Asymmetric interactions in the adenosine-binding pockets of the MS2 coat protein dimer

Background  

The X-ray structure of the MS2 coat protein-operator RNA complex reveals the existence of quasi-synmetric interactions of adenosines -4 and -10 in pockets formed on different subunits of the coat protein dimer. Both pockets utilize the same five amino acid residues, namely Val29, Thr45, Ser47, Thr59, and Lys61. We call these sites the adenosine-binding pockets.     

Crystal structures of MS2 coat protein mutants in complex with wild-type RNA operator fragments.

In MS2 assembly of phage particles results from an interaction between a coat protein dimer and a stem-loop of the RNA genome (the operator hairpin). Amino acid residues Thr45, which is universally conserved among the small RNA phages, and Thr59 are part of the specific RNA binding pocket and interact directly with the RNA; the former through a hydrogen bond, the latter through hydrophobic contacts. The crystal structures of MS2 protein capsids formed by mutants Thr45Ala and Thr59Ser, both with and without the 19 nt wild-type operator hairpin bound, are reported here. The RNA hairpin binds to these mutants in a similar way to its binding to wild-type protein. In a companion paper both mutants are shown to be deficient in RNA binding in an in vivo assay, but in vitro the equilibrium dissociation constant is significantly higher than wild-type for the Thr45Ala mutant. The change in binding affinity of the Thr45Ala mutant is probably a direct consequence of removal of direct hydrogen bonds between the protein and the RNA. The properties of the Thr59Ser mutant are more difficult to explain, but are consistent with a loss of non-polar contact.     

Combinatorial control of protein phosphatase-1.

The catalytic subunit of the type 1 Ser/Thr protein phosphatases (PP1) can interact with many different regulatory (R) subunits. These R subunits function as activity-modulators, targeting subunits and/or substrates. The specificity of the R subunits can be accounted for by their interaction with specific subsets of binding pockets on the catalytic subunit and by the presence of subcellular targeting sequences. Hormones, growth factors and metabolites control the function of PP1 holoenzymes mainly by modulating the interaction of the subunits.     

Substitutions of Thr30 provide mechanistic insight into tryptophan-mediated activation of TRAP binding to RNA

TRAP is an 11 subunit RNA binding protein that regulates expression of genes involved in tryptophan biosynthesis and transport in Bacillus subtilis. TRAP is activated to bind RNA by binding up to 11 molecules of l-tryptophan in pockets formed by adjacent subunits. The precise mechanism by which tryptophan binding activates TRAP is not known. Thr30 is in the tryptophan binding pocket. A TRAP mutant in which Thr30 is substituted with Val (T30V) does not bind tryptophan but binds RNA constitutively, suggesting that Thr30 plays a key role in the activation mechanism. We have examined the effects of other substitutions of Thr30. TRAP proteins with small beta-branched aliphatic side chains at residue 30 bind RNA constitutively, whereas those with a small polar side chain show tryptophan-dependent RNA binding. Several mutant proteins exhibited constitutive RNA binding that was enhanced by tryptophan. Although the tryptophan and RNA binding sites on TRAP are distinct and are separated by approximately 7.5 A, several substitutions of residues that interact with the bound RNA restored tryptophan binding to T30V TRAP. These observations support the hypothesis that conformational changes in TRAP relay information between the tryptophan and RNA binding sites of the protein.     

Rearrangements in a hydrophobic core region mediate cAMP action in the regulatory subunit of PKA

cAMP-dependent protein kinase (PKA) forms an inactive heterotetramer of two regulatory (R; with two cAMP-binding domains A and B each) and two catalytic (C) subunits. Upon the binding of four cAMP molecules to the R dimer, the monomeric C subunits dissociate. Based on sequence analysis of cyclic nucleotide-binding domains in prokaryotes and eukaryotes and on crystal structures of cAMP-bound R subunit and cyclic nucleotide-free Epac (exchange protein directly activated by cAMP), four amino acids were identified (Leu203, Tyr229, Arg239 and Arg241) and probed for cAMP binding to the R subunits and for R/C interaction. Arg239 and Arg241 (mutated to Ala and Glu) displayed no differences in the parameters investigated. In contrast, Leu203 (mutated to Ala and Trp) and Tyr229 (mutated to Ala and Thr) exhibited up to 30-fold reduced binding affinity for the C subunit and up to 120-fold reduced binding affinity for cAMP. Tyr229Asp showed the most severe effects, with 350-fold reduced affinity for cAMP and no detectable binding to the C subunit. Based on these results and structural data in the cAMP-binding domain, a switch mechanism via a hydrophobic core region is postulated that is comparable to an activation model proposed for Epac.     

Probing the kinetics of formation of the bacteriophage MS2 translational operator complex: identification of a protein conformer unable to bind RNA

We have investigated the kinetics of complex formation between bacteriophage MS2 coat protein subunits and synthetic RNA fragments encompassing the natural translational operator site, or the consensus sequences of three distinct RNA aptamer families, which are known to bind to the same site on the protein. Reactions were assayed using stopped-flow fluorescence spectroscopy and either the intrinsic tryptophan fluorescence of the protein or the signals from RNA fragments site-specifically substituted with the fluorescent adenosine analogue 2'-deoxy, 2-aminopurine. The kinetics observed were independent of the fluorophore being monitored or its position within the complex, indicating that the data report global events occurring during complex formation. Competition assays show that the complex being formed consists of a single coat protein dimer and one RNA molecule. The binding reaction is at least biphasic. The faster phase, constituting 80-85 % of the amplitude, is a largely diffusion driven RNA-protein interaction (k1 approximately 2x10(9) M(-1) s(-1)). The salt dependence of the forward reaction and the similarities of the on-rates of lower-affinity RNA fragments are consistent with a diffusion-controlled step dominated by electrostatic steering. The slower phase is independent of reactant concentration, and appears to correspond to isomerisation of the coat protein subunit(s) prior to RNA binding (k(iso) approximately 0.23 s(-1)). Measurements with a coat protein mutant (Pro78Asn) show that this phase is not due to cis-trans isomerisation at this residue. The conformational changes in the protein ligand during formation of an RNA-protein complex might play a role in the triggering of capsid self-assembly and a model for this is discussed.     

Calcium-dependent conformational changes in the 36-kDa subunit of intestinal protein I related to the cellular 36-kDa target of Rous sarcoma virus tyrosine kinase

Protein I from intestinal epithelium is biochemically and immunologically related to the fibroblast 36-kDa substrate of the Rous sarcoma virus-encoded tyrosine protein kinase (Gerke and Weber (1984) EMBO J. 3, 227-233). Protein I is a Ca2+-binding protein containing two copies each of a 36- and 10-kDa subunit. Denaturation/renaturation experiments show that the 36-kDa subunit is a monomer, whereas the 10-kDa subunit forms a dimer. Mixing of the subunits leads to reconstituted protein I. Physicochemical properties of protein I and its isolated subunits reveal a Ca2+-dependent conformational change in the 36-kDa subunit which involves the exposure of 1 or more tyrosine residues to a more aqueous environment. This change points to a Ca2+ binding constant of about 10(4) M-1 in the presence of 2 mM Mg2+ and induces the ability of protein I and the 36-kDa subunit to bind in vitro to F-actin and nonerythroid spectrin. The same high Ca2+ requirement has been reported for the in vitro tyrosine phosphorylation of a 35-kDa protein from A-431 carcinoma cells by the epidermal growth factor receptor kinase (Fava and Cohen (1984) J. Biol. Chem. 259, 2636-2645). Here we show that this 35-kDa substrate is biochemically and immunologically related to the 36-kDa subunit of protein I, which in turn corresponds to the substrate of the Rous sarcoma virus kinase. The protein of A-431 cells exists not only as a monomer but also as a dimer. The latter fraction contains a 10-kDa polypeptide immunologically related to the corresponding subunit of protein I. Given past results on the A-431 system, we speculate that the monomer rather than the dimer is the preferred in vitro substrate for the epidermal growth factor receptor kinase. Thus, the 10-kDa subunit, which induces dimerization of the phosphorylatable large subunit, may act as an inhibitor.     

Complementation of RNA binding site mutations in MS2 coat protein heterodimers.

The coat protein of bacteriophage MS2 functions as a symmetric dimer to bind an asymmetric RNA hairpin. This implies the existence of two equivalent RNA binding sites related to one another by a 2-fold symmetry axis. In this view the symmetric binding site defined by mutations conferring the repressor-defective phenotype is a composite picture of these two asymmetric sites. In order to determine whether the RNA ligand interacts with amino acid residues on both subunits of the dimer and in the hope of constructing a functional map of the RNA binding site, we performed heterodimer complementation experiments. Taking advantage of the physical proximity of their N- and C-termini, the two subunits of the dimer were genetically fused, producing a duplicated coat protein which folds normally and allows the construction of the functional equivalent of obligatory heterodimers containing all possible pairwise combinations of the repressor-defective mutations. The restoration of repressor function in certain heterodimers shows that a single RNA molecule interacts with both subunits of the dimer and allows the construction of a functional map of the binding site.     

Molecular dynamics simulation studies of a protein-RNA complex with a selectively modified binding interface

The RNA recognition motif (RRM) is one of the most common RNA binding domains. We have investigated the contribution of three highly conserved aromatic amino acids to RNA binding by the N-terminal RRM of the U1A protein. Recently, we synthesized a modified base (A-4CPh) in which a phenyl group is tethered to adenine using a linker of 4 methylene groups. The substitution of this base for adenine in the target RNA selectively stabilizes the complex formed with a U1A protein in which one of the conserved aromatic amino acids is replaced with Ala (Phe56Ala). In this article, we report molecular dynamics (MD) simulations that probe the structural consequences of the substitution of A-4CPh for adenine in the wild type and Phe56Ala U1A-RNA complexes and in the free RNA. The simulations suggest that A-4CPh stabilizes the complex formed with Phe56Ala by adopting a folded conformation in which the tethered phenyl group fills the site occupied by the phenyl group of Phe56 in the wild-type complex. In contrast, an extended conformation of A-4CPh is predicted to be most stable in the complex formed with the wild-type protein. The calculations indicate A-4CPh is in an extended conformation in the free RNA. Therefore, preorganizing the structure of the phenyl-tethered base for binding may improve both the affinity and specificity of the RNA containing A-4CPh for the Phe56Ala U1A protein. Taken together, the previous experimental work and the calculations reported here suggest a general design strategy for altering RRM-RNA complex stability.     

Dynamic Allostery Controls Coat Protein Conformer Switching during MS2 Phage Assembly

Previously, an RNA stem-loop (TR) encompassing 19 nt of the genome of bacteriophage MS2 was shown to act as an allosteric effector of conformational switching in the coat protein during in vitro capsid assembly. TR RNA binding to symmetric coat protein dimers results in conformational changes, principally at the FG-loop connecting the F and G β-strands in each subunit, yielding an asymmetric structure. The FG-loops define the quasi-equivalent conformers of the coat protein subunit (A, B, and C) in the T = 3 capsid. Efficient assembly of this capsid in vitro requires that both symmetrical and asymmetrical forms of the coat protein dimer be present in solution, implying that they closely resemble the quasi-equivalent dimers (A/B and C/C) seen in the final capsid. Experiments show that assembly can be triggered by a number of RNA stem-loops unrelated to TR in sequence and detailed secondary structure, suggesting that there is little sequence specificity to the allosteric effect. Since the stem-loop binding site on the coat protein dimer is distal to the FG-loops the mechanism of this switching effect needs to be investigated. We have analyzed the vibrational modes of both TR-bound and RNA-free coat protein dimers using an all-atom normal-mode analysis. The results suggest that asymmetric contacts between the A-duplex RNA phosphodiester backbone and the EF-loop in one coat protein subunit result in the FG-loop of that subunit becoming more dynamic, whilst the equivalent loop on the other monomer decreases its mobility. The increased dynamic behaviour occurs in the FG-loop of the subunit required to undergo the largest conformational change when adopting the quasi-equivalent B conformation. The free energy barrier on the pathway to form this new structure would consequently be reduced compared to the unbound subunit. Our results also imply that the allosteric effect should be independent of the base sequence of the bound stem-loop, as observed experimentally. As a test of this model, we also examined the vibrational modes of a known assembly mutant, W82R, which cannot assemble beyond dimer. This mutation leads to an increased mobility of the DE-loop rather than the FG-loop after TR binding, consistent with the non-assembling phenotype of this mutant protein.     

The crystal structure of a high affinity RNA stem-loop complexed with the bacteriophage MS2 capsid: further challenges in the modeling of ligand-RNA interactions

We have determined the structure to 2.8 A of an RNA aptamer (F5), containing 2'-deoxy-2-aminopurine (2AP) at the -10 position, complexed with MS2 coat protein by soaking the RNA into precrystallised MS2 capsids. The -10 position of the RNA is an important determinant of binding affinity for coat protein. Adenine at this position in other RNA stem-loops makes three hydrogen bonds to protein functional groups. Substituting 2AP for the -10 adenine in the F5 aptamer yields an RNA with the highest yet reported affinity for coat protein. The refined X-ray structure shows that the 2AP base makes an additional hydrogen bond to the protein compared to adenine that is presumably the principal origin of the increased affinity. There are also slight changes in phosphate backbone positions compared to unmodified F5 that probably also contribute to affinity. Such phosphate movements are common in structures of RNAs bound to the MS2 T = 3 protein shell and highlight problems for de novo design of RNA binding ligands.     

The Structure of Bacteriophage φCb5 Reveals a Role of the RNA Genome and Metal Ions in Particle Stability and Assembly

The structure of the Leviviridae bacteriophage φCb5 virus-like particle has been determined at 2.9 Å resolution and the structure of the native bacteriophage φCb5 at 3.6 Å. The structures of the coat protein shell appear to be identical, while differences are found in the organization of the density corresponding to the RNA. The capsid is built of coat protein dimers and in shape corresponds to a truncated icosahedron with T = 3 quasi-symmetry. The capsid is stabilized by four calcium ions per icosahedral asymmetric unit. One is located at the symmetry axis relating the quasi-3-fold related subunits and is part of an elaborate network of hydrogen bonds stabilizing the interface. The remaining calcium ions stabilize the contacts within the coat protein dimer. The stability of the φCb5 particles decreases when calcium ions are chelated with EDTA. In contrast to other leviviruses, φCb5 particles are destabilized in solution with elevated salt concentration. The model of the φCb5 capsid provides an explanation of the salt-induced destabilization of φCb5, since hydrogen bonds, salt bridges and calcium ions have important roles in the intersubunit interactions.Electron density of three putative RNA nucleotides per icosahedral asymmetric unit has been observed in the φCb5 structure. The nucleotides mediate contacts between the two subunits forming a dimer and a third subunit in another dimer. We suggest a model for φCb5 capsid assembly in which addition of coat protein dimers to the forming capsid is facilitated by interaction with the RNA genome. The φCb5 structure is the first example in the levivirus family that provides insight into the mechanism by which the genome-coat protein interaction may accelerate the capsid assembly and increase capsid stability.     

Adenosine 3'':5''-cyclic monophosphate-binding proteins in bovine and rat tissues.

1. At least two classes of high-affinity cyclic AMP-binding proteins have been identified: those derived from cyclic AMP-dependent protein kinases (regulatory subunits) and those that bind a wide range of adenine analogues (adenine analogue-binding proteins). 2. In fresh-tissue extracts, regulatory subunits could be further subdivided into 'type I or 'type II' depending on whether they were derived from 'type I' or 'type II' protein kinase [see Corbin et al. (1975) J. Biol. Chem. 250, 218-225]. 3. The adenine analogue-binding protein was detected in crude tissue supernatant fractions of bovine and rat liver. It differed from the regulatory subunit of cyclic AMP-dependent protein kinase in many of its properties. Under the conditions of assay used, the protein accounted for about 45% of the binding of cyclic AMP to bovine liver supernatants. 4. The adenine analogue-binding protein from bovine liver was partially purified by DEAE-cellulose and Sepharose 6B chromatography. It had mol.wt. 185000 and was trypsin-sensitive. As shown by competition and direct binding experiments, it bound adenosine and AMP in addition to cyclic AMP. At intracellular concentrations of adenine nucleotides, binding of cyclic AMP was essentially completely inhibited in vitro. Adenosine binding was inhibited by only 30% under similar conditions. 5. Rat tissues were examined for the presence of the adenine analogue-binding protein, and, of those examined (adipose tissue, heart, brain, testis, kidney and liver), significant amounts were only found in the liver. The possible physiological role of the adenine analogue-binding protein is discussed. 6. Because the adenine analogue-binding protein or other cyclic AMP-binding proteins in tissues may be products of partial proteolysis of the regulatory subunit of cyclic AMP-dependent protein kinase, the effects of trypsin and aging on partially purified protein kinase and its regulatory subunit from bovine liver were investigated. In all studies, the effects of trypsin and aging were similar. 7. In fresh preparations, the cyclic AMP-dependent protein kinase had mol.wt. 150000. Trypsin treatment converted it into a form of mol.wt 79500. 8. The regulatory subunit of the protein kinase had mol.wt. 87000. It would reassociate with and inhibit the catalytic subunit of the enzyme. Trypsin treatment of the regulatory subunit produced a species of mol.wt. 35500 which bound cyclic AMP but did not reassociate with the catalytic subunit. Trypsin treatment of the protein kinase and dissociation of the product by cyclic AMP produced a regulatory subunit of mol.wt. 46500 which reassociated with the catalytic subunit. 9. These results may be explained by at least two trypsin-sensitive sites on the regulatory subunit. A model for the effects of trypsin is described.     

Creating hetero-11-mers composed of wild-type and mutant subunits to study RNA binding to TRAP.

TRAP (trp RNA-binding attenuation protein) is an RNA-binding protein that regulates expression of the tryptophan biosynthetic genes in Bacillus subtilis by binding to RNA targets that contain multiple GAG and UAG repeats. TRAP is composed of 11 identical subunits arranged symmetrically in a ring. The secondary structure of the protein consists entirely of antiparallel beta-sheets, beta-turns, and loops. We show here that the TRAP 11-mer can be reversibly denatured into unfolded monomers by guanidine hydrochloride. Removing the denaturant allows the protein to spontaneously renature into fully functional 11-mers. Based on this finding, we developed a subunit mixing method to hybridize wild-type and mutant subunits into heteromeric 11-mers by denaturation followed by subunit mixing renaturation. This method allows the study of subunit cooperativity in protein-ligand interaction such as RNA binding. Our data further support and extend the previously proposed two-step model for RNA binding to TRAP by showing that the initiation of binding requires at least one fully active subunit in the protein combined with one fully functional repeat in the RNA. The initiation complex tethers the RNA on the protein, thus allowing cooperative interaction with the remainder of the repeats.     

Characterization of a trp RNA-binding attenuation protein (TRAP) mutant with tryptophan independent RNA binding activity

TRAP (trp RNA-binding attenuation protein) is an 11 subunit RNA-binding protein that regulates expression of genes involved in tryptophan metabolism (trp) in Bacillus subtilis in response to changes in intracellular tryptophan concentration. When activated by binding up to 11 tryptophan residues, TRAP binds to the mRNAs of several trp genes and down-regulates their expression. Recently, a TRAP mutant was found that binds RNA in the absence of tryptophan. In this mutant protein, Thr30, which is part of the tryptophan-binding site, is replaced with Val (T30V). We have compared the RNA-binding properties of T30V and wild-type (WT) TRAP, as well as of a series of hetero-11-mers containing mixtures of WT and T30V TRAP subunits. The most significant difference between the interaction of T30V and WT TRAP with RNA is that the affinity of T30V TRAP is more dependent on ionic strength. Analysis of the hetero-11-mers allowed us to examine how subunits interact within an 11-mer with regard to binding to tryptophan or RNA. Our data suggest that individual subunits retain properties similar to those observed when they are in homo-11-mers and that individual G/UAG triplets within the RNA can bind to TRAP differently.     

A novel electron transfer mechanism suggested by crystallographic studies of mitochondrial cytochrome bc1 complex.

The crystal structure of bovine mitochondrial cytochrome bc1 complex, an integral membrane protein complex of 11 different subunits with a total molecular mass of 242 kDa, demonstrated a tightly associated dimer consisting of three major regions: a matrix region primarily made of subunits core1, core2, 6, and 9; a transmembrane-helix region of 26 helices in the dimer contributed by cytochrome b, cytochrome c1, the Rieske iron-sulfur protein (ISP), subunits 7, 10, and 11; and an intermembrane-space region composed of extramembrane domains of ISP, cytochrome c1, and subunit 8. The structure also revealed the positions of and distances between irons of prosthetic groups, and two symmetry related cavities in the transmembrane-helix region upon dimerization of the bc1 complex. Extensive crystallographic studies on crystals of bc1 complexed with inhibitors of electron transfer identified binding pockets for both Qo and Qi site inhibitors. Discrete binding sites for subtypes of Qo site inhibitors have been mapped onto the Qo binding pocket, and bindings of different subtypes of Qo site inhibitors are capable of inducing dramatic conformational changes in the extramembrane domain of ISP. A novel electron transfer mechanism for the bc1 complex consistent with crystallographic observations is discussed.     

Mutagenesis of the regulatory subunit of yeast cAMP-dependent protein kinase. Isolation of site-directed mutants with altered binding affinity for catalytic subunit

Oligonucleotide-directed mutagenesis was used to produce mutants in the hinge region of the regulatory subunit (R) of the Saccharomyces cerevisiae cAMP-dependent protein kinase. The mutant proteins were expressed in Escherichia coli, purified, urea treated to produce cAMP-free regulatory (R), and analyzed in vitro for catalytic (C) subunit inhibitory activity in the presence and absence of cAMP. When assayed in the absence of cAMP, wild type R dimer inhibited C with an IC50 of 40 nM. Replacement of amino acid residue Ser-145 (the autophosphorylation site of yeast R) with Ala or Gly produced mutants which were 2-10-fold better inhibitors of C, while replacement with Glu, Asp, Lys, or Thr produced mutants which were 2-5-fold worse inhibitors of C relative to wild type R. When assayed in the presence of cAMP, all R subunits had a decreased affinity for C subunit, with Ser-145 and Thr-145 undergoing autophosphorylation. These results suggest that the amino acid at position 145 of R contributes to R-C interaction and therefore influences the equilibrium of yeast protein kinase subunits in vitro.