Maintenance of genome stability during cell division depends on establishing correct attachments between chromosomes and spindle microtubules. Correct, bi-oriented attachments are stabilized, while incorrect attachments are selectively destabilized. This process relies largely on increased phosphorylation of kinetochore substrates of Aurora B kinase at misaligned versus aligned kinetochores. Current models explain this differential phosphorylation by spatial changes in the position of substrates relative to a constant pool of kinase at the inner centromere. However, these models are based on studies in aneuploid cells. We show that normal diploid cells have a more robust error correction machinery. Aurora B is enriched at misaligned centromeres in these cells, and the dynamic range of Aurora B substrate phosphorylation at misaligned versus aligned kinetochores is increased. These findings indicate that in addition to Aurora B regulating kinetochore-microtubule binding, the kinetochore also controls Aurora B recruitment to the inner centromere. We show that this recruitment depends on both activity of Plk1, a kinetochorelocalized kinase, and activity of Aurora B itself. Our results suggest a feedback mechanism in which Aurora B both regulates and is regulated by chromosome attachment to the spindle, which amplifies the differential phosphorylation of kinetochore substrates and increases the efficiency of error correction.

W ater is a highly polar molecule, consisting of a very electronegative atom, oxygen, bonded to two weakly electropositive hydrogen atoms with two lone pairs of electrons. These features give water remarkable physical properties, some of which are anomalous, such as its lower density in the solid phase compared with the liquid phase. Its ability to serve as both a hydrogen bond donor and hydrogen bond acceptor governs its role as a solvent, a role that is of central interest for biological chemists. In this Account, we focus on water's properties as a solvent. Water dissolves a vast range of solutes with solubilities that range over 10 orders of magnitude. Differences in solubility define the fundamental dichotomy between polar, or hydrophilic, solutes and apolar, or hydrophobic, solutes. This important distinction plays a large part in the structure, stability, and function of biological macromolecules. The strength of hydrogen bonding depends on the H-O • • • O H-bond angle, and the angular distribution is bimodal. Changes in the width and frequency of infrared spectral lines and in the heat capacity of the solution provide a measure of the changes in the strength and distribution of angles of the hydrogen bonds. Polar solutes and inorganic ions increase the population of bent hydrogen bonds at the expense of the more linear population, while apolar solutes or groups have the opposite effect. We examine how protein denaturants might alter the solvation behavior of water. Urea has very little effect on water's hydrogen bond network, while guanidinium ions promote more linear hydrogen bonds. These results point to fundamental differences in the protein denaturation mechanisms of these molecules. We also suggest a mechanism of action for antifreeze (or thermal hysteresis) proteins: ordering of water around the surface of these proteins prior to freezing appears to interfere with ice formation.

▪   Heat capacity (Cp) is one of several major thermodynamic quantities commonly measured in proteins. With more than half a dozen definitions, it is the hardest of these quantities to understand in physical terms, but the richest in insight. There are many ramifications of observed Cp changes: The sign distinguishes apolar from polar solvation. It imparts a temperature (T) dependence to entropy and enthalpy that may change their signs and which of them dominate. Protein unfolding usually has a positive ΔCp, producing a maximum in stability and sometimes cold denaturation. There are two heat capacity contributions, from hydration and protein-protein interactions; which dominates in folding and binding is an open question. Theoretical work to date has dealt mostly with the hydration term and can account, at least semiquantitatively, for the major Cp-related features: the positive and negative Cp of hydration for apolar and polar groups, respectively; the convergence of apolar group h...

Electronic spectra of mesoporphyrin-substituted yeast cytochrome c peroxidase (MP-CcP) were measured as a function of pH, ionic strength, and binding of cytochrome c (cyt c) by fluorescence line narrowing (FLN) spectroscopy at 5 K. The FLN spectra provided information about the vibrational structure of the first excited singlet state of MP-CcP, the various tautomeric forms of mesoporphyrin, and the positions and widths of their 0,O bands. The composite 0,O band of MP-CcP at pH 6 could be resolved into three components with peak positions at 16 046, 16 103, and 16 203 cm-I. MP-CcP at pH 8 could be analyzed using two components with peak positions at 16 048 and 16 193 cm-l. The disappearance of the 16 103-cm-l component at alkaline pH suggests that it is due to a "chemical substate" arising from protonation of His52 in the distal side of the porphyrin. Computer simulations of the electrostatic field that CcP imposes on its porphyrin show that, in the presence of charged axial histidines His52 and His175, the electrostatic field at porphyrin nitrogens increases, especially along the normal to the heme by about 200 mV/A. Electric field effects may account for pH-dependent spectral shifts of the 0,O positions of the resolved components, although hydrogen bonding may also affect these positions. On the other hand, the peak position of the components was not affected by ionic strength or binding of cyt c, implying that the electrostatic field of the heme pocket of MP-CcP remains unchanged. Indeed, computed changes in ionic strength of the solvent show no modification of the electrostatic field at the porphyrin. The only detectable effect of ionic strength and binding of cyt c to MP-CcP is on the relative contributions of the components, suggesting some rearrangements in the vicinity of the heme. Finally, shifts in the position of the vibrational lines for MP-CcP components indicate either that the tautomers have different vibrational frequencies due to the nonsymmetry of the porphyrin and/or that tautomers experience various distortions. Comparison of the vibrational spectrum of the first excited singlet state of mesoporphyrin in CcP and horseradish peroxidase also suggests that the heme pocket in the two peroxidases provides different steric restrictions. Various experimental techniques have been used to study the structural and functional aspects of the complex between cytochrome c peroxidase (CcP)' and its native redox partner, cytochrome c (cyt c). From the structural point of view, recent X-ray data at 2.8-A resolution show that CcP and horse heart cyt c form a complex in a highly specific manner with an iron to iron distance of 30 8, (Pelletier & Kraut, 1992). Whereas .crystallography offers compelling information on the details of complex formation and the protein interfaces involved, NMR spectroscopy can reveal whether the hemes of the two proteins and amino acids in their vicinity are affected. It was found that CcP induced a significant shift in the NMR spectrum of the heme 3-methyl group resonance of cyt c (Moench et al., 1992). On the other hand, resonance Raman Preliminary data were presented in abstract form at the 37th Annual

The highly conserved, buried, Asp 26 in Escherichia coli thioredoxin has a pKa = 7.5, and its titration is associated with a sizable destabilization of the protein [Langsetmo, K., Fuchs, J., & Woodward, C. (1991) Biochemistry (preceding paper in this issue)]. A fit of the experimental pH dependence of thioredoxin stability to a theoretical expression for the pH/stability relation in proteins agrees closely with 'The work of K.L., J.A.F., and C.W. is supported by grants from the Industry-university Cooperative Research Center for Biocatalytic Processing, the Graduate School of the University of Minnesota, the Minnesota Supercomputer Institute, and the University of Minnesota Molecular Biology Computer Center. K.L. was supported by NIH Molecular Biophysics Training Grant GM08277. The work of K.A.S. is supwrted bv NIH Grants GM305,8 and GM4,37, and NSF Grant DMB

The temperature dependence of the fast internal dynamics of calcium-saturated calmodulin in complex with a peptide corresponding to the calmodulin-binding domain of the smooth muscle myosin light chain kinase is examined using 15 N and 2 H NMR relaxation methods. NMR relaxation studies of the complex were carried out at 13 temperatures that span 288-346 K. The dynamics of the backbone and over four dozen methyl-bearing side chains, distributed throughout the calmodulin molecule, were probed. The side chains show a much more variable and often considerably larger response to temperature than the backbone. A significant variation in the temperature dependence of the amplitude of motion of individual side chains is seen. The amplitude of motion of some side chains is essentially temperature-independent while many show a simple roughly linear temperature dependence. In a few cases, angular order increases with temperature, which is interpreted as arising from interactions with neighboring residues. In addition, a number of side chains display a nonlinear temperature dependence. The significance of these and other results is illuminated by several simple interpretative models. Importantly, analysis of these models indicates that changes in generalized order parameters can be robustly related to corresponding changes in residual entropy. A simple cluster model that incorporates features of cooperative or conditional motion reproduces many of the unusual features of the experimentally observed temperature dependence and illustrates that side chain interactions result in a dynamically changing environment that significantly influences the motion of internal side chains. This model also suggests that the intrinsic entropy of interacting clusters of side chains is only modestly reduced from that of independent side chain motion. Finally, estimates of protein heat capacity support the view that the major contribution to the heat capacity of protein solutions largely arises from local bond vibrations and solvent interactions and not from torsional oscillations of side chains.

Calculations of the electrostatic potentials were made around yeast elongator phenylalanine, aspartate tRNAs, and yeast initiator methionine t R N A in aqueous solution a t physiological ionic strength. The calculations were carried out with a finite difference algorithm for solving the nonlinear Poisson-Boltzmann equation that incorporates the screening effects of the electrolyte, the exclusion of ions by the molecule, the molecular shape, and the different polarizabilities of the solvent and the tRNA. The initiator 'This rescarch was supported by grants from the N I H (GM-30518 to B.H.), the N S F (DMB-8706551 to S.C.H.), and the ONR (NO001 4-86-KO483 to B.H.

All-atom, explicit water molecular dynamics simulations of calcium-loaded calmodulin complexed with a peptide corresponding to the smooth muscle myosin light chain kinase target were carried out at 295 and 346 K. Amide and side chain methyl angular generalized order parameters were calculated and analyzed in the context of the protein's structure and dynamics. The agreement between amide order parameters measured by NMR and those from the simulations was found to be good, especially at the higher temperature, indicating both better convergence for the latter and excellent transferrability of the CHARMM parameters to the higher temperature. Subtle dynamical features such as helix fraying were reproduced. A large range of order parameters for the nine calmodulin methionines was observed in the NMR, and reproduced quite well in the simulations. The major determinant of the methionine order parameter was found to be the proximity to side chains of aromatic residues. An upper bound estimate of the difference in backbone entropy between loop and helical regions was extracted from the order parameters using a model of motion in an effective potential. Although loop regions are more flexible than helical regions, it was found that the entropy loss per residue upon folding was only ∼20% less for loops than for helices. Pairwise correlated motions, which could significantly lower entropy estimates obtained from order parameter analysis alone, were found to be largely absent.

The role of the solvent matrix in affecting CO bound to ferrous horseradish peroxidase was examined by comparing band-widths of n for the protein in aqueous solutions and in trehaloseysucrose glasses. We have previously CO observed that the optical absorption band and the CO stretching mode respond to the glass transition of glyceroly water in ways that depend upon the presence of substrate (Biochemistry 40 (2001) 3483). It is now demonstrated that the CO group band-width for the protein with bound inhibitor benzhydroxamic acid is relatively insensitive to temperature or the glass transition of the solvent. In contrast, in the absence of inhibitor, the band-width varies with the temperature that the glass is formed. The results show that solvent dependent and independent motions can be distinguished, and that the presence of substrate changes the protein such that the Fe-CO site is occluded from the solvent conditions. Molecular dynamic calculations, based upon X-ray structures, showed that the presence of benzhydroxamic acid decreases the distance between His42 and Arg38 and this leads for closer distances to the O of the CO from these residues. These results are invoked to account for the observed line width changes of the CO band.

The random network model of water quantitatively describes the different hydration heat capacities of polar and apolar solutes in terms of differential distortions of the water-water hydrogen bonding angle in the first hydration shell. This method of hydration analysis is applied here to study the hydration of the wild type III thermal hysteresis protein from eel pout and three mutations at residue 16. Wild type and one mutant have full activity, the other two mutants have little or no antifreeze (thermal hysteresis) activity. The analysis reveals significant differences in the hydration structure of the ice-binding site (centered on residue 16) among four proteins. For the A16T and A16Y mutants with reduced activity, polar groups have a typical polar-like hydration. For the wild type and mutant A16C with 100% of the wild type activity, polar groups have unusual, very apolar-like hydration. In the latter case, hydrating water molecules form a more ice-like pattern of hydrogen bonding on the ice-binding face, while in the former case water-water H-bonds are more distorted and more heterogenous. Overall, the binding surface of active protein strongly enhances the water tetrahedral structure, i.e. promotes ice-like hydration. It is concluded that the specific shape, residue size and clustering of both polaryapolar groups are essential for the binding surface to recognize, and preferentially interact with nascent ice crystals forming in liquid water.

The influence of proteins and solutes on hysteresis of freezing and melting of water was measured by infrared (IR) spectroscopy. Of the solutes examined, poly-L-arginine and flounder antifreeze protein produced the largest freezing point depression of water, with little effect on the melting temperature. Poly-L-lysine, poly-Lglutamate, cytochrome c and bovine serum albumin had less effect on the freezing of water. Small compounds used to mimic non-polar (trimethylamine N-oxide, methanol), positively charged (guanidinium chloride, NH 4 Cl, urea) and negatively charged (Na acetate) groups on protein surfaces were also examined. These molecules and ions depress water's freezing point and the melting profiles became broad. Since infrared absorption measures both bulk solvent and solvent bound to the solutes, this result is consistent with solutes interacting with liquid water. The amide I absorption bands of antifreeze protein and poly-Larginine do not detectably change with the phase transition of water. An interpretation is that the antifreeze protein and poly-L-arginine order liquid water such that the water around the group is ice-like.

In order to better characterize changes in water structure induced by a hydrophobic solute the oxygen᎐oxygen and Ž Ž. Ž .. hydrogen᎐hydrogen radial distribution functions g r , g r and the hydrogen bond angle distribution function oo hh oo hh show that the angle distribution function provides the most sensitive way to analyze water structure changes associated with hydrophobic solvation.

The hydration of polar and apolar groups can be explained quantitatively, via the random network model of water, in terms of differential distortions in first hydration shell water-water hydrogen bonding angle. This method of analyzing solute induced structural distortions of water is applied to study the ice-binding type III thermal hysteresis protein. The analysis reveals subtle but significant differences in solvent structuring of the ice-binding surface, compared to non-ice binding protein surface. The major differences in hydration in the ice-binding region are (i) polar groups have a very apolar-like hydration. (ii) there is more uniform hydration structure. Overall, this surface strongly enhances the tetrahedral, or ice-like, hydration within the primary hydration shell. It is concluded that these two specific features of the hydration structure are important for this surface to recognize, and preferentially interact with nascent ice crystals forming in liquid water.

Hirudm is a good anticoagulant owing to potent inhibition of the serine protease thrombin. An a\partate-and glutamate-rich portion of hirudin plays an important part in its tight binding to thrombin through a ladder of salt bridges. and these residues have previously been mutated to asparagine or glutamine. Detailed calculations of the electrostatic contribution to changes in binding from these mutations have been performed using the finite-difference Poisson-Boltzmann method which include charge-charge interactions, salvation interactions, the residual electrostatic interaction of mutant residues, pK, shifts. and ionic strength. Single mutant effects on binding energ were close to experimental Laluex, except for the D55N mutant whose effect is overestimated, perhaps because of displacement of a bound chloride ion from the site where it binds. Multiple mutation values were generally overestimated. The effect of pK,, shifts upon the binding! is significant for one hirudin residue E58, but this appears to be due to a poor salt bridge with thrombin caused by crystal contact\. Electrostatic interaction between the acidic residues is unfavorable. However. analysis of experimental multiple mutation/single mutation data shows apparently negative interactions between these residues. from which it is concluded that structural changes can occur in the complex to relieve an unfavorable interaction when more than one acidic r&due i\ mutated. In all cases, there is a loss in stability of the complex from mutations due to loss of favorable charge-charge interaction3 with thrombin. but this is largely compensated for by reduced unt'avorable desolvation interactions. and by residual polar interactions in the Asn/Gln mutants.

The finite-difference Poisson-Boltzmann methodology was applied to a series of parallel, ␣-helical bundle models of the designed ion channel peptide Ac-(LSSLLSL) 3-CONH 2. This method is able to fully describe the current-voltage curves for this channel and quantitatively explains their cation selectivity and rectification. We examined a series of energy-minimized models representing different aggregation states, side-chain rotamers, and helical rotations, as well as an ensemble of structures from a molecular dynamics trajectory. Potential energies were computed for single, permeating K ϩ and Cl Ϫ ions at a series of positions along a central pathway through the models. A variable-electric-field Nernst-Planck electrodiffusion model was used, with two adjustable parameters representing the diffusion coefficients of K ϩ and Cl Ϫ to scale the individual ion current magnitudes. The ability of a given DelPhi potential profile to fit the experimental data depended strongly on the magnitude of the desolvation of the permeating ion. Below a pore radius of 3.8 Å, the predicted profiles showed large energy barriers, and the experimental data could be fit only with unrealistically high values for the K ϩ and Cl Ϫ diffusion coefficients. For pore radii above 3.8 Å, the desolvation energies were 2kT or less. The electrostatic calculations were sensitive to positioning of the Ser side chains, with the best fits associated with maximum exposure of the Ser side-chain hydroxyls to the pore. The backbone component was shown to be the major source of asymmetry in the DelPhi potential profiles. Only two of the energy-minimized structures were able to explain the experimental data, whereas an average of the dynamics structures gave excellent agreement with experimental results. Thus this method provides a promising approach to prediction of current-voltage curves from three-dimensional structures of ion channel proteins.

The temperature dependence of the internal dynamics of recombinant human ubiquitin has been measured using solution NMR relaxation techniques. Nitrogen-15 relaxation has been employed to obtain a measure of the amplitude of subnanosecond motion at amide N-H sites in the protein. Deuterium relaxation has been used to obtain a measure of the amplitude of motion of methyl-groups in amino-acid side chains. Data was obtained between 5 and 55°C. The majority of amide N-H and methyl groups show a roughly linear (R 2. 0.75) temperature dependence of the associated Lipari-Szabo model-free squared generalized-order parameter (O 2) describing the amplitude of motion. Interestingly, for those sites showing a linear response, the temperature dependence of the backbone is distinct from that of the methyl-bearing side chains with the former being characterized by a significantly larger L-value, where L is defined as d ln(1 À O)/d lnT. These results are comparable to the sole previous such study of the temperature dependence of protein motion obtained for a calmodulin-peptide complex. This suggests that the distinction between the main chain and methyl-bearing side chains may be general. Insight into the temperature dependence is gathered from a simple two-state step potential model.