An anti-diabetic drug targets NEET (CISD) proteins through destabilization of their [2Fe-2S] clusters.

Elevated levels of mitochondrial iron and reactive oxygen species (ROS) accompany the progression of diabetes, negatively impacting insulin production and secretion from pancreatic cells. In search for a tool to reduce mitochondrial iron and ROS levels, we arrived at a molecule that destabilizes the [2Fe-2S] clusters of NEET proteins (M1). Treatment of db/db diabetic mice with M1 improved hyperglycemia, without the weight gain observed with alternative treatments such as rosiglitazone. The molecular interactions of M1 with the NEET proteins mNT and NAF-1 were determined by X-crystallography. The possibility of controlling diabetes by molecules that destabilize the [2Fe-2S] clusters of NEET proteins, thereby reducing iron-mediated oxidative stress, opens a new route for managing metabolic aberration such as in diabetes.

Expression, purification and crystallization of CLK1 kinase - A potential target for antiviral therapy.

Cdc-like kinase 1 (CLK1) is a dual-specificity kinase capable of autophosphorylation on tyrosine residues and Ser/Thr phosphorylation of its substrates. CLK1 belongs to the CLK kinase family that regulates alternative splicing through phosphorylation of serine-arginine rich (SR) proteins. Recent studies have demonstrated that CLK1 has an important role in the replication of influenza A and chikungunya viruses. Furthermore, CLK1 was found to be relevant for the replication of HIV-1 and the West Nile virus, making CLK1 an interesting cellular candidate for the development of a host-directed antiviral therapy that might be efficient for treatment of newly emerging viruses. We describe here our attempts and detailed procedures to obtain the recombinant kinase domain of CLK1 in suitable amounts for crystallization in complex with specific inhibitors. The key solution for the reproducibility of crystals resides in devising and refining expression and purification protocols leading to homogeneous protein. Co-expression of CLK1 with λ-phosphatase and careful purification has yielded crystals of CLK1 complexed with the KH-CB19 inhibitor that diffracted to 1.65 Å. These results paved the path to the screening of more structures of CLK1 complexed compounds, leading to further optimization of their inhibitory activity. Moreover, since kinases are desired targets in numerous pathologies, the approach we report here, the co-expression of kinases with λ-phosphatase, previously used in other kinases, can be adopted as a general protocol in numerous kinase targets for obtaining reproducible and homogenic non-phosphorylated (inactive) forms suitable for biochemical and structural studies thus facilitating the development of novel inhibitors.

A point mutation in the [2Fe-2S] cluster binding region of the NAF-1 protein (H114C) dramatically hinders the cluster donor properties.

NAF-1 is an important [2Fe-2S] NEET protein associated with human health and disease. A mis-splicing mutation in NAF-1 results in Wolfram Syndrome type 2, a lethal childhood disease. Upregulation of NAF-1 is found in epithelial breast cancer cells, and suppression of NAF-1 expression by knockdown significantly suppresses tumor growth. Key to NAF-1 function is the NEET fold with its [2Fe-2S] cluster. In this work, the high-resolution structure of native NAF-1 was determined to 1.65 Šresolution (R factor = 13.5%) together with that of a mutant in which the single His ligand of its [2Fe-2S] cluster, His114, was replaced by Cys. The NAF-1 H114C mutant structure was determined to 1.58 Šresolution (R factor = 16.0%). All structural differences were localized to the cluster binding site. Compared with native NAF-1, the [2Fe-2S] clusters of the H114C mutant were found to (i) be 25-fold more stable, (ii) have a redox potential that is 300 mV more negative and (iii) have their cluster donation/transfer function abolished. Because no global structural differences were found between the mutant and the native (wild-type) NAF-1 proteins, yet significant functional differences exist between them, the NAF-1 H114C mutant is an excellent tool to decipher the underlying biological importance of the [2Fe-2S] cluster of NAF-1 in vivo.

Lipid molecules induce p38α activation via a novel molecular switch.

p38α mitogen-activated protein kinase (MAPK) is generally activated by dual phosphorylation but has also been shown to exhibit alternative activation modes. One of these modes included a direct interaction with phosphatidylinositol ether lipid analogues (PIA) inducing p38α autoactivation and apoptosis. Perifosine, an Akt inhibitor in phase II clinical trials, also showed p38α activation properties similarly to those of PIAs. The crystal structures of p38α in complex with PIA23, PIA24 and perifosine provide insights into this unique activation mode. The activating molecules bind a unique hydrophobic binding site in the kinase C'-lobe formed in part by the MAPK insert region. In addition, there are conformational changes in the short αEF/αF loop region that acts as an activation switch, inducing autophosphorylation. Structural and biochemical characterization of the αEF/αF loop identified Trp197 as a key residue in the lipid binding and in p38α catalytic activity. The lipid binding site also accommodates hydrophobic inhibitor molecules and, thus, can serve as a novel p38α-target for specific activation or inhibition, with novel therapeutic implications.

Characterization of Arabidopsis NEET reveals an ancient role for NEET proteins in iron metabolism.

The NEET family is a newly discovered group of proteins involved in a diverse array of biological processes, including autophagy, apoptosis, aging, diabetes, and reactive oxygen homeostasis. They form a novel structure, the NEET fold, in which two protomers intertwine to form a two-domain motif, a cap, and a unique redox-active labile 2Fe-2S cluster binding domain. To accelerate the functional study of NEET proteins, as well as to examine whether they have an evolutionarily conserved role, we identified and characterized a plant NEET protein. Here, we show that the Arabidopsis thaliana At5g51720 protein (At-NEET) displays biochemical, structural, and biophysical characteristics of a NEET protein. Phenotypic characterization of At-NEET revealed a key role for this protein in plant development, senescence, reactive oxygen homeostasis, and Fe metabolism. A role in Fe metabolism was further supported by biochemical and cell biology studies of At-NEET in plant and mammalian cells, as well as mutational analysis of its cluster binding domain. Our findings support the hypothesis that NEET proteins have an ancient role in cells associated with Fe metabolism.

Allostery in the ferredoxin protein motif does not involve a conformational switch.

Regulation of protein function via cracking, or local unfolding and refolding of substructures, is becoming a widely recognized mechanism of functional control. Oftentimes, cracking events are localized to secondary and tertiary structure interactions between domains that control the optimal position for catalysis and/or the formation of protein complexes. Small changes in free energy associated with ligand binding, phosphorylation, etc., can tip the balance and provide a regulatory functional switch. However, understanding the factors controlling function in single-domain proteins is still a significant challenge to structural biologists. We investigated the functional landscape of a single-domain plant-type ferredoxin protein and the effect of a distal loop on the electron-transfer center. We find the global stability and structure are minimally perturbed with mutation, whereas the functional properties are altered. Specifically, truncating the L1,2 loop does not lead to large-scale changes in the structure, determined via X-ray crystallography. Further, the overall thermal stability of the protein is only marginally perturbed by the mutation. However, even though the mutation is distal to the iron-sulfur cluster (∼20 Å), it leads to a significant change in the redox potential of the iron-sulfur cluster (57 mV). Structure-based all-atom simulations indicate correlated dynamical changes between the surface-exposed loop and the iron-sulfur cluster-binding region. Our results suggest intrinsic communication channels within the ferredoxin fold, composed of many short-range interactions, lead to the propagation of long-range signals. Accordingly, protein interface interactions that involve L1,2 could potentially signal functional changes in distal regions, similar to what is observed in other allosteric systems.

Chimeric avidin shows stability against harsh chemical conditions--biochemical analysis and 3D structure.

Avidin and its bacterial analog streptavidin have been widely used in applications in life sciences. Recently, we described a highly thermostable engineered avidin, called chimeric avidin, which is a hybrid of avidin and avidin-related protein 4. Here, we report a protocol for pilot-scale production in E. coli and the X-ray structure of chimeric avidin. The ligand-binding properties of chimeric avidin were explored with isothermal titration calorimetry. We found chimeric avidin to be more stable against various harsh organic solvents at elevated temperatures compared to avidin and streptavidin. The properties of chimeric avidin make it a potential tool for new applications in biotechnology.

Critical importance of loop conformation to avidin-enhanced hydrolysis of an active biotin ester.

The homotetrameric and biotin-binding properties of avidin and streptavidin have been exploited for a myriad of biotechnological applications and theoretical studies. Among the few differences between the two proteins is the capacity of avidin to hydrolyze biotinyl p-nitrophenyl ester (BNP), as opposed to streptavidin, which fully protects the same pseudosubstrate from hydrolysis. Combined mutagenesis and X-ray analysis have been used to attempt to understand this diametric difference in activities. It was found that a charged residue and one of the loops (L3,4) are together responsible for this difference. Recently, the avidin-related analogue AVR4 was found to have an even more pronounced BNP-hydrolysis activity than avidin. Again, the combination of charged residue(s) (Asp39 and/or Arg112) and the rigid conformation of the L3,4 loop was suggested to be responsible for the observed hydrolysis reaction. However, replacement of the latter charged residues in AVR4 resulted in only a modest reduction in hydrolytic activity at most, whereas replacement of the L3,4 loop of avidin with the rigid loop of AVR4 caused a dramatic increase in the activity of avidin. These results clearly demonstrate that the main feature responsible for the observed differences in rates of hydrolysis among the avidins is the conformational status of the L3,4 loop, which imposes conformational constraints on the pseudosubstrate, thereby rendering it susceptible to nucleophilic attack by solvent. In this context, the hydrolytic properties of the avidins reflect enzyme catalysis, in that subtleties in substrate binding are the determining features of catalytic efficiency.

Factors dictating the pseudocatalytic efficiency of avidins.

The hydrolysis of biotinyl p-nitrophenyl ester (BNP) by a series of avidin derivatives was examined. Surprisingly, a hyperthermostable avidin-related protein (AVR4) was shown to display extraordinary yet puzzling hydrolytic activity. In order to evaluate the molecular determinants that contribute to the reaction, the crystal structure of AVR4 was compared with those of avidin, streptavidin and key mutants of the two proteins in complex with biotinyl p-nitroanilide (BNA), the inert amide analogue of BNP. The structures revealed that a critical lysine residue contributes to the hydrolysis of BNP by avidin but has only a minor contribution to the AVR4-mediated reaction. Indeed, the respective rates of hydrolysis among the different avidins reflect several molecular parameters, including binding-site architecture, the availability of the ligand to solvent and the conformation of the ligand and consequent susceptibility to efficient nucleophilic attack. In avidin, the interaction of BNP with Lys111 and disorder of the L3,4 loop (and consequent solvent availability) together comprise the major driving force behind the hydrolysis, whereas in AVR4 the status of the ligand (the pseudo-substrate) is a major distinguishing feature. In the latter protein, a unique conformation of the L3,4 loop restrains the pseudo-substrate, thereby exposing the carbonyl carbon atom to nucleophilic attack. In addition, due to its conformation, the pseudo-substrate in the AVR4 complex cannot interact with the conserved lysine analogue (Lys109); instead, this function is superseded by polar interactions with Arg112. The results demonstrate that, in highly similar proteins, different residues can perform the same function and that subtle differences in the active-site architecture of such proteins can result in alternative modes of reaction.

High-resolution crystal structure of an avidin-related protein: insight into high-affinity biotin binding and protein stability.

The chicken avidin gene belongs to an extended gene family encoding seven avidin-related genes (AVRs), of which only avidin is expressed in the chicken. The sequences of AVR4 and AVR5 are identical and the common protein (AVR4) has been expressed both in insect and bacterial systems. The recombinant proteins are similarly hyperthermostable and bind biotin with similarly high affinities. AVR4 was crystallized in the apo and biotin-complexed forms and their structures were determined at high resolution. Its tertiary and quaternary structures are very similar to those of avidin and streptavidin. Its biotin-binding site shows only a few alterations compared with those of avidin and streptavidin, which account for the observed differences in binding affinities. The increased hyperthermostability can be attributed to the conformation of the critical L3,4 loop and the extensive network of 1-3 inter-monomeric interactions. The loop contains a tandem Pro-Gly sequence and an Asp-Arg ion pair that collectively induce rigidity, thus maintaining its closed and ordered conformation in both the apo and biotin-complexed forms. In addition, Tyr115 is present on the AVR4 1-3 monomer-monomer interface, which is absent in avidin and streptavidin. The interface tyrosine generates inter-monomeric interactions, i.e. a tyrosine-tyrosine pi-pi interaction and a hydrogen bond with Lys92. The resultant network of interactions confers a larger 1-3 dimer-dimer contact surface on AVR4, which correlates nicely with its higher thermostability compared with avidin and streptavidin. Several of the proposed thermostability-determining factors were found to play a role in strengthening the tertiary and quaternary integrity of AVR4.

Design and construction of highly stable, protease-resistant chimeric avidins.

The chicken avidin gene family consists of avidin and seven separate avidin-related genes (AVRs) 1-7. Avidin protein is a widely used biochemical tool, whereas the other family members have only recently been produced as recombinant proteins and characterized. In our previous study, AVR4 was found to be the most stable biotin binding protein thus far characterized (T(m) = 106.4 degrees C). In this study, we studied further the biotin-binding properties of AVR4. A decrease in the energy barrier between the biotin-bound and unbound state of AVR4 was observed when compared with that of avidin. The high resolution structure of AVR4 facilitated comparison of the structural details of avidin and AVR4. In the present study, we used the information obtained from these comparative studies to transfer the stability and functional properties of AVR4 to avidin. A chimeric avidin protein, ChiAVD, containing a 21-amino acid segment of AVR4 was found to be significantly more stable (T(m) = 96.5 degrees C) than native avidin (T(m) = 83.5 degrees C), and its biotin-binding properties resembled those of AVR4. Optimization of a crucial subunit interface of avidin by an AVR4-inspired point mutation, I117Y, significantly increased the thermostability of the avidin mutant (T(m) = 97.5 degrees C) without compromising its high biotin-binding properties. By combining these two modifications, a hyperthermostable ChiAVD(I117Y) was constructed (T(m) = 111.1 degrees C). This study provides an example of rational protein engineering in which another member of the protein family has been utilized as a source in the optimization of selected properties.

Structural elements responsible for conversion of streptavidin to a pseudoenzyme.

Avidin enhances the hydrolysis of biotinyl p-nitrophenyl ester (BNP) under mild alkaline conditions, whereas streptavidin prevents hydrolysis of BNP up to pH 12. Recently, we imposed hydrolytic activity on streptavidin by rational mutagenesis, based on the molecular elements responsible for the hydrolysis by avidin. Three mutants were designed, whereby the desired features, the distinctive L124R point mutation (M1), the L3,4 loop replacement (M2), and the combined mutation (M3), were transferred from avidin to streptavidin. The crystal structures of the mutants, in complex with biotinyl p-nitroanilide (BNA), the stable amide analogue of BNP, were determined. The results demonstrate that the point mutation alone has little effect on hydrolysis, and BNA exhibits a conformation similar to that of streptavidin. Substitution of a lengthier L3,4 loop (from avidin to streptavidin), resulted in an open conformation, thus exposing the ligand to solvent. Moreover, the amide bond of BNA was flipped relative to that of the streptavidin and M1 complexes, thus deflecting the nitro group toward Lys-121. Consequently, the leaving group potential of the nitrophenyl group of BNP is increased, and M2 hydrolyzes BNP at pH values >8.5. To better emulate the hydrolytic potential of avidin, M3 was required. The combination of loop replacement and point mutation served to further increase the leaving group potential by interaction of the nitro group with Arg-124 and Lys-121. The information derived from this study may provide insight into the design of enzymes and transfer of desired properties among homologous proteins.

Dimer-tetramer transition between solution and crystalline states of streptavidin and avidin mutants.

The biotin-binding tetrameric proteins, streptavidin from Streptomyces avidinii and chicken egg white avidin, are excellent models for the study of subunit-subunit interactions of a multimeric protein. Efforts are thus being made to prepare mutated forms of streptavidin and avidin, which would form monomers or dimers, in order to examine their effect on quaternary structure and assembly. In the present communication, we compared the crystal structures of binding site W-->K mutations in streptavidin and avidin. In solution, both mutant proteins are known to form dimers, but upon crystallization, both formed tetramers with the same parameters as the native proteins. All of the intersubunit bonds were conserved, except for the hydrophobic interaction between biotin and the tryptophan that was replaced by lysine. In the crystal structure, the binding site of the mutated apo-avidin contains 3 molecules of structured water instead of the 5 contained in the native protein. The lysine side chain extends in a direction opposite that of the native tryptophan, the void being partially filled by an adjacent lysine residue. Nevertheless, the binding-site conformation observed for the mutant tetramer is an artificial consequence of crystal packing that would not be maintained in the solution-phase dimer. It appears that the dimer-tetramer transition may be concentration dependent, and the interaction among subunits obeys the law of mass action.