Lactose repressor protein (LacI), a negative transcriptional regulator in Escherichia coli, relies on an allosteric conformational change for its function. The LacI effector isopropyl-β,D-thiogalactoside (IPTG) promotes this allosteric response and engages the side chains of residues N125 and D149 based on the crystallographic structure of LacI·IPTG. Targeted molecular dynamics (TMD) simulations have indicated involvement of these side chains during the protein structural changes in response to inducer binding. To examine this region further, we applied stochastic boundary molecular dynamics (SBMD) simulation and identified a transient interaction between residues N125 and D149. On the basis of these data, we introduced substitutions for either/both residues and analyzed their impact on protein function. The substitutions utilized were alanine to preclude hydrogen bonding or cysteine to allow disulfide bond formation, which was not observed for N125C/D149C. Minimal impacts were observed on operator affinity for all substitutions, but D149C, N125A/D149A, and N125C/D149C bound to IPTG with 5-8-fold lower affinity than wild-type LacI, and exhibited decreased allosteric amplitude (K(RI/O)/K(R/O)). Of interest, the double mutants did not exhibit an allosteric response to an alternate inducer, 2-phenylethyl-β,D-galactoside (PhEG), despite demonstration of PhEG binding. Further, the presence of the anti-inducer, o-nitrophenyl-β,D-fucoside (ONPF), enhanced operator affinity for wild-type LacI and all other mutant proteins examined, but behaved as an inducer for N125A/D149A, decreasing operator binding affinity. These results confirm the role of residues 125 and 149 in ligand binding and allosteric response and illustrate how readily the function of a regulatory protein can be altered.

Theoretical and in vitro experiments suggest that protein folding cores form early in the process of folding, and that proteins may have evolved to optimize both folding speed and native-state stability. In our previous work (Chen et al., Structure, 14 (2006) 1401), we developed a set of empirical potential functions and used them to analyze interaction energies among secondary-structure elements in two β-sandwich proteins. Our work on this group of proteins demonstrated that the predicted folding core also harbors residues that form native-like interactions early in the folding reaction. In the current work, we have tested our empirical potential functions on structurally-different proteins for which the folding cores have been revealed by protein hydrogen–deuterium exchange experiments. Using a set of 29 unrelated proteins, which have been extensively studied in the literature, we demonstrate that the average prediction result from our method is significantly better than predictions based on other computational methods. Our study is an important step towards the ultimate goal of understanding the correlation between folding cores and native structures.

Human fatty acid synthase (hFAS) thioesterase domain (TE) is an attractive drug target to treat obesity and cancer. On the basis of the recently published crystal structure of TE domain of hFAS, we performed molecular surface analysis and docking study to characterize the molecular interactions between the enzyme and its various ligands. Surface analysis identified the ligand‐binding pocket of TE domain that encompasses the catalytic triad of Ser2308, His2481, Asp2338. Docking of palmitate, the main biological product of hFAS, into this pocket revealed the ligand‐binding mode, in which the hydrophobic interactions are the dominant driving forces. The catalytic mechanism of TE domain can also be well explained based on the generated TE‐palmitate complex structure. Moreover, the comparison of the binding modes of five fatty acids with chain lengths ranging from 12 to 20 carbons confirmed that the ligand binding pocket of TE domain is a decisive factor in chain length specificity. In addition, docking of two known TE inhibitors, c75 and orlistat revealed the pharmacophore of these hFAS TE inhibitors, which will prove useful in structure‐based drug design against this important target. Proteins 2008. © 2007 Wiley‐Liss, Inc.

Dynein light chain 1 (DLC1, also known as DYNLL1, LC8, and PIN), a ubiquitously expressed and highly conserved protein, participates in a variety of essential intracellular events. Transition of DLC1 between dimer and monomer forms might play a crucial role in its function. However, the molecular mechanism(s) that control the transition remain unknown. DLC1 phosphorylation on Ser88 by p21-activated kinase 1 (Pak1), a signaling nodule, promotes mammalian cell survival by regulating its interaction with Bim and the stability of Bim. Here we discovered that phosphorylation of Ser88, which juxtapose each other at the interface of the DLC dimer, disrupts DLC1 dimer formation and consequently impairs its interaction with Bim. Overexpression of a Ser88 phosphorylation-inactive DLC1 mutant in mammary epithelium cells and in a transgenic animal model caused apoptosis and accelerated mammary gland involution, respectively, with increased Bim levels. Structural and biophysical studies suggested that phosphorylation-mimicking mutation leads to dissociation of the DLC1 dimer to a pure folded monomer. The phosphorylation-induced DLC1 monomer is incapable of binding to its substrate Bim. These findings reveal a previously unrecognized regulatory mechanism of DLC1 in which the Ser88 phosphorylation acts as a molecular switch for the transition of DLC1 from dimer to monomer, thereby modulating its interaction with substrates and consequently regulating the functions of DLC1.

In this article, we present a de novo method for predicting protein domain boundaries, called OPUS-Dom. The core of the method is a novel coarse-grained folding method, VECFOLD, which constructs low-resolution structural models from a target sequence by folding a chain of vectors representing the predicted secondary-structure elements. OPUS-Dom generates a large ensemble of folded structure decoys by VECFOLD and labels the domain boundaries of each decoy by a domain parsing algorithm. Consensus domain boundaries are then derived from the statistical distribution of the putative boundaries and three empirical sequence-based domain profiles. OPUS-Dom generally outperformed several state-of-the-art domain prediction algorithms over various benchmark protein sets. Even though each VECFOLD-generated structure contains large errors, collectively these structures provide a more robust delineation of domain boundaries. The success of OPUS-Dom suggests that the arrangement of protein domains is more a consequence of limited coordination patterns per domain arising from tertiary packing of secondary-structure segments, rather than sequence-specific constraints.

In this paper, we report a knowledge‐based potential function, named the OPUS‐Ca potential, that requires only Cα positions as input. The contributions from other atomic positions were established from pseudo‐positions artificially built from a Cα trace for auxiliary purposes. The potential function is formed based on seven major representative molecular interactions in proteins: distance‐dependent pairwise energy with orientational preference, hydrogen bonding energy, short‐range energy, packing energy, tri‐peptide packing energy, three‐body energy, and solvation energy. From the testing of decoy recognition on a number of commonly used decoy sets, it is shown that the new potential function outperforms all known Cα‐based potentials and most other coarse‐grained ones that require more information than Cα positions. We hope that this potential function adds a new tool for protein structural modeling.

This paper reports a combined computational and experimental study of the correlation between protein stability cores and folding kinetics. An empirical potential function was developed, and it was used for analyzing interaction energies among secondary structure elements. Studies on a beta sandwich protein, Pseudomonas aeruginosa azurin, showed that the computationally identified substructure with the strongest interactions in the native state is identical to the "interlocked pair" of beta strands, an invariant motif found in most sandwich-like proteins. Moreover, previous and new in vitro folding results revealed that the identified substructure harbors most residues that form native-like interactions in the folding transition state. These observations demonstrate that the potential function is effective in revealing the relative strength of interactions among various protein parts; they also strengthen the suggestion that the most stable regions in native proteins favor stable interactions early during folding.

Metastasis tumor-associated 1 short form (MTA1s) is a naturally occurring, alternatively spliced variant of MTA1 that functions as a repressor of estrogen receptor (ER) α transcriptional functions, at least in part by binding and sequestering ERα in the cytoplasm. A unique C-terminal 33-amino acid region containing a nuclear receptor (NR)-box motif (-LRILL-) mediates binding of MTA1s with ERα and is indispensable in this interaction. Here, we elucidated the solution structure of this 33-amino acid region by NMR spectroscopy. We found a predominance of the α-helical region toward the N-terminal region, which includes the NR-box motif. In silico docking and comparison studies showed similarities between the NR-box motif of MTA1s and a similar motif of coregulators, both in structure and mode of ERα binding. In MCF-7 breast cancer cells, the MTA1s peptide effectively repressed ERα transactivation function, as evidenced by the estrogen response element-luc assay and down-regulation of estrogen-induced genes. In mechanistic studies, we found that the antiestrogenic effects of the MTA1s peptide were due to its ability to compete with the coactivator recruitment to ERα. Furthermore, the peptide efficiently repressed estrogen-induced proliferation and anchorage-independent growth of MCF-7 cells. In addition, the MTA1s peptide blocked the progression of tumors formed by MCF-7 cells overexpressing an ERα coactivator in a xenograft-based assay. In brief, the characterization of structure and antiestrogenic activity of MTA1s peptide highlight its therapeutic potential.

The complex between flavin mononucleotide (FMN) and apo-flavodoxin is dominated by isoalloxazine-stacking interactions and 5′-phosphate hydrogen bonds. We show here that FMN binding to Desulfovibrio desulfuricans apo-flavodoxin is faster and the affinity is higher in the presence of inorganic phosphate as compared to in its absence (I=110 mM, pH 7, 20 °C). The transition-state of complex formation was investigated by ϕ-value analysis using Trp60Ala and Tyr98Ala apo-flavodoxin variants. We find that Tyr98 is highly involved in the FMN/protein transition state independent of inorganic phosphate, whereas the participation of Trp60 is modulated by inorganic phosphate. The ϕ-value for Trp60 is higher without phosphate, implying that at this condition stronger binding of Trp60 is required in the transition state to assure successful complex formation. Consistent with the experimental data, all-atom molecular dynamic simulations reveal that the presence of an anion in the phosphate subsite restricts the mobility of the Trp60-containing loop in terms of both backbone and side-chain movements, but has no effect on the Tyr98-containing loop. The overall thermodynamic stability of apo-flavodoxin is higher in the presence of inorganic phosphate as compared to in its absence (I=110 mM, pH 7, 20 °C). Kinetic experiments reveal that the additional stability originates in slower unfolding. The combined experimental and computational observations demonstrate that phosphate has an ordering effect on the Trp60-containing loop, which positions Trp60 favorably for FMN binding and increases the barrier for protein unfolding.

We report a novel computational procedure for determining protein native topology, or fold, by defining loop connectivity based on skeletons of secondary structures that can usually be obtained from low to intermediate-resolution density maps. The procedure primarily involves a knowledge-based geometry filter followed by an energetics-based evaluation. It was tested on a large set of skeletons covering a wide range of protein architecture, including one modeled from an experimentally determined 7.6 Å cryo-electron microscopy (cryo-EM) density map. The results showed that the new procedure could effectively deduce protein folds without high-resolution structural data, a feature that could also be used to recognize native fold in structure prediction and to interpret data in fields like structure genomics. Most importantly, in the energetics-based evaluation, it was revealed that, despite the inevitable errors in the artificially constructed structures and limited accuracy of knowledge-based potential functions, the average energy of an ensemble of structures with slightly different configurations around the native skeleton is a much more robust parameter for marking native topology than the energy of individual structures in the ensemble. This result implies that, among all the possible topology candidates for a given skeleton, evolution has selected the native topology as the one that can accommodate the largest structural variations, not the one rigidly trapped in a deep, but narrow, conformational energy well.

This paper reports a computational method for folding small helical proteins. The goal was to determine the overall topology of proteins given secondary structure assignment on sequence. In doing so, a Monte Carlo protocol, which combines coarse-grained normal modes and a Hamiltonian at a different scale, was developed to enhance sampling. In addition to the knowledge-based potential functions, a small-angle X-ray scattering (SAXS) profile was also used as a weak constraint for guiding the folding. The algorithm can deliver structural models with overall correct topology, which makes them similar to those of 5∼6 Å cryo-EM density maps. The success could contribute to make the SAXS technique a fast and inexpensive solution-phase experimental method for determining the overall topology of small, soluble, but noncrystallizable, helical proteins.

Protein recognition and binding play a fundamental role in living systems but sometimes also result in pathological aggregates. To probe the coupling between folding and binding in a homoheptameric system, we have characterized the time-resolved unfolding/disassembly mechanism of human co-chaperonin protein 10 (cpn10) by a combination of experimental and computational methods. The results from both approaches are in excellent agreement and make obvious that the kinetic process is three-state: an initial polypeptide-unfolding step, resulting in a non-native heptamer, is followed by a slower heptamer-dissociation step. This demonstrates that the barriers on the kinetic free-energy landscape are defined by thermodynamic stability. cpn10 is one of few, and the only heptameric, experimentally characterized system that follows the "fly casting scenario" of molecular recognition.