Accurate prediction by AlphaFold2 for ligand binding in a reductive dehalogenase: Implications for PFAS (per- and polyfluoroalkyl substance) biodegradation

Accurate prediction by AlphaFold2 for ligand binding in a reductive dehalogenase: Implications for PFAS (per- and polyfluoroalkyl substance) biodegradation | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Accurate prediction by AlphaFold2 for ligand binding in a reductive dehalogenase: Implications for PFAS (per- and polyfluoroalkyl substance) biodegradation Hao-Bo Guo, Vanessa Varaljay, Gary Kedziora, Kimberly Taylor, and 9 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-2057833/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 11 Mar, 2023 Read the published version in Scientific Reports → Version 1 posted 10 You are reading this latest preprint version Abstract Despite the success of AlphaFold2 (AF2), it is unclear how AF2 models accommodate for ligand binding. Here, we start with a protein sequence from Acidimicrobiaceae TMED77 (T7RdhA) with potential for catalyzing the degradation of per- and polyfluoroalkyl substances (PFASs). AF2 models and experiments identified T7RdhA as a corrinoid iron-sulfur protein (CoFeSP) which uses a norpseudo-cobalamin (BVQ) cofactor and two [4Fe4S] iron-sulfur clusters (SF4) for catalysis. Docking and molecular dynamics simulations suggest that T7RdhA uses perfluorooctanoic acetate (PFOA) as a substrate, supporting the reported defluorination activity of its homolog, A6RdhA. We showed that AF2 provides processual (dynamic) predictions for the binding pockets of ligands (cofactors and/or substrates). Because the pLDDT scores provided by AF2 reflect the protein native states in complex with ligands as the evolutionary constraints, the Evoformer network of AF2 predicts protein structures and residue flexibility in complex with the ligands, i.e., in their native states. Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction AlphaFold2 1 (AF2) has achieved near-experimental accuracy for predicting protein structures from the primary sequences. This breakthrough, together with the developments of other tools including RoseTTAFold 2 , allow us to understand the protein structure-function relationships with atomic precision. The performance of AF2, however, was found to produce contradictory results in some assessments. For instance, it is the subject of debate whether AF2 fails to predict the impact of point mutations in protein structure 3 and stability 4 ; whereas other studies indicated that the structures and phenotypic effects of the point mutations can be correctly predicted 5 or assisted 6 by AF2. With these controversies in mind, further modifications to the AF2-predicted structures are required to appropriately understand the protein functions 7 , including the addition of ligands (cofactors and/or substrates). Moreover, as proteins are not static and generally perform functions in the cell but not in crystals, it is important to examine if AF2 can capture protein dynamics in aqueous environments 8 . Over 60 years before the first glimpse of a protein structure 9 , the “key-lock” model 10 was proposed to describe how proteins perform functions via ligand binding. In this model the protein is described as a lock awaiting the ligand as a key to unleash its function. Unfitted ligands ࣧincluding water moleculesࣧwould fail to unlock the protein. In later studies, especially with the help of protein and protein-ligand complex structures that were becoming increasingly abundant, a refined “induced-fit” model 11 , 12 was proposed which accommodates the conformational changes of the protein upon ligand-binding. To illustrate the long-range (allosteric) effect of ligand binding, the conformation selection model was proposed 13 , 14 , in which the binding conformations pre-exist in the protein such that the ligand binding could spontaneously occur. Both sequences and structures (shapes, sizes, and locations) of the binding pockets in proteins are thought to evolve to facilitate association with different ligands 15 . Although only a small number of representative pockets in proteins have been estimated, ligand specificity of proteins may emerge in evolution without functional constraints 16 . In this regard most of the previously or currently recognized hard-to-degrade chemicals, including polymers 17 and per- and polyfluoroalkyl substances (PFAS) 18 , could serve as the “keys” for certain lock proteins. With AF2 it is likely that every single sequenced protein has its high-resolution 3D structure available in the database 19 , 20 or can be accurately predicted. In the Protein Data Bank 21 , 22 , however, the majority of the structures derived from experiments (crystallography, cryoEM, NMR, etc.) are complexes, including multimers (often with symmetry) 23 , as well as bound with cofactors 24 and other ligands 25 . This information is as important as the protein structures themselves for informing the protein functions and mechanisms, ever since the first solved protein structure of mioglobin 9 . However, the initial structures that we obtain from AF2 1 , AlphaFold-multimer 26 , or AF2Complex 27 are apo-proteins, i.e., proteins without ligands. Given the important roles of ligands play in the protein functions, it is crucial to determine whether AF2 is suitable for predicting structure and function of such proteins. In another word, do the apo-proteins predicted by AF2 have the proper binding pockets (cavities) for the ligand binding? Here, we used a multiple-ligand protein to answer this question. As the PFAS substrate (e.g., PFOA) can bind to the protein as a substrate, it has the potential for biodegrading the PFAS contaminants. A recent work found a bacterium Acidimicrobiaceae sp. A6 that, when cultured in the presence of either perfluorooctanoic acid (PFOA) or perfluorooctane sulfonate (PFOS), was able to defluorinate these chemicals with an observed release of fluoride ion, shorter-chain perfluorinated products, and acetate 28 . The key enzyme for defluorination of PFOA/PFOS was identified as a reductive dehalogenase subunit A (RdhA) in GenBank (id: MK358462.1) 28 . However, only partial sequence of this enzyme (A6RdhA hereafter) was available with a missing C-terminus of over 100 AA’s compared with known reductive dehalogenases including PceA 29 and NpRdhA 30 . Sequence mining starting from the partial A6RdhA sequence revealed a full protein sequence from the bacteria TMED77 in a metagenomic assembly of the Mediterranean Sea microorganisms 31 , which shares 98% sequence identity with the known part of A6RdhA protein. This protein is referred to as the T7RdhA in present work. It is worth noting that the TMED77 bacterium belongs to the same Acidimicrobiaceae family as A. sp. A6 . In the present work, we showed that T7RdhA is a PceA-like protein 29 which utilizes two [4Fe4S] iron sulfur (SF4) clusters and a norpseudo-cobalamin (BVQ) cofactor. We constructed AF2 models of T7RdhA, and for the highest-ranked model, both BVQ cofactor and SF4 clusters can be put on the binding pockets precisely. Molecular dynamics (MD) simulations were performed on this model with no ligand (apo-form), partially bound by cofactors (either BVQ or SF4), or with both cofactors, and with both cofactors and a substrate (PFOA). The results indicate that the AF2 is able to predict the binding pockets for both cofactors and substrates in the protein models, with regard to the binding pockets dynamics 32 . The model used in the MD simulation was constructed using AF2 V2.0.1 (July 2021 version). A newer version of AF2 V2.2.2 (downloaded in July 2022) was employed to construct additional 90 models and compared with the MD model. High similarity of the new models with the MD model illustrates the reproducibility of AF2. Interestingly, we show that the diversity of AF2 models resemble the MD results. We perform residue-interaction network (RIN) analyses using the MD trajectories of the model with both BVQ, SF4, and PFOA. We identified the binding pockets for both cofactors and the substrate in T7RdhA, which will help to search and design proteins for PFAS biosequestration and degradation. Results T7RdhA (and potentially A6RdhA) is a CoFeSP From a sequence similarity network (SSN) constructed using the NCBI nr database 33 , T7RdhA and the T7RdhA-like proteins comprise highly conserved residues for the binding of a corrinoid cofactor and two [4Fe4S] iron-sulfur (SF4) clusters (see Fig. S1 in the supplementary information, SI). These proteins were termed as corrinoid iron-sulfur proteins (CoFeSPs) 34 . The corrinoid cofactor or these proteins include the cobalamin (B12) in NpRdhA 30 and B12-derivatives such as the norpseudo cobalamin (BVQ) in PceA 29 . NpRdhA uses the B12 cofactor and belongs to an aerobic bacterium Nitratireductor pacificus 30 , 35 . However, PceA that uses the BVQ cofactor is carried by the anaerobic bacterium Sulfurospirillum multivorans 29 , 36 . It is likely T7RdhA uses the BVQ cofactor not only the Acidimicrobiaceae bacterial family is anaerobic 28 , but also because T7RdhA belongs to the PceA branch (Fig. S1) in the clustering of the T7RdhA-like proteins from the SSN. A cross-linked binding mode has been found in both PceA- and NpRdhA-like proteins, in which two SF4-binding motifs are required for binding of each of the two SF4 clusters (Fig. S2). Moreover, we cloned and expressed T7RdhA in Escherichia coli , and verified that T7RdhA binds both a corrinoid cofactor and two iron-sulfur clusters (Fig. S3). The network-assisted de novo structured prediction approach and experimental verifications indicate that T7RdhA is a CoFeSP. Besides the cofactors BVQ, SF4-1 and SF4-2, the PFOA substrate is also docked into T7RdhA (Methods), and Fig. 1 shows the binding of all four ligands in T7RdhA. Alphafold2 Predicts Residue Flexibility Of T7rdha Upon Ligand Binding We showed previously that the per-residue pLDDT (predicted local distance difference test) scores accompanying the predicted protein models by AF2 also anticipate the residue flexibilities for globular proteins, protein dimers and intrinsically disordered proteins 8 . However, AF2 only provides the apo-forms of the proteins or protein-multimers, and the knowledge of cofactors and/or substrates related to the protein functions can only be acquired from experiments or literature. In the case of T7RdhA, since it is likely a CoFeSP which performs the functions utilizing the corrinoid (BVQ) cofactor and two SF4 clusters, and presumably the PFOA substrate can be bound to the active site of the protein for catalysis, we asked if AF2 can predict the binding of these proposed ligands. To answer this question, we performed MD simulations five different systems: 1) T7RdhA complexed with BVQ, two SF4 clusters (SF4-1 and SF4-2) and the PFOA substrate; 2) apo-T7RdhA with no ligand; 3) T7RdhA complexed with BVQ; 4) T7RdhA complexed with two SF4 clusters; and 5) T7RdhA complexed with the BVQ cofactor and two SF4 clusters. T7RdhA is a well-folded globular protein. The binding of cofactors or ligands does not significantly alter the conformation of T7RdhA. However, the residue flexibility vary significantly for all the five systems, as shown in Fig. 2 . It has been suggested that the diversity of AF2 models would yield biological insights that might be otherwise ignored from a single snapshot of the protein structure 1 . In addition to these five systems, we constructed 90 more AF2 structures and calculated the residue fluctuations among these structures for comparison (line 6, see below). For the well-folded globular proteins, the residue flexibility profiles measured by the root-mean square fluctuation (RMSF) from MD simulations were found to be highly consistent with the AF2-scores from the AF2 predictions, which is a reverse normalization of the pLDDT scores 8 AF2 i = (pLDDT max -pLDDT i )/(pLDDT max -pLDDT min ), (1) where AF2 i is the AF2-score of the i-th residue calculated from the pLDDT-score of the Cα atom of the i-th residue 8 . For systems 1–5, 300 ns MD simulations was performed, and the RMSF profile was calculated using the last 100 ns trajectory. For all five models positive correlation between the AF2-scores and RMSFs were observed (Fig. 2 ), and the best fit comes from the T7RdhA model in complex with BVQ, two SF4 clusters and a PFOA substrate (system 1)ࣧexcept for an inconsistency at the C-terminus region for which AF2 anticipates it highly flexible but the RMSF from MD indicates it is instead relatively rigid. Nevertheless, for the important binding regions such as three β-sheets and the helices H12 to H16 (Fig. 2 ), RMSF of system 1 is highly consistent with the AF2-scores. In contrast, systems 2 and 3 indicate that without the SF4 clusters, the regions containing the binding Cys residues (H14 to H16, including the loop region between H15 and H16) are highly flexible, contradicting with the AF2 predictions. In system 4, binding of the SF4 lead to relatively small flexibility of the binding Cys residues, however, the loop region between H15 and H16 still shows significantly higher flexibility compared with the AF2 prediction. System 5 also shows better consistency between MD and AF2 prediction, slightly lower than system 1. We also measured the root-mean square deviation (RMSD) of all five systems. The mean RMSD of the last 10 ns are calculated as shown in Table 1 , and again, system 1 shows the lowest RMSD value. The RMSD profiles from the last 100 ns of all five systems are shown in Fig. S4 in the SI. Table 1 Comparing protein dynamics metrics with the AF2 predictions System Description RMSF 1 (Å) PCC 2 p-value 2 RMSD 3 (Å) 1 T7RdhA + BVQ + SF4-1/SF4-2 + PFOA 0.8 ± 0.6 0.561 5.2E-35 1.4 2 Apo-T7RdhA 0.9 ± 0.4 0.357 1.2E-13 1.7 3 T7RdhA + BVQ 1.0 ± 0.6 0.342 1.4E-12 2.4 4 T7RdhA + SF4-1/SF4-2 1.1 ± 0.7 0.355 1.6E-13 2.5 5 T7RdhA + BVQ + SF4-1/SF4-2 0.9 ± 0.5 0.426 2.7E-19 1.5 6 AF2 models 4 0.5 ± 0.2 0.873 2.7E-128 0.5 1. The root-mean-square fluctuation (RMSF) from a 100 ns MD simulation after 200 ns equilibration. 2. Pearson’s correlation coefficient (PCC) and p-values between the RMSF and the predicted AF2-score. 3. Mean RMSD from the last 10 ns trajectory referenced with the initial structure of the 100 ns MD. 4. 90 AF2 models are used to calculate the RMSF of all residues. The RMSD is averaged from all 90 models referenced to the model used in the MD simulations. Diversity Of Af2 Models Resembles The Md Simulation The original AF2 publication suggested that diversity of AF2 models (i.e., via multiple runs) may yield new biological insights by predicting alternative configurations of the proteins 1 . In this work, the T7RdhA model for the MD simulations was constructed using an old AF2 version (V2.0.1, released July 2021). We used the newer AF2 version (V2.2.2, downloaded July 2022) to construct 90 T7RdhA models (18 independent runs, each gives 5 models). These models show low RMSD to the MD model (Fig. S5 in the SI), indicating the reproducibility of the AF2 algorithm. By combining the configuration of all 90 new T7RdhA models, we also calculated the residue RMSF values and compared with the AF2 scores from the MD model (Fig. 2 ). Interestingly, this RMSF profile (black dashed line) shows a PCC = 0.873 (p = 2.7E-128) to the AF2-score profile. In this profile not only the BVQ- or SF4-binding regions, but also the dynamic N-terminus is consistent with the AF2-scores. Therefore, from the protein sequence, AF2 not only provided dynamics information of all residues via the (pLDDT or AF2-scores) 8 , it seems that multiple AF2 runs can reproduce the MD simulation, i.e., the diversity in structures from multiple AF2 runs can retrieve the residue fluctuations. Residue Interaction Network And The Cofactor/ligand Binding Modes Of T7rdha The protein residue distance maps usually defined as the distance d ij between the C β atoms (C α for Gly) of residues i and j , and there is a contact between these two residues if d ij is shorter than a criterion (e.g., 8 Å) 38 . A residue interaction network (RIN) 39 can be constructed based on the contact map in which all residues are regarded as vertices and the contacts as edges. Here, we used a modified approach to define d ij as the shortest distance between non-hydrogen atoms of two residues are measured, and a cutoff of 3.5 Å is used to identify contacts. This approach would avoid potential false contact assignments, see Methods and Fig. S6 in the SI. The distance map of system 1 obtained from the modified approach is shown in Fig. 3 a, which has very similar patterns as the predicted aligned error (PAE) map provided by AF2, as shown in Fig. 3 b. The RIN for system 1, in which the cofactors (BVQ, SF4-1 and SF4-2) and the substrate (PFOA) were treated as individual vertices, constructed from the final snapshot of the 300 ns MD is shown in Fig. 3 c. To capture the dynamics of the RIN, we constructed the RIN every 1 ns from the last 100 ns MD and monitored the residues that interact with the BVQ cofactor, both SF4 clusters, as well as the PFOA substrate. The distributions of these residues are shown in Fig. 4 . We observed more residues interact with the BVQ than those interact with the two SF4 clusters and PFOA combined. The motifs involved in BVQ binding include short helical segments H2, H3, H4, and longer helices H12, H13, H16, and H17; two strands from β-sheet A and the β-hairpin C are involved in BVQ binding (see Fig. 1 c for all motif names). The β-hairpin C also interacts with SF4-1 at the loop region via two positively charged residues H259 and K258, as Fe 4 S 4 Cys 4 carries negative charge (-1 for oxidized and − 2 for reduced states, respectively). In this simulation we used a reduced SF4-1, an oxidized SF4-2 and a reduced BVQ (with Co − 1 ), see methods. H16 is involved in the interactions for BVQ, SF4-2 and PFOA. In particular, the aromatic residue W343 that is conserved in other T7RdhA-like proteins (Fig. S2) shows interactions with BVQ (82%), PFOA (80%) and SF4-2 (15%). Y213 from H12 has been considered to mediate the reductive dehalogenation in PceA (Y246 of S. multivorans PceA) 29 , and its interactions with BVQ (98%) and PFOA (40%) may be needed for potential defluorination. F47 from a small helix H4 also interacts with BVQ (100%) and PFOA (90%). We noticed that the residues interact with PFOA (> 50%) are either aromatic (Y68, Y65, F47, W343, F64, W93, F340) or positively charged (R89), which may be a unique feature of the binding pocket for PFAS substrates. Discussion AlphaFold2 correctly predicts cofactor/ligand binding in T7RdhA AF2 opens an avenue in biology on which the functions and interactions mediated by proteins can be understood with the assistance of highly accurate atomic models. However, the structures predicted by AF2, either single-chain monomers or multi-chain oligomers, are in apo-forms, i.e., unbounded form. Even the necessary solvents are missing in the structures predicted by AF2. Cofactors play an important, sometimes essential role in protein folding and functions 40 . Folding and functions of proteins may also be assisted by the substrate that they bind 41 . We asked how reliable are the AF2 models in depicting the structures and dynamics of proteins upon cofactor and/or ligand binding? This is a critical question to answer for protein systems with cofactor/substrate, especially for understanding the interactions among them, as well as for protein-protein interactions. Previous publications discussed the above question on ligand binding 42 , peptide binding 43 , and protein-protein interactions 44 . In this work, the functional T7RdhA structure incorporates the natural corrinoid norpseudo cobalamin (BVQ) 45 , together with two [4Fe4S] iron-sulfur clusters (SF4-1 and SF4-2), which is known as the “Nature’s modular structures” 46 . We showed that when cofactors (BVQ/SF4) and substrate (PFOA) are present in the correct pockets, the residue flexibility calculated from molecular dynamics simulations can best describe the AF2-scores by AlphaFold2, which is an inverse normalization of the pLDDT scores. In the complex model, the residue distance map also mirrors the predicted aligned error map by AlphaFold2. Our results indicate that the AF2 structures already have the pre-built pockets for the correct cofactors and ligands. We also showed that multiple AF2 structures (90 T7RdhA models in the present work) can also capture the protein dynamics. The diversity of protein structures, in our opinion, originates from protein dynamics and can be recaptured by AF2 in the structure modeling. A processual view of protein structure-function relationships The protein function is determined by the protein structure. However, a static protein does not perform the function without dynamics and interactions. The processual nature of reality 47 applies to all biomolecules, including proteins. We collected the structures of different systems (System1-5 in Fig.2) during the MD simulation, together with selected AF2 models (System 6), and compared these snapshots in Fig. 5. The structures of all models after 300 ns MD are similar to the original AF2 model (Fig.5b). Without MD trajectories and the residue flexibilities, it would be difficult to tell which system has a ligand or ligands. Our results support that the protein interactions and functions are based on their intrinsic processual nature 47 . For other AF2 models such as those of missense mutations 3 , it might not be fair to make a judgement based on a static configuration. Implications for PFAS biodegradation The persistence and accumulation of per- and polyfluoroalkyl substances, or PFASs, in the environment, and their adverse effects on human health have led to the current global concern 48 . The T7RdhA sequence is highly similar to the partial sequence of A6RdhA from the Acidimicrobiaceae sp. A6 which degrade both PFOA and PFOS under anaerobic conditions 28,49 . Nevertheless, the full A6RdhA sequence and the defluorination mechanisms remain unclear. From the structure modeling and MD simulations, we confirmed the participation of both corrinoid cofactors (BVQ) and iron-sulfur clusters (SF4) by experiment. The binding mode of the cofactors and the PFOA ligand have been identified using a dynamic residue interaction network from the MD trajectories. We also showed that AF2 combined with MD simulation can help to identify proteins with targeted functions such as PFAS bioremediation. Methods And Materials Multi-sequence alignment & sequence similarity network A combined A6RdhA/T7RdhA Hidden Markov Model (HMM) was constructed from 529 non-redundant similar sequences identified via blastp from the NCBI and UniProt KB databases. Briefly, these sequences were identified to clade together (with a consensus support value of 100) with A6RdhA/T7RdhA in an amino acid tree using MAFFT v7.453) 50 multi-sequence alignment, and were then used to construct an HMM profile using the program HMMer (v3.3.2) 51 . The first portion of the NCBI non-redundant database, nr00 (8,812,511 sequences), was queried using this HMM profile using the HMMer default threshold values. The resulting 1279 (including T7RdhA) sequences were submitted to the EFI (Enzyme Function Initiative) enzyme similarity tool for generation of the sequence similarity network (SSN) with evalues ≤ 10 -5 and an alignment cutoff of 20 corresponding to an id% of ~30. 52 Network clustering and the T7RdhA clique identification was performed using the igraph package in R. 53 The multi-sequence alignment by MAFFT was visualized using WebLogo (v3.6.0) 54 . TM-align 55 was used for structure alignment and RMSD calculations. The calculated RMSD matrix was converted to phylogeny using the ape package in R 56 , and visualized by Mega-X 57 . AlphaFold 2 structure predictions The T7RdhA model used in the MD simulation and other T7RdhA-like proteins models (all 39 models in the SSN shown in Fig. S1) were constructed using AlphaFold2 V2.0.1 (installed in July 2021). 90 more T7RdhA models for the protein-structure-based RMSF profile in Fig. 2 (system 6), were built by a newer version of AlphaFold2 (V2.2.2, installed in July 2022) 1 . Molecular dynamics simulations The molecular dynamics simulations were performed using NAMD. 58 The CHARMM force field (c36m) 59,60 was employed for the protein and a modified TIP3P model 61 for the solvent water molecules. The CHARMM-format force field parameters of norpseudo-B12 (BVQ) 62 and [4Fe4S] iron sulfur cluster (SF4) 63 under different redox states have been adopted. The force field parameters of the PFOA molecule were derived from the TEAM (Transferable, Extensible, Accurate and Modular) force field in the Direct Force Field (DFF, v7.2) 64 software, and have been listed in the Appendix of the supplementary materials. The BVQ and SF4 cofactors in the crystallographic structures of PceA (e.g., 4UQU 29 ) can be superimposed very well, with the eight Cys residues precisely bound the SF4 iron atoms. All hydrogen atoms have been added using the HBuild function of CHARMM 65 . The covalent bond between SF4 cofactors and their binding Cys residues were generated using the Patch function of CHARMM 65 . The whole system was put in a solvent box with H 2 O molecules added at least 15 Å to the edge of the protein system. The solvation and neutralization (using Na+ and/or Cl-) were carried out by the Solvate and Autoionization packages of VMD 66 . A reduced BVQ (Co(I)), oxidized SF4-2 (the proximal SF4, Fe 4 S 4 (Cys) 4 - ) and reduced SF4-1 (Fe 4 S 4 (Cys) 4 2- ) were used in the MD simulations. After solvation and neutralization, the whole system was optimized by 50,000 steps under 0 K. Then the temperature of the system was “naturally” increased to 300 K with a rate of 0.001 K/timestep. A constant-pressure, constant-temperature (NPT) ensemble was used in the MD simulation with the system pressure of 1 atm and temperature of 300 K maintained by the Langevin piston controls. The SHAKE algorithm was applied to fix the bond lengths involving hydrogen atoms and a timestep of 2 fs was used for the simulations. The nonbonded interaction cutoff switching was set as between 9 and 11 Å. For the long-range interactions, the particle mesh Ewald summation with a grid spacing of 1.35 was applied. 310 ns MD simulations were performed for all systems (Fig. 2) and the last 100 ns were taken for further analysis. Residue interaction network The residue interaction network (RIN) or contact map of a protein was based on the distance map with a criterion 38,39 . A common approach, for example, is to measure the C β -C β distances (C a for Gly) and if the measured distance between residues R i and R j is shorter than 8 Å then there is a contact between R i and R j . This approach, however, we found may lead to incorrect assignment (Fig. S6). We adopted an alternative approach. Considering the hydrogen bond interaction X-H…Y (where X/Y can be C, N, O, S in proteins), the distance between X and Y for a typical H-bonds are in the range of 2 to 3 Å, and is ~3.5 Å for a C-H…O hydrogen bond in protein 67 . Here, for residues (vertices) R i and R j we define the distance d ij as the shortest distance between all heavy atoms. The distance map under this approach (Fig. 3a) agree well with the PAE map predicted by AF2 (Fig. 3b). The contact map is further defined based on the distance map: if d ij is shorter than 3.5 Å, we define an interaction (edge) between R i and R j . We then construct a binary adjacency matrix (1 for interaction and 0 for non-interaction) based this definition. The network analysis was performed using the igraph 53 package in R. The distance analysis was performed using the bio3d 68 package in R. The BVQ cofactor, the SF4-1 and SF4-2 clusters, and the PFOA substrate was treated as a residue (vertex) in the RIN. Ligand binding AutoDock Vina (V1.2.0) 69 was used for ligand docking. Using the T7RdhA-BVQ-SF4 system (system 5), after 10 ns MD equilibration, the PFOA ligand was docked into the protein complex (solvent and ions removed), and the top-score ligand was used to construct system 1. The force field parameters of the PFOA ligand can be found in the Appendix in SI. Declarations Author contributions : Conceptualization, H.-B.G., R.B.; methodology, H.-B.G., V.V., S.F., P.D., R.B.;, validation, all authors; writing—original draft preparation, H.-B.G.; writing—review and editing, all authors; project administration. N.K.-L., R.B.; funding acquisition, N.K.-L., R.B. All authors have read the final revision of this manuscript. Acknowledgements This work was supported by funding from the OUSD (R&E) ARAP Program. The structural modeling and MD simulations were performed using the DoD HPC. We appreciate Dr. Jerry Parks for the cobalamin force field parameters, and Dr. Marcel Swart for providing the [4Fe4S] iron sulfur (SF4) cluster force field parameters. We thank Dr. Peter Jaffe for helpful discussions on the Acidimicrobiaceae sp. A6 organism and the mechanism of the A6RdhA enzyme. We appreciate the MAPS TEAM from the Summit Country Day School for the 3D print of the T7RdhA-ligand model (Fig. S7). Data Availability All data generated or analyzed during this study are included in this published article and its supplementary information files. R codes for data analyses and visualizations are available upon request (H.-B.G., [email protected] ) Competing Interests The authors declare no competing interests. References Jumper, J. Evans, R.; Pritzel, A.; Green, T.; Figurnov, M.; Ronneberger, O.; Tunyasuvunakool, K.; Bates, R.; Zidek, A.; Potapenko, A.; Bridgland, A.; Meyer, C.; Kohl, S. A. A.; Ballard, A. J.; Cowie, A.; Romera-Paredes, B.; Nikolov, S.; Jain, R.; Adler, J.; Back, T.; Petersen, S.; Reiman, D.; Clancy, E.; Zielinski, M.; Steinegger, M.; Pacholska, M.; Berghammer, T.; Bodenstein, S.; Silver, D.; Vinyals, O.; Senior, A. W.; Kavukcuoglu, K.; Kohli, P.; Hassabis, D. Highly accurate protein structure prediction with AlphaFold. Nature 2021 , 596 , 583-589. Baek, M., Dimaio, F.; Anishchenko, I.; Dauparas, J.; Ovchinnikov, S.; Lee, G. R.; Wang, J.; Cong, Q.; Kinch, L. N.; Schaeffer, R. D.; Millan, C.; Park, H.; Adams, C.; Glassman, C. R.; Degiovanni, A.; Pereira, J. H.; Rodrigues, A. V.; van Dijk, A. A.; Ebrecht, A. C.; Opperman, D. J.; Sagmeister, T.; Buhlheller, C.; Pavkov-Keller, T.; Rathinaswamy, M. K.; Dalwadi, U.; Yip, C. K.; Burke, J. E.; Garcia, K. C.; Grishin, N. V.; Adams, P. D.; Read, R. J.; Baker, D. Accurate prediction of protein structures and interactions using a three-track neural network. Science 2021 , 373 , 871–876. Buel, G. R.; Walters, K. J. Can alphafold2 predict the impact of missense mutations on structure? Nature Struct. Mol. Biol . 2022 , 29 , 1-2. Pak, M. A.; Markhieva, K. A.; Novikova, M. S.; Petrov, D. S.; Vorobyev, I. S.; Maksimova, E. S.; Kondrashov, F. A.; Ivankov, D. N. Using AlphaFold to predict the impact of single mutations on protein stability and function. bioRxiv 2021 , https://doi.org/10.1101/2021.09.19.460937. McBride, J. M.; Polev, K.; Reinharz, V.; Grzybowski, B. A.; Tlusty, T. AlphaFold2 can predict structural and phenotypic effects of single mutations. arXiv 2022 :2004.068860[q-bio.BM], https://arxiv.org/abs/2204.06860. Schmidt, A.; Roner, S.; Mai, K.; Klinkhammer, H.; Kircher, M.; Ludwig, K. U. Predicting the pathogenicity of missense variants using features derived from AlphaFold2. bioRxiv 2022, https://doi.org/10.1101/2022.03.05.483091. Bagdonas, H.; Fogarty, C. A.; Fadda, E.; Agirre, J. The case for post-predictional modifications in the AlphaFold protein structure database. Nature Struct. Mol. Biol . 2021 , 28 , 869-870. Guo, H.-B.; Perminov, A.; Bekele, S.; Kedziora, G.; Farajollahi, S.; Varaljay, V.; Hinkle, K.; Molinero, V.; Meister, K.; Hung, C.; Dennis, P.; Kelley-Loughnane, N.; Berry, R. AlphaFold2 models indicate that protein sequence determines both structure and dynamics. Sci. Rep . 2022 , 12 , 10696. Kendrew, J. C.; Bodo, G.; Dintzis, H. M.; Parrish, R. G.; Wyckoff, H.; Phillips, D. C. A three-dimensional model of the myoglobin molecule obtained by X-ray analysis. Nature 1958 , 181 , 662-666. Fischer, E. Einfluss der configuration auf die wirkung der enzyme. Ber. Dtsch. Chem. Ges . 1894 , 27 , 2985-2993. Koshland, D. E. Jr. Application of a theory of enzyme specificity to protein synthesis. Proc. Natl. Acad. Sci. U. S. A. 1958 , 44 , 98-104. Koshland, D. E. Jr. The key-lock theory and the induced fit theory, Angew. Chem. Int. Ed. Engl . 1994 , 33 , 2375-2378. Monod, J.; Wyman, J.; Changeux, J.-P. On the nature of allosteric transitions: a plausible model. J. Mol. Biol . 1965 , 12 , 88-118. Changeux, J.-P.; Edelstein., S. Conformational selection or induced fit? 50 years of debate resolved. F1000 Biol. Rep. 2011 , 3 , 19. Najmanovich, R. J. Evolutionary studies of ligand binding sites in proteins. Curr. Opinion Struct. Biol . 2017 , 45 , 85-90. Skolnick, J.; Gao, M. Interplay of physics and evolution in the likely origin of protein biochemical function. Proc. Natl. Acad. Sci. U. S. A . 2013 , 110 , 9344-9349. Lu, H.; Diaz, D. J.; Czarnecki, N. J.; Zhu, C.; Kim, W.; Shroff, R.; Acosta, D. J.; Alexander, B. R.; Cole, H. O.; Zhang, Y.; Lynd, N. A.; Ellington, A. D.; Alper, H. S. Machine learning-aided engineering of hydrolases for PET depolymerization. Nature 2022 , 604 , 662-667. Wackett, L. P.; Robinson, S. L. The ever-expending limits of enzyme catalysis and biodegradation: polyaromatic, poly chlorinated, polyfluorinated and polymeric compounds. Biochem. J . 2020 , 477 , 2875-2891. Tunyasuvunakool, K.; Adler, J.; Wu, Z.; Green, T.; Zielinski, M.; Zidek, A.; Bridgland, A.; Cowie, A.; Meyer, C.; Laydon, A.; Velankar, S.; Kleywegt, G. J.; Bateman, A.; Evans, R.; Pritzel, A.; Figurnov, M.; Ronneberger, O.; Bates, R.; Kohl, S. A. A.; Potapenko, A.; Ballard, A. J.; Romera-Paredes, B.; Nikolov, S.; Jain, R.; Clancy, E.; Reiman, D.; Petersen, S.; Senior, A. W.; Kavukcuoglu, K.; Birney, E.; Kohli, P.; Jumper, J.; Hassabis, D. Highly accurate protein structure prediction for the human proteome. Nature 2021 , 596 , 590-596. Varadi, M. Anyango, S.; Deshpande, M.; Nair, S.; Natassia, C.; Yordanova, G.; Yuan, D.; Stroe, O.; Wood, G.; Laydon, A.; Zidek, A.; Green, T.; Tunyasuvunakool, K.; Petersen, S.; Jumper, J.; Clancy, E.; Green, R.; Vora, A.; Lutfi, M.; Figurnov, M.; Cowie, A.; Hobbs, N.; Kohli, P.; Kleywegt, G.; Birney, E.; Hassabis, D.; Velankar, S. AlphaFold Protein Structure Database: Massively expanding the structural coverage of protein-sequence space with high-accuracy models. Nucleic Acids Res . 2022 , 50 , D439–D444. Berman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.; Bhat, T. N.; Weissig, H.; Shindyalov, I. N.; Bourne, P. E. The protein data bank. Nucleic Acids Res . 2000 , 28 , 235-242. Goodsell, D. S.; Zardecki, C.; Costanzo, L. D.; Duarte, J. M.; Hudson, B. P.; Persikova, I.; Segura, J.; Shao, C.; Voigt, M.; Westbrook, J. D.; Young, J. Y.; Burley, S. K. RCSB protein data bank: enabling biomedical research and drug discovery. Protein Sci . 2020 , 29 , 52-65. Duarte, J. M.; Dutta, S.; Goodsell, D. S.; Burley, S. K. Exploring protein symmetry at the RCSB protein data bank. Emerging Topics Life Sci. 2022 , https://doi.org/10.1042/ETLS20210267. Mukhopadhyay, A.; Borkakoti, N.; Pravda, L.; Tyzack, J. D.; Thornton, J. M.; Velankar, S. Finding enzyme cofactors in protein data bank. Bioinformatics 2019 , 35 , 3510-3511. Stierand, S.; Rarey, M. Drawing the PDB: protein-ligand complexes in two dimensions. ACS Med. Chem. Lett . 2010 , 1 , 540-545. Evans, R.; O’Neill, M.; Pritzel, A.; Antropova, N.; Senior, A.; Green, T.; Zidek, A.; Bates, R.; Blackwell, S.; Yim, J.; Ronneberger, O.; Bodenstein, S.; Zielinski, M.; Bridgland, A.; Potapenko, A.; Cowie, A.; Tunyasuvunakool, K.; Jain, R.; Clancy, E.; Kohli, P.; Jumper, J.; Hassabis, D. Protein complex prediction with AlphaFold-Multimer. bioRxiv 2021, https://doi.org/10.1101/2021.10.04.463034 Gao, M.; An, D. N.; Parks, J. M.; Skolnick, J. AF2Complex predicts physical interactions in multimeric proteins with deep learning. Nature Commun . 2022 , 13 , 1744. Huang, S., and Jaffé, P. R. Defluorination of Perfluorooctanoic Acid (PFOA) and Perfluorooctane Sulfonate (PFOS) by Acidimicrobium sp. Strain A6. Environ. Sci. Technol . 2019 , 53 , 11410–11419. Bommer, M., Kunze, C., Fesseler, J., Schubert, T., Diekert, G., and Dobbek, H. Structural basis for organohalide respiration. Science 2014 , 346 , 455–458. Payne, K. A., Quezada, C. P., Fisher, K., Dunstan, M. S., Collins, F. A., Sjuts, H., Levy, C., Hay, S., Rigby, S. E., and Leys, D. Reductive dehalogenase structure suggests a mechanism for B12-dependent dehalogenation. Nature 2015 , 517 , 513–516. Tully, B. J., Sachdeva, R., Graham, E. D., and Heidelberg, J. F. 290 metagenome-assembled genomes from the Mediterranean Sea: a resource for marine microbiology. PeerJ 2017 , 5 , e3558. Stank, A.; Kokh, D. B.; Fuller, J. C.; Wade, R. C. Protein binding pocket dynamics. Acc. Chem. Res . 2016 , 49 , 809-815. Pruitt, K. D., Tatusova, T., and Maglott, D. R. NCBI Reference Sequence (RefSeq): a curated non-redundant sequence database of genomes, transcripts and proteins. Nucleic Acids Res . 2005 , 33 , D501–D504. Cooper, C. J., Zheng, K., Rush, K. W., Johs, A., Sanders, B. C., Pavlopoulos, G. A., Kyrpides, N. C., Podar, M., Ovchinnikov, S., Ragsdale, S. W., and Parks, J. M. Structure determination of the HgcAB complex using metagenome sequence data: insights into microbial mercury methylation. Commun Biol . 2020 , 3 , 320. Halliwell, T., Fisher, K., Payne, K. A. P., Rigby, S. E. J., and Leys, D. Catabolic reductive dehalogenase substrate complex structures underpin rational repurposing of substrate scope. Microorganisms 2020 , 8, 1344. Kunze, C., Bommer, M., Hagen, W. R., Uksa, M., Dobbek, H., Schubert, T., and Diekert, G. Cobamide-mediated enzymatic reductive dehalogenation via long-range electron transfer. Nature Commun . 2017 , 8 , 15858. Laskowski, R. A.; Thornton, J. M. PDBsum extras: SARS-CoV-2 and AlphaFold models. Prot. Sci . 2022 , 31 , 283-289. Du, Z.; Su, H.; Wang, W.; Ye, L.; Wei, Hong; Peng, Z; Anishchenko, I.; Baker, D.; Yang, J. The trRosetta server for fast and accurate protein structure prediction. Nature Protocols 2022 , 16 , 5634-5651. Clementel, D.; Del Conte, A.; Monzon, A. M.; Camagni, G. F.; Minervini, G.; Piovesan, D.; Tosatto, S. C. E. Ring 3.0: fast generation of probabilistic residue interaction networks from structural ensembles. Nucleic Acids Res . 2022 , 50 , W651-W656. Wittung-Stafshede, P. Role of cofactors in protein folding. Acc. Chem. Res . 2002 , 35 , 201-208. Xu, Q.; Guo, H.-B.; Wlodawer, A.; Guo, H. The importance of dynamics in substrate-assisted catalysis and specificity. J. Am. Chem. Soc . 2006 , 128 , 5994-5995. Hekkelman, M. L.; de Vries, I.; Joosten, R. P.; Perrakis, A. AlphaFill: enriching the AlphaFold models with ligands and co-factors. bioRxiv 2021 , https://doi.org/10.1101/2021.11.26.470110. Chang, L.; Perez, A. AlphaFold encodes the principles to identify high affinity peptide binders. bioRxiv 2022 , https://doi.org/10.1101/2022.03.18.484931. Bryant, P.; Pozzati, G.; Elofsson, A. Improved prediction of protein-protein interactions using AlphaFold2. Nature Commun . 2022 , 13 , 1265. Krautler, B.; Fieber, W.; Ostermann, S.; Fasching, M.; Ongania, K.-H.; Gruber, K.; Kratky, C.; Mikl, C.; Siebert, A.; Diekert, G. The cofactor of tetrachloroethene reductive dehalogenase of Dehalospirillum multivorans is norpseudo-B12, a new type of natural corrinoid. Helvetica 2003 , 86 , 3698-3716. Beinert, H., Holm, R. H., and Münck, E. Iron-sulfur clusters: nature’s modular, multipurpose structures. Science 1997 , 277 , 653–659. Jaeger, J.; Monk, N. Everything flows: a process perspective of life. EMBO Rep . 2015 , 16 , 1064-1067. Buck, R. C., Franklin, J., Berger, U., Conder, J. M., Cousins, I. T., de Voogt, P., Jensen, A. A., Kannan, K., Mabury, S. A., and van Leeuwen, S. P. J. Perfluoroalkyl and polyfluoroalkyl substances in the environment: terminology, classification, and origins. Integr. Environ. Assess. Manag . 2011 , 7 , 513–541. Ruiz-Uriguen, M.; Shuai, W.; Huang, S.; Jaffe, P. R. Biodegradation of PFOA in microbial electrolysis cells by Acidimicrobiaceae sp. Strain A6 . Chemosphere , 2022 , 292 , 133506. Katoh, K., and Standley, D. M. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol. Biol. Evol . 2013 , 30 , 772–780. Finn, R. D., Clements, J., and Eddy, S. R. HMMER web server: interactive sequence similarity searching. Nucleic Acids Res . 2011 , 39 , W29–W37. Zallot, R., Oberg, N., and Gerlt, J. A. The EFI Web Resource for Genomic Enzymology Tools: Leveraging Protein, Genome, and Metagenome Databases to Discover Novel Enzymes and Metabolic Pathways. Biochemistry 2019 , 58 , 4169–4182. Han, W.-S., Lee, J., Pham, M.-D., and Yu, J. X. iGraph: a framework for comparisons of disk-based graph indexing techniques. Proceedings VLDB Endowment 2010 , 3 , 449–459. Crooks, G. E., Hon, G., Chandonia, J.-M., and Brenner, S. E. WebLogo: a sequence logo generator. Genome Res . 2004 , 14 , 1188–1190. Zhang, Y.; Skolnick, J. TM-align: a protein structure alignment algorithm based on the TM-score. Nucleic Acids Res . 2005 , 33 , 2302-2309. Paradis, E.; Claude, J.; Strimmer, K. APE: analyses of phylogenetics and evolution in R language. Bioinformatics , 2004 , 20 , 289-290. Kumar, S., Stecher, G., Li, M., Knyaz, C., and Tamura, K. MEGA X: Molecular Evolutionary Genetics Analysis across Computing Platforms. Mol. Biol. Evol . 2018 , 35 , 1547–1549. Phillips, J. C., Braun, R., Wang, W., Gumbart, J., Tajkhorshid, E., Villa, E., Chipot, C., Skeel, R. D., Kalé, L., and Schulten, K. Scalable molecular dynamics with NAMD. J. Comput. Chem . 2005 , 26 , 1781–1802. Huang, J., Rauscher, S., Nawrocki, G., Ran, T., Feig, M., de Groot, B. L., Grubmüller, H., and MacKerell, A. D., Jr. CHARMM36m: an improved force field for folded and intrinsically disordered proteins. Nat. Methods 2017 , 14 , 71–73. MacKerell, A. D., Bashford, D., Bellott, M., Dunbrack, R. L., Evanseck, J. D., Field, M. J., Fischer, S., Gao, J., Guo, H., Ha, S., Joseph-McCarthy, D., Kuchnir, L., Kuczera, K., Lau, F. T., Mattos, C., Michnick, S., Ngo, T., Nguyen, D. T., Prodhom, B., Reiher, W. E., Roux, B., Schlenkrich, M., Smith, J. C., Stote, R., Straub, J., Watanabe, M., Wiórkiewicz-Kuczera, J., Yin, D., and Karplus, M. All-atom empirical potential for molecular modeling and dynamics studies of proteins. J. Phys. Chem. B 1998 , 102 , 3586–3616. Price, D. J., and Brooks, C. L., 3rd. A modified TIP3P water potential for simulation with Ewald summation. J. Chem. Phys. 2004 , 121 , 10096–10103. Pavlova, A., Parks, J. M., and Gumbart, J. C. Development of CHARMM-Compatible Force-Field Parameters for Cobalamin and Related Cofactors from Quantum Mechanical Calculations. J. Chem. Theory Comput . 2018 , 14 , 784–798. Carvalho, A. T. P., and Swart, M. Electronic structure investigation and parametrization of biologically relevant iron-sulfur clusters. J. Chem. Inf. Model . 2014 , 54 , 613–620. Direct Force Field 7.0, Aeon Technology, Inc. San Diego, 2011 . Brooks, B. R., Brooks, C. L., 3rd, Mackerell, A. D., Jr, Nilsson, L., Petrella, R. J., Roux, B., Won, Y., Archontis, G., Bartels, C., Boresch, S., Caflisch, A., Caves, L., Cui, Q., Dinner, A. R., Feig, M., Fischer, S., Gao, J., Hodoscek, M., Im, W., Kuczera, K., Lazaridis, T., Ma, J., Ovchinnikov, V., Paci, E., Pastor, R. W., Post, C. B., Pu, J. Z., Schaefer, M., Tidor, B., Venable, R. M., Woodcock, H. L., Wu, X., Yang, W., York, D. M., and Karplus, M. CHARMM: the biomolecular simulation program. J. Comput. Chem . 2009 , 30 , 1545–1614. Humphrey, W., Dalke, A., and Schulten, K. VMD: visual molecular dynamics. J. Mol. Graph . 1996 , 14 , 33–38. Guo, H.-B.; Beahm, R. F., Guo, H. Stabilization and destabilization of the C d -H…OC hydrogen bonds involving proline residues in helices. Grant, B. J., Skjaerven, L., and Yao, X.-Q. The Bio3D packages for structural bioinformatics. Protein Sci . 2021 , 30 , 20–30. Trott, O.; Olson, A. J. AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization and multithreading. J. Comput. Chem . 2010 , 31 , 455-461. Additional Declarations No competing interests reported. Supplementary Files AF2.TMED77.si2.docx Cite Share Download PDF Status: Published Journal Publication published 11 Mar, 2023 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Major revision 17 Nov, 2022 Reviews received at journal 01 Oct, 2022 Reviews received at journal 27 Sep, 2022 Reviewers agreed at journal 24 Sep, 2022 Reviewers agreed at journal 16 Sep, 2022 Reviewers invited by journal 16 Sep, 2022 Editor assigned by journal 16 Sep, 2022 Editor invited by journal 16 Sep, 2022 Submission checks completed at journal 16 Sep, 2022 First submitted to journal 12 Sep, 2022 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-2057833","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":137234616,"identity":"962d77f3-2db9-478a-b760-2fcc098cb69e","order_by":0,"name":"Hao-Bo Guo","email":"","orcid":"","institution":"Materials and Manufacturing Directorate, Air Force Research Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Hao-Bo","middleName":"","lastName":"Guo","suffix":""},{"id":137234617,"identity":"585a36e8-136d-4293-9814-3e9f6a593848","order_by":1,"name":"Vanessa Varaljay","email":"","orcid":"","institution":"Materials and Manufacturing Directorate, Air Force Research Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Vanessa","middleName":"","lastName":"Varaljay","suffix":""},{"id":137234618,"identity":"9cab2918-7efa-49fb-81d7-48927de84911","order_by":2,"name":"Gary Kedziora","email":"","orcid":"","institution":"GDIT Inc","correspondingAuthor":false,"prefix":"","firstName":"Gary","middleName":"","lastName":"Kedziora","suffix":""},{"id":137234619,"identity":"5090ebc5-51fe-48c9-b862-aa8eb68436f3","order_by":3,"name":"Kimberly Taylor","email":"","orcid":"","institution":"Materials and Manufacturing Directorate, Air Force Research Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Kimberly","middleName":"","lastName":"Taylor","suffix":""},{"id":137234620,"identity":"e0fc0a30-5a1b-4c85-a843-89c126eee414","order_by":4,"name":"Sanaz Farajollahi","email":"","orcid":"","institution":"Materials and Manufacturing Directorate, Air Force Research Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Sanaz","middleName":"","lastName":"Farajollahi","suffix":""},{"id":137234621,"identity":"89c8111e-e58e-4977-a276-445ea7f1546d","order_by":5,"name":"Nina Lombardo","email":"","orcid":"","institution":"Materials and Manufacturing Directorate, Air Force Research Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Nina","middleName":"","lastName":"Lombardo","suffix":""},{"id":137234622,"identity":"d0a43fb6-ea8e-4f87-8ee6-64e86d3d8361","order_by":6,"name":"Eric Harper","email":"","orcid":"","institution":"Materials and Manufacturing Directorate, Air Force Research Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Eric","middleName":"","lastName":"Harper","suffix":""},{"id":137234623,"identity":"048593fc-eb01-4b3a-b74f-4cd9452e3997","order_by":7,"name":"Chia Hung","email":"","orcid":"","institution":"Materials and Manufacturing Directorate, Air Force Research Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Chia","middleName":"","lastName":"Hung","suffix":""},{"id":137234624,"identity":"bfb6f59e-9116-4b71-8621-0d53d1932f74","order_by":8,"name":"Marie Gross","email":"","orcid":"","institution":"Materials and Manufacturing Directorate, Air Force Research Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Marie","middleName":"","lastName":"Gross","suffix":""},{"id":137234625,"identity":"5ab2457d-24b1-4b8b-bce4-c5db52d3910d","order_by":9,"name":"Alexander Perminov","email":"","orcid":"","institution":"Materials and Manufacturing Directorate, Air Force Research Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Alexander","middleName":"","lastName":"Perminov","suffix":""},{"id":137234626,"identity":"ba3beea3-ca7a-4817-bd2a-ddb78562d301","order_by":10,"name":"Patrick Dennis","email":"","orcid":"","institution":"Materials and Manufacturing Directorate, Air Force Research Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Patrick","middleName":"","lastName":"Dennis","suffix":""},{"id":137234627,"identity":"8d91c2d1-d349-4503-a884-ffb1bbaf6dd7","order_by":11,"name":"Nancy Kelley-Loughnane","email":"","orcid":"","institution":"Materials and Manufacturing Directorate, Air Force Research Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Nancy","middleName":"","lastName":"Kelley-Loughnane","suffix":""},{"id":137234628,"identity":"0cd6e07a-09d2-470d-ab13-8855969c9314","order_by":12,"name":"Rajiv Berry","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA90lEQVRIie2RsUoDQRCG/7BwJzjB9gQxr3BHIDbqvcocadOlESy8KtXZH4S8g+IDuHJwNqKtYLP2FvsAW7gTSLvbCu5X/MzCfDsDAyQSf5C6BWvAST0xknkLaBNQSr1XRqlVKUn+qTmsCINEVhwUBJVX/T0QxvOL7dvilhxmhOMmOKXumL3i5mfvq8XXdIOqw/QhqFyD+aXH2PREXmkxeY4p8xMjynAnytovVkenVD2zthi48IqiDE1UKT+NV8qx6ilbn+42xbJTMeVjtbR842YFqUf74y6vuvz+ydqA4u/gf9zf8EjSn0YF24VcHwoT7U0kEon/yS9NXFMFkmuIMwAAAABJRU5ErkJggg==","orcid":"","institution":"Materials and Manufacturing Directorate, Air Force Research Laboratory","correspondingAuthor":true,"prefix":"","firstName":"Rajiv","middleName":"","lastName":"Berry","suffix":""}],"badges":[],"createdAt":"2022-09-12 18:14:08","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-2057833/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-2057833/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-023-30310-x","type":"published","date":"2023-03-11T19:33:35+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":26734701,"identity":"8ab01e4b-f093-4586-9844-ba1d82a09fbe","added_by":"auto","created_at":"2022-09-20 21:42:23","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":591102,"visible":true,"origin":"","legend":"\u003cp\u003eT7RdhA structure. (\u003cstrong\u003ea\u003c/strong\u003e) The structure of T7RdhA in complex with BVQ (green) cofactor, two SF4 (purple) clusters and the PFOA (purple) substrate. In the protein cartoon α-helices are in red, β-sheets are in yellow and coils in white. (\u003cstrong\u003eb\u003c/strong\u003e) A closer view of the BVQ cofactor, SF4 clusters and the binding Cys residues, and the PFOA substrate. Fe in purple, S in yellow, Co in green, F in pink, C in cyan, N in blue and O in red. (\u003cstrong\u003ec\u003c/strong\u003e) A wire presentation of the secondary structures in T7RdhA plotted by PDBsum37. Note that the red dots on top of the amino acids indicate that the residue is involved in cofactor/substrate binding. The b-strands form three sheets (A, B and C). Positions of β- and γ-turns in the loop regions are labeled.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-2057833/v1/bb7dc6e9a03735633c76587b.png"},{"id":26734706,"identity":"668ae521-0b0e-4be8-a9f3-1def4e24a053","added_by":"auto","created_at":"2022-09-20 21:42:24","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":108504,"visible":true,"origin":"","legend":"\u003cp\u003eRoot-mean square fluctuation profiles of five systems in comparison with the AF2 scores. The spheres correspond to the Cys residues that are covalently bonded to the Fe atoms in the SF4 clusters. AF2 (0, black solid) is a reverse normalization of the per-residue pLDDT scores of the AF2 protein model. The protein in complex with the PFOA substrate, BVQ cofactor and two SF4 clusters (1, red solid) shows the best match with the AF2 scores with Pearson’s correlation coefficient (PCC) of 0.561. The other systems include apo-T7RdhA (2, light blue dashed), T7RdhA with BVQ (3, dark blue dashed), T7RdhA with two SF4 clusters (4, orange dashed) and T7RdhA with both BVQ and two SF4 clusters (5, purple dashed) are also plotted with PCC in parentheses. The RMSF calculated from 90 models (6, black dashed) matches well with the AF2-scores.\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-2057833/v1/bf1a60b3924dcd47d7d0fcab.jpeg"},{"id":26734704,"identity":"2490360d-35ef-48da-9bce-985c6bef8ac9","added_by":"auto","created_at":"2022-09-20 21:42:24","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":764061,"visible":true,"origin":"","legend":"\u003cp\u003eT7RdhA distance map, PAE map and RIN. (\u003cstrong\u003ea\u003c/strong\u003e) Distance map of T7RdhA at 100 ns compared with (\u003cstrong\u003eb\u003c/strong\u003e) the PAE map from AF2. (\u003cstrong\u003ec\u003c/strong\u003e) The RIN at 100 ns constructed from the contact map. The BVQ cofactor in sphere (dark green), the SF4 clusters in square (purple) and PFOA in rectangle (pink). Blue nodes are positively charged residues (Lys and Arg) and red are negatively charged residues (Glu and Asp); black nodes are aromatic residues (His, Tyr, Phe and Trp); and all other residues in yellow.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-2057833/v1/61f834cdff8f0427134b60cd.png"},{"id":26734703,"identity":"26dafffe-f6b3-44b6-8554-fe38b9ab7b30","added_by":"auto","created_at":"2022-09-20 21:42:23","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1109892,"visible":true,"origin":"","legend":"\u003cp\u003eThe protein-cofactor/ligand interactions. Distribution of interaction residues from 100 RINs (constructed from snapshots of a 100 ns MD) to (\u003cstrong\u003ea\u003c/strong\u003e) BVQ cofactor, (\u003cstrong\u003eb\u003c/strong\u003e) SF4-2 cluster, (\u003cstrong\u003ec\u003c/strong\u003e) SF4-1 cluster and (\u003cstrong\u003ed\u003c/strong\u003e) PFOA substrate. Number of interactions (mean ±standard deviation) detected in all RINs are in parentheses. All residues are ranked by the percentage of interactions observed in all RINs. The y-axis indicates the percentage of the interactions in all RINs. Representative clusters are shown for the interactions centering (\u003cstrong\u003ee\u003c/strong\u003e) BVQ, (\u003cstrong\u003ef\u003c/strong\u003e) SF4-1, (g) SF4-2 and (\u003cstrong\u003eh\u003c/strong\u003e) PFOA. Motifs (sequences in Fig. 1c) that contain interacting residues and some of the important residues are labeled. Cobalt is colored in silver, iron in purple, fluorine in pink, carbon in cyan, nitrogen in blue and oxygen in red.\u003c/p\u003e","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-2057833/v1/ff36dad302e7439058628357.jpeg"},{"id":26734707,"identity":"b510d09e-c09c-4b9d-af84-0057dc7e58ed","added_by":"auto","created_at":"2022-09-20 21:42:24","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":302333,"visible":true,"origin":"","legend":"\u003cp\u003eComparisons of protein structures during MD and AF2 models. (\u003cstrong\u003ea\u003c/strong\u003e) A structure-based phylogenetic tree using the protein structure snapshots during the MD at different simulation time (100 ns, 200 ns and 300 ns for snapshots 1-3) for the MD systems 1-5 and AF2 structures (system 6), see Fig. 2. The initial MD structure is sys6-1, and sys6-2 and sys6-3 has the RMSD values 0.391 and 0.937 Å to sys6-1, respectively. (\u003cstrong\u003eb\u003c/strong\u003e) Superimposed structures of systems 1-6 colored by RMSD to the initial MD structure (sys6-1). Systems1-5 were collected from the MD simulations after 300 ns. A BWR color scheme was applied with blue for low, red for high and white for in-between RMSD values. The overall RMSD for all pairs of structures plotted in the phylogenetic tree is 2.45±0.57 Å.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-2057833/v1/b156bee3587ad750f3254fa1.png"},{"id":44721837,"identity":"342b70c1-204a-44c8-a832-1a54ef8a11f6","added_by":"auto","created_at":"2023-10-16 19:40:07","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2451583,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-2057833/v1/7aad8620-d70d-4135-b4c3-990030870622.pdf"},{"id":26734975,"identity":"4351d209-f1c6-4ab7-9901-7f2de4d25a49","added_by":"auto","created_at":"2022-09-20 21:47:24","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":2201943,"visible":true,"origin":"","legend":"","description":"","filename":"AF2.TMED77.si2.docx","url":"https://assets-eu.researchsquare.com/files/rs-2057833/v1/2d60b6a672472caac19139cd.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Accurate prediction by AlphaFold2 for ligand binding in a reductive dehalogenase: Implications for PFAS (per- and polyfluoroalkyl substance) biodegradation","fulltext":[{"header":"Introduction","content":"\u003cp\u003eAlphaFold2\u003csup\u003e1\u003c/sup\u003e (AF2) has achieved near-experimental accuracy for predicting protein structures from the primary sequences. This breakthrough, together with the developments of other tools including RoseTTAFold\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, allow us to understand the protein structure-function relationships with atomic precision. The performance of AF2, however, was found to produce contradictory results in some assessments. For instance, it is the subject of debate whether AF2 fails to predict the impact of point mutations in protein structure\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e and stability\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e; whereas other studies indicated that the structures and phenotypic effects of the point mutations can be correctly predicted\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e or assisted\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e by AF2. With these controversies in mind, further modifications to the AF2-predicted structures are required to appropriately understand the protein functions\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e, including the addition of ligands (cofactors and/or substrates). Moreover, as proteins are not static and generally perform functions in the cell but not in crystals, it is important to examine if AF2 can capture protein dynamics in aqueous environments\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eOver 60 years before the first glimpse of a protein structure\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e, the \u0026ldquo;key-lock\u0026rdquo; model\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e was proposed to describe how proteins perform functions via ligand binding. In this model the protein is described as a lock awaiting the ligand as a key to unleash its function. Unfitted ligands ࣧincluding water moleculesࣧwould fail to unlock the protein. In later studies, especially with the help of protein and protein-ligand complex structures that were becoming increasingly abundant, a refined \u0026ldquo;induced-fit\u0026rdquo; model\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e was proposed which accommodates the conformational changes of the protein upon ligand-binding. To illustrate the long-range (allosteric) effect of ligand binding, the conformation selection model was proposed\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, in which the binding conformations pre-exist in the protein such that the ligand binding could spontaneously occur. Both sequences and structures (shapes, sizes, and locations) of the binding pockets in proteins are thought to evolve to facilitate association with different ligands\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Although only a small number of representative pockets in proteins have been estimated, ligand specificity of proteins may emerge in evolution without functional constraints\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. In this regard most of the previously or currently recognized hard-to-degrade chemicals, including polymers\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e and per- and polyfluoroalkyl substances (PFAS)\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e, could serve as the \u0026ldquo;keys\u0026rdquo; for certain lock proteins.\u003c/p\u003e \u003cp\u003eWith AF2 it is likely that every single sequenced protein has its high-resolution 3D structure available in the database\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e or can be accurately predicted. In the Protein Data Bank\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, however, the majority of the structures derived from experiments (crystallography, cryoEM, NMR, etc.) are complexes, including multimers (often with symmetry)\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e, as well as bound with cofactors\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e and other ligands\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. This information is as important as the protein structures themselves for informing the protein functions and mechanisms, ever since the first solved protein structure of mioglobin\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. However, the initial structures that we obtain from AF2\u003csup\u003e1\u003c/sup\u003e, AlphaFold-multimer\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e, or AF2Complex\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e are apo-proteins, i.e., proteins without ligands. Given the important roles of ligands play in the protein functions, it is crucial to determine whether AF2 is suitable for predicting structure and function of such proteins. In another word, do the apo-proteins predicted by AF2 have the proper binding pockets (cavities) for the ligand binding? Here, we used a multiple-ligand protein to answer this question. As the PFAS substrate (e.g., PFOA) can bind to the protein as a substrate, it has the potential for biodegrading the PFAS contaminants.\u003c/p\u003e \u003cp\u003eA recent work found a bacterium \u003cem\u003eAcidimicrobiaceae sp. A6\u003c/em\u003e that, when cultured in the presence of either perfluorooctanoic acid (PFOA) or perfluorooctane sulfonate (PFOS), was able to defluorinate these chemicals with an observed release of fluoride ion, shorter-chain perfluorinated products, and acetate\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. The key enzyme for defluorination of PFOA/PFOS was identified as a reductive dehalogenase subunit A (RdhA) in GenBank (id: MK358462.1)\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. However, only partial sequence of this enzyme (A6RdhA hereafter) was available with a missing C-terminus of over 100 AA\u0026rsquo;s compared with known reductive dehalogenases including PceA\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e and NpRdhA\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. Sequence mining starting from the partial A6RdhA sequence revealed a full protein sequence from the bacteria \u003cem\u003eTMED77\u003c/em\u003e in a metagenomic assembly of the Mediterranean Sea microorganisms\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, which shares 98% sequence identity with the known part of A6RdhA protein. This protein is referred to as the T7RdhA in present work. It is worth noting that the \u003cem\u003eTMED77\u003c/em\u003e bacterium belongs to the same \u003cem\u003eAcidimicrobiaceae\u003c/em\u003e family as \u003cem\u003eA. sp. A6\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eIn the present work, we showed that T7RdhA is a PceA-like protein\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e which utilizes two [4Fe4S] iron sulfur (SF4) clusters and a norpseudo-cobalamin (BVQ) cofactor. We constructed AF2 models of T7RdhA, and for the highest-ranked model, both BVQ cofactor and SF4 clusters can be put on the binding pockets precisely. Molecular dynamics (MD) simulations were performed on this model with no ligand (apo-form), partially bound by cofactors (either BVQ or SF4), or with both cofactors, and with both cofactors and a substrate (PFOA). The results indicate that the AF2 is able to predict the binding pockets for both cofactors and substrates in the protein models, with regard to the binding pockets dynamics\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. The model used in the MD simulation was constructed using AF2 V2.0.1 (July 2021 version). A newer version of AF2 V2.2.2 (downloaded in July 2022) was employed to construct additional 90 models and compared with the MD model. High similarity of the new models with the MD model illustrates the reproducibility of AF2. Interestingly, we show that the diversity of AF2 models resemble the MD results. We perform residue-interaction network (RIN) analyses using the MD trajectories of the model with both BVQ, SF4, and PFOA. We identified the binding pockets for both cofactors and the substrate in T7RdhA, which will help to search and design proteins for PFAS biosequestration and degradation.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eT7RdhA (and potentially A6RdhA) is a CoFeSP\u003c/h2\u003e \u003cp\u003eFrom a sequence similarity network (SSN) constructed using the NCBI nr database\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e, T7RdhA and the T7RdhA-like proteins comprise highly conserved residues for the binding of a corrinoid cofactor and two [4Fe4S] iron-sulfur (SF4) clusters (see Fig. S1 in the supplementary information, SI). These proteins were termed as corrinoid iron-sulfur proteins (CoFeSPs)\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. The corrinoid cofactor or these proteins include the cobalamin (B12) in NpRdhA\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e and B12-derivatives such as the norpseudo cobalamin (BVQ) in PceA\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. NpRdhA uses the B12 cofactor and belongs to an aerobic bacterium \u003cem\u003eNitratireductor pacificus\u003c/em\u003e\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. However, PceA that uses the BVQ cofactor is carried by the anaerobic bacterium \u003cem\u003eSulfurospirillum multivorans\u003c/em\u003e\u003csup\u003e\u003cem\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/em\u003e\u003c/sup\u003e. It is likely T7RdhA uses the BVQ cofactor not only the \u003cem\u003eAcidimicrobiaceae\u003c/em\u003e bacterial family is anaerobic\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e, but also because T7RdhA belongs to the PceA branch (Fig. S1) in the clustering of the T7RdhA-like proteins from the SSN. A cross-linked binding mode has been found in both PceA- and NpRdhA-like proteins, in which two SF4-binding motifs are required for binding of each of the two SF4 clusters (Fig. S2). Moreover, we cloned and expressed T7RdhA in \u003cem\u003eEscherichia coli\u003c/em\u003e, and verified that T7RdhA binds both a corrinoid cofactor and two iron-sulfur clusters (Fig. S3). The network-assisted de novo structured prediction approach and experimental verifications indicate that T7RdhA is a CoFeSP. Besides the cofactors BVQ, SF4-1 and SF4-2, the PFOA substrate is also docked into T7RdhA (Methods), and Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e shows the binding of all four ligands in T7RdhA.\u003c/p\u003e \n\u003ch2\u003eAlphafold2 Predicts Residue Flexibility Of T7rdha Upon Ligand Binding\u003c/h2\u003e\n\u003cp\u003eWe showed previously that the per-residue pLDDT (predicted local distance difference test) scores accompanying the predicted protein models by AF2 also anticipate the residue flexibilities for globular proteins, protein dimers and intrinsically disordered proteins\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. However, AF2 only provides the apo-forms of the proteins or protein-multimers, and the knowledge of cofactors and/or substrates related to the protein functions can only be acquired from experiments or literature. In the case of T7RdhA, since it is likely a CoFeSP which performs the functions utilizing the corrinoid (BVQ) cofactor and two SF4 clusters, and presumably the PFOA substrate can be bound to the active site of the protein for catalysis, we asked if AF2 can predict the binding of these proposed ligands.\u003c/p\u003e \u003cp\u003eTo answer this question, we performed MD simulations five different systems: 1) T7RdhA complexed with BVQ, two SF4 clusters (SF4-1 and SF4-2) and the PFOA substrate; 2) apo-T7RdhA with no ligand; 3) T7RdhA complexed with BVQ; 4) T7RdhA complexed with two SF4 clusters; and 5) T7RdhA complexed with the BVQ cofactor and two SF4 clusters. T7RdhA is a well-folded globular protein. The binding of cofactors or ligands does not significantly alter the conformation of T7RdhA. However, the residue flexibility vary significantly for all the five systems, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. It has been suggested that the diversity of AF2 models would yield biological insights that might be otherwise ignored from a single snapshot of the protein structure\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. In addition to these five systems, we constructed 90 more AF2 structures and calculated the residue fluctuations among these structures for comparison (line 6, see below).\u003c/p\u003e \u003cp\u003eFor the well-folded globular proteins, the residue flexibility profiles measured by the root-mean square fluctuation (RMSF) from MD simulations were found to be highly consistent with the AF2-scores from the AF2 predictions, which is a reverse normalization of the pLDDT scores\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eAF2\u003csub\u003ei\u003c/sub\u003e = (pLDDT\u003csub\u003emax\u003c/sub\u003e-pLDDT\u003csub\u003ei\u003c/sub\u003e)/(pLDDT\u003csub\u003emax\u003c/sub\u003e-pLDDT\u003csub\u003emin\u003c/sub\u003e), (1)\u003c/p\u003e \u003cp\u003ewhere AF2\u003csub\u003ei\u003c/sub\u003e is the AF2-score of the i-th residue calculated from the pLDDT-score of the Cα atom of the i-th residue\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. For systems 1\u0026ndash;5, 300 ns MD simulations was performed, and the RMSF profile was calculated using the last 100 ns trajectory. For all five models positive correlation between the AF2-scores and RMSFs were observed (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), and the best fit comes from the T7RdhA model in complex with BVQ, two SF4 clusters and a PFOA substrate (system 1)ࣧexcept for an inconsistency at the C-terminus region for which AF2 anticipates it highly flexible but the RMSF from MD indicates it is instead relatively rigid. Nevertheless, for the important binding regions such as three β-sheets and the helices H12 to H16 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), RMSF of system 1 is highly consistent with the AF2-scores. In contrast, systems 2 and 3 indicate that without the SF4 clusters, the regions containing the binding Cys residues (H14 to H16, including the loop region between H15 and H16) are highly flexible, contradicting with the AF2 predictions. In system 4, binding of the SF4 lead to relatively small flexibility of the binding Cys residues, however, the loop region between H15 and H16 still shows significantly higher flexibility compared with the AF2 prediction. System 5 also shows better consistency between MD and AF2 prediction, slightly lower than system 1. We also measured the root-mean square deviation (RMSD) of all five systems. The mean RMSD of the last 10 ns are calculated as shown in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, and again, system 1 shows the lowest RMSD value. The RMSD profiles from the last 100 ns of all five systems are shown in Fig. S4 in the SI.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eComparing protein dynamics metrics with the AF2 predictions\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSystem\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDescription\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRMSF\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e (\u0026Aring;)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePCC\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003ep-value\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eRMSD\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e (\u0026Aring;)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eT7RdhA\u0026thinsp;+\u0026thinsp;BVQ\u0026thinsp;+\u0026thinsp;SF4-1/SF4-2\u0026thinsp;+\u0026thinsp;PFOA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.8 \u0026plusmn; 0.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.561\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.2E-35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eApo-T7RdhA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.9 \u0026plusmn; 0.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.357\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.2E-13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eT7RdhA\u0026thinsp;+\u0026thinsp;BVQ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e1.0 \u0026plusmn; 0.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.342\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.4E-12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eT7RdhA\u0026thinsp;+\u0026thinsp;SF4-1/SF4-2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e1.1 \u0026plusmn; 0.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.355\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.6E-13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eT7RdhA\u0026thinsp;+\u0026thinsp;BVQ\u0026thinsp;+\u0026thinsp;SF4-1/SF4-2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.9 \u0026plusmn; 0.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.426\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.7E-19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAF2 models\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.5 \u0026plusmn; 0.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.873\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.7E-128\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003e1. The root-mean-square fluctuation (RMSF) from a 100 ns MD simulation after 200 ns equilibration. 2. Pearson\u0026rsquo;s correlation coefficient (PCC) and p-values between the RMSF and the predicted AF2-score. 3. Mean RMSD from the last 10 ns trajectory referenced with the initial structure of the 100 ns MD. 4. 90 AF2 models are used to calculate the RMSF of all residues. The RMSD is averaged from all 90 models referenced to the model used in the MD simulations.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e\n\u003ch2\u003eDiversity Of Af2 Models Resembles The Md Simulation\u003c/h2\u003e\n\u003cp\u003eThe original AF2 publication suggested that diversity of AF2 models (i.e., via multiple runs) may yield new biological insights by predicting alternative configurations of the proteins\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. In this work, the T7RdhA model for the MD simulations was constructed using an old AF2 version (V2.0.1, released July 2021). We used the newer AF2 version (V2.2.2, downloaded July 2022) to construct 90 T7RdhA models (18 independent runs, each gives 5 models). These models show low RMSD to the MD model (Fig. S5 in the SI), indicating the reproducibility of the AF2 algorithm. By combining the configuration of all 90 new T7RdhA models, we also calculated the residue RMSF values and compared with the AF2 scores from the MD model (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Interestingly, this RMSF profile (black dashed line) shows a PCC\u0026thinsp;=\u0026thinsp;0.873 (p\u0026thinsp;=\u0026thinsp;2.7E-128) to the AF2-score profile. In this profile not only the BVQ- or SF4-binding regions, but also the dynamic N-terminus is consistent with the AF2-scores. Therefore, from the protein sequence, AF2 not only provided dynamics information of all residues via the (pLDDT or AF2-scores)\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e, it seems that multiple AF2 runs can reproduce the MD simulation, i.e., the diversity in structures from multiple AF2 runs can retrieve the residue fluctuations.\u003c/p\u003e\n\u003ch2\u003eResidue Interaction Network And The Cofactor/ligand Binding Modes Of T7rdha\u003c/h2\u003e\n\u003cp\u003eThe protein residue distance maps usually defined as the distance \u003cem\u003ed\u003c/em\u003e\u003csub\u003e\u003cem\u003eij\u003c/em\u003e\u003c/sub\u003e between the C\u003csub\u003eβ\u003c/sub\u003e atoms (C\u003csub\u003eα\u003c/sub\u003e for Gly) of residues \u003cem\u003ei\u003c/em\u003e and \u003cem\u003ej\u003c/em\u003e, and there is a contact between these two residues if \u003cem\u003ed\u003c/em\u003e\u003csub\u003e\u003cem\u003eij\u003c/em\u003e\u003c/sub\u003e is shorter than a criterion (e.g., 8 \u0026Aring;)\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. A residue interaction network (RIN)\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e can be constructed based on the contact map in which all residues are regarded as vertices and the contacts as edges. Here, we used a modified approach to define \u003cem\u003ed\u003c/em\u003e\u003csub\u003e\u003cem\u003eij\u003c/em\u003e\u003c/sub\u003e as the shortest distance between non-hydrogen atoms of two residues are measured, and a cutoff of 3.5 \u0026Aring; is used to identify contacts. This approach would avoid potential false contact assignments, see Methods and Fig. S6 in the SI. The distance map of system 1 obtained from the modified approach is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea, which has very similar patterns as the predicted aligned error (PAE) map provided by AF2, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb. The RIN for system 1, in which the cofactors (BVQ, SF4-1 and SF4-2) and the substrate (PFOA) were treated as individual vertices, constructed from the final snapshot of the 300 ns MD is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec.\u003c/p\u003e \u003cp\u003eTo capture the dynamics of the RIN, we constructed the RIN every 1 ns from the last 100 ns MD and monitored the residues that interact with the BVQ cofactor, both SF4 clusters, as well as the PFOA substrate. The distributions of these residues are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e. We observed more residues interact with the BVQ than those interact with the two SF4 clusters and PFOA combined. The motifs involved in BVQ binding include short helical segments H2, H3, H4, and longer helices H12, H13, H16, and H17; two strands from β-sheet A and the β-hairpin C are involved in BVQ binding (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec for all motif names). The β-hairpin C also interacts with SF4-1 at the loop region via two positively charged residues H259 and K258, as Fe\u003csub\u003e4\u003c/sub\u003eS\u003csub\u003e4\u003c/sub\u003eCys\u003csub\u003e4\u003c/sub\u003e carries negative charge (-1 for oxidized and \u0026minus;\u0026thinsp;2 for reduced states, respectively). In this simulation we used a reduced SF4-1, an oxidized SF4-2 and a reduced BVQ (with Co\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), see methods. H16 is involved in the interactions for BVQ, SF4-2 and PFOA. In particular, the aromatic residue W343 that is conserved in other T7RdhA-like proteins (Fig. S2) shows interactions with BVQ (82%), PFOA (80%) and SF4-2 (15%). Y213 from H12 has been considered to mediate the reductive dehalogenation in PceA (Y246 of \u003cem\u003eS. multivorans\u003c/em\u003e PceA)\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e, and its interactions with BVQ (98%) and PFOA (40%) may be needed for potential defluorination. F47 from a small helix H4 also interacts with BVQ (100%) and PFOA (90%). We noticed that the residues interact with PFOA (\u0026gt;\u0026thinsp;50%) are either aromatic (Y68, Y65, F47, W343, F64, W93, F340) or positively charged (R89), which may be a unique feature of the binding pocket for PFAS substrates.\u003c/p\u003e \n \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAlphaFold2 correctly predicts cofactor/ligand binding in T7RdhA\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAF2 opens an avenue in biology on which the functions and interactions mediated by proteins can be understood with the assistance of highly accurate atomic models. However, the structures predicted by AF2, either single-chain monomers or multi-chain oligomers, are in apo-forms, i.e., unbounded form. Even the necessary solvents are missing in the structures predicted by AF2. Cofactors play an important, sometimes essential role in protein folding and functions\u003csup\u003e40\u003c/sup\u003e. Folding and functions of proteins may also be assisted by the substrate that they bind\u003csup\u003e41\u003c/sup\u003e. We asked how reliable are the AF2 models in depicting the structures and dynamics of proteins upon cofactor and/or ligand binding? This is a critical question to answer for protein systems with cofactor/substrate, especially for understanding the interactions among them, as well as for protein-protein interactions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003ePrevious publications discussed the above question on ligand binding\u003csup\u003e42\u003c/sup\u003e, peptide binding\u003csup\u003e43\u003c/sup\u003e, and protein-protein interactions\u003csup\u003e44\u003c/sup\u003e. In this work, the functional T7RdhA structure incorporates the natural corrinoid norpseudo cobalamin (BVQ)\u003csup\u003e45\u003c/sup\u003e, together with two [4Fe4S] iron-sulfur clusters (SF4-1 and SF4-2), which is known as the \u0026ldquo;Nature\u0026rsquo;s modular structures\u0026rdquo;\u003csup\u003e46\u003c/sup\u003e. We showed that when cofactors (BVQ/SF4) and substrate (PFOA) are present in the correct pockets, the residue flexibility calculated from molecular dynamics simulations can best describe the AF2-scores by AlphaFold2, which is an inverse normalization of the pLDDT scores. In the complex model, the residue distance map also mirrors the predicted aligned error map by AlphaFold2. Our results indicate that the AF2 structures already have the pre-built pockets for the correct cofactors and ligands. We also showed that multiple AF2 structures (90 T7RdhA models in the present work) can also capture the protein dynamics. The diversity of protein structures, in our opinion, originates from protein dynamics and can be recaptured by AF2 in the structure modeling.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eA processual view of protein structure-function relationships\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe protein function is determined by the protein structure. However, a static protein does not perform the function without dynamics and interactions. The processual nature of reality\u003csup\u003e47\u003c/sup\u003e applies to all biomolecules, including proteins. We collected the structures of different systems (System1-5 in Fig.2) during the MD simulation, together with selected AF2 models (System 6), and compared these snapshots in Fig. 5.\u003c/p\u003e\n\u003cp\u003eThe structures of all models after 300 ns MD are similar to the original AF2 model (Fig.5b). Without MD trajectories and the residue flexibilities, it would be difficult to tell which system has a ligand or ligands. Our results support that the protein interactions and functions are based on their intrinsic processual nature\u003csup\u003e47\u003c/sup\u003e. For other AF2 models such as those of missense mutations\u003csup\u003e3\u003c/sup\u003e, it might not be fair to make a judgement based on a static configuration.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eImplications for PFAS biodegradation \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe persistence and accumulation of per- and polyfluoroalkyl substances, or PFASs, in the environment, and their adverse effects on human health have led to the current global concern\u003csup\u003e48\u003c/sup\u003e. The T7RdhA sequence is highly similar to the partial sequence of A6RdhA from the \u003cem\u003eAcidimicrobiaceae sp. A6\u003c/em\u003e which degrade both PFOA and PFOS under anaerobic conditions\u003csup\u003e28,49\u003c/sup\u003e. Nevertheless, the full A6RdhA sequence and the defluorination mechanisms remain unclear. From the structure modeling and MD simulations, we confirmed the participation of both corrinoid cofactors (BVQ) and iron-sulfur clusters (SF4) by experiment. The binding mode of the cofactors and the PFOA ligand have been identified using a dynamic residue interaction network from the MD trajectories. We also showed that AF2 combined with MD simulation can help to identify proteins with targeted functions such as PFAS bioremediation.\u003c/p\u003e"},{"header":"Methods And Materials","content":"\u003cp\u003e\u003cstrong\u003e\u003cem\u003eMulti-sequence alignment \u0026amp; sequence similarity network\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA combined A6RdhA/T7RdhA Hidden Markov Model (HMM) was constructed from 529 non-redundant similar sequences identified via blastp from the NCBI and UniProt KB databases. \u0026nbsp;Briefly, these sequences were identified to clade together (with a consensus support value of 100) with A6RdhA/T7RdhA in an amino acid tree using MAFFT v7.453)\u003csup\u003e50\u003c/sup\u003e multi-sequence alignment, and were then used to construct an HMM profile using the program HMMer (v3.3.2)\u003csup\u003e51\u003c/sup\u003e. The first portion of the NCBI non-redundant database, nr00 (8,812,511 sequences), was queried using this HMM profile using the HMMer default threshold values. \u0026nbsp;The resulting 1279 (including T7RdhA) sequences were submitted to the EFI (Enzyme Function Initiative) enzyme similarity tool for generation of the sequence similarity network (SSN) with evalues \u0026le; 10\u003csup\u003e-5\u003c/sup\u003e and an alignment cutoff of 20 corresponding to an id% of ~30.\u003csup\u003e52\u003c/sup\u003e Network clustering and the T7RdhA clique identification was performed using the \u003cem\u003eigraph\u003c/em\u003e package in R.\u003csup\u003e53\u003c/sup\u003e The multi-sequence alignment by MAFFT was visualized using WebLogo (v3.6.0)\u003csup\u003e54\u003c/sup\u003e. TM-align\u003csup\u003e55\u003c/sup\u003e was used for structure alignment and RMSD calculations. The calculated RMSD matrix was converted to phylogeny using the \u003cem\u003eape\u003c/em\u003e package in R\u003csup\u003e56\u003c/sup\u003e, and visualized by Mega-X\u003csup\u003e57\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAlphaFold 2 structure predictions\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe T7RdhA model used in the MD simulation and other T7RdhA-like proteins models (all 39 models in the SSN shown in Fig. S1) were constructed using AlphaFold2 V2.0.1 (installed in July 2021). 90 more T7RdhA models for the protein-structure-based RMSF profile in Fig. 2 (system 6), were built by a newer version of AlphaFold2 (V2.2.2, installed in July 2022)\u003csup\u003e1\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eMolecular dynamics simulations\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe molecular dynamics simulations were performed using NAMD.\u003csup\u003e58\u003c/sup\u003e The CHARMM force field (c36m)\u003csup\u003e59,60\u003c/sup\u003e was employed for the protein and a modified TIP3P model\u003csup\u003e61\u003c/sup\u003e for the solvent water molecules. The CHARMM-format force field parameters of norpseudo-B12 (BVQ)\u003csup\u003e62\u003c/sup\u003e and [4Fe4S] iron sulfur cluster (SF4)\u003csup\u003e63\u003c/sup\u003e under different redox states have been adopted. The force field parameters of the PFOA molecule were derived from the TEAM (Transferable, Extensible, Accurate and Modular) force field in the Direct Force Field (DFF, v7.2)\u003csup\u003e64\u003c/sup\u003e software, and have been listed in the Appendix of the supplementary materials.\u003c/p\u003e\n\u003cp\u003eThe BVQ and SF4 cofactors in the crystallographic structures of PceA (e.g., 4UQU\u003csup\u003e29\u003c/sup\u003e) can be superimposed very well, with the eight Cys residues precisely bound the SF4 iron atoms. All hydrogen atoms have been added using the HBuild function of CHARMM\u003csup\u003e65\u003c/sup\u003e. The covalent bond between SF4 cofactors and their binding Cys residues were generated using the Patch function of CHARMM\u003csup\u003e65\u003c/sup\u003e. The whole system was put in a solvent box with H\u003csub\u003e2\u003c/sub\u003eO molecules added at least 15 Å to the edge of the protein system. The solvation and neutralization (using Na+ and/or Cl-) were carried out by the Solvate and Autoionization packages of VMD\u003csup\u003e66\u003c/sup\u003e. A reduced BVQ (Co(I)), oxidized SF4-2 (the proximal SF4, Fe\u003csub\u003e4\u003c/sub\u003eS\u003csub\u003e4\u003c/sub\u003e(Cys)\u003csub\u003e4\u003c/sub\u003e\u003csup\u003e-\u003c/sup\u003e) and reduced SF4-1 (Fe\u003csub\u003e4\u003c/sub\u003eS\u003csub\u003e4\u003c/sub\u003e(Cys)\u003csub\u003e4\u003c/sub\u003e\u003csup\u003e2-\u003c/sup\u003e) were used in the MD simulations.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAfter solvation and neutralization, the whole system was optimized by 50,000 steps under 0 K. Then the temperature of the system was \u0026ldquo;naturally\u0026rdquo; increased to 300 K with a rate of 0.001 K/timestep. A constant-pressure, constant-temperature (NPT) ensemble was used in the MD simulation with the system pressure of 1 atm and temperature of 300 K maintained by the Langevin piston controls. The SHAKE algorithm was applied to fix the bond lengths involving hydrogen atoms and a timestep of 2 fs was used for the simulations. The nonbonded interaction cutoff switching was set as between 9 and 11 Å. For the long-range interactions, the particle mesh Ewald summation with a grid spacing of 1.35 was applied. 310 ns MD simulations were performed for all systems (Fig. 2) and the last 100 ns were taken for further analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eResidue interaction network\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe residue interaction network (RIN) or contact map of a protein was based on the distance map with a criterion\u003csup\u003e38,39\u003c/sup\u003e. A common approach, for example, is to measure the C\u003csub\u003e\u0026beta;\u003c/sub\u003e-C\u003csub\u003e\u0026beta;\u003c/sub\u003e distances (C\u003csub\u003ea\u003c/sub\u003e for Gly) and if the measured distance between residues \u003cem\u003eR\u003csub\u003ei\u003c/sub\u003e\u003c/em\u003e and \u003cem\u003eR\u003csub\u003ej\u003c/sub\u003e\u003c/em\u003e is shorter than 8 \u0026Aring; then there is a contact between \u003cem\u003eR\u003csub\u003ei\u003c/sub\u003e\u003c/em\u003e and \u003cem\u003eR\u003csub\u003ej\u003c/sub\u003e\u003c/em\u003e. This approach, however, we found may lead to incorrect assignment (Fig. S6). We adopted an alternative approach. Considering the hydrogen bond interaction X-H\u0026hellip;Y (where X/Y can be C, N, O, S in proteins), the distance between X and Y for a typical H-bonds are in the range of 2 to 3 \u0026Aring;, and is ~3.5 \u0026Aring; for a C-H\u0026hellip;O hydrogen bond in protein\u003csup\u003e67\u003c/sup\u003e. Here, for residues (vertices) \u003cem\u003eR\u003csub\u003ei\u003c/sub\u003e\u003c/em\u003e and \u003cem\u003eR\u003csub\u003ej\u003c/sub\u003e\u003c/em\u003e we define the distance \u003cem\u003ed\u003csub\u003eij\u003c/sub\u003e\u003c/em\u003e as the shortest distance between all heavy atoms. The distance map under this approach (Fig. 3a) agree well with the PAE map predicted by AF2 (Fig. 3b). The contact map is further defined based on the distance map: if \u003cem\u003ed\u003csub\u003eij\u003c/sub\u003e\u003c/em\u003e is shorter than 3.5 \u0026Aring;, we define an interaction (edge) between \u003cem\u003eR\u003csub\u003ei\u003c/sub\u003e\u003c/em\u003e and \u003cem\u003eR\u003csub\u003ej\u003c/sub\u003e\u003c/em\u003e. We then construct a binary adjacency matrix (1 for interaction and 0 for non-interaction) based this definition. The network analysis was performed using the \u003cem\u003eigraph\u003c/em\u003e\u003csup\u003e53\u003c/sup\u003e package in R. The distance analysis was performed using the \u003cem\u003ebio3d\u003c/em\u003e\u003csup\u003e68\u003c/sup\u003e package in R. The BVQ cofactor, the SF4-1 and SF4-2 clusters, and the PFOA substrate was treated as a residue (vertex) in the RIN.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eLigand binding\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAutoDock Vina (V1.2.0)\u003csup\u003e69\u003c/sup\u003e was used for ligand docking. Using the T7RdhA-BVQ-SF4 system (system 5), after 10 ns MD equilibration, the PFOA ligand was docked into the protein complex (solvent and ions removed), and the top-score ligand was used to construct system 1. The force field parameters of the PFOA ligand can be found in the Appendix in SI.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAuthor contributions\u003c/em\u003e\u003c/strong\u003e:\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eConceptualization, H.-B.G., R.B.; methodology, H.-B.G., V.V., S.F., P.D., R.B.;, validation, all authors; writing\u0026mdash;original draft preparation, H.-B.G.; writing\u0026mdash;review and editing, all authors; project administration. N.K.-L., R.B.; funding acquisition, N.K.-L., R.B. All authors have read the final revision of this manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAcknowledgements\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by funding from the OUSD (R\u0026amp;E) ARAP Program. The structural modeling and MD simulations were performed using the DoD HPC. We appreciate Dr. Jerry Parks for the cobalamin force field parameters, and Dr. Marcel Swart for providing the [4Fe4S] iron sulfur (SF4) cluster force field parameters. We thank Dr. Peter Jaffe for helpful discussions on the \u003cem\u003eAcidimicrobiaceae sp. A6\u003c/em\u003e organism and the mechanism of the A6RdhA enzyme. We appreciate the MAPS TEAM from the Summit Country Day School for the 3D print of the T7RdhA-ligand model (Fig. S7).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eData Availability\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data generated or analyzed during this study are included in this published article and its supplementary information files. R codes for data analyses and visualizations are available upon request (H.-B.G., [email protected])\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eCompeting Interests\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eJumper, J. Evans, R.; Pritzel, A.; Green, T.; Figurnov, M.; Ronneberger, O.; Tunyasuvunakool, K.; Bates, R.; Zidek, A.; Potapenko, A.; Bridgland, A.; Meyer, C.; Kohl, S. A. A.; Ballard, A. J.; Cowie, A.; Romera-Paredes, B.; Nikolov, S.; Jain, R.; Adler, J.; Back, T.; Petersen, S.; Reiman, D.; Clancy, E.; Zielinski, M.; Steinegger, M.; Pacholska, M.; Berghammer, T.; Bodenstein, S.; Silver, D.; Vinyals, O.; Senior, A. W.; Kavukcuoglu, K.; Kohli, P.; Hassabis, D. Highly accurate protein structure prediction with AlphaFold. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e2021\u003c/strong\u003e, \u003cem\u003e596\u003c/em\u003e, 583-589.\u003c/li\u003e\n\u003cli\u003eBaek, M., Dimaio, F.; Anishchenko, I.; Dauparas, J.; Ovchinnikov, S.; Lee, G. R.; Wang, J.; Cong, Q.; Kinch, L. N.; Schaeffer, R. D.; Millan, C.; Park, H.; Adams, C.; Glassman, C. R.; Degiovanni, A.; Pereira, J. H.; Rodrigues, A. V.; van Dijk, A. A.; Ebrecht, A. C.; Opperman, D. J.; Sagmeister, T.; Buhlheller, C.; Pavkov-Keller, T.; Rathinaswamy, M. K.; Dalwadi, U.; Yip, C. K.; Burke, J. E.; Garcia, K. C.; Grishin, N. V.; Adams, P. D.; Read, R. J.; Baker, D. Accurate prediction of protein structures and interactions using a three-track neural network. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e2021\u003c/strong\u003e, \u003cem\u003e373\u003c/em\u003e, 871–876.\u003c/li\u003e\n\u003cli\u003eBuel, G. R.; Walters, K. J. Can alphafold2 predict the impact of missense mutations on structure? \u003cem\u003eNature Struct. Mol. Biol\u003c/em\u003e. \u003cstrong\u003e2022\u003c/strong\u003e, \u003cem\u003e29\u003c/em\u003e, 1-2.\u003c/li\u003e\n\u003cli\u003ePak, M. A.; Markhieva, K. A.; Novikova, M. S.; Petrov, D. S.; Vorobyev, I. S.; Maksimova, E. S.; Kondrashov, F. A.; Ivankov, D. N. Using AlphaFold to predict the impact of single mutations on protein stability and function. bioRxiv \u003cstrong\u003e2021\u003c/strong\u003e, https://doi.org/10.1101/2021.09.19.460937.\u003c/li\u003e\n\u003cli\u003eMcBride, J. M.; Polev, K.; Reinharz, V.; Grzybowski, B. A.; Tlusty, T. AlphaFold2 can predict structural and phenotypic effects of single mutations. arXiv \u003cstrong\u003e2022\u003c/strong\u003e :2004.068860[q-bio.BM], https://arxiv.org/abs/2204.06860.\u003c/li\u003e\n\u003cli\u003eSchmidt, A.; Roner, S.; Mai, K.; Klinkhammer, H.; Kircher, M.; Ludwig, K. U. Predicting the pathogenicity of missense variants using features derived from AlphaFold2. bioRxiv 2022, https://doi.org/10.1101/2022.03.05.483091.\u003c/li\u003e\n\u003cli\u003eBagdonas, H.; Fogarty, C. A.; Fadda, E.; Agirre, J. The case for post-predictional modifications in the AlphaFold protein structure database. \u003cem\u003eNature Struct. Mol. Biol\u003c/em\u003e. \u003cstrong\u003e2021\u003c/strong\u003e, \u003cem\u003e28\u003c/em\u003e, 869-870.\u003c/li\u003e\n\u003cli\u003eGuo, H.-B.; Perminov, A.; Bekele, S.; Kedziora, G.; Farajollahi, S.; Varaljay, V.; Hinkle, K.; Molinero, V.; Meister, K.; Hung, C.; Dennis, P.; Kelley-Loughnane, N.; Berry, R. AlphaFold2 models indicate that protein sequence determines both structure and dynamics. \u003cem\u003eSci. Rep\u003c/em\u003e. \u003cstrong\u003e2022\u003c/strong\u003e, \u003cem\u003e12\u003c/em\u003e, 10696.\u003c/li\u003e\n\u003cli\u003eKendrew, J. C.; Bodo, G.; Dintzis, H. M.; Parrish, R. G.; Wyckoff, H.; Phillips, D. C. A three-dimensional model of the myoglobin molecule obtained by X-ray analysis. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e1958\u003c/strong\u003e, \u003cem\u003e181\u003c/em\u003e, 662-666.\u003c/li\u003e\n\u003cli\u003eFischer, E. Einfluss der configuration auf die wirkung der enzyme. \u003cem\u003eBer. Dtsch. Chem. Ges\u003c/em\u003e. \u003cstrong\u003e1894\u003c/strong\u003e, \u003cem\u003e27\u003c/em\u003e, 2985-2993.\u003c/li\u003e\n\u003cli\u003eKoshland, D. E. Jr. Application of a theory of enzyme specificity to protein synthesis. \u003cem\u003eProc. Natl. Acad. Sci. U. S. A.\u003c/em\u003e \u003cstrong\u003e1958\u003c/strong\u003e, \u003cem\u003e44\u003c/em\u003e, 98-104.\u003c/li\u003e\n\u003cli\u003eKoshland, D. E. Jr. The key-lock theory and the induced fit theory, \u003cem\u003eAngew. Chem. Int. Ed. Engl\u003c/em\u003e. \u003cstrong\u003e1994\u003c/strong\u003e, \u003cem\u003e33\u003c/em\u003e, 2375-2378.\u003c/li\u003e\n\u003cli\u003eMonod, J.; Wyman, J.; Changeux, J.-P. On the nature of allosteric transitions: a plausible model. \u003cem\u003eJ. Mol. Biol\u003c/em\u003e. \u003cstrong\u003e1965\u003c/strong\u003e, \u003cem\u003e12\u003c/em\u003e, 88-118.\u003c/li\u003e\n\u003cli\u003eChangeux, J.-P.; Edelstein., S. Conformational selection or induced fit? 50 years of debate resolved. \u003cem\u003eF1000 Biol. Rep.\u003c/em\u003e \u003cstrong\u003e2011\u003c/strong\u003e, \u003cem\u003e3\u003c/em\u003e, 19.\u003c/li\u003e\n\u003cli\u003eNajmanovich, R. J. Evolutionary studies of ligand binding sites in proteins. \u003cem\u003eCurr. Opinion Struct. Biol\u003c/em\u003e. \u003cstrong\u003e2017\u003c/strong\u003e, \u003cem\u003e45\u003c/em\u003e, 85-90.\u003c/li\u003e\n\u003cli\u003eSkolnick, J.; Gao, M. Interplay of physics and evolution in the likely origin of protein biochemical function. \u003cem\u003eProc. Natl. Acad. Sci. U. S. A\u003c/em\u003e. \u003cstrong\u003e2013\u003c/strong\u003e, \u003cem\u003e110\u003c/em\u003e, 9344-9349.\u003c/li\u003e\n\u003cli\u003eLu, H.; Diaz, D. J.; Czarnecki, N. J.; Zhu, C.; Kim, W.; Shroff, R.; Acosta, D. J.; Alexander, B. R.; Cole, H. O.; Zhang, Y.; Lynd, N. A.; Ellington, A. D.; Alper, H. S. Machine learning-aided engineering of hydrolases for PET depolymerization. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e2022\u003c/strong\u003e, \u003cem\u003e604\u003c/em\u003e, 662-667.\u003c/li\u003e\n\u003cli\u003eWackett, L. P.; Robinson, S. L. The ever-expending limits of enzyme catalysis and biodegradation: polyaromatic, poly chlorinated, polyfluorinated and polymeric compounds. \u003cem\u003eBiochem. J\u003c/em\u003e. \u003cstrong\u003e2020\u003c/strong\u003e, \u003cem\u003e477\u003c/em\u003e, 2875-2891.\u003c/li\u003e\n\u003cli\u003eTunyasuvunakool, K.; Adler, J.; Wu, Z.; Green, T.; Zielinski, M.; Zidek, A.; Bridgland, A.; Cowie, A.; Meyer, C.; Laydon, A.; Velankar, S.; Kleywegt, G. J.; Bateman, A.; Evans, R.; Pritzel, A.; Figurnov, M.; Ronneberger, O.; Bates, R.; Kohl, S. A. A.; Potapenko, A.; Ballard, A. J.; Romera-Paredes, B.; Nikolov, S.; Jain, R.; Clancy, E.; Reiman, D.; Petersen, S.; Senior, A. W.; Kavukcuoglu, K.; Birney, E.; Kohli, P.; Jumper, J.; Hassabis, D. Highly accurate protein structure prediction for the human proteome. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e2021\u003c/strong\u003e, \u003cem\u003e596\u003c/em\u003e, 590-596.\u003c/li\u003e\n\u003cli\u003eVaradi, M. Anyango, S.; Deshpande, M.; Nair, S.; Natassia, C.; Yordanova, G.; Yuan, D.; Stroe, O.; Wood, G.; Laydon, A.; Zidek, A.; Green, T.; Tunyasuvunakool, K.; Petersen, S.; Jumper, J.; Clancy, E.; Green, R.; Vora, A.; Lutfi, M.; Figurnov, M.; Cowie, A.; Hobbs, N.; Kohli, P.; Kleywegt, G.; Birney, E.; Hassabis, D.; Velankar, S. AlphaFold Protein Structure Database: Massively expanding the structural coverage of protein-sequence space with high-accuracy models. \u003cem\u003eNucleic Acids Res\u003c/em\u003e. \u003cstrong\u003e2022\u003c/strong\u003e, \u003cem\u003e50\u003c/em\u003e, D439–D444.\u003c/li\u003e\n\u003cli\u003eBerman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.; Bhat, T. N.; Weissig, H.; Shindyalov, I. N.; Bourne, P. E. The protein data bank. \u003cem\u003eNucleic Acids Res\u003c/em\u003e. \u003cstrong\u003e2000\u003c/strong\u003e, \u003cem\u003e28\u003c/em\u003e, 235-242.\u003c/li\u003e\n\u003cli\u003eGoodsell, D. S.; Zardecki, C.; Costanzo, L. D.; Duarte, J. M.; Hudson, B. P.; Persikova, I.; Segura, J.; Shao, C.; Voigt, M.; Westbrook, J. D.; Young, J. Y.; Burley, S. K. RCSB protein data bank: enabling biomedical research and drug discovery. \u003cem\u003eProtein Sci\u003c/em\u003e. \u003cstrong\u003e2020\u003c/strong\u003e, \u003cem\u003e29\u003c/em\u003e, 52-65.\u003c/li\u003e\n\u003cli\u003eDuarte, J. M.; Dutta, S.; Goodsell, D. S.; Burley, S. K. Exploring protein symmetry at the RCSB protein data bank. Emerging Topics Life Sci. \u003cstrong\u003e2022\u003c/strong\u003e, https://doi.org/10.1042/ETLS20210267.\u003c/li\u003e\n\u003cli\u003eMukhopadhyay, A.; Borkakoti, N.; Pravda, L.; Tyzack, J. D.; Thornton, J. M.; Velankar, S. Finding enzyme cofactors in protein data bank. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e2019\u003c/strong\u003e, \u003cem\u003e35\u003c/em\u003e, 3510-3511.\u003c/li\u003e\n\u003cli\u003eStierand, S.; Rarey, M. Drawing the PDB: protein-ligand complexes in two dimensions. \u003cem\u003eACS Med. Chem. Lett\u003c/em\u003e. \u003cstrong\u003e2010\u003c/strong\u003e, \u003cem\u003e1\u003c/em\u003e, 540-545.\u003c/li\u003e\n\u003cli\u003eEvans, R.; O’Neill, M.; Pritzel, A.; Antropova, N.; Senior, A.; Green, T.; Zidek, A.; Bates, R.; Blackwell, S.; Yim, J.; Ronneberger, O.; Bodenstein, S.; Zielinski, M.; Bridgland, A.; Potapenko, A.; Cowie, A.; Tunyasuvunakool, K.; Jain, R.; Clancy, E.; Kohli, P.; Jumper, J.; Hassabis, D. Protein complex prediction with AlphaFold-Multimer. bioRxiv 2021, https://doi.org/10.1101/2021.10.04.463034\u003c/li\u003e\n\u003cli\u003eGao, M.; An, D. N.; Parks, J. M.; Skolnick, J. AF2Complex predicts physical interactions in multimeric proteins with deep learning. \u003cem\u003eNature Commun\u003c/em\u003e. \u003cstrong\u003e2022\u003c/strong\u003e, \u003cem\u003e13\u003c/em\u003e, 1744.\u003c/li\u003e\n\u003cli\u003eHuang, S., and Jaffé, P. R. Defluorination of Perfluorooctanoic Acid (PFOA) and Perfluorooctane Sulfonate (PFOS) by Acidimicrobium sp. Strain A6. \u003cem\u003eEnviron. Sci. Technol\u003c/em\u003e. \u003cstrong\u003e2019\u003c/strong\u003e, \u003cem\u003e53\u003c/em\u003e, 11410–11419.\u003c/li\u003e\n\u003cli\u003eBommer, M., Kunze, C., Fesseler, J., Schubert, T., Diekert, G., and Dobbek, H. Structural basis for organohalide respiration. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e2014\u003c/strong\u003e, \u003cem\u003e346\u003c/em\u003e, 455–458.\u003c/li\u003e\n\u003cli\u003ePayne, K. A., Quezada, C. P., Fisher, K., Dunstan, M. S., Collins, F. A., Sjuts, H., Levy, C., Hay, S., Rigby, S. E., and Leys, D. Reductive dehalogenase structure suggests a mechanism for B12-dependent dehalogenation. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e2015\u003c/strong\u003e, \u003cem\u003e517\u003c/em\u003e, 513–516.\u003c/li\u003e\n\u003cli\u003eTully, B. J., Sachdeva, R., Graham, E. D., and Heidelberg, J. F. 290 metagenome-assembled genomes from the Mediterranean Sea: a resource for marine microbiology. \u003cem\u003ePeerJ\u003c/em\u003e \u003cstrong\u003e2017\u003c/strong\u003e, \u003cem\u003e5\u003c/em\u003e, e3558.\u003c/li\u003e\n\u003cli\u003eStank, A.; Kokh, D. B.; Fuller, J. C.; Wade, R. C. Protein binding pocket dynamics. \u003cem\u003eAcc. Chem. Res\u003c/em\u003e. \u003cstrong\u003e2016\u003c/strong\u003e, \u003cem\u003e49\u003c/em\u003e, 809-815.\u003c/li\u003e\n\u003cli\u003ePruitt, K. D., Tatusova, T., and Maglott, D. R. NCBI Reference Sequence (RefSeq): a curated non-redundant sequence database of genomes, transcripts and proteins. \u003cem\u003eNucleic Acids Res\u003c/em\u003e. \u003cstrong\u003e2005\u003c/strong\u003e, \u003cem\u003e33\u003c/em\u003e, D501–D504.\u003c/li\u003e\n\u003cli\u003eCooper, C. J., Zheng, K., Rush, K. W., Johs, A., Sanders, B. C., Pavlopoulos, G. A., Kyrpides, N. C., Podar, M., Ovchinnikov, S., Ragsdale, S. W., and Parks, J. M. Structure determination of the HgcAB complex using metagenome sequence data: insights into microbial mercury methylation. \u003cem\u003eCommun Biol\u003c/em\u003e. \u003cstrong\u003e2020\u003c/strong\u003e, \u003cem\u003e3\u003c/em\u003e, 320.\u003c/li\u003e\n\u003cli\u003eHalliwell, T., Fisher, K., Payne, K. A. P., Rigby, S. E. J., and Leys, D. Catabolic reductive dehalogenase substrate complex structures underpin rational repurposing of substrate scope. \u003cem\u003eMicroorganisms\u003c/em\u003e \u003cstrong\u003e2020\u003c/strong\u003e, 8, 1344.\u003c/li\u003e\n\u003cli\u003eKunze, C., Bommer, M., Hagen, W. R., Uksa, M., Dobbek, H., Schubert, T., and Diekert, G. Cobamide-mediated enzymatic reductive dehalogenation via long-range electron transfer. \u003cem\u003eNature Commun\u003c/em\u003e. \u003cstrong\u003e2017\u003c/strong\u003e, \u003cem\u003e8\u003c/em\u003e, 15858.\u003c/li\u003e\n\u003cli\u003eLaskowski, R. A.; Thornton, J. M. PDBsum extras: SARS-CoV-2 and AlphaFold models. \u003cem\u003eProt. Sci\u003c/em\u003e. \u003cstrong\u003e2022\u003c/strong\u003e, \u003cem\u003e31\u003c/em\u003e, 283-289.\u003c/li\u003e\n\u003cli\u003eDu, Z.; Su, H.; Wang, W.; Ye, L.; Wei, Hong; Peng, Z; Anishchenko, I.; Baker, D.; Yang, J. The trRosetta server for fast and accurate protein structure prediction. \u003cem\u003eNature Protocols\u003c/em\u003e \u003cstrong\u003e2022\u003c/strong\u003e, \u003cem\u003e16\u003c/em\u003e, 5634-5651.\u003c/li\u003e\n\u003cli\u003eClementel, D.; Del Conte, A.; Monzon, A. M.; Camagni, G. F.; Minervini, G.; Piovesan, D.; Tosatto, S. C. E. Ring 3.0: fast generation of probabilistic residue interaction networks from structural ensembles. \u003cem\u003eNucleic Acids Res\u003c/em\u003e. \u003cstrong\u003e2022\u003c/strong\u003e, \u003cem\u003e50\u003c/em\u003e, W651-W656.\u003c/li\u003e\n\u003cli\u003eWittung-Stafshede, P. Role of cofactors in protein folding. \u003cem\u003eAcc. Chem. Res\u003c/em\u003e. \u003cstrong\u003e2002\u003c/strong\u003e, \u003cem\u003e35\u003c/em\u003e, 201-208.\u003c/li\u003e\n\u003cli\u003eXu, Q.; Guo, H.-B.; Wlodawer, A.; Guo, H. The importance of dynamics in substrate-assisted catalysis and specificity. \u003cem\u003eJ. Am. Chem. Soc\u003c/em\u003e. \u003cstrong\u003e2006\u003c/strong\u003e, \u003cem\u003e128\u003c/em\u003e, 5994-5995.\u003c/li\u003e\n\u003cli\u003eHekkelman, M. L.; de Vries, I.; Joosten, R. P.; Perrakis, A. AlphaFill: enriching the AlphaFold models with ligands and co-factors. bioRxiv \u003cstrong\u003e2021\u003c/strong\u003e, https://doi.org/10.1101/2021.11.26.470110.\u003c/li\u003e\n\u003cli\u003eChang, L.; Perez, A. AlphaFold encodes the principles to identify high affinity peptide binders. bioRxiv \u003cstrong\u003e2022\u003c/strong\u003e, https://doi.org/10.1101/2022.03.18.484931.\u003c/li\u003e\n\u003cli\u003eBryant, P.; Pozzati, G.; Elofsson, A. Improved prediction of protein-protein interactions using AlphaFold2. \u003cem\u003eNature Commun\u003c/em\u003e. \u003cstrong\u003e2022\u003c/strong\u003e, \u003cem\u003e13\u003c/em\u003e, 1265.\u003c/li\u003e\n\u003cli\u003eKrautler, B.; Fieber, W.; Ostermann, S.; Fasching, M.; Ongania, K.-H.; Gruber, K.; Kratky, C.; Mikl, C.; Siebert, A.; Diekert, G. The cofactor of tetrachloroethene reductive dehalogenase of Dehalospirillum multivorans is norpseudo-B12, a new type of natural corrinoid. \u003cem\u003eHelvetica\u003c/em\u003e \u003cstrong\u003e2003\u003c/strong\u003e, \u003cem\u003e86\u003c/em\u003e, 3698-3716.\u003c/li\u003e\n\u003cli\u003eBeinert, H., Holm, R. H., and Münck, E. Iron-sulfur clusters: nature’s modular, multipurpose structures. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e1997\u003c/strong\u003e, \u003cem\u003e277\u003c/em\u003e, 653–659.\u003c/li\u003e\n\u003cli\u003eJaeger, J.; Monk, N. Everything flows: a process perspective of life. \u003cem\u003eEMBO Rep\u003c/em\u003e. \u003cstrong\u003e2015\u003c/strong\u003e, \u003cem\u003e16\u003c/em\u003e, 1064-1067.\u003c/li\u003e\n\u003cli\u003eBuck, R. C., Franklin, J., Berger, U., Conder, J. M., Cousins, I. T., de Voogt, P., Jensen, A. A., Kannan, K., Mabury, S. A., and van Leeuwen, S. P. J. Perfluoroalkyl and polyfluoroalkyl substances in the environment: terminology, classification, and origins. Integr. \u003cem\u003eEnviron. Assess. Manag\u003c/em\u003e. \u003cstrong\u003e2011\u003c/strong\u003e, \u003cem\u003e7\u003c/em\u003e, 513–541.\u003c/li\u003e\n\u003cli\u003eRuiz-Uriguen, M.; Shuai, W.; Huang, S.; Jaffe, P. R. Biodegradation of PFOA in microbial electrolysis cells by \u003cem\u003eAcidimicrobiaceae sp. Strain A6\u003c/em\u003e. \u003cem\u003eChemosphere\u003c/em\u003e, \u003cstrong\u003e2022\u003c/strong\u003e, \u003cem\u003e292\u003c/em\u003e, 133506.\u003c/li\u003e\n\u003cli\u003eKatoh, K., and Standley, D. M. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. \u003cem\u003eMol. Biol. Evol\u003c/em\u003e. \u003cstrong\u003e2013\u003c/strong\u003e, \u003cem\u003e30\u003c/em\u003e, 772–780.\u003c/li\u003e\n\u003cli\u003eFinn, R. D., Clements, J., and Eddy, S. R. HMMER web server: interactive sequence similarity searching. \u003cem\u003eNucleic Acids Res\u003c/em\u003e. \u003cstrong\u003e2011\u003c/strong\u003e, \u003cem\u003e39\u003c/em\u003e, W29–W37.\u003c/li\u003e\n\u003cli\u003eZallot, R., Oberg, N., and Gerlt, J. A. The EFI Web Resource for Genomic Enzymology Tools: Leveraging Protein, Genome, and Metagenome Databases to Discover Novel Enzymes and Metabolic Pathways. \u003cem\u003eBiochemistry\u003c/em\u003e \u003cstrong\u003e2019\u003c/strong\u003e, \u003cem\u003e58\u003c/em\u003e, 4169–4182.\u003c/li\u003e\n\u003cli\u003eHan, W.-S., Lee, J., Pham, M.-D., and Yu, J. X. iGraph: a framework for comparisons of disk-based graph indexing techniques. \u003cem\u003eProceedings VLDB Endowment\u003c/em\u003e \u003cstrong\u003e2010\u003c/strong\u003e, \u003cem\u003e3\u003c/em\u003e, 449–459.\u003c/li\u003e\n\u003cli\u003eCrooks, G. E., Hon, G., Chandonia, J.-M., and Brenner, S. E. WebLogo: a sequence logo generator. \u003cem\u003eGenome Res\u003c/em\u003e. \u003cstrong\u003e2004\u003c/strong\u003e, \u003cem\u003e14\u003c/em\u003e, 1188–1190.\u003c/li\u003e\n\u003cli\u003eZhang, Y.; Skolnick, J. TM-align: a protein structure alignment algorithm based on the TM-score. \u003cem\u003eNucleic Acids Res\u003c/em\u003e. \u003cstrong\u003e2005\u003c/strong\u003e, \u003cem\u003e33\u003c/em\u003e, 2302-2309.\u003c/li\u003e\n\u003cli\u003eParadis, E.; Claude, J.; Strimmer, K. APE: analyses of phylogenetics and evolution in R language. \u003cem\u003eBioinformatics\u003c/em\u003e, \u003cstrong\u003e2004\u003c/strong\u003e, \u003cem\u003e20\u003c/em\u003e, 289-290.\u003c/li\u003e\n\u003cli\u003eKumar, S., Stecher, G., Li, M., Knyaz, C., and Tamura, K. MEGA X: Molecular Evolutionary Genetics Analysis across Computing Platforms. \u003cem\u003eMol. Biol. Evol\u003c/em\u003e. \u003cstrong\u003e2018\u003c/strong\u003e, \u003cem\u003e35\u003c/em\u003e, 1547–1549. \u003c/li\u003e\n\u003cli\u003ePhillips, J. C., Braun, R., Wang, W., Gumbart, J., Tajkhorshid, E., Villa, E., Chipot, C., Skeel, R. D., Kalé, L., and Schulten, K. Scalable molecular dynamics with NAMD. \u003cem\u003eJ. Comput. Chem\u003c/em\u003e. \u003cstrong\u003e2005\u003c/strong\u003e, \u003cem\u003e26\u003c/em\u003e, 1781–1802.\u003c/li\u003e\n\u003cli\u003eHuang, J., Rauscher, S., Nawrocki, G., Ran, T., Feig, M., de Groot, B. L., Grubmüller, H., and MacKerell, A. D., Jr. CHARMM36m: an improved force field for folded and intrinsically disordered proteins. Nat. Methods \u003cstrong\u003e2017\u003c/strong\u003e, \u003cem\u003e14\u003c/em\u003e, 71–73.\u003c/li\u003e\n\u003cli\u003eMacKerell, A. D., Bashford, D., Bellott, M., Dunbrack, R. L., Evanseck, J. D., Field, M. J., Fischer, S., Gao, J., Guo, H., Ha, S., Joseph-McCarthy, D., Kuchnir, L., Kuczera, K., Lau, F. T., Mattos, C., Michnick, S., Ngo, T., Nguyen, D. T., Prodhom, B., Reiher, W. E., Roux, B., Schlenkrich, M., Smith, J. C., Stote, R., Straub, J., Watanabe, M., Wiórkiewicz-Kuczera, J., Yin, D., and Karplus, M. All-atom empirical potential for molecular modeling and dynamics studies of proteins. \u003cem\u003eJ. Phys. Chem. B\u003c/em\u003e \u003cstrong\u003e1998\u003c/strong\u003e, \u003cem\u003e102\u003c/em\u003e, 3586–3616.\u003c/li\u003e\n\u003cli\u003ePrice, D. J., and Brooks, C. L., 3rd. A modified TIP3P water potential for simulation with Ewald summation. \u003cem\u003eJ. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e2004\u003c/strong\u003e, \u003cem\u003e121\u003c/em\u003e, 10096–10103.\u003c/li\u003e\n\u003cli\u003ePavlova, A., Parks, J. M., and Gumbart, J. C. Development of CHARMM-Compatible Force-Field Parameters for Cobalamin and Related Cofactors from Quantum Mechanical Calculations. \u003cem\u003eJ. Chem. Theory Comput\u003c/em\u003e. \u003cstrong\u003e2018\u003c/strong\u003e, \u003cem\u003e14\u003c/em\u003e, 784–798.\u003c/li\u003e\n\u003cli\u003eCarvalho, A. T. P., and Swart, M. Electronic structure investigation and parametrization of biologically relevant iron-sulfur clusters. \u003cem\u003eJ. Chem. Inf. Model\u003c/em\u003e. \u003cstrong\u003e2014\u003c/strong\u003e, \u003cem\u003e54\u003c/em\u003e, 613–620.\u003c/li\u003e\n\u003cli\u003eDirect Force Field 7.0, Aeon Technology, Inc. San Diego, \u003cstrong\u003e2011\u003c/strong\u003e.\u003c/li\u003e\n\u003cli\u003eBrooks, B. R., Brooks, C. L., 3rd, Mackerell, A. D., Jr, Nilsson, L., Petrella, R. J., Roux, B., Won, Y., Archontis, G., Bartels, C., Boresch, S., Caflisch, A., Caves, L., Cui, Q., Dinner, A. R., Feig, M., Fischer, S., Gao, J., Hodoscek, M., Im, W., Kuczera, K., Lazaridis, T., Ma, J., Ovchinnikov, V., Paci, E., Pastor, R. W., Post, C. B., Pu, J. Z., Schaefer, M., Tidor, B., Venable, R. M., Woodcock, H. L., Wu, X., Yang, W., York, D. M., and Karplus, M. CHARMM: the biomolecular simulation program. \u003cem\u003eJ. Comput. Chem\u003c/em\u003e. \u003cstrong\u003e2009\u003c/strong\u003e, \u003cem\u003e30\u003c/em\u003e, 1545–1614.\u003c/li\u003e\n\u003cli\u003eHumphrey, W., Dalke, A., and Schulten, K. VMD: visual molecular dynamics. \u003cem\u003eJ. Mol. Graph\u003c/em\u003e. \u003cstrong\u003e1996\u003c/strong\u003e, \u003cem\u003e14\u003c/em\u003e, 33–38.\u003c/li\u003e\n\u003cli\u003eGuo, H.-B.; Beahm, R. F., Guo, H. Stabilization and destabilization of the C\u003csub\u003ed\u003c/sub\u003e-H…OC hydrogen bonds involving proline residues in helices.\u003c/li\u003e\n\u003cli\u003eGrant, B. J., Skjaerven, L., and Yao, X.-Q. The Bio3D packages for structural bioinformatics. \u003cem\u003eProtein Sci\u003c/em\u003e. \u003cstrong\u003e2021\u003c/strong\u003e, \u003cem\u003e30\u003c/em\u003e, 20–30.\u003c/li\u003e\n\u003cli\u003eTrott, O.; Olson, A. J. AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization and multithreading. \u003cem\u003eJ. Comput. Chem\u003c/em\u003e. \u003cstrong\u003e2010\u003c/strong\u003e, \u003cem\u003e31\u003c/em\u003e, 455-461.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-2057833/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-2057833/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eDespite the success of AlphaFold2 (AF2), it is unclear how AF2 models accommodate for ligand binding. Here, we start with a protein sequence from \u003cem\u003eAcidimicrobiaceae TMED77\u003c/em\u003e (T7RdhA) with potential for catalyzing the degradation of per- and polyfluoroalkyl substances (PFASs). AF2 models and experiments identified T7RdhA as a corrinoid iron-sulfur protein (CoFeSP) which uses a norpseudo-cobalamin (BVQ) cofactor and two [4Fe4S] iron-sulfur clusters (SF4) for catalysis. Docking and molecular dynamics simulations suggest that T7RdhA uses perfluorooctanoic acetate (PFOA) as a substrate, supporting the reported defluorination activity of its homolog, A6RdhA. We showed that AF2 provides processual (dynamic) predictions for the binding pockets of ligands (cofactors and/or substrates). Because the pLDDT scores provided by AF2 reflect the protein native states in complex with ligands as the evolutionary constraints, the Evoformer network of AF2 predicts protein structures and residue flexibility in complex with the ligands, i.e., in their native states.\u003c/p\u003e","manuscriptTitle":"Accurate prediction by AlphaFold2 for ligand binding in a reductive dehalogenase: Implications for PFAS (per- and polyfluoroalkyl substance) biodegradation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2022-09-20 21:42:21","doi":"10.21203/rs.3.rs-2057833/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Major revision","date":"2022-11-17T13:17:04+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2022-10-01T14:55:48+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2022-09-27T05:04:46+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"e1f3deb8-2e24-4ec1-860a-7aaaa4e8e581","date":"2022-09-25T02:30:38+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"790eae1f-fe42-4059-93a0-5b7df0a3303e","date":"2022-09-16T22:57:41+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2022-09-16T14:25:17+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2022-09-16T14:22:25+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2022-09-16T14:06:02+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2022-09-16T12:24:04+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2022-09-12T18:01:31+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"0694493f-e0dc-4cc6-8c65-78caf8d6667a","owner":[],"postedDate":"September 20th, 2022","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2023-10-16T19:36:58+00:00","versionOfRecord":{"articleIdentity":"rs-2057833","link":"https://doi.org/10.1038/s41598-023-30310-x","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2023-03-11 19:33:35","publishedOnDateReadable":"March 11th, 2023"},"versionCreatedAt":"2022-09-20 21:42:21","video":"","vorDoi":"10.1038/s41598-023-30310-x","vorDoiUrl":"https://doi.org/10.1038/s41598-023-30310-x","workflowStages":[]},"version":"v1","identity":"rs-2057833","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-2057833","identity":"rs-2057833","version":["v1"]},"buildId":"J0_U0BvcaRcwD8yVFaRlm","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}