Keehyoung Joo

Improving physical realism, stereochemistry, and side-chain accuracy in homology modeling: Four approaches that performed well in CASP8.

A correct alignment is an essential requirement in homology modeling. Yet in order to bridge the structural gap between template and target, which may not only involve loop rearrangements, but also shifts of secondary structure elements and repacking of core residues, high‐resolution refinement methods with full atomic details are needed. Here, we describe four approaches that address this “last mile of the protein folding problem” and have performed well during CASP8, yielding physically realistic models: YASARA, which runs molecular dynamics simulations of models in explicit solvent, using a new partly knowledge‐based all atom force field derived from Amber, whose parameters have been optimized to minimize the damage done to protein crystal structures. The LEE‐SERVER, which makes extensive use of conformational space annealing to create alignments, to help Modeller build physically realistic models while satisfying input restraints from templates and CHARMM stereochemistry, and to remodel the side‐chains. ROSETTA, whose high resolution refinement protocol combines a physically realistic all atom force field with Monte Carlo minimization to allow the large conformational space to be sampled quickly. And finally UNDERTAKER, which creates a pool of candidate models from various templates and then optimizes them with an adaptive genetic algorithm, using a primarily empirical cost function that does not include bond angle, bond length, or other physics‐like terms. Proteins 2009.

Structural studies of a bacterial condensin complex reveal ATP-dependent disruption of intersubunit interactions.

Condensins are key mediators of chromosome condensation across organisms. Like other condensins, the bacterial MukBEF condensin complex consists of an SMC family protein dimer containing two ATPase head domains, MukB, and two interacting subunits, MukE and MukF. We report complete structural views of the intersubunit interactions of this condensin along with ensuing studies that reveal a role for the ATPase activity of MukB. MukE and MukF together form an elongated dimeric frame, and MukFs C-terminal winged-helix domains (C-WHDs) bind MukB heads to constitute closed ring-like structures. Surprisingly, one of the two bound C-WHDs is forced to detach upon ATP-mediated engagement of MukB heads. This detachment reaction depends on the linker segment preceding the C-WHD, and mutations on the linker restrict cell growth. Thus ATP-dependent transient disruption of the MukB-MukF interaction, which creates openings in condensin ring structures, is likely to be a critical feature of the functional mechanism of condensins.

Prediction of protein tertiary structure using PROFESY, a novel method based on fragment assembly and conformational space annealing

A novel method for ab initio prediction of protein tertiary structures, PROFESY (PROFile Enumerating SYstem), is proposed. This method utilizes the secondary structure prediction information of a query sequence and the fragment assembly procedure based on global optimization. Fifteen‐residue‐long fragment libraries are constructed using the secondary structure prediction method PREDICT, and fragments in these libraries are assembled to generate full‐length chains of a query protein. Tertiary structures of 50 to 100 conformations are obtained by minimizing an energy function for proteins, using the conformational space annealing method that enables one to sample diverse low‐lying local minima of the energy. We apply PROFESY for benchmark tests to proteins with known structures to demonstrate its feasibility. In addition, we participated in CASP5 and applied PROFESY to four new‐fold targets for blind prediction. The results are quite promising, despite the fact that PROFESY was in its early stages of development. In particular, PROFESY successfully provided us the best model‐one structure for the target T0161. Proteins 2004.

High accuracy template based modeling by global optimization.

For high‐accuracy template‐based‐modeling of CASP7 targets, we have applied a procedure based on the rigorous optimization of score functions at three stages: multiple alignment, chain building, and side‐chain modeling. We applied the conformational space annealing method to a newly developed consistency based score function for multiple alignment. For chain building, we optimized the MODELLER energy function. For side‐chain modeling, we optimized a SCWRL‐like energy function using a rotamer library constructed specifically for a given target sequence. By rigorous optimization, we have achieved significant improvement in backbone as well as side‐chain modeling for TBM and TBM/HA targets. For most TBM/HA targets (17/26), the predicted model was more accurate than the model one can construct from the best template in a posteriori fashion. It appears that the current method can extract relevant information out of multiple templates. Proteins 2007.

Sann: Solvent accessibility prediction of proteins by nearest neighbor method

We present a method to predict the solvent accessibility of proteins which is based on a nearest neighbor method applied to the sequence profiles. Using the method, continuous real‐value prediction as well as two‐state and three‐state discrete predictions can be obtained. The method utilizes the z‐score value of the distance measure in the feature vector space to estimate the relative contribution among the k‐nearest neighbors for prediction of the discrete and continuous solvent accessibility. The Solvent accessibility database is constructed from 5717 proteins extracted from PISCES culling server with the cutoff of 25% sequence identities. Using optimal parameters, the prediction accuracies (for discrete predictions) of 78.38% (two‐state prediction with the threshold of 25%), 65.1% (three‐state prediction with the thresholds of 9 and 36%), and the Pearson correlation coefficient (between the predicted and true RSAs for continuous prediction) of 0.676 are achieved An independent benchmark test was performed with the CASP8 targets where we find that the proposed method outperforms existing methods. The prediction accuracies are 80.89% (for two state prediction with the threshold of 25%), 67.58% (three‐state prediction), and the Pearson correlation coefficient of 0.727 (for continuous prediction) with mean absolute error of 0.148. We have also investigated the effect of increasing database sizes on the prediction accuracy, where additional improvement in the accuracy is observed as the database size increases. The SANN web server is available at http://lee.kias.re.kr/∼newton/sann/.Proteins 2012;

Necessary and sufficient conditions for the asymmetric synthesis of chiral amines using ω‐aminotransferases

The half reactions of ω‐aminotransferase (ω‐AT) from Vibrio fluvialis JS17 (ω‐ATVf) were carried out using purified pyridoxal 5′‐phosphate‐enzyme (PLP‐Enz) and pyridoxamine 5′‐phosphate‐enzyme (PMP‐Enz) complexes to investigate the relative activities of substrates. In the reaction generating PMP‐Enz from PLP‐Enz using L‐alanine as an amine donor, L‐alanine showed about 70% of the initial reaction rate of (S)‐α‐methylbenzylamine ((S)‐α‐MBA). However, in the subsequent half reaction recycling PLP‐Enz from PMP‐Enz using acetophenone as an amine acceptor, acetophenone showed nearly negligible reactivity compared to pyruvate. These results indicate that the main bottleneck in the asymmetric synthesis of (S)‐α‐MBA lies not in the amination of PLP by alanine, but in the amination of acetophenone by PMP‐Enz, where conformational restraints of the enzyme structure is likely to be the main reason for limiting the amine group transfer from PMP‐Enz to acetophenone. Based upon those half reaction experiments using the two amino acceptors of different activity, it appears that the relative activities of the two amine donors and the two acceptors involved in the ω‐AT reactions can roughly determine the asymmetric synthesis yield of the target chiral amine compound. Predicted conversion yields of several target chiral amines were calculated and compared with the experimental conversion yields. Approximately, a positive linear correlation (Pearsons correlation coefficient = 0.92) was observed between the calculated values and the experimental conversion yields. To overcome the low (S)‐α‐MBA productivity of ω‐ATVf caused by the possible disadvantageous structural constraints for acetophenone, new ω‐ATs showing higher affinity to benzene ring of acetophenone than ω‐ATVf were computationally screened using comparative modeling and protein‐ligand docking. ω‐ATs from Streptomyces avermitilis MA‐4680 (SAV2612) and Agrobacterium tumefaciens str. C58 (Atu4761) were selected, and the two screened ω‐ATs showed higher asymmetric synthesis reaction rate of (S)‐α‐MBA and lower (S)‐α‐MBA degradation reaction rate than ω‐ATVf. To verify the higher conversion yield of the variants of ω‐ATs, the reaction with 50 mM acetophenone and 50 mM alanine was performed with coupling of lactate dehydrogenase and two‐phase reaction system. SAV2612 and Atu4761 showed 70% and 59% enhanced yield in the synthesis of (S)‐α‐MBA compared to that of ω‐ATVf, respectively. Biotechnol. Bioeng. 2011;108: 253–263.

All‐atom chain‐building by optimizing MODELLER energy function using conformational space annealing

We have investigated the effect of rigorous optimization of the MODELLER energy function for possible improvement in protein all‐atom chain‐building. For this we applied the global optimization method called conformational space annealing (CSA) to the standard MODELLER procedure to achieve better energy optimization than what MODELLER provides. The method, which we call MODELLERCSA, is tested on two benchmark sets. The first is the 298 proteins taken from the HOMSTRAD multiple alignment set. By simply optimizing the MODELLER energy function, we observe significant improvement in side‐chain modeling, where MODELLERCSA provides about 10.7% (14.5%) improvement for χ1 (χ1 + χ2) accuracy compared to the standard MODELLER modeling. The improvement of backbone accuracy by MODELLERCSA is shown to be less prominent, and a similar improvement can be achieved by simply generating many standard MODELLER models and selecting lowest energy models. However, the level of side‐chain modeling accuracy by MODELLERCSA could not be matched either by extensive MODELLER strategies, side‐chain remodeling by SCWRL3, or copying unmutated rotamers. The identical procedure was successfully applied to 100 CASP7 template base modeling domains during the prediction season in a blind fashion, and the results are included here for comparison. From this study, we observe a good correlation between the MODELLER energy and the side‐chain accuracy. Our findings indicate that, when a good alignment between a target protein and its templates is provided, thorough optimization of the MODELLER energy function leads to accurate all‐atom models. Proteins 2009.

Protein structure modeling for CASP10 by multiple layers of global optimization.

In the template‐based modeling (TBM) category of CASP10 experiment, we introduced a new protocol called protein modeling system (PMS) to generate accurate protein structures in terms of side‐chains as well as backbone trace. In the new protocol, a global optimization algorithm, called conformational space annealing (CSA), is applied to the three layers of TBM procedure: multiple sequence‐structure alignment, 3D chain building, and side‐chain re‐modeling. For 3D chain building, we developed a new energy function which includes new distance restraint terms of Lorentzian type (derived from multiple templates), and new energy terms that combine (physical) energy terms such as dynamic fragment assembly (DFA) energy, DFIRE statistical potential energy, hydrogen bonding term, etc. These physical energy terms are expected to guide the structure modeling especially for loop regions where no template structures are available. In addition, we developed a new quality assessment method based on random forest machine learning algorithm to screen templates, multiple alignments, and final models. For TBM targets of CASP10, we find that, due to the combination of three stages of CSA global optimizations and quality assessment, the modeling accuracy of PMS improves at each additional stage of the protocol. It is especially noteworthy that the side‐chains of the final PMS models are far more accurate than the models in the intermediate steps. Proteins 2014; 82(Suppl 2):188–195.

Multiple Sequence Alignment by Conformational Space Annealing

We present a new method for multiple sequence alignment (MSA), which we call MSACSA. The method is based on the direct application of a global optimization method called the conformational space annealing (CSA) to a consistency-based score function constructed from pairwise sequence alignments between constituting sequences. We applied MSACSA to two MSA databases, the 82 families from the BAliBASE reference set 1 and the 366 families from the HOMSTRAD set. In all 450 cases, we obtained well optimized alignments satisfying more pairwise constraints producing, in consequence, more accurate alignments on average compared with a recent alignment method SPEM. One of the advantages of MSACSA is that it provides not just the global minimum alignment but also many distinct low-lying suboptimal alignments for a given objective function. This is due to the fact that conformational space annealing can maintain conformational diversity while searching for the conformations with low energies. This characteristics can help us to alleviate the problem arising from using an inaccurate score function. The method was the key factor for our success in the recent blind protein structure prediction experiment.

Heterozygous mutations in cyclic AMP phosphodiesterase-4D (PDE4D) and protein kinase A (PKA) provide new insights into the molecular pathology of acrodysostosis

Acrodysostosis without hormone resistance is a rare skeletal disorder characterized by brachydactyly, nasal hypoplasia, mental retardation and occasionally developmental delay. Recently, loss-of-function mutations in the gene encoding cAMP-hydrolyzing phosphodiesterase-4D (PDE4D) have been reported to cause this rare condition but the pathomechanism has not been fully elucidated. To understand the pathogenetic mechanism of PDE4D mutations, we conducted 3D modeling studies to predict changes in the binding efficacy of cAMP to the catalytic pocket in PDE4D mutants. Our results indicated diminished enzyme activity in the two mutants we analyzed (Gly673Asp and Ile678Thr; based on PDE4D4 residue numbering). Ectopic expression of PDE4D mutants in HEK293 cells demonstrated this reduction in activity, which was identified by increased cAMP levels. However, the cells from an acrodysostosis patient showed low cAMP accumulation, which resulted in a decrease in the phosphorylated cAMP Response Element-Binding Protein (pCREB)/CREB ratio. The reason for this discrepancy was due to a compensatory increase in expression levels of PDE4A and PDE4B isoforms, which accounted for the paradoxical decrease in cAMP levels in the patient cells expressing mutant isoforms with a lowered PDE4D activity. Skeletal radiographs of 10-week-old knockout (KO) rats showed that the distal part of the forelimb was shorter than in wild-type (WT) rats and that all the metacarpals and phalanges were also shorter in KO, as the name acrodysostosis implies. Like the G-protein α-stimulatory subunit and PRKAR1A, PDE4D critically regulates the cAMP signal transduction pathway and influences bone formation in a way that activity-compromising PDE4D mutations can result in skeletal dysplasia. We propose that specific inhibitory PDE4D mutations can lead to the molecular pathology of acrodysostosis without hormone resistance but that the pathological phenotype may well be dependent on an over-compensatory induction of other PDE4 isoforms that can be expected to be targeted to different signaling complexes and exert distinct effects on compartmentalized cAMP signaling.