SP3; urethanase

Introduction

Proteins are biomolecules that play a pivotal role in biology and biotechnology. Some of these biomolecules can promote chemical reactions in an efficient manner, i.e. enzymes, while others support the function of other biological macromolecules through signal transmission or complex stabilization, i.e. structural proteins. 1 Therefore, understanding the determinants of protein function is of the utmost interest for several fields, such as health and industry. 2 Early on, researchers with different backgrounds dedicated their work to revealing the way these biomolecules function. Due to their inherent complexity, this task has been approached from different perspectives. Most of the approaches have required the use of biomolecular or biochemical techniques that give information about a given chemical or physical characteristic. 3 In some assays, the interpretation of this information was straightforward, whilst in others, it was more difficult. Throughout the years, researchers collected significant data that promoted important steps for solving the challenging puzzle of understanding protein structure, stability, and function.

Despite the noteworthy advances, experimental methods faced a significant challenge regarding the study of real-time protein dynamical behaviour. 4 Protein dimensions are in nm while their range of motions span fs (atomic fluctuations) to s (folding/unfolding). 5 As such, studying these macromolecules and their dynamic behaviour presents a significant challenge. However, with the advancements in structural determination methods (e.g., X-ray crystallography, nuclear magnetic resonance, and cryogenic electron microscopy), researchers were able to resolve the 3-D structures of proteins. 6 , 7 , 8 This was an important cornerstone, as it allowed the observation of the tertiary and quaternary structure of these biomolecules. In addition, the structural information of the resolved proteins started to be used also to predict the structural arrangement of their close unresolved homologues. 9

Rapidly, structural data was collected for many proteins. However, studying their dynamic behaviour remained a challenge. Furthermore, obtaining mechanistic insights into enzymatic catalytic mechanisms was also a considerable endeavour. The development and use of computational methods, such as molecular dynamics (MD), and quantum mechanics/molecular mechanics (QM/MM), helped researchers overcoming these hurdles. 10 , 11

Since the 1950s, computer simulations started to be applied to condensed matter systems. An important foundation stone for computer simulations was the development of MD, 12 in which, the Newtonian motion equation of the classical particles within a given system was solved by numerical methods. This allowed the study of the time evolution of a given system, allowing one to observe the dynamical properties of that system. Even though this may seem trivial, the equations of statistical mechanics are analytically unsolvable for complex systems. 13 Consequently, the advancement of computer power dictated the progression speed of computer simulations. In addition, the design of efficient numerical methods was essential, as the early computers were considerably hindered by speed and memory. 14 , 15 As computation became more powerful and widespread, the first realistic model systems started to be simulated. Moreover, as researchers extended on the premises of the MD idea, the method became more versatile and powerful. In 1977, Karplus et al. published the first MD simulation of a protein system. 16 This breakthrough propelled the field of biochemical simulations. This area continued to advance with the discovery and implementation of several methods that allowed simulations to be carried out with smaller time steps and in different thermodynamic ensembles. 17 , 18 , 19 , 20 Nowadays, it is possible to simulate complex protein systems, with or without ligands and/or cofactors, for hundreds of ns to µs, which means it is possible to study a wide range of motions and dynamic behaviours. This poses a significant complement to experimental assays, as several concepts can be better consolidated, allowing for a better understanding of the properties observed experimentally at macroscale. 21 , 22 , 23

Even though MD simulations were a breakthrough in the study of protein motions, they did not allow the study of chemical reactions due to their inherent approximations. Understanding how enzymes catalyse these reactions is of paramount importance. In 1972, it was proposed to combine empirical force fields with electronic structure theory to study chemical processes. 24 Subsequently, Warshel et al. built upon this idea by developing an approach that used an all-valence electron semiempirical method. 25 This approach was later named hybrid QM/MM. Despite the potential shown by this methodology, it took around 20 years for the QM/MM methods to become the scheme of choice for studying chemical reactions. 26 Since then, the scientific community has improved and developed this approach, increasing its applicability. Allied with advances in computation, nowadays QM/MM methodologies present a powerful technique that allows the study of complex chemical processes occurring in biological macromolecules, e.g. enzymes. 27 , 28 , 29

When studying proteins, it should be borne in mind that these biological macromolecules are not rigid; they constantly change their shape and conformation. Therefore, chemical reactions catalysed by enzymes are strongly influenced by the dynamic behaviour of the given biochemical system, particularly in catalytically relevant regions such as the active site. 30 Consequently, it is common for computational studies to combine both methods mentioned above, i.e. MD and QM/MM. Usually, the information obtained by MD simulations (e.g., structural stability, conformational space explored, intra- and intermolecular interactions) is used as an important input for subsequent QM/MM studies. 31 , 32 , 33

Herein, we aim to demonstrate the potential of combining MD simulations with experimental data to shed light on the study of proteins. In 2023, Branson et al. discovered three enzymes that were capable of degrading efficiently the carbamate bond of small polyurethane (PU) fragments. 34 These enzymes were combined in a chemoenzymatic process that culminated in the degradation of a PU foam. One of these enzymes was UMG-SP3, which was reported to have an optimum temperature of 35 °C, whereas the other two enzymes functioned best at 70 °C. Curiously, after incubating the enzyme at 20 °C for 12 h, the authors observed a total activity loss. This meant that UMG-SP3 had a narrower functional temperature range than the other enzymes. Despite that, this enzyme presented the second highest activity and shared 53 % and 83 % identity with UMG-SP1 and UMG-SP2, respectively. 34 Rotilio et al. also tested the activity of these three enzymes on an industrial-like PU monomer at 25 °C. 22 Once again, UMG-SP3 exhibited the second-highest PU-degrading activity. However, after 24 h of incubation, UMG-SP3 suffered a significant activity loss, unlike the other two enzymes. Intriguingly, despite the resemblance shared with its counterparts, UMG-SP3 seemed unstable. Rotilio et al. resolved the apo, inhibitor-bound (phenylmethyl sulfonate, PMS), and product-bound UMG-SP3 crystallographic structures. 22 This paved the way for applying computational methods to understand the underlying cause of this instability.

To showcase the power of computational methods, we simulated a UMG-SP3-PU monomer complex at various temperatures (20 °C, 25 °C, 35 °C, 40 °C, and 70 °C) and compared the results from each simulation. Ultimately, we were able to propose a rationale for the results obtained experimentally, underscoring the critical value of integrating computational techniques with experimental assays.

Computational methods

System preparation

We used the coordinates of the X-ray UMG-SP3 structure (PDB ID: 9FW1) with a covalently-bound inhibitor (PMS). 22 As it is known that the enzyme is active in the monomeric form, chain B was omitted in this study. The enzyme titratable residues pK a was predicted at the UMG-SP3’s reported optimum pH (pH 10) using the webserver ProteinPrepare and PROPKA 3.0. 35 , 36 The ionization state of the titratable residues was defined based on the result of the predictions and on the visual inspection of their surrounding chemical environment. Correspondingly, Lys76 was set to its deprotonated form (–NH2). Additionally, His93/142/181/255/256/361/424 were set to their neutral form with the Nε protonated, while His96/200 were set to their neutral form with the Nδ protonated.

As the objective of this work was to simulate the enzyme-substrate complex at different temperatures, we modelled a PU monomer (di-urethane ethylene methylenedianiline, DUE-MDA), which resembled the ones used in industrial-like applications. DUE-MDA has a polyester tail connected through a urethane bond to the MDA core (Fig. 1). This was the same substrate used by Rotilio et al. in the experimental activity assays, 22 and by Paiva et al. to study the catalytic mechanism of UMG-SP2. 33

Fig. 1: Structure of the modelled PU monomer diurethane ethylene methylenedialine (DUE-MDA).
Fig. 1:

Structure of the modelled PU monomer diurethane ethylene methylenedialine (DUE-MDA).

To parameterize and build DUE-MDA, we employed the Gaussian 09 package, using HF/6-31G(d):FF14SB as the level of theory. 37 To model the UMG-SP3-DUE-MDA complex, we used the Open-Source PyMOL (version 2.3.0), in which we aligned the monomer’s carbonyl group with the inhibitor’s S and O atoms (Fig. 2). 38 This allowed the monomer to maintain the urethane bond with the same orientation shown by the covalently-bound inhibitor. Furthermore, we observed that UMG-SP3’s active site accommodated DUE-MDA and that the latter resembled the pose observed in UMG-SP2 (Fig. 2). 33 , 39

Fig. 2: Alignment of the modelled DUE-MDA substrate and the crystallographic inhibitor, PMS (left) and the UMG-SP2 DUE-MDA’s reactants state (right). In both images, UMG-SP3’s key active site residues are represented. The enzyme’s carbon atoms are shown in cyan, the PMS’s and UMG-SP2 DUE-MDA’s carbon atoms are shown in lime, and the modelled DUE-MDA’S carbon atoms are shown in magenta. Hydrogens were omitted for image clarity.
Fig. 2:

Alignment of the modelled DUE-MDA substrate and the crystallographic inhibitor, PMS (left) and the UMG-SP2 DUE-MDA’s reactants state (right). In both images, UMG-SP3’s key active site residues are represented. The enzyme’s carbon atoms are shown in cyan, the PMS’s and UMG-SP2 DUE-MDA’s carbon atoms are shown in lime, and the modelled DUE-MDA’S carbon atoms are shown in magenta. Hydrogens were omitted for image clarity.

The AMBER FF14SB forcefield was used to generate the bonded and nonbonded parameters of UMG-SP3, while GAFF2 was used to generate the parameters for DUE-MDA. 40 , 41 The system was solvated with 19,110 TIP3P water molecules in an isometric box of 12 Å. To neutralize the system, 17 Na+ counter ions were added.

System minimization

The parameterized UMG-SP3-DUE-MDA complex was subjected to a three-step minimization protocol. In the first step, the solvent was minimized (positional restraints were applied to non-solvent atoms). Subsequently, in the second step, positional restraints were only applied to the enzyme’s backbone and DUE-MDA’s atoms. In the third and final step, no positional restraints were applied. All positional restraints were employed with a force constant of 2000 kJ mol−1.nm−2. For each minimization step, the steepest descent algorithm was employed through a maximum of 50 000 cycles. In addition, long-range electrostatic interactions were treated by the Particle Mesh-Ewald method, 42 with a cutoff of 1.0 nm.

MD simulation

The minimized system was subject to a four-step MD simulation protocol: i) NVT heating (100 ps); ii) NPT solvent equilibration (2 ns); iii) NPT protein (except for the active site) and solvent equilibration (2 ns); and iv) free production (400 ns). All simulations were carried out using GROMACS. 43 As the purpose of this work was to evaluate the impact of different temperatures on UMG-SP3, the minimized structure was simulated at 20, 25, 35, 40, and 70 °C. Therefore, initially, the minimized system was heated to one of the desired temperatures for 100 ps, with positional restraints being applied to the enzyme’s and DUE-MDA’s atoms (the forces constant remains equal to the ones used in the minimization protocol). We employed a modified Berendsen thermostat to control the temperature. 44 Subsequently, each system underwent a solvent equilibration of 2 ns (the positional restraints were maintained). For this step, the pressure was maintained constant at 1 bar by an isotropic position scaling, a pressure relaxation time of 2 ps, and using a Berendsen barostat. 44 The heating step’s temperature control method was maintained. In the third step, the system was equilibrated for an extra 2 ns. However, positional restraints were only applied to DUE-MDA’s atoms and the enzyme’s active site residues (Lys76, Ser151, Ile172, Gly173, and Ser175). The active site positional restraint force constant was 3000 kJ mol−1 nm−2. The pressure and temperature control were identical to the ones used in the previous equilibration. Finally, each equilibrated system was subjected to a restraints-free 400 ns production, where coordinates were saved every 100 ps. The temperature control method was maintained, whilst the pressure control method was altered. For pressure control, in the production, the isotropic position scaling was used, with a pressure relaxation time of 1 ps, and the Parrinello-Rahman barostat. 45 All steps used an integration time step of 2 fs and the Particle Mesh-Ewald method, with a cutoff of 1.0 nm, to treat long-range electrostatic interactions. The MD protocol was repeated two additional times for each temperature, resulting in 3 independent replicas, i.e. a total of 1.2 μs, per temperature window.

Differential scanning fluorimetry

We measured UMG-SP3 unfolding profiles on Prometeus Panta (NanoTemper). Purified protein (0.5 mg/ml final concentration) was assayed in a range of pH and Urea concentrations. Temperature was increased 1.0 °C per minute from 20 °C to 90 °C. Protein onset temperature and inflection point were estimated using the Panta analysis software from NanoTemper.

Results and discussion

To understand the difference in behaviour shown by UMG-SP3, we simulated the UMG-SP3-DUE-MDA complex at different temperatures: i) 20 °C, resulting in a slow inactivation of the protein; ii) 25 °C, with activity assays reported by Rotilio et al.; iii) 35 °C, the optimum temperature reported for SP3; iv) 40 °C, which is close to the optimum temperature; v) 70 °C, the optimum temperature reported for UMG-SP1/2. 22 , 34 Subsequently, several metrics were analysed to rationalize the activity loss shown by UMG-SP3. These included tracking the evolution of the root mean squared deviation RMSd, or fluctuation RMSf, and the solvent-accessible surface area SASA of UMG-SP3 throughout each temperature’s three independent 400 ns simulations. Additionally, the key catalytic distances were closely monitored.

Fig. 3 shows a tendency for an increase in the mobility of the backbone atoms with an increase in the temperature. As expected, higher temperatures caused higher structural deviations compared to the structural arrangement adopted by the equilibrated structure. Despite this tendency, we observed that from 20 to 25 °C, the enzyme’s backbone atoms suffered lower deviations. This meant that this small rise in temperature was not sufficient to increase UMG-SP3’s backbone mobility. Surprisingly, the system showed a slight tendency to become more stable with this temperature rise, as observed by the average RMSd (1.59 ± 0.06 Å vs. 1.49 ± 0.17 Å, respectively). It is known that the protein stability curves resemble a parabola, meaning that each protein has two unfolding temperatures (one at a higher and one at a lower temperature). 46 The decrease in temperature causes non-polar protein groups to interact with the solvent (the hydration of these groups is favoured by the temperature decrease). Consequently, the tightly packed protein chain starts to unfold, further exposing non-polar groups, culminating in the loss of its hydrophobic core. 47 This means that low temperatures can also promote protein instability and unfolding. Accordingly, UMG-SP3 was reported to lose its activity after 12 h incubation at 20 °C. Therefore, this increase in stability, caused by the rise in temperature, corroborates the experimental observation. The increase from 25 to 35 °C promoted a higher backbone deviation (average RMSd of 1.64 ± 0.26 Å). As expected, the subsequent temperature windows (40 and 70 °C) promoted a higher backbone adjustment. This was evidenced by their average RMSd of 1.87 ± 0.10 Å and 1.94 ± 0.08 Å.

Fig. 3: Evolution of the enzyme’s backbone RMSd throughout each temperature’s three independent 400 ns replicas: 20 °C (black), 25 °C (cyan), 35 °C (violet), 40 °C (red), and 70 °C (magenta). Both reported functional temperatures were compared with each non-active temperature. Therefore, 25 °C and 35 °C were compared with 20 °C (A), 40 °C (B), and 70 °C (C).
Fig. 3:

Evolution of the enzyme’s backbone RMSd throughout each temperature’s three independent 400 ns replicas: 20 °C (black), 25 °C (cyan), 35 °C (violet), 40 °C (red), and 70 °C (magenta). Both reported functional temperatures were compared with each non-active temperature. Therefore, 25 °C and 35 °C were compared with 20 °C (A), 40 °C (B), and 70 °C (C).

As a higher structural backbone deviation could indicate a less stable initial structural arrangement UMG-SP3, we decided to observe the impact of the temperature on the RMSf of the enzyme to observe the effect on residue mobility (Fig. 4). Accordingly, as expected, we observed that the increase in temperature enhanced the flexibility of some UMG-SP3 residues. This effect was particularly heightened at the 70 °C window, with the residue section 310–320 suffering the most significant fluctuations. This region consists of a random coil that connects two α-helices, which is also highly exposed to the solvent. Thus, as the solvent molecules‘ mobility is influenced by the temperature, this section became more mobile at a higher temperature.

Fig. 4: Evolution of UMG-SP3 RMSf throughout each temperature’s MD production: 20 °C (black), 25 °C (cyan), 35 °C (violet), 40 °C (red), and 70 °C (magenta). Both reported functional temperatures were compared with each non-active temperature. Therefore, 25 °C and 35 °C were compared with 20 °C (A), 40 °C (B), and 70 °C (C).
Fig. 4:

Evolution of UMG-SP3 RMSf throughout each temperature’s MD production: 20 °C (black), 25 °C (cyan), 35 °C (violet), 40 °C (red), and 70 °C (magenta). Both reported functional temperatures were compared with each non-active temperature. Therefore, 25 °C and 35 °C were compared with 20 °C (A), 40 °C (B), and 70 °C (C).

Up to this point, we observed that higher temperatures contributed to an overall increase in the mobility of the enzyme, which in turn could decrease its stability. The exception was the simulation at 20 °C, which presented a slightly less stable backbone than the one exhibited at 25 °C. As this is the lowest temperature window simulated, the promotion of mobility through temperature increase could not explain this observation. Therefore, we decided to analyse the evolution of the SASA throughout each temperature’s simulation. SASA has been established as an important factor in protein folding and stability assays. 48 , 49 SASA gives information about proteins’ surface area, which is accessible to the solvent, allowing the classification of enzyme residues as buried or solvent-exposed. Moreover, changes in SASA values can indicate structural rearrangements. 49 Therefore, this method can provide useful information concerning the impact of each temperature on the overall conformation of UMG-SP3.

Interestingly, we observed that the rise in temperature from 20 to 25 °C induced a trend of decreasing UMG-SP3’s SASA (Fig. 5). This was evidenced by their average SASA of 16.942 × 103 ± 0.124 × 103 and 16.930 × 103 ± 0.155 × 103 Å2. Moreover, the same tendency was observed from 25 to 35 °C (average SASA decreased to 16.893 × 103 ± 0.138 × 103 Å2). Nevertheless, when the temperature surpassed the optimum, we observed a tendency inversion. From 35 to 40 °C, the SASA increased (average value of 16.938 × 103 ± 0.205 × 103 Å2) and the same was observed for the 70 °C (17.138 × 103 ± 0.058 × 103 Å2). This change in the tendency indicated that the temperature was impacting the overall fold of UMG-SP3. Moreover, it seemed that temperatures outside the optimal one (35 °C) were increasing the solvent accessibility, meaning the enzyme was becoming more exposed to the solvent. DUE-MDA has a bulky and hydrophobic section that is stabilized by the enzyme active site. Thus, the increase in solvent exposure could indicate a reduction in the ability for the enzyme to accommodate DUE-MDA. Ultimately, this might hinder the UMG-SP3’s catalytic activity.

Fig. 5: Evolution of the enzyme SASA throughout each temperature’s three independent 400 ns replicas: 20 °C (black), 25 °C (cyan), 35 °C (violet), 40 °C (red), and 70 °C (magenta). Both reported functional temperatures were compared with each non-active temperature. Therefore, 25 °C and 35 °C were compared with 20 °C (A), 40 °C (B), and 70 °C (C).
Fig. 5:

Evolution of the enzyme SASA throughout each temperature’s three independent 400 ns replicas: 20 °C (black), 25 °C (cyan), 35 °C (violet), 40 °C (red), and 70 °C (magenta). Both reported functional temperatures were compared with each non-active temperature. Therefore, 25 °C and 35 °C were compared with 20 °C (A), 40 °C (B), and 70 °C (C).

To complement the simulation results, we measured the unfolding profile of UMG-SP3 at different values of pH and different urea concentrations (Fig. 6 and Table 1). The melting point increased continuously with the pH basicity rise, but reached a maximum at a pH 9. However, in line with the observations of Branson et al., 34 the onset temperature also increased with the pH basicity, reaching a maximum at pH 10 (46.0 ± 0.4 °C). Assuming that the enzyme loses its functionality as soon as it starts unfolding, this implies that it should remain catalytically functional up to higher temperatures at the higher pH value, confirming that this is the optimum pH value within the tested range. With the urea concentration increase, the enzyme became more unstable. This was evidenced by the decrease in the onset temperatures and inflection points with increasing urea concentration (Table 1).

Fig. 6: Effect of pH and urea on UMG-SP3 stability. 1) Unfolding profiles of UMG-SP3 under different pH conditions. 2) Unfolding profiles of UMG-SP3 under different urea concentrations assayed at pH 10.
Fig. 6:

Effect of pH and urea on UMG-SP3 stability. 1) Unfolding profiles of UMG-SP3 under different pH conditions. 2) Unfolding profiles of UMG-SP3 under different urea concentrations assayed at pH 10.

Table 1:

Onset temperature and inflection point of UMG-SP3 under different pH and urea concentration conditions. Standard deviation is calculated on two individual measurements.

Condition Onset temperature (°C) Inflection point (°C)
pH 5 – –
pH 7 15.0 ± 0.1 41.0 ± 0.2
pH 8 37.5 ± 1.4 47.8 ± 0.4
pH 9 37.8 ± 0.2 58.1 ± 0.2
pH 10 46.0 ± 0.4 53.7 ± 0.1
Urea 0.1 M (pH 10) 41.6 ± 0.1 50.5 ± 0.1
Urea 0.5 M (pH 10) 41.6 ± 0.1 50.5 ± 0.1
Urea 1 M (pH 10) 40.7 ± 0.8 49.8 ± 0.6
Urea 2 M (pH 10) 26.0 ± 3.1 42.1 ± 0.8
Urea 4 M (pH 10) 20.7 ± 3.3 36.3 ± 2.5

The unfolding profile of the enzyme, at a pH value of 10, and the corresponding protein melting temperature (53.7 ± 0.1 °C, Fig. 6), corroborated the results obtained in the simulations. Accordingly, in the MD simulations, at 70 °C, we observed a tendency for the enzyme to exhibit difficulties in maintaining the initial structural arrangement (Fig. 3) and to become more flexible (Fig. 4). Additionally, there was a tendency for an increase in SASA, indicating that this temperature impacted the overall fold of UMG-SP3 (Fig. 5). The simulations at a temperature of 40 °C were also in line with the obtained unfolding profile. At this temperature, the enzyme became more mobile (Fig. 4) and solvent-exposed (Fig. 5). Due to the proximity to the onset temperature (46.0 ± 0.4 °C), it is expected that the overall fold of the enzyme starts to be impacted by the temperature. Therefore, all tendencies shown by our simulations were in line with the experimentally determined unfolding profile.

UMG-SP3 is an enzyme that catalyses the hydrolysis of the urethane bond of polyurethane substrates. It is known that for an enzyme to exert its catalytic activity on a given substrate, the enzyme must be able to accommodate the substrate within the active site in a favourable orientation. Therefore, we wanted to study the impact of the structural rearrangements on the overall enzyme-substrate complex. Due to the similarities shared with UMG-SP2, we predict that UMG-SP3 will follow a similar mechanism. 33 This means that for the start of the catalysis, the catalytic residues, Lys76, Ser151 and Ser175, are required to establish a hydrogen bond network, facilitating the mandatory proton transfers. In addition, the target substrate’s carbonyl oxygen needs to establish hydrogen bonds with the backbone nitrogens of the oxyanion hole-participating residues, Ile172 and Gly173, ensuring the correct positioning of the substrate for the start of catalysis. Fig. 7 shows that the equilibrated structure obtained at the optimum temperature obeyed these conditions.

Fig. 7: Arrangement of UMG-SP3 active site equilibrated at 35 °C. The enzyme’s carbon atoms are shown in cyan, while the DUE-MDA’s ones are shown in magenta. Some atoms are not shown for clarity purposes. Key interatomic distances are given in Å.
Fig. 7:

Arrangement of UMG-SP3 active site equilibrated at 35 °C. The enzyme’s carbon atoms are shown in cyan, while the DUE-MDA’s ones are shown in magenta. Some atoms are not shown for clarity purposes. Key interatomic distances are given in Å.

We inspected the impact of the temperature on the two most relevant interatomic distances for the start of catalysis: i) the nucleophilic attack conducted by Ser175 side chain Oγ to the substrate’s carbonyl carbon; ii) the proton transfer from Ser175 to Ser151.

Fig. 8 shows that the tendency observed for the SASA evolution was also spotted in the interatomic distance between Ser175 Oγ and DUE-MDA carbonyl carbon. From 20 to 25 °C, this distance was maintained at a catalytic feasible value (≤3.2 Å) throughout a higher percentage of time, i.e. 16 vs 27 %, respectively. This was also evidenced by the average distance for 25 °C (3.44 ± 0.11 Å) being lower than for 20 °C (3.56 ± 0.04 Å). Interestingly, the average distance at 35 °C was similar to the one obtained at 25 °C (3.45 ± 0.06 Å). The same can be reported about the percentage of time with a catalytic feasible distance, around 25 %. When the enzyme-substrate complex was simulated at higher temperatures, i.e. 40 and 70 °C, the average distance increased, while the percentage of time lowered: for 40 °C the average was 3.52 ± 0.06 Å and the percentage was 18 %; for 70 °C the average was 4.73 ± 2.07 Å and the percentage was 13 %. Altogether, the observed tendency indicated that temperatures outside the optimum window lowered the ability of the DUE-MDA to present a catalytically favourable orientation. In addition, only at 70 °C does this distance increase to >5.00 Å, suggesting that this temperature can hinder the active site’s ability for accommodating DUE-MDA.

Fig. 8: Evolution of the interatomic distance between Ser175 Oγ and DUE-MDA carbonyl carbon: 20 °C (black), 25 °C (cyan), 35 °C (violet), 40 °C (red), and 70 °C (magenta). Both reported functional temperatures were compared with each non-active temperature. Therefore, 25 and 35 °C were compared with 20 °C (A), 40 °C (B), and 70 °C (C).
Fig. 8:

Evolution of the interatomic distance between Ser175 Oγ and DUE-MDA carbonyl carbon: 20 °C (black), 25 °C (cyan), 35 °C (violet), 40 °C (red), and 70 °C (magenta). Both reported functional temperatures were compared with each non-active temperature. Therefore, 25 and 35 °C were compared with 20 °C (A), 40 °C (B), and 70 °C (C).

To confirm if the temperature was also impacting the H-bond network of the active site, specifically between the catalytic serine residues, we studied the evolution of the interatomic distance between Ser175 side chain H and Ser151 side chain Oγ. This is a key catalytic distance because Ser175 needs to transfer its proton to the adjacent Ser151 to promote the nucleophilic attack. Fig. 9 shows that, in line with the nucleophilic attack distance, the average distance decreased from 20 to 25 °C (3.29 ± 0.10 vs 3.11 ± 0.17 Å). On the other hand, the percentage of catalytically favourable distance increased, with the temperature rise, from 14 % to 23 %. When the temperature was raised to 35 °C, the average distance continued to decrease (3.08 ± 0.18 Å), while the percentage remained nearly identical to the one exhibited at 25 °C (24 %). Curiously, unlike the nucleophilic attack distance, at 40 °C both the average distance (3.11 ± 0.20 Å) and time percentage (23 %) were identical to those obtained at 25 °C. At 70 °C, we observed that the average distance increased to 3.48 ± 0.35 Å. The time percentage decreased to 12 %. The evolution of this distance indicated that only the extreme simulated temperatures (20 and 70 °C) had a notable impact on the stability of this catalytically required hydrogen bond.

Fig. 9: Evolution of the interatomic distance between Ser175’s side chain Hγ and Ser175 Oγ: 20 (black), 25 °C (cyan), 35 °C (violet), 40 °C (red), and 70 °C (magenta). Both reported functional temperatures were compared with each non-active temperature. Therefore, 25 °C and 35 °C were compared with 20 °C (A), 40 °C (B), and 70 °C (C).
Fig. 9:

Evolution of the interatomic distance between Ser175’s side chain Hγ and Ser175 Oγ: 20 (black), 25 °C (cyan), 35 °C (violet), 40 °C (red), and 70 °C (magenta). Both reported functional temperatures were compared with each non-active temperature. Therefore, 25 °C and 35 °C were compared with 20 °C (A), 40 °C (B), and 70 °C (C).

Altogether, our results revealed that the tested temperatures impacted the enzyme differently. At 20 °C, we observed the enzyme structural arrangement was slightly affected by an increase in solvent exposure (evidenced by the SASA analysis). In addition, the ability of the enzyme to start the catalysis was hindered, as the enzyme had difficulties in positioning correctly DUE-MDA for the nucleophilic attack. Moreover, the important hydrogen bond between Ser175 and Ser151 was less stable. At 25 °C and 35 °C, the enzyme appeared to show higher stability. Additionally, the UMG-SP3 maintained the catalytic-favourable orientation of its active site and DUE-MDA more easily. Furthermore, the enzyme was slightly more stable at 35 °C, which was expected as this is the optimum temperature. Nevertheless, even at the optimum temperature, we observed that the enzyme was not able to maintain the catalytically favourable arrangement. Consistent with this finding, experimental results revealed that the enzyme lacked stability even under conditions that were otherwise favourable. Therefore, our simulations were in line with the experimental results. At 40 and 70 °C, we observed that the enzyme was considerably more mobile and underwent structural rearrangements, in contrast to the behaviour at the optimum temperature. Furthermore, the experimentally measured unfolding profile was in accordance with these observations. In addition, the active site arrangement was also negatively impacted by both temperatures (40 and 70 °C). The ability to position the DUE-MDA in catalytic favourable position was compromised, as evidenced by the nucleophilic attack distance.

Conclusions

In this work, we highlighted the role of molecular dynamics in understanding protein function and stability, particularly in combination with experimental approaches. Using MD simulations, we provided a rationale for the instability shown by the urethanase UMG-SP3 across different temperatures.

Our simulations showed that at the lower temperature of 20 °C, the enzyme’s structural arrangement was slightly impacted, leading to less frequent and suboptimal catalytic orientations. At the optimal temperature range of 25–35 °C, UMG-SP3 showed improved stability and was able to better maintain an active conformation for a higher percentage of time. Nevertheless, it still exhibited some difficulties in sustaining a catalytic favourable arrangement. Interestingly, the enzyme was more stable at 35 °C, reported to be the optimal temperature, aligning with the experimental observations. However, even at this temperature, UMG-SP3 struggled to maintain a favourable active site configuration, mirroring experimental findings of limited stability. At higher temperatures of 40 and 70 °C, structural rearrangements impacted the enzyme’s mobility and active site orientation. The ability to position DUE-MDA was compromised, particularly at 70 °C, explaining the reason behind the loss of activity.

Overall, our findings reinforce the importance of integrating computational and experimental approaches to better understand the enzyme atomistic behaviour, offering valuable insights into protein dynamics and stability.

Data and software availability

The cartesian coordinates of the modelled complex and equilibrated structures, along with the force field parameters, are provided in the Supporting Information. The initial PDB structure (9FW1) is available for download at https://www.rcsb.org/structure/9FW1. System preparation, minimization, and molecular dynamics simulations were performed using the AMBER 18 package, which can be purchased at https://ambermd.org/. The pK a prediction was carried out using the ProteinPrepare web server (https://open.playmolecule.org/tools/proteinprepare) and PROPKA 3.0, which is available for download at https://propka.readthedocs.io/en/latest. Visualization of the system was done using VMD, which can be downloaded from https://www.ks.uiuc.edu/Research/vmd/. Parameterization of DUE-MDA was performed using Gaussian 09 D01, which can be purchased at https://gaussian.com/. The enzyme-substrate complex was modelled using PyMOL, available at https://www.pymol.org/.

Supporting Information

The supporting information is available and provides access to:

  1. Parameter file for the modelled UMG-SP3-DUE-MDA complex and restart files for the equilibrated structures (ZIP)

Corresponding author: Maria J. Ramos, LAQV/REQUIMTE, Departamento de Química e Bioquímica, Faculdade de Ciências Universidade do Porto, Rua do Campo Alegre, s/n, 4169-007, Porto, Portugal; and EnZync Center for Enzymatic Deconstruction of Thermoset Plastics, Aarhus , Denmark, e-mail:
Article note: A collection of invited papers to celebrate the UN’s proclamation of 2025 as the International Year of Quantum Science and Technology.

Funding source: Novo Nordisk Fonden

Award Identifier / Grant number: NNF22OC0072891

Funding source: Fundação para a Ciência e a Tecnologia

Award Identifier / Grant number: UID/50006

Acknowledgments

LMCT, PP, PF, PAF and MJR would like to thank the European High-Performance Computing Joint Undertaking (EuroHPC JU) that granted the access to MeluXina, the petascale Euro-HPC supercomputer located in Bissen, Luxembourg, under proposal EHPC-REG2023R03-163. PF thanks FCT for funding (Ref. CEECINST/00136/2021/CP2820/CT0002 DOI 10.54499/CEECINST/00136/2021/CP2820/CT0002. LMCT would also like to thank FCT/MECI for funding his PhD project through the grant 2024.05090.BD.

  1. Research ethics: Not applicable.

  2. Informed consent: Not applicable.

  3. Author contributions: Conceptualization: LMCT, PP, PF, LR, JPM, DEO, PAF, and MJR; Methodology: LMCT, PP, PF, LR, PAF, and MJR; Formal Analysis: LMCT and PP; Investigation: LMCT, PP, and LR; Writing – Original Draft: LMCT; Writing – Review & Editing: LMCT, PP, PF, LR, JPM, DEO, PAF, and MJR; Project Administration: JPM, DEO, PAF, and MJR; Funding Acquisition: PAF and MJR.

  4. Use of Large Language Models, AI and Machine Learning Tools: None declared.

  5. Conflict of interest: The authors state no conflict of interest.

  6. Research funding: The authors would like to thank the EnZync consortium, and the financial support received by the Novo Nordisk Fonden, under project NNF22OC0072891 (Challenge Programme 2022 - Recycling or a Sustainable Society). This work received financial support from the PT national funds (FCT/MECI, Fundação para a Ciência e Tecnologia and Ministério da Educação, Ciência e Inovação) through the project UID/50006 -Laboratório Associado para a Química Verde - Tecnologias e Processos Limpos.

  7. Data availability: The cartesian coordinates of the modelled complex and equilibrated structures, along with the force field parameters, are provided in the Supporting Information.

  8. Software availability: The initial PDB structure (9FW1) is available for download at https://www.rcsb.org/structure/9FW1. System preparation, minimization, and molecular dynamics simulations were performed using the AMBER 18 package, which can be purchased at https://ambermd.org/. The pKa prediction was carried out using the ProteinPrepare web server (https://open.playmolecule.org/tools/proteinprepare) and PROPKA 3.0, which is available for download at https://propka.readthedocs.io/en/latest. Visualization of the system was done using VMD, which can be downloaded from https://www.ks.uiuc.edu/Research/vmd/. Parameterization of DUE-MDA was performed using Gaussian 09 D01, which can be purchased at https://gaussian.com/. The enzyme-substrate complex was modelled using PyMOL, available at https://www.pymol.org/.

References

1. Bruce Alberts, A. J.; Lewis, Julian; Raff, Martin; Roberts, Keith; Walter, Peter Molecular Biology of the Cell; Garland Science: New York, 2002.Search in Google Scholar

2. Gregory, A.; Petsko, D. R. Protein Structure and Function; New Science Press: Oxford, 2004.Search in Google Scholar

3. Lesk, A. Introduction to Protein Science: Architecture, Function, and Genomics; Oxford University Press: Oxford, 2016.10.1093/hesc/9780198716846.001.0001Search in Google Scholar

4. Nam, K.; Wolf-Watz, M. Protein Dynamics: The Future Is Bright and Complicated. Struct. Dyn. 2023, 10, 014301; https://doi.org/10.1063/4.0000179.Search in Google Scholar PubMed PubMed Central

5. Schirò, G.; Weik, M. Role of Hydration Water in the Onset of Protein Structural Dynamics. J. Phys. Condens. Matter. 2019, 31, 463002; https://doi.org/10.1088/1361-648X/ab388a.Search in Google Scholar PubMed

6. Maveyraud, L.; Mourey, L. Protein X-Ray Crystallography and Drug Discovery. Molecules 2020, 25; https://doi.org/10.3390/molecules25051030.Search in Google Scholar PubMed PubMed Central

7. Hu, Y.; Cheng, K.; He, L.; Zhang, X.; Jiang, B.; Jiang, L.; Li, C.; Wang, G.; Yang, Y.; Liu, M. NMR-Based Methods for Protein Analysis. Anal. Chem. 2021, 93, 1866–1879; https://doi.org/10.1021/acs.analchem.0c03830.Search in Google Scholar PubMed

8. de Oliveira, T. M.; van Beek, L.; Shilliday, F.; Debreczeni, J. É.; Phillips, C. Cryo-E. M. The Resolution Revolution and Drug Discovery. SLAS Discovery 2021, 26, 17–31. https://doi.org/10.1177/2472555220960401.Search in Google Scholar PubMed

9. Hameduh, T.; Haddad, Y.; Adam, V.; Heger, Z. Homology Modeling in the Time of Collective and Artificial Intelligence. Comput. Struct. Biotechnol. J. 2020, 18, 3494–3506. https://doi.org/10.1016/j.csbj.2020.11.007.Search in Google Scholar PubMed PubMed Central

10. Hollingsworth, S. A.; Dror, R. O. Molecular Dynamics Simulation for All. Neuron 2018, 99, 1129–1143. https://doi.org/10.1016/j.neuron.2018.08.011.Search in Google Scholar PubMed PubMed Central

11. Levine, I. N. Quantum Chemistry; Pearson: New Jersey, 2014.Search in Google Scholar

12. Ciccotti, G.; Dellago, C.; Ferrario, M.; Hernández, E. R.; Tuckerman, M. E. Molecular Simulations: Past, Present, and Future (A Topical Issue in EPJB). Eur. Phys. J. B 2022, 95, 3; https://doi.org/10.1140/epjb/s10051-021-00249-x.Search in Google Scholar

13. Tuckerman, M. E. Statistical Mechanics: Theory and Molecular Simulation; Oxford University Press: Oxford, 2009.Search in Google Scholar

14. Verlet, L. Computer “Experiments” on Classical Fluids. I. Thermodynamical Properties of Lennard-Jones Molecules. Phys. Rev. 1967, 159, 98–103; https://doi.org/10.1103/PhysRev.159.98.Search in Google Scholar

15. Michael Griebel, G. Z.; Knapek, S. Numerical Simulation in Molecular Dynamics: Numerics, Algorithms, Parallelization, Applications; Springer Science & Business Media: Berlin, 2007.Search in Google Scholar

16. McCammon, J. A.; Gelin, B. R.; Karplus, M. Dynamics of Folded Proteins. Nature 1977, 267, 585–590; https://doi.org/10.1038/267585a0.Search in Google Scholar PubMed

17. Ryckaert, J.-P.; Ciccotti, G.; Berendsen, H. J. C. Numerical Integration of the Cartesian Equations of Motion of a System with Constraints: Molecular Dynamics of N-Alkanes. J. Comput. Phys. 1977, 23, 327–341. https://doi.org/10.1016/0021-9991(77)90098-5.Search in Google Scholar

18. Andersen, H. C. Molecular Dynamics Simulations at Constant Pressure and/or Temperature. J. Chem. Phys. 1980, 72, 2384–2393; https://doi.org/10.1063/1.439486.Search in Google Scholar

19. Nosé, S. A Unified Formulation of the Constant Temperature Molecular Dynamics Methods. J. Chem. Phys. 1984, 81, 511–519; https://doi.org/10.1063/1.447334.Search in Google Scholar

20. Barth, E.; Kuczera, K.; Leimkuhler, B.; Skeel, R. D. Algorithms for Constrained Molecular Dynamics. J. Comput. Chem. 1995, 16, 1192–1209. https://doi.org/10.1002/jcc.540161003.Search in Google Scholar

21. Umegawa, Y.; Yamamoto, T.; Dixit, M.; Funahashi, K.; Seo, S.; Nakagawa, Y.; Suzuki, T.; Matsuoka, S.; Tsuchikawa, H.; Hanashima, S.; Oishi, T.; Matsumori, N.; Shinoda, W.; Murata, M. Amphotericin B Assembles into Seven-Molecule Ion Channels: An NMR and Molecular Dynamics Study. Sci. Adv. 2022, 8, eabo2658; https://doi.org/10.1126/sciadv.abo2658.Search in Google Scholar PubMed PubMed Central

22. Rotilio, L.; Bayer, T.; Meinert, H.; Teixeira, L. M. C.; Johansen, M. B.; Sommerfeldt, A.; Petersen, A. R.; Sandahl, A.; Keller, M. B.; Holck, J.; Paiva, P.; Otzen, D. E.; Bornscheuer, U. T.; Wei, R.; Fernandes, P. A.; Ramos, M. J.; Westh, P.; Morth, J. P. Structural and Functional Characterization of an Amidase Targeting a Polyurethane for Sustainable Recycling. Angew Chem. Int. Ed. Engl. 2025, 64, e202419535; https://doi.org/10.1002/anie.202419535.Search in Google Scholar PubMed

23. Aranda-García, D.; Stepniewski, T. M.; Torrens-Fontanals, M.; García-Recio, A.; Lopez-Balastegui, M.; Medel-Lacruz, B.; Morales-Pastor, A.; Peralta-García, A.; Dieguez-Eceolaza, M.; Sotillo-Nuñez, D.; Ding, T.; Drabek, M.; Jacquemard, C.; Jakowiecki, J.; Jespers, W.; Jiménez-Rosés, M.; Jun-Yu-Lim, V.; Nicoli, A.; Orzel, U.; Shahraki, A.; Tiemann, J. K. S.; Ledesma-Martin, V.; Nerín-Fonz, F.; Suárez-Dou, S.; Canal, O.; Pándy-Szekeres, G.; Mao, J.; Gloriam, D. E.; Kellenberger, E.; Latek, D.; Guixà-González, R.; Gutiérrez-de-Terán, H.; Tikhonova, I. G.; Hildebrand, P. W.; Filizola, M.; Babu, M. M.; Di Pizio, A.; Filipek, S.; Kolb, P.; Cordomi, A.; Giorgino, T.; Marti-Solano, M.; Selent, J. Large Scale Investigation of GPCR Molecular Dynamics Data Uncovers Allosteric Sites and Lateral Gateways. Nat. Commun. 2025, 16, 2020; https://doi.org/10.1038/s41467-025-57034-y.Search in Google Scholar PubMed PubMed Central

24. Warshel, A.; Karplus, M. Calculation of Ground and Excited State Potential Surfaces of Conjugated Molecules. I. Formulation and Parametrization. J. Am. Chem. Soc. 1972, 94, 5612–5625; https://doi.org/10.1021/ja00771a014.Search in Google Scholar

25. Warshel, A.; Levitt, M. Theoretical Studies of Enzymic Reactions: Dielectric, Electrostatic and Steric Stabilization of the Carbonium Ion in the Reaction of Lysozyme. J. Mol. Biol. 1976, 103, 227–249. https://doi.org/10.1016/0022-2836(76)90311-9.Search in Google Scholar PubMed

26. Liu, M.; Wang, Y.; Chen, Y.; Field, M. J.; Gao, J. QM/MM Through the 1990s: The First Twenty Years of Method Development and Applications. Isr. J. Chem. 2014, 54, 1250–1263. https://doi.org/10.1002/ijch.201400036.Search in Google Scholar PubMed PubMed Central

27. van der Kamp, M. W.; Mulholland, A. J. Combined Quantum Mechanics/Molecular Mechanics (QM/MM) Methods in Computational Enzymology. Biochemistry 2013, 52, 2708–2728; https://doi.org/10.1021/bi400215w.Search in Google Scholar PubMed

28. Teixeira, L. M. C.; Coimbra, J. T. S.; Ramos, M. J.; Fernandes, P. A. Transmembrane Protease Serine 2 Proteolytic Cleavage of the SARS-CoV-2 Spike Protein: A Mechanistic Quantum Mechanics/Molecular Mechanics Study to Inspire the Design of New Drugs to Fight the COVID-19 Pandemic. J. Chem. Inf. Model. 2022, 62, 2510–2521; https://doi.org/10.1021/acs.jcim.1c01561.Search in Google Scholar PubMed PubMed Central

29. Makurat, S.; Neves, R. P. P.; Ramos, M. J.; Rak, J. Q. M/M. M. MD Study on the Reaction Mechanism of Thymidine Phosphorylation Catalyzed by the Enzyme Thermotoga Maritima Thymidine Kinase 1. ACS Catal. 2024, 14, 9840–9849; https://doi.org/10.1021/acscatal.4c01867.Search in Google Scholar

30. Kohen, A. Role of Dynamics in Enzyme Catalysis: Substantial versus Semantic Controversies. Acc. Chem. Res. 2015, 48, 466–473; https://doi.org/10.1021/ar500322s.Search in Google Scholar PubMed PubMed Central

31. Duran, C.; Casadevall, G.; Osuna, S. Harnessing Conformational Dynamics in Enzyme Catalysis to Achieve Nature-like Catalytic Efficiencies: the Shortest Path Map Tool for Computational Enzyme Redesign. Faraday Discuss. 2024, 252, 306–322; https://doi.org/10.1039/D3FD00156C.Search in Google Scholar

32. Krishna, N. B.; Roopa, L.; Pravin Kumar, R.; S, G. T. Computational Studies on the Catalytic Potential of the Double Active Site for Enzyme Engineering. Sci. Rep. 2024, 14, 17892; https://doi.org/10.1038/s41598-024-60824-x.Search in Google Scholar PubMed PubMed Central

33. Paiva, P.; Teixeira, L. M. C.; Wei, R.; Liu, W.; Weber, G.; Morth, J. P.; Westh, P.; Petersen, A. R.; Johansen, M. B.; Sommerfeldt, A.; Sandahl, A.; Otzen, D. E.; Fernandes, P. A.; Ramos, M. J. Unveiling the Enzymatic Pathway of UMG-SP2 Urethanase: Insights into Polyurethane Degradation at the Atomic Level. Chem. Sci. 2025, 16, 2437–2452; https://doi.org/10.1039/D4SC06688J.Search in Google Scholar PubMed PubMed Central

34. Branson, Y.; Söltl, S.; Buchmann, C.; Wei, R.; Schaffert, L.; Badenhorst, C. P. S.; Reisky, L.; Jäger, G.; Bornscheuer, U. T. Urethanases for the Enzymatic Hydrolysis of Low Molecular Weight Carbamates and the Recycling of Polyurethanes. Angew Chem. Int. Ed. Engl. 2023, 62, e202216220; https://doi.org/10.1002/anie.202216220.Search in Google Scholar PubMed

35. Martínez-Rosell, G.; Giorgino, T.; De Fabritiis, G. PlayMolecule ProteinPrepare: A Web Application for Protein Preparation for Molecular Dynamics Simulations. J. Chem. Inf. Model. 2017, 57, 1511–1516; https://doi.org/10.1021/acs.jcim.7b00190.Search in Google Scholar PubMed

36. Olsson, M. H. M.; Søndergaard, C. R.; Rostkowski, M.; Jensen, J. H. PROPKA3: Consistent Treatment of Internal and Surface Residues in Empirical pKa Predictions. J. Chem. Theory Comput. 2011, 7, 525–537; https://doi.org/10.1021/ct100578z.Search in Google Scholar PubMed

37. Gaussian 09, Revision D; Gaussian, Inc.: Wallingford CT, 2016.Search in Google Scholar

38. The PyMOL Molecular Graphics System; Schrödinger, LLC: New York, 2015.Search in Google Scholar

39. Li, Z.; Han, X.; Cong, L.; Singh, P.; Paiva, P.; Branson, Y.; Li, W.; Chen, Y.; Jaradat, D. M. M.; Lennartz, F.; Bayer, T.; Schmidt, L.; Garscha, U.; You, S.; Fernandes, P. A.; Ramos, M. J.; Bornscheuer, U. T.; Weber, G.; Wei, R.; Liu, W. Structure-Guided Engineering of a Versatile Urethanase Improves its Polyurethane Depolymerization Activity. Adv. Sci., 2416019. https://doi.org/10.1002/advs.202416019.Search in Google Scholar PubMed PubMed Central

40. AMBER 2018; University of California: San Francisco, 2018.Search in Google Scholar

41. Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman, P. A. A Second Generation Force Field for the Simulation of Proteins, Nucleic Acids, and Organic Molecules. J. Am. Chem. Soc. 1995, 117, 5179–5197; https://doi.org/10.1021/ja00124a002.Search in Google Scholar

42. Essmann, U.; Perera, L.; Berkowitz, M. L.; Darden, T.; Lee, H.; Pedersen, L. G. A Smooth Particle Mesh Ewald Method. J. Chem. Phys. 1995, 103, 8577–8593; https://doi.org/10.1063/1.470117.Search in Google Scholar

43. Abraham, M. J.; Murtola, T.; Schulz, R.; Páll, S.; Smith, J. C.; Hess, B.; Lindahl, E. GROMACS: High Performance Molecular Simulations through Multi-Level Parallelism from Laptops to Supercomputers. SoftwareX 2015, 1-2, 19–25. https://doi.org/10.1016/j.softx.2015.06.001.Search in Google Scholar

44. Berendsen, H. J. C.; Postma, J. P. M.; van Gunsteren, W. F.; DiNola, A.; Haak, J. R. Molecular Dynamics with Coupling to an External Bath. J. Chem. Phys. 1984, 81, 3684–3690; https://doi.org/10.1063/1.448118.Search in Google Scholar

45. Parrinello, M.; Rahman, A. Polymorphic Transitions in Single Crystals: A New Molecular Dynamics Method. J. Appl. Phys. 1981, 52, 7182–7190; https://doi.org/10.1063/1.328693.Search in Google Scholar

46. Sanfelice, D.; Temussi, P. A. Cold Denaturation as a Tool to Measure Protein Stability. Biophys. Chem. 2016, 208, 4–8; https://doi.org/10.1016/j.bpc.2015.05.007.Search in Google Scholar PubMed PubMed Central

47. Privalov, P. L. Cold Denaturation of Proteins. Crit. Rev. Biochem. Mol. Biol. 1990, 25, 281–305; https://doi.org/10.3109/10409239009090612.Search in Google Scholar PubMed

48. Connolly, M. L. Solvent-Accessible Surfaces of Proteins and Nucleic Acids. Science 1983, 221, 709–713; https://doi.org/10.1126/science.6879170.Search in Google Scholar PubMed

49. Ali, S. A.; Hassan, M. I.; Islam, A.; Ahmad, F. A Review of Methods Available to Estimate Solvent-Accessible Surface Areas of Soluble Proteins in the Folded and Unfolded States. Curr. Protein Pept. Sci. 2014, 15, 456–476; https://doi.org/10.2174/1389203715666140327114232.Search in Google Scholar PubMed

Supplementary Material

This article contains supplementary material (https://doi.org/10.1515/pac-2025-0489).

Received: 2025-04-22
Accepted: 2025-08-08
Published Online: 2025-08-27
Published in Print: 2025-10-27

© 2025 IUPAC & De Gruyter

Articles in the same Issue

  1. Frontmatter
  2. Review Articles
  3. Minimum energy path methods and reactivity for enzyme reaction mechanisms: a perspective
  4. The quantum revolution in enzymatic chemistry: combining quantum and classical mechanics to understand biochemical processes
  5. A quantum chemical perspective of photoactivated biological functions
  6. Does chemistry need more physics?
  7. Rotational dynamics of ATP synthase: mechanical constraints and energy dissipative channels
  8. Transforming dreams into reality: a fairy-tale wedding of chemistry with quantum mechanics
  9. The quantum chemistry revolution and the instrumental revolution as evidenced by the Nobel Prizes in chemistry
  10. Influence of symmetry on the second-order NLO properties: insights from the few state approximations
  11. The dichotomy between chemical concepts and numbers after almost 100 years of quantum chemistry: conceptual density functional theory as a case study
  12. How ‘de facto variational’ are fully iterative, approximate iterative, and quasiperturbative coupled cluster methods near equilibrium geometries?
  13. Electronic structure of methyl radical photodissociation
  14. Bridging experiment and theory: a computational exploration of UMG-SP3 dynamics
  15. Research Articles
  16. O–Li⋯O and C–Li⋯C lithium bonds in small closed shell and open shell systems as analogues of hydrogen bonds
  17. Metal–ligand bonding and noncovalent interactions of mutated myoglobin proteins: a quantum mechanical study
https://doi.org/10.1515/pac-2025-0489
Keywords for this article
Enzyme stability; molecular dynamics; protein dynamics; quantum science and technology; UMG-SP3; urethanase

Articles in the same Issue

  1. Frontmatter
  2. Review Articles
  3. Minimum energy path methods and reactivity for enzyme reaction mechanisms: a perspective
  4. The quantum revolution in enzymatic chemistry: combining quantum and classical mechanics to understand biochemical processes
  5. A quantum chemical perspective of photoactivated biological functions
  6. Does chemistry need more physics?
  7. Rotational dynamics of ATP synthase: mechanical constraints and energy dissipative channels
  8. Transforming dreams into reality: a fairy-tale wedding of chemistry with quantum mechanics
  9. The quantum chemistry revolution and the instrumental revolution as evidenced by the Nobel Prizes in chemistry
  10. Influence of symmetry on the second-order NLO properties: insights from the few state approximations
  11. The dichotomy between chemical concepts and numbers after almost 100 years of quantum chemistry: conceptual density functional theory as a case study
  12. How ‘de facto variational’ are fully iterative, approximate iterative, and quasiperturbative coupled cluster methods near equilibrium geometries?
  13. Electronic structure of methyl radical photodissociation
  14. Bridging experiment and theory: a computational exploration of UMG-SP3 dynamics
  15. Research Articles
  16. O–Li⋯O and C–Li⋯C lithium bonds in small closed shell and open shell systems as analogues of hydrogen bonds
  17. Metal–ligand bonding and noncovalent interactions of mutated myoglobin proteins: a quantum mechanical study
Sign up now to receive a 20% welcome discount
Institutional Access
How does access work?
Winner of the OpenAthens UX Award 2026
Winner of the OpenAthens UX Award 2026
Have an idea on how to improve our website?
Please write us.
© 2026 De Gruyter Brill
Downloaded on 9.3.2026 from https://www.degruyterbrill.com/document/doi/10.1515/pac-2025-0489/html
Scroll to top button