A combinatorial approach toward analyzing functional elements of the Escherichia coli hemolysin signal sequence

Secretion of hemolysin is directed by a signal sequence located within its C-terminal 60 amino acids. Deletion analyses have indicated that the extreme end of this C-terminus is critical for transport; however, it is not known if this region contains structural features necessary for function. In this study, we have used a combinatorial approach to generate two contiguous 8-residue random libraries (Cterm1 and Cterm2) in the signal sequence to investigate the functional specificity of the last 16 residues. The large number of variants generated had provided us with a rich data set to determine if a restricted subset of sequences was actually required for function in the extreme C-terminus. We observed that over 90% of the random sequences in the Cterm1 region were secreted at close to wild-type level, while the Cterm2 region was more restricted with only 50% of the random sequences supporting wild-type-like transport. It appeared that, in the Cterm2 region, the relative lack of positive charge is favored for function. These findings, along with previous results, allow us to propose a model for recognition and transport of hemolysin that emphasizes secondary structure and general biophysical properties over primary sequence. This model may have implications for understanding the broad substrate specificity common among ATP-binding cassette transporters.     

3D structure and significance of the GPhiXXG helix packing motif in tetramers of the E1beta subunit of pyruvate dehydrogenase from the archeon Pyrobaculum aerophilum.

As part of a structural genomics project, we have determined the 2.0 A structure of the E1beta subunit of pyruvate dehydrogenase from Pyrobaculum aerophilum (PA), a thermophilic archaeon. The overall fold of E1beta from PA is closely similar to the previously determined E1beta structures from humans (HU) and P. putida (PP). However, unlike the HU and PP structures, the PA structure was determined in the absence of its partner subunit, E1alpha. Significant structural rearrangements occur in E1beta when its E1alpha partner is absent, including rearrangement of several secondary structure elements such as helix C. Helix C is buried by E1alpha in the HU and PP structures, but makes crystal contacts in the PA structure that lead to an apparent beta(4) tetramer. Static light scattering and sedimentation velocity data are consistent with the formation of PA E1beta tetramers in solution. The interaction of helix C with its symmetry-related counterpart stabilizes the tetrameric interface, where two glycine residues on the same face of one helix create a packing surface for the other helix. This GPhiXXG helix-helix interaction motif has previously been found in interacting transmembrane helices, and is found here at the E1alpha-E1beta interface for both the HU and PP alpha(2)beta(2) tetramers. As a case study in structural genomics, this work illustrates that comparative analysis of protein structures can identify the structural significance of a sequence motif.     

Structural requirements for interaction of peroxisomal targeting signal 2 and its receptor PEX7

The import of a subset of peroxisomal matrix proteins is mediated by the peroxisomal targeting signal 2 (PTS2). The results of our sequence and physical property analysis of known PTS2 signals and of a mutational study of the least characterized amino acids of a canonical PTS2 motif indicate that PTS2 forms an amphipathic helix accumulating all conserved residues on one side. Three-dimensional structural modeling of the PTS2 receptor PEX7 reveals a groove with an evolutionarily conserved charge distribution complementary to PTS2 signals. Mammalian two-hybrid assays and cross-complementation of a mutation in PTS2 by a compensatory mutation in PEX7 confirm the interaction site. An unstructured linker region separates the PTS2 signal from the core protein. This additional information on PTS2 signals was used to generate a PTS2 prediction algorithm that enabled us to identify novel PTS2 signals within human proteins and to describe KChIP4 as a novel peroxisomal protein.     

Mutational analysis of the linker region of EnvZ, an osmosensor in Escherichia coli.

EnvZ, a transmembrane signal transducer, is composed of a periplasmic sensor domain, transmembrane domains, and a cytoplasmic signaling domain. Between the second transmembrane domain and the cytoplasmic signaling domain there is a linker domain consisting of approximately 50 residues. In this study, we investigated the functional role of the EnvZ linker domain with respect to signal transduction. Amino acid sequence alignment of linker regions among various bacterial signal transducer proteins does not show a high sequence identity but suggests a common helix 1-loop-helix 2 structure. Among several mutations introduced in the EnvZ linker region, it was found that hydrophobic-to-charged amino acid substitutions in helix 1 and helix 2 and deletions in helix 1, loop, and helix 2 (delta14, delta8, and delta7) resulted in constitutive OmpC expression. In the linker mutant EnvZ x delta7, both kinase and phosphatase activities were significantly reduced but the ratio of kinase to phosphatase activity increased, consistent with the constitutive OmpC expression. In contrast, the purified cytoplasmic fragment of EnvZ x delta7 possessed both kinase and phosphatase activities at levels similar to those of the cytoplasmic fragment of wild-type EnvZ. In addition, the linker mutations had no direct effect on EnvZ C-terminal dimerization. These results together with previous data suggest that the linker region is not directly involved in EnvZ enzymatic activities and that it may have a crucial role in propagating a conformational change to ensure correct positioning of two EnvZ molecules within a dimer during the transmembrane signaling.     

Crystallographic investigation of peptide binding sites in the N-domain of the ClpA chaperone

Escherichia coli ClpA, an Hsp100/Clp chaperone and an integral component of the ATP-dependent ClpAP protease, participates in the dissolution and degradation of regulatory proteins and protein aggregates. ClpA consists of three functional domains: an N-terminal domain and two ATPase domains, D1 and D2. The N-domain is attached to D1 by a mobile linker and is made up of two tightly bound, identically folded alpha-helical bundles related by a pseudo 2-fold symmetry. Between the halves of the pseudo-dimer is a large flexible acidic loop that becomes better ordered upon binding of the small adaptor protein, ClpS. We have identified a number of structural features in the N-domain, including a Zn(++) binding motif, several interfaces for binding to ClpS, and a prominent hydrophobic surface area that binds peptides in different configurations. These structural motifs may contribute to binding of protein or peptide substrates with weak affinity and broad specificity. Kinetic studies comparing wild-type ClpA to a mutant ClpA with its N-domain deleted show that the N-domains contribute to the binding of a non-specific protein substrate but not of a folded substrate with the specific SsrA recognition tag. A functional model is proposed in which the N-domains in ClpA function as tentacles to weakly hold on to proteins thereby enhancing local substrate concentration.     

Statistical models for discerning protein structures containing the DNA-binding helix-turn-helix motif

A method for discerning protein structures containing the DNA-binding helix-turn-helix (HTH) motif has been developed. The method uses statistical models based on geometrical measurements of the motif. With a decision tree model, key structural features required for DNA binding were identified. These include a high average solvent-accessibility of residues within the recognition helix and a conserved hydrophobic interaction between the recognition helix and the second alpha helix preceding it. The Protein Data Bank was searched using a more accurate model of the motif created using the Adaboost algorithm to identify structures that have a high probability of containing the motif, including those that had not been reported previously.     

Common active site architecture and binding strategy of four phenylpropanoid P450s from Arabidopsis thaliana as revealed by molecular modeling

Despite extensive primary sequence diversity, crystal structures of several bacterial cytochrome P450 monooxygenases (P450s) and a single eukaryotic P450 indicate that these enzymes share a structural core of alpha-helices and beta-sheets and vary in the loop regions contacting individual substrates. To determine the extent to which individual structural features are conserved among divergent P450s existing in a single biosynthetic pathway, we have modeled the structures of four highly divergent P450s (CYP73A5, CYP84A1, CYP75B1, CYP98A3) in the Arabidopsis phenylpropanoid pathway synthesizing lignins, flavonoids and anthocyanins. Analysis of these models has indicated that, despite primary sequence identities as low as 13%, the structural cores and several loop regions of these P450s are highly conserved. Substrate docking indicated that all four enzymes employ a common strategy to identify their substrates in that their cinnamate-derived substrates align along helix I with their aromatic ring positioned towards the C-terminus of this helix and their aliphatic tails positioned towards the N-terminus. Further similarity was observed in the way the substrates contact the consensus P450 substrate recognition sites (SRS). Residues predicted to contact the aromatic ring region exist in SRS5, SRS6 and the C-terminal portion of SRS4 and residues contacting the distal end of each substrate exist in SRS1, SRS2 and the N-terminal portion of SRS4. Alignments of the regions contacting the aromatic ring region indicate that SRS4, SRS5 and SRS6 share higher degrees of sequence conservation than found in SRS1, SRS2 or the full-length protein.     

How bacterial ADP-ribosylating toxins recognize substrates

ExoS and ExoT are bifunctional type III cytotoxins of Pseudomonas aeruginosa that contain an N-terminal RhoGAP domain and a C-terminal ADP-ribosylation domain. Although they share 76% amino acid identity, ExoS and ExoT ADP-ribosylate different substrates. Using protein modeling and site-directed mutagenesis, the regions of ExoS and ExoT that define substrate specificity were determined. Regions B (active site loop), C (ARTT motif) and E (PN loop) on ExoS are necessary and sufficient to recognize ExoS targets, whereas regions B, C and E on ExoT are necessary but not sufficient to recognize ExoT targets, such as the Crk proteins. A specific Crk recognition motif on ExoT was defined as region A (helix alpha1). The electrostatic properties of regions A, B, C and E define the substrate specificity of ExoS and ExoT and these interactions can explain how other bacterial ADP-ribosylating toxins recognize their unique substrates.     

When simple sequence comparison fails: the cryptic case of the shared domains of the bacterial replication initiation proteins DnaB and DnaD

DnaD and DnaB are essential DNA-replication-initiation proteins in low-G+C content Gram-positive bacteria. Here we use sensitive Hidden Markov Model-based techniques to show that the DnaB and DnaD proteins share a common structure that is evident across all their structural domains, termed DDBH1 and DDBH2 (DnaD DnaB Homology 1 and 2). Despite strong sequence divergence, many of the DNA-binding and oligomerization properties of these domains have been conserved. Although eluding simple sequence comparisons, the DDBH2 domains share the only strong sequence motif; an extremely highly conserved YxxxIxxxW sequence that contributes to DNA binding. Sequence alignments of DnaD alone fail to identify another key part of the DNA-binding module, since it includes a poorly conserved sequence, a solvent-exposed and somewhat unstable helix and a mobile segment. We show by NMR, in vitro mutagenesis and in vivo complementation experiments that the DNA-binding module of Bacillus subtilis DnaD comprises the YxxxIxxxW motif, the unstable helix and a portion of the mobile region, the latter two being essential for viability. These structural insights lead us to a re-evaluation of the oligomerization and DNA-binding properties of the DnaD and DnaB proteins.     

Evaluation of the RNA determinants for bacterial and yeast RNase III binding and cleavage

Bacterial double-stranded RNA-specific RNase III recognizes the A-form of an RNA helix with little sequence specificity. In contrast, baker yeast RNase III (Rnt1p) selectively recognizes NGNN tetraloops even when they are attached to a B-form DNA helix. To comprehend the general mechanism of RNase III substrate recognition, we mapped the Rnt1p binding signal and directly compared its substrate specificity to that of both Escherichia coli RNase III and fission yeast RNase III (PacI). Rnt1p bound but did not cleave long RNA duplexes without NGNN tetraloops, whereas RNase III indiscriminately cleaved all RNA duplexes. PacI cleaved RNA duplexes with some preferences for NGNN-capped RNA stems under physiological conditions. Hydroxyl radical footprints indicate that Rnt1p specifically interacts with the NGNN tetraloop and its surrounding nucleotides. In contrast, Rnt1p interaction with GAAA-capped hairpins was weak and largely unspecific. Certain duality of substrate recognition was exhibited by PacI but not by bacterial RNase III. E. coli RNase III recognized RNA duplexes longer than 11 bp with little specificity, and no specific features were required for cleavage. On the other hand, PacI cleaved long, but not short, RNA duplexes with little sequence specificity. PacI cleavage of RNA stems shorter than 27 bp was dependent on the presence of an UU-UC internal loop two nucleotides upstream of the cleavage site. These observations suggest that yeast RNase IIIs have two recognition mechanisms, one that uses specific structural features and another that recognizes general features of the A-form RNA helix.     

Solution structures of the core light-harvesting alpha and beta polypeptides from Rhodospirillum rubrum: implications for the pigment-protein and protein-protein interactions

We have determined the solution structures of the core light-harvesting (LH1) alpha and beta-polypeptides from wild-type purple photosynthetic bacterium Rhodospirillum rubrum using multidimensional NMR spectroscopy. The two polypeptides form stable alpha helices in organic solution. The structure of alpha-polypeptide consists of a long helix of 32 amino acid residues over the central transmembrane domain and a short helical segment at the N terminus that is followed by a three-residue loop. Pigment-coordinating histidine residue (His29) in the alpha-polypeptide is located near the middle of the central helix. The structure of beta-polypeptide shows a single helix of 32 amino acid residues in the membrane-spanning region with the pigment-coordinating histidine residue (His38) at a position close to the C-terminal end of the helix. Strong hydrogen bonds have been identified for the backbone amide protons over the central helical regions, indicating a rigid property of the two polypeptides. The overall structures of the R.rubrum LH1 alpha and beta-polypeptides are different from those previously reported for the LH1 beta-polypeptide of Rhodobacter sphaeroides, but are very similar to the structures of the corresponding LH2 alpha and beta-polypeptides determined by X-ray crystallography. A model constructed for the structural subunit (B820) of LH1 complex using the solution structures reveals several important features on the interactions between the LH1 alpha and beta-polypeptides. The significance of the N-terminal regions of the two polypeptides for stabilizing both B820 and LH1 complexes, as clarified by many experiments, may be attributed to the interactions between the short N-terminal helix (Trp2-Gln6) of alpha-polypeptide and a GxxxG motif in the beta-polypeptide.     

Distinct structural elements of rab5 define its functional specificity.

Members of the rab family of small GTPases are localized to distinct cellular compartments and function as specific regulators of vesicle transport between organelles. Overexpression of rab5, which is associated with early endosomes and the plasma membrane, increases the rate of endocytosis [Bucci et al. (1992) Cell, 70, 715-728]. From sequence alignments and molecular modelling we identified structural elements that might contribute to the definition of the functional specificity of rab5. To test the role of these elements experimentally, we transplanted them onto rab6, which is associated with the Golgi complex. The chimeric proteins were assayed for intracellular localization and stimulation of endocytosis. First, we found that the C-terminus of rab5 could target rab6 to the plasma membrane and early endosomes but it did not confer rab5-like stimulation of endocytosis. Further replacement of other regions revealed that the N-terminus, helix alpha 2/loop 5 and helix alpha 2/loop 7 were all required to functionally convert rab6 into rab5. Reciprocal hybrids of rab5 containing these regions replaced with those of rab6 were inactive, demonstrating that each region is essential for rab5 function. These results indicate that distinct structural elements specify the localization, membrane association and regulatory function of rab5.     

Molecular cloning of the rat integrin alpha 1-subunit: a receptor for laminin and collagen

Integrin heterodimers mediate a variety of adhesive interactions, including neuronal attachment to and process outgrowth on laminin. We report here the cloning and primary sequence of an M-200 kD integrin alpha subunit that associates with the integrin beta 1 subunit to form a receptor for both laminin and collagen. Similarities in ligand-binding specificity, relative molecular mass and NH2-terminal sequence make this a strong candidate for the rat homologue of the alpha subunit of the human integrin VLA-1. The full-length rat alpha 1 cDNAs encode a protein containing a purative signal sequence and a mature polypeptide of 1,152 amino acids, with extracellular, transmembrane and cytoplasmic domains. Several structural features are conserved with other integrin alpha chains, including (a) a sequence motif repeated seven times in the NH2-terminal half; (b) potential Ca2+/Mg2+ binding sites in repeats 5, 6, and 7, and (c) alignment of at least 14 of 23 cysteine residues. This rat alpha 1 sequence also contains a 206-amino acid I domain, inserted between repeats 2 and 3, that is homologous to I domains found in the same position in the alpha subunits of several integrins (VLA-2, Mac-1, LFA-1, p150). The rat alpha 1 and human VLA-2 apha subunits share greater than 50% sequence identity in the seven repeats and I domain, suggesting that these sequence identities may underlie some of their similar ligand-binding specificities. However, the rat integrin alpha 1 subunit has several unique features, including a 38-residue insert between two Ca2+/Mg2+ binding domains, and a divergent 15-residue cytoplasmic sequence, that may potentially account for unique functions of this integrin.     

Substrate specificity of plant UDP-dependent glycosyltransferases predicted from crystal structures and homology modeling

Plant family 1 UDP-dependent glycosyltransferases (UGTs) catalyze the glycosylation of a plethora of bioactive natural products. In Arabidopsis thaliana, 120 UGT encoding genes have been identified. The crystal-based 3D structures of four plant UGTs have recently been published. Despite low sequence conservation, the UGTs show a highly conserved secondary and tertiary structure. The sugar acceptor and sugar donor substrates of UGTs are accommodated in the cleft formed between the N- and C-terminal domains. Several regions of the primary sequence contribute to the formation of the substrate binding pocket including structurally conserved domains as well as loop regions differing both with respect to their amino acid sequence and sequence length. In this review we provide a detailed analysis of the available plant UGT crystal structures to reveal structural features determining substrate specificity. The high 3D structural conservation of the plant UGTs render homology modeling an attractive tool for structure elucidation. The accuracy and utility of UGT structures obtained by homology modeling are discussed and quantitative assessments of model quality are performed by modeling of a plant UGT for which the 3D crystal structure is known. We conclude that homology modeling offers a high degree of accuracy. Shortcomings in homology modeling are also apparent with modeling of loop regions remaining as a particularly difficult task.     

Stable proline box motif at the N-terminal end of alpha-helices.

We describe a novel N-terminal alpha-helix local motif that involves three hydrophobic residues and a Pro residue (Pro-box motif). Database analysis shows that when Pro is the N-cap of an alpha-helix the distribution of amino acids in adjacent positions changes dramatically with respect to the average distribution in an alpha-helix, but not when Pro is at position N1. N-cap Pro residues are usually associated to Ile and Leu, at position N', Val at position N3 and a hydrophobic residue (h) at position N4. The side chain of the N-cap Pro packs against Val, while the hydrophobic residues at positions N' and N4 make favorable interactions. To analyze the role of this putative motif (sequence fingerprint hPXXhh), we have synthesized a series of peptides and analyzed them by circular dichroism (CD) and NMR. We find that this motif is formed in peptides, and that the accompanying hydrophobic interactions contribute up to 1.2 kcal/mol to helix stability. The fact that some of the residues in this fingerprint are not good N-cap and helix formers results in a small overall stabilization of the alpha-helix with respect to other peptides having Gly as the N-cap and Ala at N3 and N4. This suggests that the Pro-box motif will not specially contribute to protein stability but to the specificity of its fold. In fact, 80% of the sequences that contain the fingerprint sequence in the protein database are adopting the described structural motif, and in none of them is the helix extended to place Pro at the more favorable N1 position.     

Structure of human MTH1, a Nudix family hydrolase that selectively degrades oxidized purine nucleoside triphosphates

Oxygen radicals generated through normal cellular respiration processes can cause mutations in genomic and mitochondrial DNA. Human MTH1 hydrolyzes oxidized purine nucleoside triphosphates, such as 8-oxo-dGTP and 2-hydroxy-dATP, to monophosphates, thereby preventing the misincorporation of these oxidized nucleotides during replication. Here we present the solution structure of MTH1 solved by multidimensional heteronuclear NMR spectroscopy. The protein adopts a fold similar to that of Escherichia coli MutT, despite the low sequence similarity between these proteins outside the conserved Nudix motif. The substrate-binding pocket of MTH1, deduced from chemical shift perturbation experiments, is located at essentially the same position as in MutT; however, a pocket-forming helix is largely displaced in MTH1 (approximately 9 A) such that the shape of the pocket differs between the two proteins. Detailed analysis of the pocket-forming residues enabled us to identify Asn33 as one of the key residues in MTH1 for discriminating the oxidized form of purine, and mutation of this residue modifies the substrate specificity. We also show that MTH1 catalyzes hydrolysis of 8-oxo-dGTP through nucleophilic substitution of water at the beta-phosphate.     

The U1 small nuclear ribonucleoprotein (snRNP) 70K protein is transported independently of U1 snRNP particles via a nuclear localization signal in the RNA-binding domain.

Expression of the recombinant human U1-70K protein in COS cells resulted in its rapid transport to the nucleus, even when binding to U1 RNA was debilitated. Deletion analysis of the U1-70K protein revealed the existence of two segments of the protein which were independently capable of nuclear localization. One nuclear localization signal (NLS) was mapped within the U1 RNA-binding domain and consists of two typically separated but interdependent elements. The major element of this NLS resides in structural loop 5 between the beta 4 strand and the alpha 2 helix of the folded RNA recognition motif. The C-terminal half of the U1-70K protein which was capable of nuclear entry contains two arginine-rich regions, which suggests the existence of a second NLS. Site-directed mutagenesis of the RNA recognition motif NLS demonstrated that the U1-70K protein can be transported independently of U1 RNA and that its association with the U1 small nuclear ribonucleoprotein particle can occur in the nucleus.     

Distant segments of Saccharomyces cerevisiae scR1 RNA promote assembly and function of the signal recognition particle

The conserved signal recognition particle targets ribosomes synthesizing presecretory proteins to the endoplasmic reticulum membrane. Key to the activity of SRP is its ability to bind the ribosome at distant locations, the signal sequence exit and elongation factor-binding sites. These contacts are made by the S and Alu domains of SRP, respectively. We tested earlier secondary structure predictions of the Saccharomyces cerevisiae SRP RNA, scR1, and provide and test a consensus structure. The structure contains four non-conserved insertions, helices 9-12, into the core SRP RNA fold, and an extended helix 7. Using a series of scR1 mutants lacking part or all of these structural elements, we find that they are important for the RNA in both function and assembly of the RNP. About 20% of the RNA, corresponding to the outer regions of these helices, is dispensable for function. Further, we examined the role of several features within the S-domain section of the core, helix 5, and find that its length and flexibility are important for proper SRP function and become essential in the absence of helix 10, 11 and/or 7 regions. Overall, the genetic data indicate that regions of scR1 distant in both primary sequence and secondary structure have interrelated roles in the function of the complex, and possibly mediate communication between Alu and S domains during targeting.