Structural basis of hierarchical multiple substates of a protein. III: Side chain and main chain local conformations

An analysis is carried out of differences in the minimum energy conformations obtained in the previous paper by energy minimization starting from conformations sampled by a Monte Carlo simulation of conformational fluctuations in the native state of a globular protein, bovine pancreatic trypsin inhibitor. Main conformational differences in each pair of energy minima are found usually localized in several side chains and in a few local main chain segments. Such side chains and local main chain segments are found to take a few distinct local conformations in the minimum energy conformations. Energy minimum conformations can thus be described in terms of combinations of these multiple local conformations.     

The impact of protein flexibility on protein-protein docking

Accounting for protein flexibility in protein-protein docking algorithms is challenging, and most algorithms therefore treat proteins as rigid bodies or permit side-chain motion only. While the consequences are obvious when there are large conformational changes upon binding, the situation is less clear for the modest conformational changes that occur upon formation of most protein-protein complexes. We have therefore studied the impact of local protein flexibility on protein-protein association by means of rigid body and torsion angle dynamics simulation. The binding of barnase and barstar was chosen as a model system for this study, because the complexation of these 2 proteins is well-characterized experimentally, and the conformational changes accompanying binding are modest. On the side-chain level, we show that the orientation of particular residues at the interface (so-called hotspot residues) have a crucial influence on the way contacts are established during docking from short protein separations of approximately 5 A. However, side-chain torsion angle dynamics simulations did not result in satisfactory docking of the proteins when using the unbound protein structures. This can be explained by our observations that, on the backbone level, even small (2 A) local loop deformations affect the dynamics of contact formation upon docking. Complementary shape-based docking calculations confirm this result, which indicates that both side-chain and backbone levels of flexibility influence short-range protein-protein association and should be treated simultaneously for atomic-detail computational docking of proteins.     

Incorporation of side chain flexibility into protein binding pockets using MTflex

The plasticity of active sites plays a significant role in drug recognition and binding, but the accurate incorporation of ‘receptor flexibility’ remains a significant computational challenge. Many approaches have been put forward to address receptor flexibility in docking studies by generating relevant ensembles on the energy surface. Herein, we describe the Movable Type with flexibility (MT_flex) method that generates ensembles on the more relevant free energy surface in a computationally tractable manner. This novel approach enumerates conformational states based on side chain flexibility and then estimates their relative free energies using the MT methodology. The resultant conformational states can then be used in subsequent docking and scoring exercises. In particular, we demonstrate that using the MT_flex ensembles improves MT’s ability to predict binding free energies over docking only to the crystal structure.     

On the chain flexibility of arabinoxylans and other beta-(1-->4) polysaccharides

The recent paper by Dervilly-Pinel and co-workers (Carbohydr. Res. 2001, 330, 365-372) presents a complete macromolecular characterisation of a series of de-esterified arabinoxylans extracted and fractionated from wheat flour. From their measurements, they were able to extract parameters related to chain flexibility, such as the Mark-Houwink exponent a and the chain persistence length. However, the estimate they obtain for the latter parameter is rather larger than would be expected, since the arabinoxylan backbone is beta-(1-->4) like cellulose and the galactomannans. By treating their data in an alternative, but well accepted manner, we are able to obtain a lower value of persistence length, which agrees well with estimates for this parameter for cellulose, in the literature, and our own recent measurements for a series of differently substituted galactomannans. These results suggest that the parameters obtained may be applicable to other beta-(1-->4) polysaccharides.     

On the flexibility of poly(γ-benzyl L-glutamate) in helicogenic solvents

The molecular-weight dependence of the rms radius of gyration of poly(γ-benzyl L -glutamate) (PBLG) in helicogenic solvents shows negative and positive deviations from expectations for an intact and rigid α-helix in the higher and lower molecular-weight ranges, respectively. In order to study the reason for both deviations, we compare the extant experimental data of with those computed for wormlike chain, freely jointed rod, and a rigid rod having random-coil portions at both ends. The computation for the freely jointed rod and the rigid rod having frayed ends is carried out by a simulation method of Muroga. From the Zimm and Bragg theory and the above comparisons, it is concluded that both deviations can be self-consistently explained if PBLG in helicogenic solvents has an essentially intact α-helical structure with some flexibility arising from random fluctuations in hydrogen bond length. This flexibility explains the negative deviations in the high molecular weight region. The positive deviations in the low molecular weight region result from the tendency of helices to unwind at the ends. © 1998 John Wiley & Sons, Inc. Biopoly 45: 281–288, 1998     

Preferred conformations and dynamics of five core structures of mucin typeO-glycans determined by NMR spectroscopy and force field calculations

Glycosyltransferases acting onO-glycans have been shown to exhibit distinct specificity for the carbohydrate and the peptide moiety of their substrates. As an approach to study the 3-dimensional interactions between enzymes andO-glycan substrates, we determined the preferred conformations of five oligosaccharide-core structures of mucin type glycoproteins by NMR spectroscopy and by static and dynamic force field calculations. Seven oligosaccharides, representing five basic core structures, were investigated: Gal(1–3)GalNAcBzl (1, core 1), GlcNAc(1–6)[Gal(1–3)]GalNAcBzl (2, core 2), GlcNAc(1–3)GalNacBzl (3, core 3), GlcNAc(1–6)[GlcNAc(1–3)]GalNAcBzl (4, core 4), GlcNAc(1–6)GalNAcBzl (5, core 6), the elongated core 2, Gal(1–4)GlcNAc(1–6)[Gal(1–3)]GalNAcpNp (6) and GalNAc-Bzl (7). The dynamic behaviour of the molecules was studied by Metropolis Monte Carlo (MMC) simulations. Experimental coupling constants, chemical shift changes, and NOEs were compared with results from static energy minimizations and dynamic MMC simulations and show a good agreement. MMC simulations show that the (1–6) linkage is much more flexible than the (1–3) or the (1–4) linkages. The preferred conformations of the disaccharides (1) and (3) show only slight differences due to the additionalN-acetyl group in (3). The conformational equilibrium of (1–3) glycosidic bonds of1 and3 was not affected by attaching a (1–6) linked GlcNAc unit to the GalNAc residue in2 and4. However, experimental and theoretical data show that the (1–6) linkages of the trisaccharides2 and4, which carry an additional (1–3) linked glycosyl residue, change their preferred conformations when compared with (5). The 6-branch also shows significant interactions with the benzyl aglycon altering the preferred conformation of the hydroxymethyl group of the GalNAc to a higher proportion of the gt conformer. The (1–6) linkage of2, 4, and6 can have two different families of conformations of which the lower energy state is populated only to about 20% of the time whereas the other state with a relative enthalpy of 4 kcal mol^–1 is populated to 80%. This fact demonstrates that the two conformational states have different entropy contents. Entropy is implicitly included in MMC simulations but cannot be derived from energy minimizations.Abbreviations Bzl benzyl - COSY correlation spectroscopy - Gal d-galactose - GalNAc N-acetyl-d-galactosamine - GalNAc-ol N-acetylgalactosaminitol - GlcNAc N-acetyl-d-glucosamine - HOHAHA homonuclear Hartmann-Hahn-spectroscopy - MMC metropolis Monte Carlo - NOE nuclear Overhauser enhancement - pNp p-nitrophenyl - ROESY rotating frame Overhauser enhancement spectroscopy - TOCSY totally correlated spectroscopy     

A straightforward chemical synthesis of 17-ketosteroids by cleavage of the C-17-dihydroxy acetone side chain in corticosteroids

A facile and convenient approach to 17-ketosteroids is described. Treatment of steroids containing the C-17-dihydroxy acetone side chain with an excess of sodium methoxide in dry 1,4-dioxane under reflux, affords high yields of the corresponding 17-ketosteroids that are recovered as pure products, without the need of further purification.     

Synthesis and antibacterial activity of novel ketolides with 11,12-quinoylalkyl side chains

A series of quinoylalkyl side chains was designed and synthesized, followed by introduction into ketolides by coupling with building block 6 or 32. The corresponding targets 7a–n, 33b, and 33e were tested for their in vitro activities against a series of macrolide-sensitive and macrolide-resistant pathogens. Some of them showed a similar antibacterial spectrum and comparable activity to telithromycin. Among them, two C2-F ketolides, compounds 33b and 33e, displayed excellent activities against macrolide-sensitive and macrolide-resistant pathogens.     

Prediction and evaluation of side-chain conformations for protein backbone structures

A common approach to protein modeling is to propose a backbone structure based on homology or threading and then to attempt to build side chains onto this backbone. A fast algorithm using the simple criteria of atomic overlap and overall rotamer probability is proposed for this purpose. The method was first tested in the context of exhaustive searches of side chain configuration space in protein cores and was then applied to all side chains in 49 proteins of known structure, using simulated annealing to sample space. The latter procedure obtains the correct rotamer for 57% and the correct χ₁ value for 74% of the 6751 residues in the sample. When low-temperature Monte-Carlo simulations are initiated from the results of the simulated-annealing processes, consensus configurations are obtained which exhibit slightly more accurate predictions. The Monte-Carlo procedure also allows converged side chain entropies to be calculated for all residues. These prove to be accurate indicators of prediction reliability. For example, the correct rotamer is obtained for 79% and the correct χ₁ value is obtained for 84% of the half of the sample residues exhibiting the lowest entropies. Side chain entropy and predictability are nearly completely uncorrelated with solvent-accessible area. Some precedents for and implications of this observation are discussed. © 1996 Wiley-Liss, Inc.     

An analysis of side chain interactions and pair correlations within antiparallel β-sheets: The differences between backbone hydrogen-bonded and non-hydrogen-bonded residue pairs

Cross-strand pair correlations are calculated for residue pairs in antiparallel β-sheet for two cases: pairs whose backbone atoms are hydrogen bonded together (H-bonded site) and pairs which are not (non-H-bonded site). The statistics show that this distinction is important. When glycine is located on the edge of a sheet, it shows a 3:1 preference for the H-bonded site. Thestrongest observed correlations are for pairs of disulfide-bonded cystines, many of which adopt a close-packed conformation with each cystine in a spiral conformation of opposite chirality to its partner. It is likely that these pairs are a signature for the family of small, cystine-rich proteins. Most other strong positive and negative correlations involve charged and polar residues. It appears that electrostatic compatibility is the strongest factor affecting pair correlation. Significant correlations are observed for β- and γ-branched residues inthe non-H-bonded site. An examination of the structures showsa directionality in side chain packing. There is a correlation between (1) the directionality in the packing interactions of non-H-bonded β- and γ-branched residue pairs, (2) the handedness of the observed enantiomers of chiral β-branched side chains, and (3) the handedness of the twist of β-sheet. These findings have implications for the formation of β-sheets during protein folding and the mechanism by which the sheet becomes twisted. © 1995 Wiley-Liss, Inc.     

Structural optimization and evaluation of butenolides as potent antifouling agents: modification of the side chain affects the biological activities of compounds

A recent global ban on the use of organotin compounds as antifouling agents has increased the need for safe and effective antifouling compounds. In this study, a series of new butenolide derivatives with various amine side chains was synthesized and evaluated for their anti-larval settlement activities in the barnacle, Balanus amphitrite. Side chain modification of butenolide resulted in butenolides 3c-3d, which possessed desirable physico-chemical properties and demonstrated highly effective non-toxic anti-larval settlement efficacy. A structure-activity relationship analysis revealed that varying the alkyl side chain had a notable effect on anti-larval settlement activity and that seven to eight carbon alkyl side chains with a tert-butyloxycarbonyl (Boc) substituent on an amine terminal were optimal in terms of bioactivity. Analysis of the physico-chemical profile of butenolide analogues indicated that lipophilicity is a very important physico-chemical parameter contributing to bioactivity.     

Characterization of threonine side chain dynamics in an antifreeze protein using natural abundance 13C NMR spectroscopy

The dynamics of threonine side chains of the Tenebrio molitor antifreeze protein (TmAFP) were investigated using natural abundance (13)C NMR. In TmAFP, the array of threonine residues on one face of the protein is responsible for conferring its ability to bind crystalline ice and inhibit its growth. Heteronuclear longitudinal and transverse relaxation rates and the [(1)H]-(13)C NOE were determined in this study. The C alpha H relaxation measurements were compared to the previously measured (15)N backbone parameters and these are found to be in agreement. For the analysis of the threonine side chain motions, the model of restricted rotational diffusion about the chi(1) dihedral angle was employed [London and Avitabile (1978) J. Am. Chem. Soc., 100, 7159-7165]. We demonstrate that the motion experienced by the ice binding threonine side chains is highly restricted, with an approximate upper limit of less than +/-25 degrees.     

Folding of the polypeptide chain(s), conformational flexibility and reactivity of the metal active site of hemocyanin and tyrosinase

Summary The comparative accessibility of the active sites of hemocyanin and tyrosinase, two proteins containing a binuclear type-3 copper site, has been investigated. The approaches were: (a) the kinetic study of the reaction of hemocyanin with cyanide in the presence of conformation perturbants; (b) the comparison between the kinetic parameters of the cyanide reaction on hemocyanin and tyrosinase; (c) the study of the efficiency and reaction mechanism of hemocyanin interaction with a typical tyrosinase substrate like catechol. The results indicate that the active site of tyrosinase is much more exposed than that of hemocyanin.     

Structure and dynamics of the homodimeric dynein light chain km23

km23 (96 residues, 11 kDa) is the mammalian ortholog of Drosophila roadblock, the founding member of LC7/robl/km23 class of dynein light chains. km23 has been shown to be serine-phosphorylated following TGFbeta receptor activation and to bind the dynein intermediate chain in response to such phosphorylation. Here, we report the three-dimensional solution structure of km23, which is shown to be that of a homodimer, similar to that observed for the heterodimeric complex formed between p14 and MP1, two distantly related members of the MglB/robl superfamily, but distinct from the LC8 and Tctex-1 classes of dynein light chains, which also adopt homodimeric structures. The conserved surface residues of km23, including three serine residues, are located predominantly on a single face of the molecule. Adjacent to this face is a large cleft formed by the incomplete overlap of loops from opposite monomers. As shown by NMR relaxation data collected at two fields, several cleft residues are flexible on the ns-ps and ms-mus timescales. Based on these observations, we propose that the patch of conserved residues on the central face of the molecule corresponds to the site at which km23 binds the dynein intermediate chain and that the flexible cleft formed between the overlap of loops from the two monomers corresponds to the site at which km23 binds other partners, such as the TGFbeta type II receptor or Smad2.     

Cellular distribution and bioactivity of the key steroidogenic enzyme,cytochrome P450side chain cleavage,in sensory neural pathways

Neurosteroids are steroids produced within the nervous system. Based on behavioural responses evoked in animals by synthetic steroid injections, several studies suggested neurosteroid involvement in important neurophysiological processes. These observations should be correlated only to neuroactive effects of the injected steroids. Neurosteroids mostly control the CNS activity through allosteric modulation of neurotransmitter receptors within concentration ranges used by neurotransmitters themselves. Therefore, neurosteroid production within pathways controlling a neurophysiological process is necessary to consider neurosteroid involvement in that process. Because of the increasing speculation about pain modulation by neurosteroids based on pharmacological observations, we decided to clarify the situation by investigating neurosteroidogenesis occurrence in sensory pathways, particularly in nociceptive structures. We studied the presence and activity of cytochrome P450side chain cleavage (P450scc) in rat pain pathways. P450scc-immunoreactive cells were localized in dorsal root ganglia (DRG), spinal cord (SC) dorsal horn, nociceptive supraspinal nuclei (SSN) and somatosensory cortex. Incubation of DRG, SSN or SC tissue homogenates with [3H]cholesterol yielded the formation of radioactive metabolites including [3H]pregnenolone of which the synthesis was reduced in presence of aminogluthetimide, a P450scc inhibitor. These first neuroanatomical and neurochemical results demonstrate the occurrence of neurosteroidogenesis in nociceptive pathways and strongly suggest that neurosteroids may control pain mechanisms.     

Refined structure of melittin bound to perdeuterated dodecylphosphocholine micelles as studied by 2D-NMR and distance geometry calculation

In our previous paper we reported the conformation of melittin bound to perdeuterated dodecylphosphocholine micelles as studied by 1H NMR experiment and distance geometry calculation. No hydrogen bonds were taken into consideration explicitly in the calculation. However, mostly alpha-helical conformations were obtained as results of the calculation even with no explicitly assumed hydrogen bonds. In the present paper we refined the distance geometry calculation by incorporating hydrogen bonds suggested by the previous calculation. As a result, we obtained the conformation of melittin, which was consistent with both NMR data and the additional hydrogen bonding data. The alpha-helical rod in the refined conformation also has a kink at Thr-11 and Gly-12, but its bent angle is now a bit narrowly distributed in 135 degrees +/- 15 degrees. In the present study another distortion at Trp-19 and IIe-20 becomes conspicuous. The average root-mean-square displacement of atoms is now much smaller and is 1.5 A for all backbone atoms. In the present paper side chain conformations are also analyzed.     

Extension of the fragment method to calculate amino acid zwitterion and side chain partition coefficients

The fragment method of calculating partition coefficients (P) has been extended to include the common amino acids (AAs). The results indicate that polar and charged side chains influence the hydrophobicity of atoms in the side chain in a predictable manner. Field effects, as evidenced through polar proximity factors and bond factors, need to be considered for accurate estimation of transfer phenomena. The calculated log P and delta G degree ' values of the 20 AAs agree well with the observed values. Pro calculates to be more hydrophilic than the observed log P. Hydrophobicity scales for peptide side chain residues are compared and evaluated in terms of suitability. Calculated pi values for nonpolar side chain residues agree well with the observed values; calculated values for uncharged polar side chain residues deviate by about 0.6 log units except for Gln and Cys; and polar side chain residues with charged side chains calculate as too hydrophilic. Reasons for the differences are explored. We also suggest that tightly bound water to polar moieties in amino acids and peptides may be transferred into the octanol phase during partitioning experiments. A quantitative methodology is presented which characterizes the thermodynamic partitioning of groups and individual atoms in amino acids and proteins.     

Model selection for the interpretation of protein side chain methyl dynamics

A number of different dynamics models are considered for fitting ¹³C and ²H side chain methyl relaxation rates. It is shown that in cases where nanosecond time scale dynamics are present the extended Lipari–Szabo model which is explicitly parameterized to include the effects of slow motions can produce wide distributions of fitting parameters even in cases where the errors are relatively small and large numbers of relaxation rates are considered. In contrast, fits of ¹⁵N backbone dynamics using this model are far more robust. The origin of this difference is analyzed and can be explained by the different functional forms of the spectral density in these two cases. The utility of a number of models for the analysis of methyl side chain dynamics is presented.     

3d Hcch-Cosy-Tocsy experiment for the assignment of ribose and amino acid side chains in 13C labeled RNA and protein

A new 3D HCCH-COSY-TOCSY experiment is presented for the assignment of RNA sugar and protein side chains. The experiment, which combines COSY and TOCSY units, is more powerful than the sum of individual HCCH-COSY and HCCH-TOCSY pulse sequences. The experiment was applied to a 13C, 15N-labeled 26 mer RNA complexed with the antibiotic tobramycin, and a 12 kDa 13C, 15N-labeled FKBP12 protein sample. The power of HCCH-COSY-TOCSY is demonstrated through complete spin system assignments of sugars in the 26 mer RNA sample, which could not be assigned using a combination of HCCH-COSY, HCCH-TOCSY and 13C-edited NOESY experiments.     

Characterization of cDNAs encoding cholesterol side chain cleavage and 3beta-hydroxysteroid dehydrogenase in the freshwater stingray Potamotrygon motoro

The interrenal gland (adrenocortical homolog) of elasmobranchs produces a unique steroid, 1α-hydroxycorticosterone (1α-B). The synthesis of this and most other steroids requires both cholesterol side chain cleavage (CYP11A) and 3β-hydroxysteroid dehydrogenase (HSD3). To facilitate the study of elasmobranch steroidogenesis, we isolated complementary DNAs encoding CYP11A and HSD3 from the freshwater stingray Potamotrygon motoro. The P. motoro CYP11A (2182 bp total length) and HSD3 (2248 bp total length) cDNAs harbor open reading frames that encode proteins of 542 and 376 amino acids (respectively) that are similar (CYP11A: 39–61% identical; HSD3: 36–53% identical) to their homologs from other vertebrates. In molecular phylogenetic analysis, P. motoro CYP11A segregates with CYP11A proteins (and not with related CYP11B proteins) and P. motoro HSD3 segregates with steroidogenic HSD3 proteins from other fishes. CYP11A and HSD3 mRNA is found only in interrenal and gonadal tissues, indicating de novo steroidogenesis is restricted to these tissues. Because 1α-B is thought to act in the elasmobranch response to hydromineral disturbances, we examined the effect of adapting P. motoro to 10 ppt seawater on mRNAs encoding steroidogenic genes. The P. motoro response to this salinity challenge does not include interrenal hypertrophy or an increase in the levels of interrenal CYP11A, HSD3 or steroidogenic acute regulatory protein (StAR) mRNA. This study is the first to isolate full length cDNAs encoding elasmobranch CYP11A and HSD3 and the first to examine the regulation of steroidogenic genes in elasmobranch interrenal cells.