Three-dimensional structure and dynamics of a brain specific growth inhibitory factor: metallothionein-3

The brain specific member of the metallothionein (MT) family of proteins, metallothionein-3, inhibits the growth and survival of neurons, in contrast to the ubiquitous mammalian MT isoforms, MT-1 and MT-2, that are found in most tissues and are thought to function in metal ion homeostasis and detoxification. Solution NMR was utilized to determine the structural and dynamic differences of MT-3 from MT-1 and 2. The high-resolution solution structure of the C-terminal alpha-domain of recombinant mouse MT-3 revealed a tertiary fold very similar to MT-1 and 2, except for a loop that accommodates an acidic insertion relative to these isoforms. This loop was distinguished from the rest of the domain by dynamics of the backbone on the nano- to picosecond time-scale shown by (15)N relaxation studies and was identified as a possible interaction site with other proteins. The N-terminal beta-domain contains the region responsible for the growth inhibitory activity, a CPCP tetrapeptide close to the N-terminus. Because of exchange broadening of a large number of the NMR signals from this domain, homology modeling was utilized to calculate models for the beta-domain and suggested that while the backbone fold of the MT-3 beta-domain is identical to MT-1 and 2, the second proline responsible for the activity, Pro9, may show structural heterogeneity. (15)N relaxation analyses implied fast internal motions for the beta-domain. On the basis of these observations, we conclude that the growth inhibitory activity exhibited by MT-3 is a result of a combination of local structural differences and global dynamics in the beta-domain.     

Different protective roles in vitro of alpha- and beta-domains of growth inhibitory factor (GIF) on neuron injuries caused by oxygen free radicals.

It was well known that beta-amyloid (Abeta) and tau protein play an important role in pathological procedure of Alzheimer's disease (AD), a senile dementia. The growth inhibitory factor (GIF, also named metallothionein-3, MT-3) had been demonstrated to inhibit the outgrowth of cortex neurons in the medium with extract of the AD patient brain. In our experiments, it was found that the neurons of cortex and the PC12 (pheochromocytoma) cells could be protected from the cytotoxicity of beta-amyloid 25-35 in presence of GIF and its domains. Additionally, GIF can scavenge the hydroxyl radical efficiently in CytC-VitC radical producing system and its alpha-domain shown more effective potentials than its beta-domain. The electron paramagnetic resonance spectra also show that the alpha-domain has more potential ability for eliminating reactive oxygen free radicals than its beta-domain. The results suggest that GIF could act as an efficient scavenger against free radicals in vitro and the alpha-domain in GIF molecule shows more potential in protecting against reactive oxygen species injury than the beta-domain.     

Residue-level NMR view of the urea-driven equilibrium folding transition of SUMO-1 (1-97): native preferences do not increase monotonously

SUMO-1 (1-97) is a crucial protein in the machinery of post-translational modifications. We observed by circular dichroism and fluorescence spectroscopy that urea-induced unfolding of this protein is a complex process with the possibility of occurrence of detectable intermediates along the way. The tertiary structure is completely lost around approximately 4.5 M urea with a transition mid-point at 2.53 M urea, while the secondary structure unfolding seems to show two transitions, with mid-points at 2.42 M and 5.69 M urea. We have elucidated by systematic urea titration, the equilibrium residue level structural and dynamics changes along the entire folding/unfolding transition by multidimensional NMR. With urea dilution, the protein is seen to progressively lose most of the broad beta-domain structural preferences present at 8 M urea, acquire some helical propensities at 5 M urea, and lose some of them again on further dilution of urea. Between 3 M and 2 M urea, the protein starts afresh to acquire native structural features. These observations are contrary to the conventional notion that proteins fold with monotonously increasing native-type preferences. For folding below approximately 3 M urea, the region around the alpha1 helix appears to be a potential folding initiation site. The folding seems to start with a collapse into native-like topologies, at least in parts, and is followed by formation of secondary and tertiary structure, perhaps by cooperative rearrangements. The motional characteristics of the protein show sequence-dependent variation as the concentration of urea is progressively reduced. At the sub-nanosecond level, the features are extremely unusual for denatured states, and only certain segments corresponding to the flexible regions in the native protein display these motions at the different concentrations of urea.     

Effect of the two conserved prolines of human growth inhibitory factor (metallothionein-3) on its biological activity and structure fluctuation: comparison with a mutant protein

Human neuronal growth inhibitory factor, a metalloprotein classified as metallothionein-3 (MT-3), impairs the survival and the neurite formation of cultured neurons. In these studies the double P7S/P9A mutant (mutMT-3) and single mutants P7S and P9A of human Zn(7)-MT-3 were generated, and their effects on the biological activity and the structure of the protein were examined. The biological results clearly established the necessity of both proline residues for the inhibitory activity, as even single mutants were found to be inactive. Using electronic absorption, circular dichroism (CD), magnetic CD (MCD), and (113)Cd NMR spectroscopy, the structural features of the metal-thiolate clusters in the double mutant Cd(7)-mutMT-3 were investigated and compared with those of wild-type Cd(7)-MT-3 [Faller, P., Hasler, D. W., Zerbe, O., Klauser, S., Winge, D. R., and Vasák, M. (1999) Biochemistry 38, 10158] and the well characterized Cd(7)-MT-2a from rabbit liver. Similarly to (113)Cd(7)-MT-3 the (113)Cd NMR spectrum of (113)Cd(7)-mutMT-3 at 298 K revealed four major and three minor resonances (approximately 20% of the major ones) between 590 and 680 ppm, originating from a Cd(4)S(11) cluster in the alpha-domain and a Cd(3)S(9) cluster in the beta-domain, respectively. Due to the presence of dynamic processes in the structure of MT-3 and mutMT-3, all resonances showed the absence of resolved homonuclear [(113)Cd-(113)Cd] couplings and large apparent line widths (between 140 and 350 Hz). However, whereas in (113)Cd(7)-mutMT-3 the temperature rise to 323 K resulted in a major recovery of the originally NMR nondetectable population of the Cd(3)S(9) cluster resonances, no such temperature effect was observed in (113)Cd(7)-MT-3. To account for the observed NMR features, a dynamic structural model for the beta-domain is proposed, which involves a folded and a partially unfolded state. It is suggested that in the partially unfolded state a slow cis/trans isomerization of Cys-Pro(7) or Cys-Pro(9) amide bonds in (113)Cd(7)-MT-3 takes place and that this process represents a rate-limiting step in a correct domain refolding. In addition, closely similar apparent stability constants of human MT-3, mutMT-3, and rabbit MT-2a with Cd(II) and Zn(II) ions were found. These results suggest that specific structural features dictated by the repetitive (Cys-Pro)(2) sequence in the beta-domain of MT-3 and not its altered metal binding affinity compared to MT-1/MT-2 isoforms are responsible for the biological activity of this protein.     

Local structural preferences and dynamics restrictions in the urea-denatured state of SUMO-1: NMR characterization

We have investigated by multidimensional NMR the structural and dynamic characteristics of the urea-denatured state of activated SUMO-1, a 97-residue protein belonging to the growing family of ubiquitin-like proteins involved in post-translational modifications. Complete backbone amide and 15N resonance assignments were obtained in the denatured state by using HNN and HN(C)N experiments. These enabled other proton assignments from TOCSY-HSQC spectra. Secondary Halpha chemical shifts and 1H-1H NOE indicate that the protein chain in the denatured state has structural preferences in the broad beta-domain for many residues. Several of these are seen to populate the (phi,psi) space belonging to polyproline II structure. Although there is no evidence for any persistent structures, many contiguous stretches of three or more residues exhibit structural propensities suggesting possibilities of short-range transient structure formation. The hetero-nuclear 1H-15N NOEs are extremely weak for most residues, except for a few at the C-terminal, and the 15N relaxation rates show sequence-wise variation. Some of the regions of slow motions coincide with those of structural preferences and these are interspersed by highly flexible residues. The implications of these observations for the early folding events starting from the urea-denatured state of activated SUMO-1 have been discussed.     

Three-dimensional structure of the flavoenzyme acyl-CoA oxidase-II from rat liver, the peroxisomal counterpart of mitochondrial acyl-CoA dehydrogenase

Acyl-CoA oxidase (ACO) catalyzes the first and rate-determining step of the peroxisomal beta-oxidation of fatty acids. The crystal structure of ACO-II, which is one of two forms of rat liver ACO (ACO-I and ACO-II), has been solved and refined to an R-factor of 20.6% at 2.2-A resolution. The enzyme is a homodimer, and the polypeptide chain of the subunit is folded into the N-terminal alpha-domain, beta-domain, and C-terminal alpha-domain. The X-ray analysis showed that the overall folding of ACO-II less C-terminal 221 residues is similar to that of medium-chain acyl-CoA dehydrogenase (MCAD). However, the N-terminal alpha- and beta-domains rotate by 13 with respect to the C-terminal alpha-domain compared with those in MCAD to give a long and large crevice that accommodates the cofactor FAD and the substrate acyl-CoA. FAD is bound to the crevice between the beta- and C-terminal domains with its adenosine diphosphate portion interacting extensively with the other subunit of the molecule. The flavin ring of FAD resides at the active site with its si-face attached to the beta-domain, and is surrounded by active-site residues in a mode similar to that found in MCAD. However, the residues have weak interactions with the flavin ring due to the loss of some of the important hydrogen bonds with the flavin ring found in MCAD. The catalytic residue Glu421 in the C-terminal alpha-domain seems to be too far away from the flavin ring to abstract the alpha-proton of the substrate acyl-CoA, suggesting that the C-terminal domain moves to close the active site upon substrate binding. The pyrimidine moiety of flavin is exposed to the solvent and can readily be attacked by molecular oxygen, while that in MCAD is protected from the solvent. The crevice for binding the fatty acyl chain is 28 A long and 6 A wide, large enough to accommodate the C23 acyl chain.     

NMR structural study of two-disulfide variant of hen lysozyme: 2SS[6-127, 30-115]--a disulfide intermediate with a partly unfolded structure

The 15N-labeled recombinant hen lysozyme and two species of two-disulfide variants, denoted as 2SS[6-127, 30-115] and 2SS[64-80, 76-94], were studied by means of NMR spectroscopy. The former variant contains two disulfide bridges in the alpha-domain, while the latter has one disulfide bridge in the beta-domain and the other one at the interface between two domains. Resonance assignments were performed using 3D TOCSY-HSQC and NOESY-HSQC spectra. The 15N-1H-HSQC spectrum of 2SS[6-127, 30-115] was similar to that of recombinant lysozyme as a whole, although a number of cross-peaks disappeared. On the other hand, the HSQC spectrum of 2SS[64-80, 76-94] was characteristic of unfolded proteins. The structure of 2SS[6-127, 30-115] was thoroughly examined on the basis of NOE contacts determined by NMR spectroscopy. The structure of the alpha-domain was quite similar to that of authentic lysozyme, while the beta-domain was largely unstructured. However, NMR data clearly demonstrated that some residual structures exist in the beta-domain. The beta1 and beta2 strands were maintained stably as an antiparallel beta-sheet. In addition, the residues 55 and 56 were located in the vicinity of the end of the B-helix. Further, the C-helix was properly set with side-chains of I88, V92, K96, and V99 facing toward the hydrophobic core in the alpha-domain. These residual structures inherent in the amino acid sequence were evaluated concerning the folding process of lysozyme. Our experiments imply that the establishment of the backbone conformation ranging from residues 76-99 plays a key role in attaining the cooperativity between two domains required for the folding transition.     

Structural polymorphism of chromodomains in Chd1

Chromodomain from heterochromatin protein 1 and polycomb protein is known to be a lysine-methylated histone H3 tail-binding module. Chromo-helicase/ATPase DNA-binding protein 1 (CHD1) is an ATP-dependent chromatin remodeling factor, containing two tandem chromodomains. In human CHD1, both chromodomains are essential for specific binding to a K4 methylated histone H3 (H3 MeK4) peptide and are found to bind cooperatively in the crystal structure. For the budding yeast homologue, Chd1, the second but not the first chromodomain was once reported to bind to an H3 MeK4 peptide. Here, we reveal that neither the second chromodomain nor a region containing tandem chromodomains from yeast Chd1 bind to any lysine-methylated or arginine-methylated histone peptides that we examined. In addition, we examined the structures of the chromodomains from Chd1 by NMR. Although the tertiary structure of the region containing tandem chromodomains could not be obtained, the secondary structure deduced from NMR is well conserved in the tertiary structures of the corresponding first and second chromodomains determined individually by NMR. Both chromodomains of Chd1 demonstrate a structure similar to that of the corresponding part of CHD1, consisting of a three-stranded beta-sheet followed by a C-terminal alpha-helix. However, an additional helix between the first and second beta-strands, which is found in both of the first chromodomains of Chd1 and CHD1, is positioned in an entirely different manner in Chd1 and CHD1. In human CHD1 this helix forms the peptide-binding site. The amino acid sequences of the chromodomains could be well aligned on the basis of these structures. The alignment showed that yeast Chd1 lacks several key functional residues, which are responsible for specific binding to a methylated lysine residue in other chromodomains. Chd1 is likely to have no binding affinity for any H3 MeK peptide, as found in other chromodomain proteins.     

Construction of alpha-alpha domains structure in recombinant monkey metallothionein-1

The differences in metal-thiolate coordination and reactivity of mammalian metallothionein (MT) domains are closely related to their distinct, highly conservative cysteine number and position. Monkey metallothionein-1, containing a beta-domain with Cd(3)S(9) cluster and an alpha-domain with Cd(4)S(11) cluster, was used to evaluate the role of cysteine residues in the formation of MT's metal-thiolate clusters. The possible influence of cysteine residues on the binding and stability of MT domains has been examined with the metallothionein mutants: N4C, T27C and N4C/T27C, which possess ten or eleven cysteine residues in the re-constructed beta-domain, respectively. Assisted by study of UV, CD and electrospray ionization mass spectroscopy (ESI-MS) and their reactivity with DTNB (5,5'-dithiobis (2-nitrobenzoic acid)), we found that besides the original alpha-domain, some kinds of new domain containing 4-cadmium-thiolate clusters were formed in the N4C and N4C/T27C mutants of mkMT1. These new domains displayed metal binding and kinetic reactivity with DTNB similar to the alpha-domain. However, the thermal stability of the mutants was less stable than that of WT mkMT1. This might result from the disturbance of the inter-domains hydrogen bonds and of the non-cysteine amino acid residue arrangement.     

Crystal structure of a novel bacterial cell-surface flagellin binding to a polysaccharide

Bacterial flagellins are generally self-assembled into extracellular flagella for cell motility. However, the flagellin homologue p5 is found on the cell surface of Sphingomonas sp. A1 (strain A1) and binds tightly to the alginate polysaccharide. To assimilate alginate, strain A1 forms a mouthlike pit on the cell surface and concentrates the polymer in the pit. p5 is a candidate receptor that recognizes extracellular alginate and controls pit formation. To improve our understanding of the structure and function of p5, we determined the crystal structure of truncated p5 (p5DeltaN53C45) at 2.0 A resolution. This, to our knowledge, is the first structure of flagellin_IN motif-containing flagellin. p5DeltaN53C45 consists of two domains: an alpha-domain rich in alpha-helices that forms the N- and C-terminal regions and a beta-domain rich in beta-strands that constitutes the central region. The alpha-domain is structurally similar to the D1 domain of Salmonella typhimurium flagellin, while the beta-domain is structurally similar to the finger domain of the bacteriophage T4 baseplate protein that is important for intermolecular interactions between baseplate and a long or short tail fiber. Results from the deletion mutant analysis suggest that residues 20-40 and 353-363 are responsible for alginate binding. Truncated N- and C-terminal regions are thought to constitute alpha-helices extending from the alpha-domain. On the basis of the size and surface charge, the cleft in extended alpha-helices is proposed as an alginate binding site of p5. Structural similarity in the beta-domain suggests that the beta-domain is involved in the proper localization and/or orientation of p5 on the cell surface.     

Three-dimensional solution structure of mouse [Cd7]-metallothionein-1 by homonuclear and heteronuclear NMR spectroscopy

Sequential 1H-NMR assignments of mouse [Cd7]-metallothionein-1 (MT1) have been carried out by standard homonuclear NMR methods and the use of an accordion-heteronuclear multiple quantum correlation (HMQC) experiment for establishing the metal, 113Cd2+, to cysteine connectivities. The three-dimensional structure was then calculated using the distance constraints from two-dimensional nuclear Overhauser effect (NOE) spectroscopy spectra and the Cys-Cd connectivities as input for a distance geometry-dynamical simulated annealing protocol in X-PLOR 3.851. Similar to the mammalian MT2 isoforms, the homologous primary structure of MT1 suggested two separate domains, each containing one metal cluster. Because there were no interdomain constraints, the structure calculation for the N-terminal beta- and the C-terminal alpha-domain were carried out separately. The structures are based on 409 NMR constraints, consisting of 381 NOEs and 28 cysteine-metal connectivities. The only elements of regular secondary structure found were two short stretches of 3(10) helices along with some half-turns in the alpha-domain. Structural comparison with rat liver MT2 showed high similarity, with the beta-domain structure in mouse MT1 showing evidence of increased flexibility compared to the same domain in MT2. The latter was reflected by the presence of fewer interresidue NOEs, no slowly exchanging backbone amide protons, and enhanced cadmium-cadmium exchange rates found in the beta-domain of MT1.     

Redox labile site in a Zn4 cluster of Cu4,Zn4-metallothionein-3

Human metallothionein-3 (MT-3) is a neuronal inhibitory factor mainly expressed in brain and downregulated in Alzheimer's disease. The neuroinhibitory activity has been established for native Cu(4),Zn(3)-MT-3 and recombinant Zn(7)-MT-3. However, there is only limited knowledge about the structure and properties of the former metalloform. We have now generated native-like MT-3 through direct Cu(I) and Zn(II) incorporation into the recombinant apoprotein. Its characterization revealed monomeric Cu(4),Zn(4)-MT-3 containing metal-thiolate clusters located in two mutually interacting protein domains, a Cu(4) cluster in the beta-domain and a Zn(4) cluster in the alpha-domain. Using the PC12 cell line, the nontoxic nature of the protein was demonstrated. The results of electronic absorption and Cu(I) luminescence at 77 K showed that the Cu(4) cluster possesses an unprecedented stability in air. In contrast, the Zn(4) cluster is air sensitive. Its oxidation results in the release of one Zn(II) and the formation of a Zn(3) cluster, i.e., Cu(4),Zn(3)-MT-3. This process can be prevented or reversed under reducing conditions. The determined apparent stability constant for the Zn(4) cluster of 2.4 x 10(11) M(-1) is similar to that obtained for other zinc-containing MTs. This suggests that a substantially increased nucleophilic reactivity of specific thiolate ligands is responsible for this effect. Thus, the Zn(4) cluster in MT-3 may play a redox-dependent regulatory role.     

The Solution Structure of the Regulatory Domain of Tyrosine Hydroxylase

Tyrosine hydroxylase (TyrH) catalyzes the hydroxylation of tyrosine to form 3,4-dihydroxyphenylalanine in the biosynthesis of the catecholamine neurotransmitters. The activity of the enzyme is regulated by phosphorylation of serine residues in a regulatory domain and by binding of catecholamines to the active site. Available structures of TyrH lack the regulatory domain, limiting the understanding of the effect of regulation on structure. We report the use of NMR spectroscopy to analyze the solution structure of the isolated regulatory domain of rat TyrH. The protein is composed of a largely unstructured N-terminal region (residues 1–71) and a well-folded C-terminal portion (residues 72–159). The structure of a truncated version of the regulatory domain containing residues 65–159 has been determined and establishes that it is an ACT domain. The isolated domain is a homodimer in solution, with the structure of each monomer very similar to that of the core of the regulatory domain of phenylalanine hydroxylase. Two TyrH regulatory domain monomers form an ACT domain dimer composed of a sheet of eight strands with four α-helices on one side of the sheet. Backbone dynamic analyses were carried out to characterize the conformational flexibility of TyrH_65–159. The results provide molecular details critical for understanding the regulatory mechanism of TyrH.     

Molecular basis of Bcl-xL's target recognition versatility revealed by the structure of Bcl-xL in complex with the BH3 domain of Beclin-1

Beclin-1, originally identified as a Bcl-2 binding protein, is an evolutionarily conserved protein required for autophagy. The direct interaction between Beclin-1 and Bcl-2 or Bcl-xL provides a potential convergence point for apoptosis and autophagy, two programmed cell death processes. Given the functional significance of the interaction between Beclin-1 and Bcl-2/Bcl-xL, we performed detailed biochemical and structural characterizations of this interaction. We demonstrated that the Bcl-xL-binding domain of Beclin-1 contains a BH3 domain. Therefore, Beclin-1 is a new member of the BH3-only family proteins. The structure of Bcl-xL in complex with the Beclin-1 BH3 domain was determined at high resolution by NMR spectroscopy. Although similar to other known BH3 domains, the Beclin-1 BH3 domain displays its own distinct features in the complex with Bcl-xL. Systematic analysis of all known Bcl-xL/BH3 domain complexes helped us to identify the molecular basis underlying the capacity of Bcl-xL to recognize diverse target sequences.     

Evidence for non-isostructural replacement of Zn(2+) with Cd(2+) in the beta-domain of brain-specific metallothionein-3

Metallothionein-3 (MT-3) is a brain-specific MT, which is downregulated in Alzheimer's disease. The N-terminal region of CdMT-3 is highly dynamic and has escaped structural characterization by nuclear magnetic resonance. We have used electrospray ionization mass spectrometry to probe conformational states of cadmium- and zinc-substituted metalloforms of MT-3 and can demonstrate that the N-terminal beta-domain of MT-3 filled with Cd(2+) has a more open conformation than that filled with Zn(2+). The results suggest that the larger Cd(2+) ions cannot isostructurally replace zinc in the beta-domain of MT-3 whereas in the case of MT-1 and MT-2 the replacement is isostructural. Specific metal binding properties of the beta-domain of MT-3 may be essential for fulfilling the specific role of MT-3 in the brain.     

NMR analysis of a kinetically trapped intermediate of a disulfide-deficient mutant of the starch-binding domain of glucoamylase

A thermally unfolded disulfide-deficient mutant of the starch-binding domain of glucoamylase refolds into a kinetically trapped metastable intermediate when subjected to a rapid lowering of temperature. We attempted to characterise this intermediate using multidimensional NMR spectroscopy. The ¹H-¹⁵N heteronuclear single quantum coherence spectrum after a rapid temperature decrease (the spectrum of the intermediate) showed good chemical shift dispersion but was significantly different from that of the native state, suggesting that the intermediate adopts a nonnative but well-structured conformation. Large chemical shift changes for the backbone amide protons between the native and the intermediate states were observed for residues in the β-sheet consisting of strands 2, 3, 5, 6, and 7 as well as in the C-terminal region. These residues were found to be in close proximity to aromatic residues, suggesting that the chemical shift changes are mainly due to ring current shifts caused by the aromatic residues. The two-dimensional nuclear Overhauser enhancement (NOE) spectroscopy experiments showed that the intermediate contained substantial, native-like NOE connectivities, although there were fewer cross peaks in the spectrum of the intermediate compared with that of the native state. It was also shown that there were native-like interresidue NOEs for residues buried in the protein, whereas many of the NOE cross peaks were lost for the residues involved in a surface-exposed aromatic cluster. These results suggest that, in the intermediate, the aromatic cluster at the surface is structurally less organised, whereas the interior of the protein has relatively rigid, native-like side-chain packing.     

Characterization of the unfolded state of bovine alpha-lactalbumin and comparison with unfolded states of homologous proteins

The unfolded states of three homologous proteins with a very similar fold have been investigated by heteronuclear NMR spectroscopy. Secondary structure propensities as derived from interpretation of chemical shifts and motional restrictions as evidenced by heteronuclear (15)N relaxation rates have been analyzed in the reduced unfolded states of hen lysozyme and the calcium-binding proteins bovine alpha-lactalbumin and human alpha-lactalbumin. For all three proteins, significant deviations from random-coil predictions can be identified; in addition, the unfolded states also differ from each other, despite the fact that they possess very similar structures in their native states. Deviations from random-coil motional properties are observed in the alpha- and the beta-domain in bovine alpha-lactalbumin and lysozyme, while only regions within the alpha-domain deviate in human alpha-lactalbumin. The motional restrictions and residual secondary structure are determined both by the amino acid sequence of the protein and by residual long-range interactions. Even a conservative single point mutation from I to L in a highly conserved region between the two alpha-lactalbumins results in considerable differences in the motional properties. Given the differences in oxidative folding between hen lysozyme and alpha-lactalbumin, the results obtained on the unfolded states suggest that residual long-range interactions, i.e., those between the alpha- and the beta-domain of lysozyme, may act as nucleation sites for protein folding, while this property of residual structure is replaced by the calcium-binding site between the domains in alpha-lactalbumin.     

The Effects of CapZ Peptide (TRTK-12) Binding to S100B-Ca as Examined by NMR and X-ray Crystallography

Structure-based drug design is underway to inhibit the S100B-p53 interaction as a strategy for treating malignant melanoma. X-ray crystallography was used here to characterize an interaction between Ca²⁺-S100B and TRTK-12, a target that binds to the p53-binding site on S100B. The structures of Ca²⁺-S100B (1.5-Å resolution) and S100B-Ca²⁺-TRTK-12 (2.0-Å resolution) determined here indicate that the S100B-Ca²⁺-TRTK-12 complex is dominated by an interaction between Trp7 of TRTK-12 and a hydrophobic binding pocket exposed on Ca²⁺-S100B involving residues in helices 2 and 3 and loop 2. As with an S100B-Ca²⁺-p53 peptide complex, TRTK-12 binding to Ca²⁺-S100B was found to increase the protein's Ca²⁺-binding affinity. One explanation for this effect was that peptide binding introduced a structural change that increased the number of Ca²⁺ ligands and/or improved the Ca²⁺ coordination geometry of S100B. This possibility was ruled out when the structures of S100B-Ca²⁺-TRTK-12 and S100B-Ca²⁺ were compared and calcium ion coordination by the protein was found to be nearly identical in both EF-hand calcium-binding domains (RMSD = 0.19). On the other hand, B-factors for residues in EF2 of Ca²⁺-S100B were found to be significantly lowered with TRTK-12 bound. This result is consistent with NMR ¹⁵N relaxation studies that showed that TRTK-12 binding eliminated dynamic properties observed in Ca²⁺-S100B. Such a loss of protein motion may also provide an explanation for how calcium-ion-binding affinity is increased upon binding a target. Lastly, it follows that any small-molecule inhibitor bound to Ca²⁺-S100B would also have to cause an increase in calcium-ion-binding affinity to be effective therapeutically inside a cell, so these data need to be considered in future drug design studies involving S100B.     

Energetics of structural domains in alpha-lactalbumin.

alpha-Lactalbumin is a small, globular protein that is stabilized by four disulfide bonds and contains two structural domains. One of these domains is rich in alpha-helix (the alpha-domain) and has Cys 6-Cys 120 and Cys 28-Cys 111 disulfide bonds. The other domain is rich in beta-sheet (the beta-domain), has Cys 61-Cys 77 and Cys 73-Cys 91 disulfide bonds, and includes one calcium binding site. To investigate the interaction between domains, we studied derivatives of bovine alpha-lactalbumin differing in the number of disulfide bonds, using calorimetry and CD at different temperatures and solvent conditions. The three-disulfide form, having a reduced Cys 6-Cys 120 disulfide bond with carboxymethylated cysteines, is similar to intact alpha-lactalbumin in secondary and tertiary structure as judged by its ellipticity in the near and far UV. the two-disulfide form of alpha-lactalbumin, having reduced Cys 6-Cys 120 and Cys 28-Cys 111 disulfide bonds with carboxymethylated cysteines, retains about half the secondary and tertiary structure of the intact alpha-lactalbumin. The remaining structure is able to bind calcium and unfolds cooperatively upon heating, although at lower temperature and with significantly lower enthalpy and entropy. We conclude that, in the two disulfide form, alpha-lactalbumin retains its calcium-binding beta-domain, whereas the alpha-domain is unfolded. It appears that the beta-domain does not require alpha-domain to fold, but its structure is stabilized significantly by the presence of the adjacent folded alpha-domain.     

Differences in the electrostatic surfaces of the type III secretion needle proteins PrgI, BsaL, and MxiH

Gram-negative bacteria use a needle-like protein assembly, the type III secretion apparatus, to inject virulence factors into target cells to initiate human disease. The needle is formed by the polymerization of approximately 120 copies of a small acidic protein that is conserved among diverse pathogens. We previously reported the structure of the BsaL needle monomer from Burkholderia pseudomallei by nuclear magnetic resonance (NMR) spectroscopy and others have determined the crystal structure of the Shigella flexneri MxiH needle. Here, we report the NMR structure of the PrgI needle protein of Salmonella typhimurium, a human pathogen associated with food poisoning. PrgI, BsaL, and MxiH form similar two helix bundles, however, the electrostatic surfaces of PrgI differ radically from those of BsaL or MxiH. In BsaL and MxiH, a large negative area is on a face formed by the helix alpha1-alpha2 interface. In PrgI, the major negatively charged surface is not on the "face" but instead is on the "side" of the two-helix bundle, and only residues from helix alpha1 contribute to this negative region. Despite being highly acidic proteins, these molecules contain large basic regions, suggesting that electrostatic contacts are important in needle assembly. Our results also suggest that needle-packing interactions may be different among these bacteria and provide the structural basis for why PrgI and MxiH, despite 63% sequence identity, are not interchangeable in S. typhimurium and S. flexneri.