Search

This study describes a method to computationally assess the function of homologous enzymes through small molecule binding interaction energy. Three experimentally determined X-ray structures and four enzyme models from ornithine cyclo-deaminase, alanine dehydrogenase, and mu-crystallin were used in combination with nine small molecules to derive a function score (FS) for each enzyme-model combination. While energy values varied for a single molecule-enzyme combination due to differences in the active sites, we observe that the binding energies for the entire pathway were proportional for each set of small molecules investigated. This proportionality of energies for a reaction pathway appears to be dependent on the amino acids in the active site and their direct interactions with the small molecules, which allows a function score (FS) to be calculated to assess the specificity of each enzyme. Potential of mean force (PMF) calculations were used to obtain the energies, and the resulting FS values demonstrate that a measurement of function may be obtained using differences between these PMF values. Additionally, limitations of this method are discussed based on: (a) larger substrates with significant conformational flexibility; (b) low homology enzymes; and (c) open active sites. This method should be useful in accurately predicting specificity for single enzymes that have multiple steps in their reactions and in high throughput computational methods to accurately annotate uncharacterized proteins based on active site interaction analysis.

The INK4 (inhibitor of cyclin-dependent kinase 4) family consists of four tumor-suppressor proteins: p15INK4B, p16INK4A, p18INK4C, and p19INK4D. While their sequences and structures are highly homologous, they show appreciable differences in conformational flexibility, stability, and aggregation tendency. Here, p16 and p18 were first compared directly by NMR for line broadening and disappearance, then investigated by three different approaches in search of the causes of these differences. From denaturation experiments it was found that both proteins are marginally stable with low denaturation stability (1.94 and 2.98 kcal/mol, respectively). Heteronuclear 1H-15N nuclear Overhauser enhancement measurements revealed very limited conformational flexibility on the pico- to nanosecond time-scale for both p16 and p18. H/2H exchange of amide protons monitored by NMR on three proteins (p16, p18 as well as p15), however, revealed markedly different rates in the order p18

Since the structures of several ankyrin-repeat proteins including the INK4 (inhibitor of cyclin-dependent kinase 4) family have been reported recently, the detailed structures and the functional roles of the loops have drawn considerable interest. This paper addresses the potential importance of the loops of ankyrin-repeat proteins in three aspects. First, the solution structure of p18INK4C was determined by NMR, and the loop structures were analyzed in detail. The loops adapt nascent antiparallel beta-sheet structures, but the positions are slightly different from those in the crystal structure. A detailed comparison between the solution structures of p16 and p18 has also been presented. The determination of the p18 solution structure made such detailed comparisons possible for the first time. Second, the [1H,15N]HSQC NMR experiment was used to probe the interactions between p18INK4C and other proteins. The results suggest that p18INK4C interacts very weakly with dna K and glutathione S-transferase via the loops. The third aspect employed site-specific mutagenesis and functional assays. Three mutants of p18 and 11 mutants of p16 were constructed to test functional importance of loops and helices. The results suggest that loop 2 is likely to be part of the recognition surface of p18INK4C or p16INK4A for CDK4, and they provide quantitative functional contributions of specific residues. Overall, our results enhance understanding of the structural and functional roles of the loops in INK4 tumor suppressors in particular and in ankyrin-repeat proteins in general.

The tumor suppressor p16INK4A with eight N-terminal amino acids deleted (p16/delta 1-8) was expressed in Escherichia coli without any fusion artifacts and purified. The integrity of p16/delta 1-8 was confirmed by mass spectrometry, and its activity was demonstrated by in vitro cdk4 inhibition assay. Various physical methods were used to characterize the molecular and structural properties of p16/delta 1-8. The protein was found to oligomerize in vitro, as demonstrated by gel electrophoresis, mass spectrometry, and NMR. Various approaches, including changes of concentration and pH, additions of salts, detergents, and various organic solvents, and construction of a C-terminal deletion mutant and a cysteine mutant were used to try to reduce the extent of oligomerization. Only decreasing the protein concentration was found to reduce oligomerization. The affinity between p16 molecules in vivo was demonstrated by the yeast two-hybrid system. The protein was found to be very unstable on the basis of urea- and guanidinium chloride-induced denaturation studies monitored by NMR and CD, respectively. Despite these unfavorable properties, total NMR assignments were accomplished with uniform 13C and 15N isotope labeling. All multidimensional NMR experiments were performed at a very low concentration of 0.2 mM. The secondary structure was then determined from the NMR data. The results of NMR and CD studies indicate that the protein is highly alpha-helical, and the ankyrin repeat sequences show helix-turn-helix structures. This is the first structural information obtained for the important motif of ankyrin repeats. Overall, p16/delta 1-8 appears to be conformationally flexible. In order to understand the structural basis of the functional changes for some mutants existing in tumor cells, several missense mutants of p16/delta 1-8 were constructed. Four of them were expressed at high levels and purified. The molecular and structural properties of these mutants were analyzed by CD and NMR and compared with the corresponding properties of wild-type p16/delta 1-8. The results suggest that the functional changes in P114L and G101W are likely to be related to global conformational changes. In addition, we have demonstrated that the tendency of aggregation increases significantly by a single D84H mutation.

The unimolecular reactions of protonated and metalated polyglycols with kiloelectronvolt translational energies have been studied by collisionally activated dissociation and neutralization-reionization mass spectrometry. The former method provides information on the ionic dissociation products, whereas the latter allows for the identification of the complementary neutral losses. Protonated linear polyglycols mainly undergo charge-initiated decompositions that lead to eliminations of smaller oligomers. On the other hand, protonated crown ethers (“cyclic” polyglycols) favor charge-induced reactions that proceed by cleavages of two ethylene oxide units in the form of 1,4-dioxane. Replacement of one O by NH in the crown ether dramatically changes its unimolecular chemistry; now, charge-remote 1,4-eliminations from ring-opened isomers are preferred. Charge-remote reactions are also the major decomposition channels of all metalated precursors studied. The linear polyglycols decompose primarily by 1,4-H2 eliminations and to a lesser extent by homolytic cleavages near chain ends. The reverse is true for metalated crown ethers, which preferentially produce distonic radical cations by the loss of saturated radicals; these reactions are proposed to involve prior rearrangement to open-chain isomers. The nature of the metal ion (Li+, Na+, or K+) does not greatly affect the unimolecular chemistry of the cationized polyglycol. In general, metalated precursors form many abundant fragment ions over the entire mass range; hence, collisional activation of such ions at high kinetic energy should be particularly useful for structure elucidations.

This study describes a method to computationally assess the function of homologous enzymes through small molecule binding interaction energy. Three experimentally determined X-ray structures and four enzyme models from ornithine cyclo-deaminase, alanine dehydrogenase, and mu-crystallin were used in combination with nine small molecules to derive a function score (FS) for each enzyme-model combination. While energy values varied for a single molecule-enzyme combination due to differences in the active sites, we observe that the binding energies for the entire pathway were proportional for each set of small molecules investigated. This proportionality of energies for a reaction pathway appears to be dependent on the amino acids in the active site and their direct interactions with the small molecules, which allows a function score (FS) to be calculated to assess the specificity of each enzyme. Potential of mean force (PMF) calculations were used to obtain the energies, and the resulting FS values demonstrate that a measurement of function may be obtained using differences between these PMF values. Additionally, limitations of this method are discussed based on: (a) larger substrates with significant conformational flexibility; (b) low homology enzymes; and (c) open active sites. This method should be useful in accurately predicting specificity for single enzymes that have multiple steps in their reactions and in high throughput computational methods to accurately annotate uncharacterized proteins based on active site interaction analysis.

Phosphatidylinositol-specific phospholipase Cs (PLCs) are a family of phosphodiesterases that catalyze the cleavage of the P-O bond via transesterification using the internal hydroxyl group of the substrate as a nucleophile, generating the five-membered cyclic inositol phosphate as an intermediate or product. To better understand the role of calcium in the catalytic mechanism of PLCs, we have determined the X-ray crystal structure of an engineered PLC enzyme from Bacillus thuringiensis to 2.1 A resolution. The active site of this enzyme has been altered by substituting the catalytic arginine with an aspartate at position 69 (R69D). This single-amino acid substitution converted a metal-independent, low-molecular weight enzyme into a metal ion-dependent bacterial PLC with an active site architecture similar to that of the larger metal ion-dependent mammalian PLC. The Ca(2+) ion shows a distorted square planar geometry in the active site that allows for efficient substrate binding and transition state stabilization during catalysis. Additional changes in the positions of the catalytic general acid/general base (GA/GB) were also observed, indicating the interrelation of the intricate hydrogen bonding network involved in stabilizing the active site amino acids. The functional information provided by this X-ray structure now allows for a better understanding of the catalytic mechanism, including stereochemical effects and substrate interactions, which facilitates better inhibitor design and sheds light on the possibilities of understanding how protein evolution might have occurred across this enzyme family.

Publisher Summary This chapter discusses about robotics for automated crystal formation and analysis. The choice of crystallization and imaging systems depends on user needs and requirements. Stand-alone workstations will be most appropriate for laboratory efforts or for systems that require added flexibility. Integrated robotic systems will be necessary for more HT industrial needs but with a concurrent limitation on system flexibility and cost. Large-scale structure determination programs, such as the NIH-sponsored structural genomics consortia and other commercial ventures, are relying on the parallel processing of a large number of protein samples that can be completed with this newly developed automation. This ensures that the three-dimensional structure determination can be completed as rapidly as possible for the largest number of targets. Moreover, as with other scientific equipment, the systems described in this chapter will be improved further and the costs can be reduced. Finally, this chapter concludes that small-scale structure determination projects clearly benefit from using these or next-generation robotic workstations, not only for the reliability, lowered cost, and reproducibility that such systems provide, but also for their use in enabling larger explorations of structural space.

Structural genomics is emerging as a principal approach to define protein structure–function relationships. To apply this approach on a genomic scale, novel methods and technologies must be developed to determine large numbers of structures. We describe the design and implementation of a high-throughput structural genomics pipeline and its application to the proteome of the thermophilic bacterium Thermotoga maritima . By using this pipeline, we successfully cloned and attempted expression of 1,376 of the predicted 1,877 genes (73%) and have identified crystallization conditions for 432 proteins, comprising 23% of the T. maritima proteome. Representative structures from TM0423 glycerol dehydrogenase and TM0449 thymidylate synthase-complementing protein are presented as examples of final outputs from the pipeline.

Within the tumor suppressor protein INK4 ~inhibitor of cyclin-dependent kinase 4! family, p15 INK4B is the smallest and the only one whose structure has not been determined previously, probably due to the protein’s conformational flexibility and instability. In this work, multidimensional NMR studies were performed on this protein. The first tertiary structure was built by comparative modeling with p16 INK4A as the template, followed by restrained energy minimization with NMR constraints ~NOE and H-bonds!. For this purpose, the solution structure of p16 INK4A , whose quality was also limited by similar problems, was refined with additional NMR experiments conducted on an 800 MHz spectrometer and by structure-based iterative NOE assignments. The nonhelical regions showed major improvement with root-meansquare deviation ~RMSD! improved from 1.23 to 0.68 A for backbone heavy atoms. The completion of p15 INK4B coupled with refinement of p16 INK4A made it possible to compare the structures of the four INK4 members in depth, and to compare the structures of p16 INK4A in the free form and in the p16 INK4A -CDK6 complex. This is an important step toward a comprehensive understanding of the precise functional roles of each INK4 member.

We report the first detailed structure -function analyses of p18 INK4C (p18), which is a homologue of the important tumor suppressor p16 INK4A (p16). Twenty-four mutants were designed rationally. The global conformations of the mutants were characterized by NMR, while the function was assayed by inhibition of cyclin-dependent kinase 4 (CDK4). Most of these mutants have unperturbed global structures, thus the changes in their inhibitory abilities can be attributed to the mutated residues. The important results are summarized as follows: (a) some residues at loops 1 and 2, but not 3, are important for the inhibitory function of p18, similar to the results for p16; (b) two residues at the first helix-turn-helix motif and two at the third are important for inhibition; (c) while the results generally agree with the prediction based on the crystal structures of p16-CDK6 and p19-CDK6 binary complexes, there are significant differences in a few residues, suggesting that the interactions in the binary complexes may not accurately represent the interactions in the ternary complexes (in the presence of cyclin D2); (d) most importantly, the extra loop of p18 appears to contribute to the function of p18, even though the crystal structure of the p19 INK4D -CDK6 complex indicates no interactions involving this loop; (e) detailed analyses of the crystal structures and the functional results suggest that there are notable differences in the interactions between different members of the INK4 family and CDKs.