Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2025 Oct 8;10(41):47879–47893. doi: 10.1021/acsomega.5c01556

Synthetic Multidomain Proteins Containing Unstructured and α‑Helical Linkers Reveal a Differential Impact of Molecular Crowding on Catalytic Activity and Conformation

Asfia Sultana 1, Anupriya M Geethakumari 1, Wesam S Ahmed 1, Kabir H Biswas 1,*
PMCID: PMC12547749  PMID: 41141769

Abstract

Molecular crowding has been shown to impact both enzymatic activity and protein conformation individually. However, a simultaneous assessment of its effect on the two parameters, especially in multidomain proteins, has not yet been reported. Here, utilizing multidomain proteins containing the mNeonGreen (mNG) fluorescent protein and the NanoLuc (NLuc) luciferase fused by either unstructured or α-helical linkers of different lengths, we report a differential impact of molecular crowding on the enzymatic activity and protein conformation in a linker and crowder size-dependent manner. Specifically, Gaussian accelerated molecular dynamics (GaMD) simulations with representative unstructured and α-helical linkers revealed differences in their structural dynamics. Simultaneous monitoring of enzymatic activity through NLuc bioluminescence and protein conformation through Bioluminescence Resonance Energy Transfer between NLuc (donor) and mNG (acceptor) revealed both polyethylene glycol molecular weight-dependent and linker length-dependent impacts on NLuc enzymatic activity and conformation of an unstructured linker-containing multidomain protein. Further, multidomain proteins containing α-helical linkers of different lengths revealed a pronounced impact of molecular crowding on NLuc enzymatic activity and a differential impact on protein conformation. Overall, through simultaneous monitoring of the impact of molecular crowding on enzymatic activity and protein conformation, we reveal a differential impact of molecular crowding on multidomain proteins containing different linkers and thus aid in further understanding the impact of molecular crowding on multidomain protein structure and function.

graphic file with name ao5c01556_0007.jpg

graphic file with name ao5c01556_0005.jpg

Introduction

The intracellular environment is densely packed with biomolecules like proteins, RNA, DNA, and small molecules, , with concentrations ranging from 50 to 200 mg/ml of molecules, leading to the phenomenon of molecular crowding. ,, Molecular crowding inevitably results in an excluded volume effect, leading to alterations in enzymatic activity, protein conformation, and other protein-related phenomena. , For instance, molecular crowding influences enzymatic activity by modulating substrate accessibility, enzyme–substrate interactions, and reaction kinetics. Some enzymes may exhibit enhanced activity due to a favorable effect of molecular crowding, whereas others may experience reduced activity due to reduced encounters between enzymes and their substrates. , Overall, the impact of molecular crowding on enzymatic activity is multifaceted, with the net effect depending on the specific enzyme and cellular context. Understanding the interplay between molecular crowding and enzymatic function is crucial for elucidating the regulatory mechanisms underlying cellular metabolism and signaling pathways.

Molecular crowding influences enzymatic activity and protein conformation through steric constraints and excluded volume effects. Studies on human arginase-1 and α-chymotrypsin reported decreased activity due to crowding-induced inhibition, while DNA ligases from rat liver nuclei exhibited enhanced catalytic efficiency. Crowding also modulates protein folding, stabilizing specific conformations or promoting structural transitions, as observed in yeast protein disulfide isomerase and hemoglobin. Previous studies have reported biosensors, studying the impact of crowded environments on intrinsically disordered regions and protein dynamics. , Besides, multidomain proteins, characterized by two or more distinct domains, often possess linker regions that connect these domains. Traditionally, these linkers have been perceived as structural elements that primarily serve as “spacers” to maintain the spatial arrangement of domains in the protein and can be classified as either unstructured or α-helical based on their secondary structure. However, recent studies have revealed that these seemingly inert linkers play a significant regulatory role in modulating the impact of molecular crowding on these proteins. , Gaining insights into the folding and interactions of proteins within populated cellular environments is essential for comprehending cellular functions, the roles of proteins, and the mechanisms underlying diseases. , Analyzing protein structure in crowded environments offers valuable perspectives into proteins’ actual structural and functional attributes within living systems. ,

A range of biophysical and biochemical methods has been employed in determining enzymatic activity under crowded conditions. Various approaches, such as structural modulation via domain displacements through FRET, have been used to study the function of enzymatic activity. , Other techniques, including neutron and X-ray scattering, isothermal calorimetry, and FTIR, have been employed in studying the enzymatic activity under crowded conditions. , Genetically encoded FRET-based sensors have been used to determine crowding in vivo and in vitro. , However, these methods could determine either the biochemical activity or the conformational change, but not both. To address the challenge of simultaneously monitoring both the biochemical rate and conformation of a protein under molecular crowding conditions, we decided to utilize the genetically encoded, bioluminescence resonance energy transfer (BRET)-based − synthetic multidomain proteins consisting of a luciferase enzyme and a fluorescent protein. In these synthetic multidomain proteins, the luciferase enzyme, which emits light through oxidizing its substrate, can be used as a reporter for enzyme activity. On the other hand, excitation of the fluorescent protein through resonance energy transfer, which is dependent upon the distance and relative orientation of the luciferase and the fluorescent protein, can be utilized as a reporter of protein conformation. BRET has been successfully used in generating a variety of biosensors used in detecting small molecule changes and in studying protein–protein interactions and is preferred over FRET because of the high signal-to-noise ratio. While several luciferases and fluorescent protein pairs have been utilized for developing BRET-based sensors, the fluorescent protein mNeonGreen (mNG) and the small, bright, and highly stable luciferase protein NanoLuc (NLuc) have gained popularity due to their brightness, spectral overlap, and thermal stability.

In the current article, for simultaneous monitoring of the impact of molecular crowding on the synthetic two-domain proteins linked with the unstructured and α-helical linker, we performed Gaussian accelerated molecular dynamics (GaMD) simulations to characterize the structural dynamics of the 21 AAs-long unstructured and (EAAAK)3 α-helical linkers, which revealed the differential structural dynamics of the two linkers. A twenty-one (21) AAs-long unstructured linker largely showed coil and turn, with some residues showing an α-helical, secondary structure, while the (EAAAK)3 α-helical linker exhibited largely an α-helical structure and fluctuations in the number of hydrogen bonds formed. Further, we performed in vitro bioluminescence and BRET assays using poly­(ethylene glycol) (PEG) as a molecular crowding agent. Experiments with both the 21 AAs-long unstructured linker and (EAAAK)3 α-helical linker containing multidomain proteins revealed PEG molecular weight- and concentration-dependent decrease in NLuc bioluminescence activity. On the other hand, BRET was found to decrease in the case of PEG 400, while it was found to increase with all other higher-molecular-weight PEGs. Importantly, these effects of molecular crowding were also dependent on the length of both the unstructured and α-helical linker-containing multidomain proteins, with a steeper decline in enzymatic activity observed for the shorter 17 and 12 AAs-long unstructured linkers and the (EAAAK)2 and (EAAAK) α-helical linker-containing multidomain protein. While both the shorter linker-containing multidomain proteins showed a negligible change in protein conformation in the presence of lower-molecular-weight PEG, they showed significant BRET changes in the presence of higher-molecular-weight PEGs (20,000 and 35,000 Da). Overall, we believe that our results demonstrate that molecular crowding effects are not uniform across different protein designs and emphasize the importance of considering these factors in protein engineering and understanding protein function in cellular contexts.

Results

Molecular crowding, as commonly encountered within living cells, has been reported to influence both enzymatic activity , as well as protein conformation. , While the two have been studied individually, we aimed to monitor the impact of molecular crowding on enzymatic activity and protein conformation simultaneously in a single assay. For this, we decided to utilize synthetic multidomain proteins containing mNG at the N-terminus and NLuc at the C-terminus, linked through either a 21 AAs-long unstructured (GTENLYAVLQSGFRGSGGSAAA) or an α-helical (GTEAAAKEAAAKEAAAKAAA) linker (Figure A). In these multidomain proteins, the bioluminescence activity of NLuc, which oxidizes its substrate and emits blue light in the process, served as a reporter of enzymatic activity and allowed us to monitor changes in the enzymatic activity of the multidomain proteins under molecular crowding (Figure B). , On the other hand, the efficiency of BRET from the donor NLuc to the acceptor mNG in these multidomain proteins allowed us to monitor changes in their conformation under molecular crowding (Figure B). , BRET efficiency is highly sensitive to the distance and relative orientation between the donor and the acceptor, and a number of multidomain biosensors have been successfully developed by utilizing the phenomenon. Thus, monitoring the NLuc bioluminescence and BRET from NLuc to mNG in our experimental setup (Figure B) allowed us to simultaneously monitor enzymatic activity and protein conformation under molecular crowding conditions.

1.

1

Open in a new tab

Multidomain protein structure and schematic representation of simultaneous monitoring of reaction rate and conformational change, monitoring the impact of molecular crowding on synthetic multidomain proteins. (A) Structural models of the 21 AAs-long unstructured linker and (EAAAK)3 α-helical linker-containing multidomain proteins consisting of an N-terminal mNG and a C-terminal NLuc. Sequences of the 21 AAs-long unstructured and the (EAAAK)3 α-helical linkers are shown below. (B) Schematic representation of a multidomain protein in the presence of low (left panels) and high (right panels) concentrations of a low (top row) and high (bottom row) molecular weight crowder. Molecular crowding is induced using increasing concentrations of PEG, ranging from 0 to 25% and increasing molecular weights of PEG, ranging from 400 Da to 35,000 Da. The size of the blue and green arrows indicates the bioluminescence activity of NLuc (enzymatic activity) and the efficiency of BRET from NLuc to mNG (conformation of the protein), respectively.

MD Simulation Reveals Differential Structural Dynamics of the 21 AAs-Long Unstructured Linker and (EAAAK)3 α-Helical Linker

For monitoring the impact of molecular crowding on the conformation of multidomain proteins, we planned to utilize an unstructured (GTENLYAVLQSGFRGSGGSAAA) and an α-helical (GTEAAAKEAAAKEAAAKAAA) linker between mNG and NLuc. The unstructured linker sequence has been characterized computationally and has been found to be largely nonstructured. , On the other hand, the (EAAAK)3 linker sequence has been reported to be α-helical and has been widely used in engineered proteins because of its strong conformational stability under diverse conditions, including molecular crowding ,

To gain insights into the structural dynamics of the unstructured and the α-helical linkers, we generated structural models of the linkers using MODELLER , (Figure A) and performed a 1 μs-long GaMD simulation, , in which a harmonic boost potential that follows a Gaussian distribution was applied to provide an enhanced sampling of protein conformational states to accelerate conformational transitions of the linkers using NAMD software. , Following the GaMD simulation, we analyzed the trajectories for a number of parameters (Figure ; Videos S1 and S2). The root-mean-square deviation (RMSD) analysis revealed a generally fluctuating value (6.7 ± 1.3 Å; mean ± standard deviation (SD)) over the 1 μs simulation of the 21 AAs-long unstructured linker (Figure B, left panel). On the other hand, RMSD analysis of the (EAAAK)3 α-helical linker showed large fluctuations, including very small values during some simulation times (4.7 ± 2.4 Å; Figure B, right panel), indicating large structural fluctuations in the peptide. Root-mean-squared fluctuation (RMSF) analysis revealed higher fluctuations in the case of the 21 AAs-long unstructured linker (Figure C, left panel) compared to the (EAAAK)3 α-helical linker (Figure C, right panel), suggesting that individual residues in the (EAAAK)3 α-helical linker remained more stable for the simulation. We then analyzed the trajectories for the radius of gyration (Rg) of linkers and found the 21 AAs-long unstructured linker to show more fluctuations (9.8 ± 1.6 Å; Figure D, left panel) compared to the (EAAAK)3 α-helical one (9.7 ± 1.43 Å; Figure D, right panel). We then analyzed the secondary structure of the linkers over the simulation trajectories by determining the propensity of each residue in the linkers for various secondary structural elements, including coils, α-helix, β-sheets, turns, and others and determined their frequencies (Figure E). This analysis revealed the 21 AAs-long unstructured linker to assume an α-helix, β-sheets, turns, and coils with similar frequencies (∼0.4) over the length of the linker (Figure E, left panel). The (EAAAK)3 linker, which is predicted to be α-helical, largely assumes an α-helical structure, although some residues also assume turn and coil secondary structures (Figure E, right panel), suggesting that the (EAAAK)3 α-helical linker can undergo structural reorganization that might lead to a change in the conformation of the α-helical linker containing a synthetic multidomain protein.

2.

2

Open in a new tab

GaMD simulations reveal differential structural dynamics of 21 AAs-long unstructured and (EAAAK)3 α-helical linkers. (A) Snapshots from a 1000 ns-long GaMD simulation, acquired every 100 ns, of the 21 AAs-long unstructured (left panel) and the (EAAAK)3 α-helical linker (right panel). (B–H) Graph showing RMSD (B), RMSF (C), Rg (D), probability of the various indicated secondary structural elements explored by individual residues (E), number of hydrogen bonds (F), number of salt bridges (G), and end-to-end distance between the N- and C-termini (H) determined from 1 μs-long GaMD simulations of the 21 AAs-long unstructured (left panels) and (EAAAK)3 α-helical (right panels) linkers. (I) Heatmaps showing the dynamic cross-correlation (DCC) scores determined from 1 μs-long GaMD simulations of the 21 AAs-long unstructured (left panels) and (EAAAK)3 α-helical linkers (right panel).

We then analyzed the trajectories for intraprotein hydrogen bond (H-bond) and salt bridge interactions (Figure F,G). The 21 AAs-long unstructured linker showed a lesser number of hydrogen bonds with a mean value of (2.43 ± 1.77) (Figure F, left panel), as compared to the (EAAAK)3 α-helical linker (3.79 ± 2.48) (Figure F, right panel), in addition to more regular fluctuations in the number of hydrogen bonds (Figure F). Importantly, the (EAAAK)3 α-helical linker showed transitions between a state with a lower number of hydrogen bonds and another with a higher number of hydrogen bonds (Figure F, right panel), suggesting conformational plasticity, in agreement with the secondary structure analysis (Figure S1). Salt bridge interaction analysis revealed a greater number of such interactions in the (EAAAK)3 α-helical linker compared to the 21 AAs-long unstructured linker (Figure G). Additionally, salt bridge-forming residue pairs E1-R16 and E5-R16 in the 21 AAs-long unstructured linker showed large fluctuations (19.5 ± 8.1 Å vs 18.0 ± 7.1 Å, respectively) (Figure S2A,B). Similarly, fluctuations were observed in the distance between salt bridge-forming residues in the (EAAAK)3 α-helical linker in most cases, except for E3-K7 (17.5 ± 4.2 Å) and E13-K12 (15.4 ± 3.5 Å), which indicated higher fluctuations (Figure S2C–K). End-to-end distance analysis revealed a shorter separation in the case of the 21 AAs-long unstructured linker as compared to the (EAAAK)3 α-helical linker (20.7 ± 8.9 Å vs 24.9 ± 8.1 Å, respectively) (Figure H, left panel). Importantly, while the 21 AAs-long unstructured linker showed a generally fluctuating end-to-end distance, the (EAAAK)3 α-helical linker showed long-lived transitions in the end-to-end distance, suggestive of a “slow” conformational sampling (Figure H, right panel). Finally, we performed DCC analysis of the trajectories to determine any long-range interaction between the residues in the peptides (Figure I). This revealed no large correlated motions in the unstructured peptide, while the α-helical peptide exhibited stronger, short-range correlated motions. This is likely due to structural constraints in the α-helical linker that reduce the conformational entropy and restrict local fluctuations, thereby enhancing the propagation of motion across nearby residues. These findings suggest that the linker peptide could promote a more cohesive dynamic network of residues, which could potentially modulate the response of the α-helical linker-containing multidomain protein to molecular crowding. Overall, the GaMD simulations suggest that the 21 AAs-long unstructured linker fluctuates between different short-lived, fast-exchanging conformational states, while the (EAAAK)3 α-helical linker shows long-lasting, slow-exchanging conformational states.

Simultaneous Monitoring of Enzymatic Activity and the Conformation of the 21 AAs-Long Unstructured Linker-Containing Synthetic Multidomain Protein

After characterizing the structural dynamics of both the 21 AAs-long unstructured and (EAAAK)3 α-helical linkers, we aimed to investigate the impact of PEG-induced molecular crowding on both the bioluminescence activity of NLuc and the conformational change of the protein simultaneously (Figure B). PEG has been utilized as a crowding agent in a large number of studies and is known to impact enzymatic activity , and is also known to impede biomolecular movements, making intermolecular interactions more challenging. , For this, we first monitored the expression of the protein through Western blot analysis using an anti-His-tag antibody and lysates prepared from HEK293T cells transfected with a plasmid expressing the unstructured linker-containing multidomain protein (expected molecular weight of ∼50 kDa) (Figure S3). We then incubated lysates containing the 21 AAs-long unstructured linker-containing multidomain protein with a range of concentrations (0 to 25%) of PEG of different molecular weights (400, 2000, 4000, 8000, 20,000, and 35,000 Da) to comprehensively monitor the influence of molecular crowding in vitro (Figure B). We first determined the bioluminescence activity of the 21 AAs-long unstructured linker-containing synthetic multidomain protein over time and determined (Figure A) the rate of decline in bioluminescence activity (k biolum; NLuc enzymatic activity/bioluminescence rate constant) and plotted them against the concentration of PEG used in the reaction (Figure B). The data were fitted using a one-phase decay model, and the resulting k values revealed a general decline in the NLuc bioluminescence activity with an increasing concentration of PEG, although the rate of this decline was dependent on the molecular weight of PEG rather than variability inherent to lysate components. Despite the expected decrease in bioluminescence with increasing PEG concentrations, the bioluminescence values remained high, indicating that the protein remained soluble and functional. Further, the preservation of fluorescence intensity supports that the conformational integrity of the protein was maintained across all crowding conditions (Figure C).

3.

3

Open in a new tab

Simultaneous monitoring of enzymatic activity, fluorescence, and conformation of unstructured linker-containing multidomain proteins under different crowding conditions. (A–D) Graphs showing bioluminescence activity over time (A), rate of bioluminescence decay (B), fluorescence (C), and BRET of the (D) unstructured linker-containing multidomain protein in the presence of a range of concentrations (0–25%; (w/v)) of different molecular weights (400, 2000, 4000, 8000, 20,000 and 35,000 Da) PEG. Graphs showing BRET against the rate of bioluminescence decay of the 21 (E), 17 (F), and 12 (G) AAs-long unstructured linker-containing multidomain proteins in the presence of the indicated concentrations of the various molecular weight PEGs. Data shown are mean ± SD obtained from three independent experiments. p-values shown were obtained from correlation analyses.

For instance, in the presence of the smaller molecular weight PEG 400, the 21 AAs-long unstructured linker-containing multidomain protein exhibited a faster decrease in the reaction rate, as indicated by the k biolum value corresponding to 0.76 ± 0.50 CPS.mL.s–1.g–1 ( Figure B, first panel). This may be attributed to the lesser impact of crowding due to its smaller size. Similar effects can be observed in the presence of PEG 2000, with a k biolum value of 0.52 ± 0.30 CPS.mL.s–1.g–1 (Figure B, second panel), indicating an intermediate effect of crowding. The impact of crowding in the presence of PEG 4000 demonstrated a k biolum value of 0.12 ± 0.64 CPS.mL.s–1.g–1 (Figure B, third panel). The reaction rates in the presence of higher-molecular-weight PEGs, specifically in the presence of PEG 8000, revealed a k biolum value of 0.03 ± 0.07 CPS.mL.s–1.g–1 ( Figure B, fourth panel), while PEG 20,000 and 35,000 showed a k biolum value of 0.11 ± 0.1 (Figure B, fifth panel) and 0.21 ± 0.06 CPS. mL.s-1.g–1, respectively. This analysis showed a molecular weight-dependent decrease in the reaction rate. However, upon comparing, no significant difference in the reaction rates of 21 AAs-long unstructured and (EAAAK)3 α-helical linker-containing multidomain protein across increasing concentrations of PEG was observed (Table S1).

We then focused our attention on the confirmation of the synthetic multidomain protein containing a 21 AAs-long unstructured linker in the presence of the crowding agent. For this, we monitored (ratio of emission at 467 and 533 nm) between NLuc (donor) and mNG (acceptor) as a change in either the distance or relative orientation of the two proteins can lead to a change in the BRET. While BRET is a ratiometric measurement, i.e., ratio of emission of mNG and that of NLuc, a change in the mNG fluorescence can potentially complicate the interpretation of the results. Therefore, we determined the fluorescence of mNG under different molecular crowding conditions through direct excitation with 480 nm light and found that mNG fluorescence remained largely unaltered in the presence of PEG (Figure C). Although PEG at higher concentrations can promote molecular crowding and potential precipitation, the absence of any loss of mNG fluorescence supported the conclusion that protein remains soluble and folded, even under the higher crowded conditions tested. Additionally, we determined the bioluminescence spectra of the synthetic multidomain protein containing a 21 AAs-long unstructured linker and found two emission peaks, one corresponding to NLuc at ∼467 nm and another corresponding to mNG at ∼533 nm, as determined from the two Gaussian fits of the spectral data (Figure S4). The presence of the mNG emission peak indicates a significant efficiency of resonance energy transfer (BRET) between NLuc and mNG, and therefore, the multidomain protein construct is suitable for monitoring the conformation of the protein.

Incubation of the synthetic multidomain protein containing the 21 AAs-long unstructured linker in the presence of increasing concentrations of PEG 400 resulted in a decrease in the BRET with a rate (k BRET) of decrease of 0.12 ± 0.04 (data fit to a one-phase decay model) (Figure D), suggesting that the small-molecular-weight PEG is causing either an increase in the distance between NLuc and mNG or a change in the relative orientation of the two proteins. This effect was not observed in the presence of other PEG molecules. In fact, we observed an increase in the BRET at higher concentrations of PEG 2000 (k BRET = 0.10 ± 0.14) and 4000 (k BRET = 0.12 ± 0.64) (Figure D). The increase in BRET was more pronounced with PEG 8000 (k BRET = 0.75 ± 1.03), PEG 20,000 (k BRET = 0.85 ± 0.42), and PEG 35,000 (k BRET = 0.81 ± 0.25). These results suggest that the molecular crowding created by larger-molecular-weight PEG molecules is causing either a decrease in the distance between NLuc and mNG or changing their orientation to increase the BRET of the synthetic multidomain protein containing the 21 AAs-long unstructured linker. Comparative analysis of the impact of molecular crowding on the conformational change of a 21 AAs-long unstructured linker and (EAAAK)3 α-helical linker containing a multidomain protein showed significance in the presence of PEG 8000, PEG 20000, and PEG 35000 (Table S2).

Taking advantage of the synthetic multidomain proteins to determine the impact of molecular crowding on enzymatic activity and protein conformation simultaneously and the relation thereof, we plotted the rate of bioluminescence against the BRET observed under a range of concentrations of various PEG molecules (Figure E). This revealed that the NLuc bioluminescence activity did not correlate with BRET with increasing concentrations of PEG 400 (Figure F), suggesting that the enzymatic activity of the multidomain protein did not show any specific trend with its conformation in the presence of a small-molecular-weight crowding agent. However, the protein appeared to show a trend in which the enzymatic activity of the multidomain protein decreased with an increase in the level of BRET with increasing concentrations of PEG 2000. Similar observations were made with a PEG 4000. Interestingly, the protein showed a more correlated trend in its enzymatic activity and BRET with increasing PEG 8000, PEG 20,000, and PEG 35,000 concentrations. Overall, these results revealed a complex interplay between the molecular weight, enzymatic activity, and conformation and highlighted the utility of the synthetic multidomain protein containing a 21 AAs-long unstructured linker in the simultaneous monitoring of the effect of the molecular crowding agent, PEG. Correlation analysis was performed to assess the relationship between the reaction rate and conformational change, and the analysis showed a statistically significant effect in the presence of higher-molecular-weight PEG 8000 (p = 0.1247), PEG 20000 (p = 0.0117), and PEG 35000 (p = 0.059) (Figure E).

Further, to investigate the linker-dependent impact of molecular crowding on the reaction rate and conformational changes in the unstructured linker-containing multidomain proteins, we generated two additional proteins with 17 and 12 AAs-long unstructured linkers and two short α-helical linker-containing multidomain proteins (EAAAK)2 and (EAAAK). ,, Western blot analysis revealed the expression of these proteins (Figure S3). We determined the bioluminescence spectra of the synthetic multidomain protein containing both 17 and 12 AAs-long unstructured linker-containing multidomain protein (Figure S5) and the short α-helical linker-containing protein (EAAAK)2 and (EAAAK) (Figure S6) and observed two emission peaks corresponding to mNG and NLuc. Additionally, basal (in the absence of any PEG) BRET measurements of the unstructured linker-containing multidomain proteins revealed an increase in BRET with the reduction in linker length (Figure S7A). We then performed experiments to monitor the impact of molecular crowding on the reaction rate and protein conformation with these shorter unstructured linker-containing multidomain proteins.

The experiments did not reveal a significant decrease in the NLuc bioluminescence with increasing PEG concentrations when compared with the 21 AAs-long unstructured linker-containing multidomain protein (Figures S8A, S9A) (Table S3). The data were fitted to a one-phase decay model, and the reaction rate of 17 and 12-AAs-long unstructured linker-containing multidomain protein (Figures S8B, S9B) obtained from the reaction rate k values showed a decrease with increasing molecular weight of PEG. In the presence of PEG 400, the k biolum of 17 and 12-AAs long unstructured linker-containing multidomain protein was 0.56 ± 0.52, and the k biolum of 0.85 ± 0.79. The k biolum values were 0.49 ± 0.15 and 0.31 ± 0.36 in the presence of PEG 2000. In the presence of PEG 4000, the 17 and 12-AAs long unstructured linker-containing multidomain protein represented a k biolum of 0.99 ± 0.58 and 0.80 ± 0.57 CPS. CPS.mL.s–1.g–1. The higher-molecular-weight PEG 8000, 20,000, and 35,000 Da showed a decrease in the bioluminescence rate with k biolum of 0.49 ± 0.28 and 0.39 ± 0.24, 0.44 ± 0.18 and 0.84 ± 0.24, and 0.62 ± 0.23 and 0.27 ± 0.18 CPS.mL.s–1.g–1. PEG-molecular-weight-dependent impact of molecular crowding on the reaction rate of different unstructured linker-containing multidomain proteins showed that the 21 AAs-long unstructured linker containing a multidomain protein in the presence of PEG 2000 was significantly different from other higher-molecular-weight PEG (Two-way ANOVA PEG 2000 vs 8000, p = 0.0305; 2000 vs 20,000, p = 0.0019; 2000 vs 35,000, p < 0.0001). The 17 AAs-long unstructured linker-containing multidomain protein showed a significantly different impact in the presence of PEG 400 in comparison with the other molecular weight PEG (Two-way ANOVA PEG 400 vs 2000, p < 0.0001; PEG 400 vs 4000, p < 0.0001; PEG 400 vs 8000, p < 0.0001; PEG 400 vs 20,000, p < 0.0001; PEG 400 vs 35,000, p < 0.0001) (Table S3).

However, no decrease in the fluorescence in the presence of increasing molecular weights and concentrations of PEG of both the 17 and 12 AAs-long unstructured linker-containing multidomain protein was observed (Figures S8C, S9C). Analyzing the BRET of both the 17 and 12 AAs-long unstructured linker-containing multidomain protein showed that smaller-molecular-weight PEGs (400, 2000, 4000, and 8000) showed a decreasing trend in BRET. The decrease in BRET of both the 17 and 12 AAs-long unstructured linker-containing multidomain protein in the presence of PEG 400 in comparison with the 21 AAs-long unstructured linker-containing multidomain protein showed a significant difference across all concentrations, with a faster decrease in BRET, while the 12 AAs-long multidomain protein in the presence of PEG 2000 showed an increasing trend (Table S4). However, k BRET value from the one-phase association model fitted data showed that higher-molecular-weight PEGs 20,000 and 35,000 of both 17 and 12 AAs-long unstructured linkers containing multidomain showed an increase in BRET with increasing PEG concentrations, with a k BRET value of 0.92 ± 0.66 and 0.54 ± 0.40, and k BRET values of 0.18 ± 0.05 and 0.32 ± 0.10 (Figures S8D, S9D). Molecular weight-dependent impact of molecular crowding on the BRET showed that a 21 AAs-long unstructured linker containing a multidomain protein differed significantly across all concentrations and molecular weight PEG (Two-way ANOVA p < 0.0001). The 17AA and 12 AAs-long unstructured linkers containing multidomain proteins also show a significantly different impact of PEG molecular weight, dependent on the impact of crowding on BRET in the presence of smaller-molecular-weight PEGs (Table S4).

Simultaneous assessment of the impact of molecular crowding on both the enzymatic activity and the conformation of both the 17 and 12 AAs-long linker-containing multidomain proteins showed a decrease in BRET and the rate of bioluminescence in the presence of smaller-molecular-weight PEGs (400, 2000, 4000, and 8000 Da). In contrast, the presence of PEG 2000, a 12 AAs-long unstructured linker containing a multidomain protein, had a different impact, with a gradual increase in BRET and a decrease in the rate of bioluminescence with increasing PEG concentrations. The observed change in trend is opposite to the impact observed in the 21 AAs-long unstructured linker-containing multidomain protein in the presence of the same-molecular-weight PEG and concentrations, although the change was gradual. A shift in the trend of BRET and bioluminescence rate changes of both the 17 and 12 AAs-long unstructured linker-containing multidomain protein in the presence of higher-molecular-weight PEG (20,000 and 35,000 Da), resulting in a decrease in enzymatic activity with an increase in BRET, representing the impact of molecular crowding, was observed (Figure G). p-values corresponding to the correlation analysis are shown in the graphs (Figure F,G).

Impact of Molecular Crowding on the Enzymatic Activity and Protein Conformation of the (EAAAK)3 α-Helical Linker-Containing Multidomain Protein

Having simultaneously monitored the enzymatic activity and conformation of the synthetic multidomain protein containing the unstructured linker-containing multidomain proteins, we performed similar experiments with the (EAAAK)3 α-helical linker-containing multidomain protein. The NLuc bioluminescence rate decreased with increasing concentrations of various molecular weights of PEG, similar to the 21 AAs-long unstructured linker-containing multidomain protein (Figure A,B). The data were fitted using a one-phase association model, and the resulting k biolum values of PEG 400 showed a k biolum of 0.44 ± 0.04 CPS.mL.s–1.g–1, and PEG 2000 showed a value of 0.89 ± 0.47 CPS.mL.s–1.g–1. The k biolum values of PEG 4000 and 8000 corresponded to 0.44 ± 1.55 and 0.75 ± 0.65 CPS.mL.s–1.g–1. The higher-molecular-weight PEG 8000, 20,000, and 35,000 showed a k biolum value of 1.24 ± 0.8 and 0.64 ± 0 CPS.mL.s–1.g–1, respectively. The rate values of the (EAAAK)3 α-helical linker indicate a similar trend toward higher rates compared to the 21 AAs-long unstructured linker-containing multidomain protein and are higher. Based on this result, it would be appropriate to conclude that the impact of PEG-mediated molecular crowding appeared to be more pronounced with the (EAAAK)3 α-helical linker-containing synthetic multidomain protein in the presence of a higher molecular weight compared to the 21 AAs-long unstructured linker-containing multidomain protein (Figure A,B).

4.

4

Open in a new tab

Simultaneous monitoring of enzymatic activity, fluorescence, and conformation of α-helical linker-containing multidomain proteins under different crowding conditions. (A–D) Graphs showing bioluminescence activity over time (A), rate of bioluminescence decay (B), fluorescence (C), and BRET (D) of the (EAAAK)3 α-helical linker-containing multidomain protein in the presence of a range of concentrations (0–25%; (w/v)) of different molecular weights (400, 2000, 4000, 8000, 20,000 and 35,000 Da) PEG. (E–G) Graphs showing BRET against the rate of bioluminescence decay of the (EAAAK)3 (E), (EAAAK)2 (F), and EAAAK (G) α-helical linker-containing multidomain proteins in the presence of the indicated concentrations of the various molecular weight PEGs. Data shown are mean ± SD obtained from three independent experiments. p-values shown were obtained from correlation analyses.

Similar to the 21 AAs-long unstructured linker multidomain protein, mNG fluorescence of the (EAAAK)3 α-helical linker-containing multidomain protein largely remained unchanged with increasing concentrations of various molecular weight PEG (Figure C), thus enabling us to monitor BRET. BRET analysis of the (EAAAK)3 α-helical linker-containing multidomain protein revealed a decrease in the BRET with a k BRET of 0.16 ± 0.05 with PEG 400 (Figure D), suggesting that incubation of the multidomain protein with the small-molecular-weight PEG is leading to either an increase in the distance between NLuc and mNG or a change in their relative orientation in a way that is leading to a decrease in BRET. On the other hand, the decrease in BRET was much reduced with PEG 2000 as compared to that observed with PEG 400, although the k BRET was found to be higher (0.29 ± 0.02 vs 0.16 ± 0.05 in the presence of PEG 400 and PEG 2000, respectively) (Figure D). Again, the (EAAAK)3 α-helical linker-containing multidomain protein showed an increase in BRET with increasing concentrations of PEG 4000, PEG 8000, PEG 20,000, and PEG 35,000, with k BRET values of 0.25 ± 0.20, 1.12 ± 0.58, 0.88 ± 0.17, and 0.33 ± 0.02, respectively.

Analyzing the rate of bioluminescence (NLuc enzymatic activity) against BRET (protein conformation) observed with the (EAAAK)3 α-helical linker-containing multidomain protein under increasing concentrations of various PEG molecules (Figure E) showed a “biphasic” behavior, with NLuc bioluminescence rates decreased with a decrease in BRET with increasing PEG 400 concentrations (Figure E). A similar effect was observed with PEG 2000, although the changes were more gradual (Figure E). Interestingly, a transition was observed with PEG 4000 wherein the NLuc bioluminescence rates decreased with increasing BRET as well as the concentration of PEG (Figure E). Also, NLuc bioluminescence rates and BRET appeared to change more “linearly” in the case of PEG 4000, as compared to either PEG 400 or PEG 2000. Similarly, we observed that the NLuc bioluminescence rates changed almost linearly with BRET with increasing concentrations of PEG 8000, PEG 20,000, and PEG 35,000 (Figure E). These observations are different from those made with the 21 AAs-long unstructured linker-containing multidomain protein, in which case a “saturation” was observed with respect to the NLuc bioluminescence rate and BRET at high concentrations (6.25, 12.5, and 25%) of PEG 8000, PEG 20,000, and PEG 35,000. p-values corresponding to the correlation analysis of the (EAAAK)3 α-helical linker containing a multidomain protein are shown in the graphs (Figure E). Upon observing the differential impact of PEG on unstructured and α-helical linkers containing multidomain proteins, we determined the percentage change in BRET, comparing the two different linker peptides containing multidomain proteins subjected to a range of concentrations and molecular weight PEGs, and fitted the data to a simple linear regression model. It is observed that a 21 AAs long unstructured linker-containing multidomain protein showed more effect with increasing molecular weight PEG, except PEG 2000, compared to the (EAAAK)3 α-linker-containing multidomain protein (Figure S10).

Further, shorter α-helical linker-containing multidomain proteins (EAAAK)2 and (EAAAK) were generated to determine the linker-dependent effect on the multidomain proteins under crowded conditions. The bioluminescence activity of the (EAAAK)2 and (EAAAK) linker-containing multidomain protein showed a decreasing trend with the increasing molecular weights of PEG (Figures S11A, S12A). No significant difference in the reaction rates of (EAAAK)2 and (EAAAK) α-helical linker containing multidomain protein compared to (EAAAK)3 α-helical linker containing multidomain protein was observed using multiple t-tests (Table S5). The rate of enzymatic activity of the (EAAAK)2 and (EAAAK) α-helical linker-containing multidomain protein, from the rate of reaction from one-phase decay model fitted data, represented k biolum values of 0.52 ± 0.19 and 0.80 ± 0.76 CPS.mL.s–1.g–1 in the presence of PEG 400. With PEG 2000 and 4000, the multidomain protein showed k biolum values of 1 ± 0.33, 0.57 ± 0.42, and k biolum values of 1.18 ± 1.28 and 0.57 ± 0.30 CPS.mL.s–1.g–1. In the presence of higher-molecular-weight PEG (8000, 20,000, and 35,000) (EAAAK)2 α-helical linker containing a multidomain protein represented a faster enzymatic activity with k biolum values of 1.05 ± 0.62, 0.61 ± 0.17, and 0.60 ± 0.17 CPS.mL.s–1.g–1. Similarly, the (EAAAK) linker containing a multidomain protein showed k biolum of 1.86 ± 0.53, 1.10 ± 1.2, and 1.13 ± 0.72 CPS.mL.s.g–1 (Figures S11B,S12B). PEG molecular weight-dependent impact on reaction rate showed significant effect on the reaction rate of (EAAAK)2 and (EAAAK) α-helical linker containing multidomain protein in the presence of PEG 400 compared to other PEGs (Two-way ANOVA PEG 400 vs 2000, p < 0.0001; PEG 400 vs 8000, p < 0.0001; PEG 400 vs 20,000, p < 0.0001; and PEG 400 vs 35,000, p < 0.0001) and PEG (PEG 400 vs 2000, p < 0.0003; PEG 400 vs 4000, p < 0.0001; PEG 400 vs 8000, p < 0.0001; and PEG 400 vs 20,000, p < 0.0001; and PEG 400 vs 35,000, p < 0.0014). PEG 8000 also had a significant impact compared to other molecular weight PEG (Table S5).

Before determining the impact on BRET, no change in the fluorescence of both the (EAAAK)2 and (EAAAK) α-helical linker containing multidomain protein was observed (Figures S11C, S12C). Monitoring the conformational changes of both the (EAAAK)2 and (EAAAK) α-helical linker-containing multidomain protein showed a differential trend in BRET (Table S6). An increase in BRET with increasing concentrations of higher-molecular-weight PEGs (8000, 20,000, and 35,000 Da) of (EAAAK)2 α-helical linker containing multidomain protein with k BRET values of 1.20 ± 0.88, 0.86 ± 0.94, and 0.18 ± 0.19 was observed (Figure S11D). However, the (EAAAK) α-helical linker-containing multidomain protein in the presence of different molecular weight PEG (8000, 20,000, and 35,000) showed negligible change in the BRET (Figure S12D). The (EAAAK)2 α-helical linker containing multidomain protein showed a significant impact of PEG molecular weight-dependent change in BRET across different molecular weight PEGs (Two-way ANOVA p < 0.0001). The (EAAAK) α-helical linker containing multidomain proteins showed a decrease in BRET across most PEG conditions (Table S6). Overall, the results suggest a linker-dependent impact, with the differential trend implying compaction or misalignment of the domains. These observations indicate that even a modest increase in the helical linker length can significantly influence structural alignment and energy transfer.

Determining the effect of enzymatic activity and the conformational change of the (EAAAK)2 and (EAAAK) linker-containing multidomain protein simultaneously under crowded conditions showed a decrease in BRET with decreasing bioluminescence rate in the presence of smaller-molecular-weight PEGs (400, 2000, and 4000 Da) (Figure F,G). A shift in the impact of the molecular crowding of (EAAAK)2 and (EAAAK) linker-containing multidomain protein, with an increase in BRET and a decrease in the reaction rate in the presence of higher molecular weight PEGs 8000, 20,000, and 35,000, with increasing PEG concentrations, was observed. However, (EAAAK) in the presence of PEG 35000 showed a reverse effect. The p values showing the correlation between the reaction rate and conformational change are shown in the graphs (Figure F,G). These varied outcomes can be attributed to the linker length, which might prevent the multidomain proteins from achieving closer spatial proximity, altering the multidomain proteins’ orientation.

Having determined the impact of molecular crowding mediated by different molecular weight PEG, which could assume extended conformations in solution, we decided to determine the impact of molecular crowding generated by a protein, bovine serum albumin (BSA), , which has a well-defined structure. This is particularly important as the behavior of proteins under crowded conditions hinges on the interaction between the protein and the crowding agent, , and the choice of crowding agent employed greatly influences the dynamics of protein structure and the excluded volume effect. Thus, the impact of these interactions, particularly upon a protein employed as the crowding agent to induce molecular crowding, , remains a subject of ongoing debate.

To determine the effect of protein (BSA)-mediated molecular crowding on the NLuc bioluminescence activity and BRET on both the 21 AAs-long unstructured and (EAAAK)3 α-helical linker-containing synthetic multidomain proteins, we incubated the protein with a range of concentrations of BSA similar to that of PEG and monitored emissions at 467 and 533 nm. As seen under PEG-mediated molecular crowding conditions, both the 21 AAs-long unstructured and the (EAAAK)3 α-helical linker-containing multidomain protein showed a decrease in the rate of NLuc bioluminescence activity with increasing concentrations of BSA (Figure S13A,B), with k biolum values of 0.15 ± 0.06 and 0.04 ± 0.08 CPS.mL.s–1.g–1 for the 21 AAs-long unstructured and the (EAAAK)3 α-helical linker-containing multidomain protein, respectively. No significant change in the reaction rate of 21 AAs long unstructured linker and (EAAAK)3 α-helical linker-containing multidomain protein was observed (Table S7). We then compared the reaction rate of both the unstructured and α-helical linker-containing multidomain protein incubated with BSA with the reaction rates of the same multidomain proteins subjected to different molecular weights and concentrations of PEG. The analysis showed no significant change in the reaction rates of unstructured linker-containing multidomain protein, while the α-helical linker-containing multidomain protein showed significance at higher concentrations in the presence of higher-molecular-weight PEG 8000, 20000, and 35,000 (Table S8).

In order to determine and compare BRET measurements, we determined the mNG fluorescence of the synthetic multidomain protein in increasing concentrations of BSA, which revealed significant increases of 12.5 and 25% of BSA (Figure S13C). These results agree with previous reports where BSA was found to increase GFP fluorescence, and the effects were attributed to hydrophobic interactions between the GFP and BSA. , BRET measurements revealed a BSA concentration-dependent increase in BRET for both the 21 AAs-long unstructured and (EAAAK)3 α-helical linker-containing multidomain proteins (Figure S13D) (Table S9). The data were fitted by using a one-phase association model, and the resulting k BRET values reflect the rate of change in BRET with increasing BSA concentration. Notably, this effect was more pronounced in the case of the 21 AAs-long unstructured linker-containing multidomain protein (k BRET value of 0.36 ± 0.5), in comparison with the one containing the (EAAAK)3 α-helical linker (k BRET value of 0.02 ± 0.1). We note that a part of the BRET increases seen with 12.5 and 25% BSA could be contributed by the increase in the mNG fluorescence. Plotting of rate of bioluminescence (NLuc enzymatic activity) against BRET (protein conformation) observed with the two multidomain proteins under increasing concentrations of BSA (Figure S13E), we observed a somewhat “linear” behavior between the two in the case of the 21 AAs-long unstructured linker-containing multidomain protein (Figure S13F). However, a nonlinear response was observed in the case of the (EAAAK)3 α-helical linker-containing multidomain protein (Figure S13G).

Further, the conformational change was analyzed by determining the percentage change in BRET of a 21 AAs long unstructured linker and a (EAAAK)3 α-helical linker-containing multidomain protein. The analysis showed that the unstructured linker containing a multidomain protein showed a higher impact of BSA (slope of 0.66 ± 0.25) (Figure S14). Further, we compared the impact of BSA on both the 21 AAs-long unstructured and the (EAAAK)3 α-helical linker containing multidomain proteins incubated with different molecular weight PEG by determining the percentage change in BRET. Data were fitted using an exponential fit, using a one-phase association function. The (EAAAK)3 α-helical linker showed a higher impact compared to 21 AAs-long unstructured linker-containing multidomain protein in the presence of BSA, compared to PEG. The impact of molecular crowding on the conformational changes observed in the presence of BSA was comparable to that induced by higher-molecular-weight PEGs (PEG 20,000 and 35,000) (Figure S15). However, PEG 2000, 4000, and 8000 showed measurable impacts, but BSA could not reproduce or show the impact of the intermediate molecular weight PEGs; this indicates that PEG could be a more effective molecular crowder in this system. Overall, the results obtained with BSA as the crowding revealed the subtle differences in the impact of molecular crowding generated using PEG and BSA.

Discussion

In this study, we employed synthetic multidomain proteins, where mNG and NLuc are fused via either unstructured or α-helical linkers, to investigate how molecular crowding influences NLuc bioluminescence and protein conformation, as assessed by BRET. Primarily, we choose to perform the linker peptide simulation without the attached folded domains. The two folded domains may contribute to steric constraints and long-range interactions that influence the dynamics and conformational behavior of the adjacent linker region. By the inclusion of these structural elements, there is the possibility that the simulation becomes oversimplified and may not accurately reflect the physiological representation of motion or accessibility. Consequently, the current approach may not provide a full view of the system’s behavior, underscoring the need for future studies incorporating full-length models to achieve a more comprehensive understanding. Our MD simulations showed that the unstructured linker exhibited conformational heterogeneity, including the formation of β-sheets and other secondary structural elements. On the other hand, the (EAAAK)3 α-helical linker exhibited slow structural transitions characterized by partial α-helix formation as well as complete disruption of the α-helix, deviating from the commonly held assumption of its stable helical conformation. However, our findings, primarily based on PEG, an inert polymer with high polarity, low toxicity, and hydrophilic properties, demonstrate linker-dependent effects of crowding. While PEG is commonly used to simulate molecular crowding through excluded volume effects, existing studies suggest potential interactions with biomolecules, though their precise mechanisms remain unclear.

Both unstructured and α-helical linker-containing multidomain proteins exhibited decreased bioluminescence with increasing molecular weight and concentrations of PEG. The reaction rates of 21 AAs-long linker-containing multidomain protein are generally higher at lower concentrations but show a gradual decrease with an increase in PEG concentration. In contrast, the (EAAAK)3 α-helical linker-containing multidomain protein demonstrated a steeper decline in the reaction rates in the presence of higher-molecular-weight PEGs, indicating greater sensitivity to molecular crowding. This difference suggests that the 21 AAs-long unstructured linker-containing multidomain protein provides more structural flexibility, potentially allowing for better accessibility for protein interactions under crowded conditions. However, the (EAAAK)3 α-helical linker-containing multidomain protein likely imposes structural constraints, making the system more susceptible to steric hindrance and reduced diffusion caused by increasing PEG concentrations. This highlights the critical role of the linker length or type in modulating the effects of molecular crowding on biochemical reactions.

Our findings have significant implications for understanding the behavior of multidomain proteins in cellular environments. The observations suggest that proteins may adopt different conformations in vivo that differ from those typically observed under dilute, in vitro conditions. Unstructured linkers may allow for greater adaptability to crowded environments, potentially enabling proteins to maintain functionality under varying cellular conditions. Conversely, α-helical linkers might provide more stability or help maintain specific domain orientations, even under molecular crowding conditions. In our study, we observed an increase in BRET with increasing concentration and molecular weight of PEG, particularly with the higher-molecular-weight PEGs (PEG 8000 and PEG 35000), indicating the preservation of an optimal donor–acceptor distance and orientation. Conversely, lower-molecular-weight PEGs, with less steric hindrance, induced milder crowding effects, which might have a less pronounced impact on protein conformation. This could also be due to other phenomena like excluded volume and enthalpic effects involving noncovalent interactions. Furthermore, the formation of additional hydrogen bonds in a 21 AAs-long unstructured linker-containing multidomain protein suggests that crowding might affect the protein shape and influence internal stabilizing interactions. This could affect protein stability, folding, and potentially even allostery.

Reducing the length of the 21 AAs-long linker-containing multidomain protein had a distinct impact on the biochemical activity and conformation of the protein. The 17 and 12-AAs-long linker-containing multidomain proteins exhibited similar orr decreased reaction rates compared to the 21-AAs-long multidomain protein. However, they did not show much impact of the crowding on their conformational change. Only the higher molecular weight represented an increase in BRET. These results might reflect competition between the crowding-induced structural changes and the enzymatic kinetics. Higher-molecular-weight PEG and higher concentration promoted structural conformations, simultaneously hindering enzymatic activity. In contrast, the lower-molecular-weight PEGs allowed higher rates but reduced the structural arrangement. Notably, the short, linker-containing multidomain protein also exhibited varied responses to molecular crowding. (EAAAK)2 linker-containing multidomain proteins, for instance, in the presence of smaller-molecular-weight PEG molecules, were not influenced by the impact of crowding, probably due to the insufficient force bringing the two multidomain proteins into closer proximity. Further reduced amino acid linker-containing multidomain protein (EAAAK) under crowded conditions could have attained more inflexibility, preventing the protein domains from approaching each other in closer proximity. These findings suggest that the effects of molecular crowding are influenced by factors beyond linker length and flexibility, such as the size and mechanical properties of the crowding agents, demonstrating the complexity of the impact of molecular crowding on multidomain proteins with different linkers.

Conclusion

To conclude, our study establishes a distinctive approach for simultaneous monitoring of both protein conformation and enzymatic activity under molecular crowding conditions. We achieved this using synthetic multidomain proteins containing NLuc and mNG proteins and connected by either an unstructured or an α-helical linker (EAAAK)3. Computational GaMD simulations revealed conformational plasticity in both the unstructured and the α-helical linker, with the unstructured linker being largely unstructured, while the α-helical linker largely preserves its α-helical structure but with slow exchange between helical and nonhelical states, making it suitable for comparison with the unstructured linker. Simultaneous monitoring of bioluminescence and BRET measurements further revealed linker-dependent differences in the impact of PEG-mediated molecular crowding on NLuc bioluminescence activity (enzymatic activity) and BRET (protein conformation). In addition to providing insights into the impact of molecular crowding on protein activity and conformation, we believe that the experimental systems reported here will be useful in further delineating the complexities of the role played by molecular crowding on the multidomain protein function in living cells.

Experimental Procedures

Homology Modeling of the Linker Peptides

The crystal structure of the N-terminal peptide substrate complexed with SARS-CoV main protease H41A mutant (PDB:2Q6G, Chain D, AAs sequence: TSAVLQSGFRK) was used as a template for generating the structural model for the 21 AAs-long unstructured linker (AAs sequence: EFGTENLYAVLQSGFRGSGGS). The 3D model for the (EAAAK)3 α-helical linker (AAs sequence: GTEAAAKEAAAKEAAAKAAA) was generated using the crystal structure of the MBPolo-Tim21 fusion protein with a 17-residue helical linker (PDB: 6K7F, Chain A, AA sequence: RIEAAAKEAAAKEAAAKAAD) as a template. Models were generated using MODELLER (10.1 release, Mar. 18, 2021). Briefly, the unstructured and (EAAAK)3 α-helical linker sequences were aligned with the templates in the PIR format. For each peptide, 100 models were initially generated using the “Automodel” function and “very-slow” MD refining mode. Scoring functions such as molpdf, DOPE, and GA34 were used to assess the generated models. The stereochemical quality of the final model was assessed using PROCHECK. Structural modeling of the synthetic multidomain proteins with either a 21 AAs-long unstructured or a (EAAAK)3 α-helical linker was performed using the protein structure prediction software, trRosetta (https://yanglab.qd.sdu.edu.cn/trRosetta/) , and visualized using PyMOL, a molecular visualization software (Schrödinger, LLC).

MD Simulation of Linkers and Trajectory Analysis

The peptide’s termini were capped with N-acetyl and N-methyl amid the capping groups to neutralize the positive and negative charges on the N-and C-termini, respectively. Topology and parameter files were generated using the CHARMM-GUI web server. The biomolecular simulation systems included the peptide model, with all hydrogens added, solvated in a TIP3P water cubic water box with a 10 Å minimum distance between the edge of the box and any of the peptide atoms. Charges were neutralized by adding 0.15 M NaCl to the solvated system. The total number of atoms was 15,480 and 13,906 for the 21 AAs-long unstructured and (EAAAK)3 α-helical linkers, respectively. MD simulations were performed using NAMD 2.13 with the CHARMM36m force field as described previously. A 2 fs integration time step was set for all simulations performed. First, energy minimization was performed on each system in 1000 steps. Following energy minimization, the system was slowly heated from 60 to 310 at 1 K intervals to reach the 310 K equilibrium temperature by using a temperature ramp that runs 500 steps after each temperature increment. Following thermalization, the temperature was maintained at 310 K using Langevin temperature control and at 1.0 atm using Nose–Hoover Langevin piston pressure control. The system was then equilibrated for 500,000 steps (1 ns) with Periodic Boundary Conditions applied. The NAMD output structure was then used as an input for GaMD simulation utilizing the integrated GaMD module in NAMD and its default parameters, , which included 2 ns of conventional molecular dynamics (cMD) equilibration run in GaMD, to collect potential statistics required for calculating the GaMD acceleration parameters, and another 50 ns equilibration run in GaMD after adding the boost potential, and finally, GaMD production runs for 1000 ns. Both equilibration steps in GaMD were preceded by 0.4 ns preparatory runs. All GaMD simulations were run at the “dual-boost” level by setting the reference energy to the lower bound, i.e., E = V max. One boost potential is applied to the dihedral energetic term and the other to the total potential energetic term. The details for calculating the boost potential, including the equations used, have been described previously. , The upper limits of the SD of the dihedral and total potential boosts in the GaMD were set to 6.0 kcal/mol. All GaMD simulations were performed using similar and constant temperature and pressure parameters. For all simulations, short-range nonbonded interactions were defined at a 12 Å cutoff with a 10 Å switching distance, while the Particle-mesh Ewald (PME) scheme was used to handle long-range electrostatic interactions at 1 Å PME grid spacing. Trajectory frames were saved every 10,000 steps (20 ps), and trajectory analysis was performed using the available tools in Visual Molecular Dynamics (VMD) software. Trajectory movies were compiled based on 1000 frames using Video Mach (http://gromad.videomachmach/) to generate 33 s movies (30 fps) in AVI format. Analyses of the MD simulation trajectories, including root-mean-square deviation (RMSD), root-mean-square fluctuation (RMSF), radius of gyration (Rg), hydrogen bond (H-bond) formation, salt bridge formation, secondary structure analysis, and center-of-mass distance, were all performed using VMD software after aligning the trajectory frames on the backbone atoms of the original structure. The Bio3D package in R software , was used to perform DCC analysis of the trajectories. The DCC analysis was based on the position of Cα atoms, and values were plotted as a heatmap that indicated the degree of correlation from +1 (highly correlated motions) to 0 (no correlated motions) to −1 (highly anticorrelated motions).

Synthetic Multidomain Protein Expression Plasmid Generation

The previously reported BRET-based Mpro protease activity sensor containing an unstructured linker sequence (EFGTENLYAVLQSGFRGSGGS) between mNG and NLuc was utilized as the unstructured linker sequence (pmNG-Mpro-Nter-auto-NLuc) (sequences in Supporting Information text). Shorter sequences of 17 (ENLYAVLQSGFRGSGGS) and 12 (VLQSGFRGSGGS) AAs, unstructured linker-containing multidomain proteins were generated through gene fragment synthesis and subcloned into pmNG-Mpro-Nter-auto-NLuc using BstXI and XhoI sites restriction enzymes (GenScript, Singapore). For generating the plasmid for expressing the synthetic multidomain protein containing a (EAAAK)3 α-helical linker sandwiched between mNG and NLuc, the gene fragment BstXI-mNG-(EAAAK)3 linker-NLuc-XhoI (Integrated DNA Technologies, IDT; Iowa, USA) was synthesized and subcloned into pmNeonGreen-DEVD-NLuc plasmid (Addgene: 98287) using BstXI and XhoI restriction enzymes (sequences in Supporting Information text). Shorter α-helical sequences with 2 repeats (EAAAK)2 and one of the EAAAK linker coding gene fragments were synthesized and subcloned in the pmNG-(EAAAK)3-3x-NLuc using BstXI and XhoI restriction enzymes (GenScript, Singapore). One Shot TOP10 Competent Escherichia coli cells were transformed with 10 μL of the ligation reaction and plated in LB agar plates supplemented with 100 μg/mL ampicillin. Clones were confirmed by both restriction digestion and Sanger sequencing using the CMV-F primer (Macrogen).

Cell Culture and Transfection

All experiments described in the manuscript were conducted using HEK 293T cells cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum and 1% penicillin–streptomycin. The cells were maintained at 37 °C in a 5% CO2 incubator. Transfections were carried out using polyethylenimine (PEI) lipid according to the manufacturer’s protocol. Briefly, HEK 293T cells were seeded onto 10 cm Petri plates before 24 h of transfection. The plasmid DNA (biosensor), Opti-MEM (Invitrogen), and 1.25 μg/well of PEI lipid (Sigma-Aldrich) were mixed using pipetting and incubated at room temperature for 30 min before being added to cells as droplets.

Cell Lysate Preparation

The cell lysates containing the synthetic multidomain protein with unstructured linkers (21, 17, and 12 AAs-long) along with α-helical linkers (EAAAK)3, (EAAAK)2, and (EAAAK) were prepared from HEK 293T cells transfected with either pmNG-(EAAAK)3 linker-NLuc or pmNG-Mpro-Nter-auto-NLuc plasmids, respectively. After 48 h of transfection, cells were washed with chilled Dulbecco’s phosphate-buffered saline. Cells were lysed in a buffer containing 50 mM HEPES (pH 7.5), 50 mM NaCl, 0.1% Triton-X 100, 1 mM dithiothreitol (DTT), and 1 mM ethylenediamine tetra-acetic acid (EDTA) on ice using a sonicator with an amplitude of 30% and a pulse of 10 s (3 cycles). Cell lysates were collected in a 1.5 mL Eppendorf tube and centrifuged at 4 °C for 1 h at 14,000 rotations per minute (rpm), following which supernatants were collected, aliquoted, and stored at −80 °C until further usage.

Fluorescence, Bioluminescence, and BRET Measurements

Fluorescence, bioluminescence, and BRET measurements under various molecular crowding conditions were performed by incubating cell lysates containing synthetic multidomain proteins with either the 21 AAs-long unstructured or the (EAAAK)3 α-helical linker and their different short linker-containing multidomain protein with a range of concentrations (0 to 25%) of PEG of different molecular weights (400, 2000, 4000, 8000, 20,000, and 35,000 Da) (Sigma-Aldrich) in tris-buffered saline (TBS). Fluorescence measurements of mNG were performed at 37 °C and 530 nm upon excitation with 480 nm light using a Tecan SPARK multimode microplate reader. Bioluminescence measurements were performed at 37 °C by monitoring emission at 467 and 533 nm wavelengths upon the addition of a luciferase substrate at a dilution of 1:200. Emissions at 467 nm were used for reporting NLuc biochemical activity, while the ratio of emissions at 533 and 467 nm was used for reporting BRET between NLuc and mNG.

Western Blot Analysis

For Western blot analysis, cell lysates were prepared by transfecting HEK 293T cells with plasmids expressing either the unstructured (21, 17, and 12 AAs-long) or α-helical linker-containing multidomain (EAAAK)3, (EAAAK)2, (EAAAK) proteins. Samples were prepared in 50 μL of 5× Laemmli sample buffer (50 mM Tris-Cl, pH 6.8, 1.6% SDS, 8% glycerol, 4% β-mercaptoethanol, and 0.04% bromophenol blue) (heated to 95 °C before addition). Equal volumes of the cell lysates (20 μL) were separated by 12% SDS-PAGE using running buffer (25 mM Tris, 192 mM glycine, 0.1% SDS) at a constant voltage of 100 V for 1.5 h, following which proteins were transferred onto nitrocellulose membranes. Membranes were blocked in Tris-buffered saline (TBS) containing 0.1% Tween-20 (TBS-T) with skimmed milk (5%) for 1 h at room temperature. Blots were incubated either with anti-His antibody (6×-His Tag Monoclonal Antibody (HIS.H8), Alexa Fluor 488; ThermoFisher Scientific-MA1-21315-A488; 1:5000) overnight at 4 °C in dilution buffer (TBS-T containing 5% BSA. Secondary antimouse IgG HRP (Anti-Mouse Ig: HRP Donkey pAb; ECM biosciences- MS3001; 1:10,000 diluted in TBS-T) and an anti-α-tubulin antibody was used for α-tubulin (mouse IgG2b mAB, Proteintech, USA - 66031-1; 1:200000) and same secondary antibody (anti-Mouse IgG: HRP Donkey pAb; ECM Biosciences, USA - MS3001; 1:20000 dilution) was used, and the bots were developed using a chemiluminescent substrate in a ChemiDoc Imaging System (BIO-RAD, Germany) to detect the expression of both the 21 AAs-long unstructured and (EAAAK)3 α-helical linker-containing multidomain proteins.

Data Analysis and Figure Preparation

GraphPad Prism (GraphPad Software, La Jolla, CA, USA; www.graphpad.com), in combination with Microsoft Excel, was used for data analysis and graph preparation. Figures were assembled using Adobe Illustrator.

Supplementary Material

ao5c01556_si_001.pdf (2.1MB, pdf)
ao5c01556_si_002.xlsx (39.9KB, xlsx)
Download video file (8.6MB, avi)
Download video file (68.4MB, avi)

Acknowledgments

This work is supported by internal funding from the College of Health & Life Sciences (CHLS), Hamad Bin Khalifa University (HBKU), a member of the Qatar Foundation. A.M.G. is supported by a postdoctoral fellowship, and A.S. and W.S.A. are supported by scholarships from CHLS/HBKU. Some of the computational research work reported in the manuscript was performed using high-performance computer resources and services provided by the Research Computing group at Texas A&M University in Qatar. Research Computing is funded by the Qatar Foundation for Education, Science, and Community Development (http://www.qf.org.qa).

All relevant information related to this study is included in the main text and the Supporting Information files.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.5c01556.

  • MD simulations, enzymatic activity, and BRET assay results (PDF)

  • Statistical test results (XLSX)

  • GaMD simulation trajectory of the unstructured peptide (AVI)

  • GaMD simulation trajectory of the α-helical peptide (AVI)

K.H.B. conceived the experiments. A.S., A.M.G., W.S.A., and K.H.B. performed experiments, analyzed data, prepared figures, and wrote the manuscript. All authors reviewed the manuscript.

The authors declare no competing financial interest.

References

  1. Kuznetsova I. M., Turoverov K. K., Uversky V. N.. What macromolecular crowding can do to a protein. Int. J. Mol. Sci. 2014;15(12):23090–23140. doi: 10.3390/ijms151223090. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Kim J. S., Yethiraj A.. Effect of macromolecular crowding on reaction rates: a computational and theoretical study. Biophys. J. 2009;96(4):1333–1340. doi: 10.1016/j.bpj.2008.11.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Christiansen A.. et al. Effects of macromolecular crowding agents on protein folding in vitro and in silico. Biophys. Rev. 2013;5(2):137–145. doi: 10.1007/s12551-013-0108-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bhattacharya A., Kim Y. C., Mittal J.. Protein-protein interactions in a crowded environment. Biophys. Rev. 2013;5(2):99–108. doi: 10.1007/s12551-013-0111-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chebotareva N. A., Kurganov B. I., Livanova N. B.. Biochemical effects of molecular crowding. Biochemistry (Moscow) 2004;69(11):1239–1251. doi: 10.1007/s10541-005-0070-y. [DOI] [PubMed] [Google Scholar]
  6. Dong H., Qin S., Zhou H. X.. Effects of macromolecular crowding on protein conformational changes. PLoS Comput. Biol. 2010;6(7):e1000833. doi: 10.1371/journal.pcbi.1000833. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. den Hamer A.. et al. Bright Bioluminescent BRET Sensor Proteins for Measuring Intracellular Caspase Activity. ACS Sens. 2017;2(6):729–734. doi: 10.1021/acssensors.7b00239. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Vöpel T., Makhatadze G. I.. Enzyme activity in the crowded milieu. PLoS One. 2012;7(6):e39418. doi: 10.1371/journal.pone.0039418. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Das N.. et al. Macromolecular crowding effects on protein dynamics. Int. J. Biol. Macromol. 2024;281:136248. doi: 10.1016/j.ijbiomac.2024.136248. [DOI] [PubMed] [Google Scholar]
  10. Akabayov B., Akabayov S. R., Lee S. J., Wagner G., Richardson C. C.. Impact of macromolecular crowding on DNA replication. Nat. Commun. 2013;4:1615. doi: 10.1038/ncomms2620. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Cuevas-Velazquez C. L., Vellosillo T., Guadalupe K., Schmidt H. B., Yu F., Moses D., Brophy J. A. N., Cosio-Acosta D., Das A., Wang L.. et al. Intrinsically disordered protein biosensor tracks the physical-chemical effects of osmotic stress on cells. Nat. Commun. 2021;12(1):5438. doi: 10.1038/s41467-021-25736-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Siddiqui S., Cong W. N., Daimon C. M., Martin B., Maudsley S.. BRET Biosensor Analysis of Receptor Tyrosine Kinase Functionality. Front. Endocrinol. 2013;4:46. doi: 10.3389/fendo.2013.00046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Chen X., Zaro J. L., Shen W. C.. Fusion protein linkers: property, design and functionality. Adv. Drug Delivery Rev. 2013;65(10):1357–1369. doi: 10.1016/j.addr.2012.09.039. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. George R. A., Heringa J.. An analysis of protein domain linkers: their classification and role in protein folding. Protein Eng., Des. Sel. 2002;15(11):871–879. doi: 10.1093/protein/15.11.871. [DOI] [PubMed] [Google Scholar]
  15. Saha S.. et al. The linker region in receptor guanylyl cyclases is a key regulatory module: mutational analysis of guanylyl cyclase C. J. Biol. Chem. 2009;284(40):27135–27145. doi: 10.1074/jbc.M109.020032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Pittas T., Zuo W., Boersma A. J.. Engineering crowding sensitivity into protein linkers. Methods Enzymol. 2021;647:51–81. doi: 10.1016/bs.mie.2020.09.007. [DOI] [PubMed] [Google Scholar]
  17. Li M.. et al. Disordered linkers in multidomain allosteric proteins: Entropic effect to favor the open state or enhanced local concentration to favor the closed state? Protein Sci. 2018;27(9):1600–1610. doi: 10.1002/pro.3475. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Zhou H. X.. Influence of crowded cellular environments on protein folding, binding, and oligomerization: biological consequences and potentials of atomistic modeling. FEBS Lett. 2013;587(8):1053–1061. doi: 10.1016/j.febslet.2013.01.064. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Pielak G. J., Miklos A. C.. Crowding and function reunite. Proc. Natl. Acad. Sci. U.S.A. 2010;107(41):17457–17458. doi: 10.1073/pnas.1013095107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gnutt D., Ebbinghaus S.. The macromolecular crowding effect – from in vitro into the cell. Biol. Chem. 2016;397(1):37–44. doi: 10.1515/hsz-2015-0161. [DOI] [PubMed] [Google Scholar]
  21. Wu Y., Jiang T.. Developments in FRET- and BRET-Based Biosensors. Micromachines. 2022;13(10):1789. doi: 10.3390/mi13101789. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Ma B., Nussinov R.. Structured crowding and its effects on enzyme catalysis. Top. Curr. Chem. 2013;337:123–137. doi: 10.1007/128_2012_316. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Leopold H. J.. et al. Crowding Effects on Energy-Transfer Efficiencies of Hetero-FRET Probes As Measured Using Time-Resolved Fluorescence Anisotropy. J. Phys. Chem. B. 2019;123(2):379–393. doi: 10.1021/acs.jpcb.8b09829. [DOI] [PubMed] [Google Scholar]
  24. Peters J.. et al. Effects of Crowding and Cosolutes on Biomolecular Function at Extreme Environmental Conditions. Chem. Rev. 2023;123(23):13441–13488. doi: 10.1021/acs.chemrev.3c00432. [DOI] [PubMed] [Google Scholar]
  25. Rajapaksha A., Stanley C., Todd B.. Effects of Macromolecular Crowding on the Structure of a Protein Complex: A Small-Angle Scattering Study of Superoxide Dismutase. Biophys. J. 2015;108(4):967–974. doi: 10.1016/j.bpj.2014.12.046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Boersma A. J., Zuhorn I. S., Poolman B.. A sensor for quantification of macromolecular crowding in living cells. Nat. Methods. 2015;12(3):227–229. doi: 10.1038/nmeth.3257. [DOI] [PubMed] [Google Scholar]
  27. Mouton S. N., Veenhoff L. M., Boersma A. J.. Macromolecular Crowding Measurements with Genetically Encoded Probes Based on Förster Resonance Energy Transfer in Living Cells. Methods Mol. Biol. 2020;2175:169–180. doi: 10.1007/978-1-0716-0763-3_12. [DOI] [PubMed] [Google Scholar]
  28. Altamash T.. et al. Intracellular Ionic Strength Sensing Using NanoLuc. Int. J. Mol. Sci. 2021;22(2):677. doi: 10.3390/ijms22020677. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Jan Z.. et al. Protegrin-2, a potential inhibitor for targeting SARS-CoV-2 main protease M­(pro) Comput. Struct. Biotechnol. J. 2023;21:3665–3671. doi: 10.1016/j.csbj.2023.07.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Moovarkumudalvan B.. et al. Structure-Based Virtual Screening and Functional Validation of Potential Hit Molecules Targeting the SARS-CoV-2 Main Protease. Biomolecules. 2022;12(12):1754. doi: 10.3390/biom12121754. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Mo X. L., Fu H.. BRET: NanoLuc-Based Bioluminescence Resonance Energy Transfer Platform to Monitor Protein-Protein Interactions in Live Cells. Methods Mol. Biol. 2016;1439:263–271. doi: 10.1007/978-1-4939-3673-1_17. [DOI] [PubMed] [Google Scholar]
  32. Masser A. E.. et al. Luciferase NanoLuc as a reporter for gene expression and protein levels in Saccharomyces cerevisiae. Yeast. 2016;33(5):191–200. doi: 10.1002/yea.3155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Geethakumari A. M.. et al. A Genetically encoded BRET-based SARS-CoV-2 Mpro protease activity sensor. bioRxiv. 2022:2022.01.31.478460. doi: 10.1101/2022.01.31.478460. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Miao Y., Feher V. A., McCammon J. A.. Gaussian Accelerated Molecular Dynamics: Unconstrained Enhanced Sampling and Free Energy Calculation. J. Chem. Theory Comput. 2015;11(8):3584. doi: 10.1021/acs.jctc.5b00436. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Xue X.. et al. Structures of two coronavirus main proteases: implications for substrate binding and antiviral drug design. J. Virol. 2008;82(5):2515–2527. doi: 10.1128/JVI.02114-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Boersma A. J., Zuhorn I. S., Poolman B.. A sensor for quantification of macromolecular crowding in living cells. Nat. Methods. 2015;12(3):227–229. doi: 10.1038/nmeth.3257. [DOI] [PubMed] [Google Scholar]
  37. Kobayashi M.. et al. Structure of the DNA-bound BRCA1 C-terminal region from human replication factor C p140 and model of the protein-DNA complex. J. Biol. Chem. 2010;285(13):10087–10097. doi: 10.1074/jbc.M109.054106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Eswar N.. et al. Protein Structure Modeling With MODELLER. Methods Mol. Biol. 2008;426:145–159. doi: 10.1007/978-1-60327-058-8_8. [DOI] [PubMed] [Google Scholar]
  39. Laskowski R. A.. et al. PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 1993;26(2):283–291. doi: 10.1107/S0021889892009944. [DOI] [Google Scholar]
  40. Alford R. F.. et al. The Rosetta All-Atom Energy Function for Macromolecular Modeling and Design. J. Chem. Theory Comput. 2017;13(6):3031–3048. doi: 10.1021/acs.jctc.7b00125. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Du Z.. et al. The trRosetta server for fast and accurate protein structure prediction. Nat. Protoc. 2021;16(12):5634–5651. doi: 10.1038/s41596-021-00628-9. [DOI] [PubMed] [Google Scholar]
  42. Seeliger D., de Groot B. L.. Ligand docking and binding site analysis with PyMOL and Autodock/Vina. J. Comput.-Aided Mol. Des. 2010;24(5):417–422. doi: 10.1007/s10822-010-9352-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Jo S.. et al. CHARMM-GUI: a web-based graphical user interface for CHARMM. J. Comput. Chem. 2008;29(11):1859–1865. doi: 10.1002/jcc.20945. [DOI] [PubMed] [Google Scholar]
  44. Jorgensen W. L.. et al. Comparison of simple potential functions for simulating liquid water. J. Chem. Phys. 1983;79(2):926–935. doi: 10.1063/1.445869. [DOI] [Google Scholar]
  45. Phillips J. C.. et al. Scalable Molecular Dynamics With NAMD. J. Comput. Chem. 2005;26(16):1781–1802. doi: 10.1002/jcc.20289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Huang J.. et al. CHARMM36m: an improved force field for folded and intrinsically disordered proteins. Nat. Methods. 2017;14(1):71–73. doi: 10.1038/nmeth.4067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Geethakumari A. M., Ahmed W. S., Rasool S., Fatima A., Nasir Uddin S. M., Aouida M., Biswas K. H.. A Genetically encoded BRET-based SARS-CoV-2 Mpro protease activity sensor. Commun. Chem. 2022;5(1):117. doi: 10.1038/s42004-022-00731-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Feller S. E.. et al. Constant pressure molecular dynamics simulation: the Langevin piston method. J. Chem. Phys. 1995;103(11):4613–4621. doi: 10.1063/1.470648. [DOI] [Google Scholar]
  49. Pang Y. T.. et al. Gaussian Accelerated Molecular Dynamics in NAMD. J. Chem. Theory Comput. 2017;13(1):9. doi: 10.1021/acs.jctc.6b00931. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Miao Y.. et al. Accelerated Molecular Dynamics Simulations of Protein Folding. J. Comput. Chem. 2015;36(20):1536. doi: 10.1002/jcc.23964. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Miao Y., McCammon J. A.. Gaussian Accelerated Molecular Dynamics: Theory, Implementation, and Applications. Annu. Rep. Comput. Chem. 2017;13:231. doi: 10.1016/bs.arcc.2017.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Humphrey W., Dalke A., Schulten K.. VMD: visual molecular dynamics. J. Mol. Graphics. 1996;14(1):33. doi: 10.1016/0263-7855(96)00018-5. [DOI] [PubMed] [Google Scholar]
  53. Skjærven L., Yao X. Q., Scarabelli G., Grant B. J.. Integrating protein structural dynamics and evolutionary analysis with Bio3D. BMC Bioinf. 2014;15(1):399. doi: 10.1186/s12859-014-0399-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Grant B. J.. et al. Bio3d: an R package for the comparative analysis of protein structures. Bioinf. 2006;22(21):2695–2696. doi: 10.1093/bioinformatics/btl461. [DOI] [PubMed] [Google Scholar]
  55. Jiang L.. et al. Whole-genome sequence data and analysis of Saccharibacillus sp. ATSA2 isolated from Kimchi cabbage seeds. Data Brief. 2019;26:104465. doi: 10.1016/j.dib.2019.104465. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Longo P. A.. et al. Transient mammalian cell transfection with polyethylenimine (PEI) Methods Enzymol. 2013;529:227–240. doi: 10.1016/B978-0-12-418687-3.00018-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Itoh M., Bissell M. J.. The organization of tight junctions in epithelia: implications for mammary gland biology and breast tumorigenesis. J. Mammary Gland Biol. Neoplasia. 2003;8(4):449–462. doi: 10.1023/B:JOMG.0000017431.45314.07. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Rastogi H., Chowdhury P. K.. Understanding enzyme behavior in a crowded scenario through modulation in activity, conformation and dynamics. Biochim. Biophys. Acta. 2021;1869(11):140699. doi: 10.1016/j.bbapap.2021.140699. [DOI] [PubMed] [Google Scholar]
  59. Liebau J.. et al. Polyethylene Glycol Impacts Conformation and Dynamics of Escherichia coli Prolyl-tRNA Synthetase Via Crowding and Confinement Effects. Biochemistry. 2024;63(13):1621–1635. doi: 10.1021/acs.biochem.3c00719. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Nolan V., Sánchez J. M., Perillo M. A.. PEG-induced molecular crowding leads to a relaxed conformation, higher thermal stability and lower catalytic efficiency of Escherichia coli β-galactosidase. Colloids Surf., B. 2015;136:1202–1206. doi: 10.1016/j.colsurfb.2015.11.003. [DOI] [PubMed] [Google Scholar]
  61. England C. G., Ehlerding E. B., Cai W.. NanoLuc: A Small Luciferase Is Brightening Up the Field of Bioluminescence. Bioconjugate Chem. 2016;27(5):1175–1187. doi: 10.1021/acs.bioconjchem.6b00112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Nemergut M., Pluskal D., Horackova J., Sustrova T., Tulis J., Barta T., Baatallah R., Gagnot G., Novakova V., Majerova M.. et al. Illuminating the mechanism and allosteric behavior of NanoLuc luciferase. Nat. Commun. 2023;14(1):7864. doi: 10.1038/s41467-023-43403-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Kirchhofer S. B., Kurz C., Geier L., Krett A. L., Krasel C., Bünemann M.. Dynamics of agonist-evoked opioid receptor activation revealed by FRET- and BRET-based opioid receptor conformation sensors. Commun. Biol. 2025;8(1):198. doi: 10.1038/s42003-025-07630-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Saca V., Burdette C., Sakmar T.. GPCR Biosensors to Study Conformational Dynamics and Signaling in Drug Discovery. Annu. Rev. Pharmacol. Toxicol. 2025;65:7–28. doi: 10.1146/annurev-pharmtox-061724-080836. [DOI] [PubMed] [Google Scholar]
  65. Geethakumari A. M.. et al. A BRET-based Mpro biosensor containing a nanobody and tandem cleavage sites shows an increased cleavage rate. Sens. Acutators Rep. 2025;9:100315. doi: 10.1016/j.snr.2025.100315. [DOI] [Google Scholar]
  66. Liu B.. et al. Design and Properties of Genetically Encoded Probes for Sensing Macromolecular Crowding. Biophys. J. 2017;112(9):1929–1939. doi: 10.1016/j.bpj.2017.04.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Ostrowska N., Feig M., Trylska J.. Modeling Crowded Environment in Molecular Simulations. Front. Mol. Biosci. 2019;6:86. doi: 10.3389/fmolb.2019.00086. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Webb B., Sali A.. Comparative Protein Structure Modeling Using MODELLER. Curr Protoc Bioinformatics. 2016;54:1–5. doi: 10.1002/cpbi.3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Phillips J. C., Hardy D. J., Maia J. D. C., Stone J. E., Ribeiro J. V., Bernardi R. C., Buch R., Fiorin G., Hénin J., Jiang W.. et al. Scalable molecular dynamics on CPU and GPU architectures with NAMD. J. Chem. Phys. 2020;153(4):044130. doi: 10.1063/5.0014475. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Phillips J. C.. et al. Scalable Molecular Dynamics with NAMD. J. Comput. Chem. 2005;26(16):1781. doi: 10.1002/jcc.20289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Chong W. L.. et al. Insight into free energy and dynamic cross-correlations of residue for binding affinity of antibody and receptor binding domain SARS-CoV-2. Heliyon. 2023;9(1):e12667. doi: 10.1016/j.heliyon.2022.e12667. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Guseman A. J.. et al. Protein shape modulates crowding effects. Proc. Natl. Acad. Sci. U.S.A. 2018;115(43):10965–10970. doi: 10.1073/pnas.1810054115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Rawat S., Raman Suri C., Sahoo D. K.. Molecular mechanism of polyethylene glycol mediated stabilization of protein. Biochem. Biophys. Res. Commun. 2010;392(4):561–566. doi: 10.1016/j.bbrc.2010.01.067. [DOI] [PubMed] [Google Scholar]
  74. Tyrrell J., Weeks K. M., Pielak G. J.. Challenge of Mimicking the Influences of the Cellular Environment on RNA Structure by PEG-Induced Macromolecular Crowding. Biochemistry. 2015;54(42):6447–6453. doi: 10.1021/acs.biochem.5b00767. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Parray Z. A.. et al. Amphiphilic nature of polyethylene glycols and their role in medical research. Polym. Test. 2020;82:106316. doi: 10.1016/j.polymertesting.2019.106316. [DOI] [Google Scholar]
  76. Karamitros C. S., Murray K., Winemiller B., Lamb C., Stone E. M., D’Arcy S., Johnson K. A., Georgiou G.. Leveraging intrinsic flexibility to engineer enhanced enzyme catalytic activity. Proc. Natl. Acad. Sci. U. S. A. 2022;119(23):e2118979119. doi: 10.1073/pnas.2118979119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Basak S.. et al. Probing Interdomain Linkers and Protein Supertertiary Structure In Vitro and in Live Cells with Fluorescent Protein Resonance Energy Transfer. J. Mol. Biol. 2021;433(5):166793. doi: 10.1016/j.jmb.2020.166793. [DOI] [PMC free article] [PubMed] [Google Scholar]
  78. Ota C., Takano K.. Behavior of Bovine Serum Albumin Molecules in Molecular Crowding Environments Investigated by Raman Spectroscopy. Langmuir. 2016;32(29):7372–7382. doi: 10.1021/acs.langmuir.6b01228. [DOI] [PubMed] [Google Scholar]
  79. Ota C., Takano K.. Behavior of Bovine Serum Albumin Molecules in Molecular Crowding Environments Investigated by Raman Spectroscopy. Langmuir. 2016;32(29):7372–7382. doi: 10.1021/acs.langmuir.6b01228. [DOI] [PubMed] [Google Scholar]
  80. Sasahara K., McPhie P., Minton A. P.. Effect of Dextran on Protein Stability and Conformation Attributed to Macromolecular Crowding. J. Mol. Biol. 2003;326(4):1227–1237. doi: 10.1016/S0022-2836(02)01443-2. [DOI] [PubMed] [Google Scholar]
  81. Dong X.. et al. Preferential Interactions of a Crowder Protein with the Specific Binding Site of a Native Protein Complex. J. Phys. Chem. Lett. 2022;13(3):792–800. doi: 10.1021/acs.jpclett.1c03794. [DOI] [PMC free article] [PubMed] [Google Scholar]
  82. Gong Z.. et al. Preferential Regulation of Transient Protein–Protein Interaction by the Macromolecular Crowders. J. Phys. Chem. B. 2022;126(26):4840–4848. doi: 10.1021/acs.jpcb.2c02713. [DOI] [PMC free article] [PubMed] [Google Scholar]
  83. Majumdar B. B., Mondal J.. Impact of Inert Crowders on Host–Guest Recognition Process. J. Phys. Chem. B. 2022;126(23):4200–4215. doi: 10.1021/acs.jpcb.2c01539. [DOI] [PubMed] [Google Scholar]
  84. Gu X., Ji F., Jiang D., Liang W., Feng Y., Niu S., Zhao Q.. The development and application of glyco-probes for bovine serum albumin. J. Chem. Res. 2023;47(3):17475198231177954. doi: 10.1177/17475198231177954. [DOI] [Google Scholar]
  85. Morikawa T. J., Fujita H., Kitamura A., Horio T., Yamamoto J., Kinjo M., Sasaki A., Machiyama H., Yoshizawa K., Ichimura T.. et al. Dependence of fluorescent protein brightness on protein concentration in solution and enhancement of it. Sci. Rep. 2016;6(1):22342. doi: 10.1038/srep22342. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ao5c01556_si_001.pdf (2.1MB, pdf)
ao5c01556_si_002.xlsx (39.9KB, xlsx)
Download video file (8.6MB, avi)
Download video file (68.4MB, avi)

Data Availability Statement

All relevant information related to this study is included in the main text and the Supporting Information files.

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES