The tobacco mosaic virus assembly origin RNA. Functional characteristics defined by directed mutagenesis

The in vitro reassembly of tobacco mosaic virus (TMV) begins with the specific recognition by the viral coat protein disk aggregate of an internal TMV RNA sequence, known as the assembly origin (Oa). This RNA sequence contains a putative stem-loop structure (loop 1), believed to be the target for disk binding in assembly initiation, which has the characteristic sequence AAGAAGUCG exposed as a single strand at its apex. We show that a 75-base RNA sequence encompassing loop 1 is sufficient to direct the encapsidation by TMV coat protein disks of a heterologous RNA fragment. This RNA sequence and structure, which is sufficient to elicit TMV assembly in vitro, was explored by site-directed mutagenesis. Structure analysis of the RNA identified mutations that appear to effect assembly via a perturbation in RNA structure, rather than by a direct effect on coat protein binding. The binding of the loop 1 apex RNA sequence to coat protein disks was shown to be due primarily to its regularly repeated G residues. Sequences such as (UUG)3 and (GUG)3 are equally effective at initiating assembly, indicating that the other bases are less functionally constrained. However, substitution of the sequences (CCG)3, (CUG)3 or (UCG)3 reduced the assembly initiation rate, indicating that C residues are unfavourable for assembly. Two additional RNA sequences within the 75-base Oa sequence, both of the form (NNG)3, may play subsidiary roles in disk binding. RNA structure plays an important part in permitting selective protein-RNA recognition, since altering the RNA folding close to the apex of the loop 1 stem reduces the rate of disk binding, as does shortening the stem itself. Whereas the RNA sequence making up the hairpin does not in general affect the specificity of the protein-RNA interaction, it is required to present the apex signal sequence in a special conformation. Mechanisms for this are discussed.     

Self-assembly of tobacco mosaic virus: the role of an intermediate aggregate in generating both specificity and speed

The tobacco mosaic virus (TMV) particle was the first macromolecular structure to be shown to self-assemble in vitro, allowing detailed studies of the mechanism. Nucleation of TMV self-assembly is by the binding of a specific stem-loop of the single-stranded viral RNA into the central hole of a two-ring sub-assembly of the coat protein, known as the 'disk'. Binding of the loop onto its specific binding site, between the two rings of the disk, leads to melting of the stem so more RNA is available to bind. The interaction of the RNA with the protein subunits in the disk cause this to dislocate into a proto-helix, rearranging the protein subunits in such a way that the axial gap between the rings at inner radii closes, entrapping the RNA. Assembly starts at an internal site on TMV RNA, about 1 kb from its 3'-terminus, and the elongation in the two directions is different. Elongation of the nucleated rods towards the 5'-terminus occurs on a 'travelling loop' of the RNA and, predominantly, still uses the disk sub-assembly of protein subunits, consequently incorporating approximately 100 further nucleotides as each disk is added, while elongation towards the 3'-terminus uses smaller protein aggregates and does not show this 'quantized' incorporation.     

Assembly of tobacco mosaic virus in vitro. Improved model for the elongation process by protein subunits.

The in vitro assembly reaction of tobacco mosaic virus (TMV), especially the elongation process of partially reconstituted RNA (PRR) by protein subunits, was observed by electron microscopy. After addition of TMV-protein subunits, the PRR appeared as rods with a clump at one end, believed to be a complex between added protein subunits and the RNA tail protruding from PRR. The subunits entrapped on the RNA tails in the forms of clumps were progressively incorporated into the growing rods on incubation, ending with the formation of completely reconstituted rods. The clumps were also observed after addition of cucumber green mottle mosaic virus (CGMMV) protein subunits to rods partially reconstituted from RNA and TMV-protein. In this case, the protein subunits, seen as clumps, did not become incorporated to form elongating rods. An improved model for the elongation of TMV rods is proposed. The elongation process is composed of two steps, with the first step being the interaction of protein subunits with the RNA tail protruding from the growing rod. Any protein having a specific binding site for TMV-rna, not limited to TMV-protein, will react in the first step. The second step is the incorporation of the protein on the RNA tail into a rod-shaped structure, with consequent elongation of the growing rod. It appears that only protein homologous with that in the partially reconstituted rods can partake in the second step.     

Assembly of the Particle of Tobacco Mosaic Virus from RNA and Disks of Protein

The reconstitution of TMV does not proceed by the stepwise addition of single protein subunits, but by the addition of preformed disks to the growing rod. The assembly is initiated by the interaction of a disk with a special sequence of about fifty nucleotides at the 5′ end of the TMV RNA. This is the basis of the selectivity for viral over other RNAs.     

RNA-protein interactions in the assembly of tobacco mosaic virus.

Assembly of tobacco mosaic virus is initiated by the binding of a specific loop of the RNA into the central hole of the disk aggregate of protein subunits. Since the nucleation loop is located about five-sixths along the RNA molecule, subsequent elongation must be bidirectional. We have now measured the rates of elongation in the two directions by determining the lengths of RNA protected from nuclease digestion at different times and using either intact TMV rNA, or RNA with most of the longer tail removed. Comparison of the rates with the protein supplied as either a mixture of disks with A-protein (a mixture of less aggregated states) or just A-protein, shows that different mechanisms and protein aggregates are used for the most rapid growth. When disks are present, they add more rapidly along the longer RNA tail but do not appear to add directly on the shorter tail. In contrast, smaller aggregates (A-protein) can add at both ends of the rod, but do so more slowly. Mechanisms for these processes are discussed. Preliminary results on the binding of the specific hexanucleotide AAGAAG to the disk are given and compared with the known changes on binding nonspecific hexanucleotides or the trinucleotide AAG.     

Location and encapsidation of the coat protein cistron of tobacco mosaic virus. A bidirectional elongation of the nucleoprotein rod.

The coat protein cistron of tobacco mosaic virus has been located on the viral RNA starting between 975 and 1050 nucleotides from the 3'-hydroxyl end. This locates its 5' end close to the origin for virus assembly, where the first protein disk interacts with RNA. It also means that the coat protein mRNA must have a short 5'-untranslated tail and a long (over 500 nucleotides) 3' one. The recovery of characteristic oligonucleotides in nuclease-protected rods during the growth from RNA and a protein disk preparation shows that elongation of the nucleated rods proceeds independently in both directions though, on average, much more rapidly along the longer 5' tail than the shorter 3' tail. Protected RNA of length equal to that in the complete virion is first seen within 6 min, showing that the most rapidly elongated particles are substantially complete by this time.     

Coordinated two-disk nucleation, growth and properties, of virus-like particles assembled from tobacco-mosaic-virus capsid protein with poly(A) or oligo(A) of different length

Assembly of nucleoprotein rods from tobacco mosaic virus (TMV) coat protein and poly(A) depends on the presence of 20S disks in a manner very similar to nucleation and growth of virions in reconstitution with TMV RNA. Products assembled with (A) approximately equal to 5000 appear to have the same buoyant density in CsCl, the same nucleotide/protein ratio and the same nuclease stability, as reconstituted and native TMV. Their rate of formation is very similar to the rate of reconstitution with TMV RNA when high-molecular-mass (A) approximately equal to 5000 is used, but becomes a function of chain length particularly with (A) less than or equal to 185. The composition of assembly products can be described sufficiently with the relation between number of capsid polypeptide monomers/particle, np, to the number of nucleotide residues/chain, nnt, of np = 1/3 (nnt + 50) with two important restrictions: (1) particles of less than four turns of helically arranged capsid subunits are unstable, and (2) particles with about 150 or less nucleotides per chain deviate in structure from mature virus and virus-like (= longer) assembly products. This is indicated by changes in both buoyant density in CsCl and optical properties, while 'dislocation' of the disk to the helical arrangement of capsid subunits ('helicalization') and nuclease stability already become established with chains as short as (A) approximately equal to 58 +/- 20. Consequently, we suggest that assembly proceeds through three distinct phases: (1) nucleation (resulting in helicalization) by interaction of nucleic acid with the first disk; (2) stabilization of the primary (unstable!) nucleation complex by addition of a second disk and formation of a four-turn virus-like and stable nucleoprotein helix, which is then fit for (3) elongation by addition of further disks. The question of what makes the TMV protein disk select specifically TMV RNA during virion assembly is discussed in some detail.     

The region of tobacco mosaic virus RNA involved in the nucleation of assembly.

The interaction of TMV RNA with the disk aggregate of TMV protein at the initiation of assembly has been studied by using the techniques of RNA sequencing. The 5' end group has been identified, and shown not to be protected in the early stages of assembly from accessibility to nuclease digestion. A population of RNA fragments of average length 250 nucleotides, originating from a unique region of TMV RNA, is encapsidated by limited assembly, and sufficient sequence information is available to identify certain unusual features. The protected region does not contain highly reiterated simple repeating sequences, but may contain more complicated repeats. The length and complexity of the nucleation region may reflect adaptation to the efficient mediation of the conformational change from disk to helix of TMV protein, besides a requirement for binding to the disk, and this may be an important part of the mechanism of specificity in the nucleation of assembly.     

Mechanism of RNA-protein interactions in tobacco mosaic virus: analysis of the pH stability of virus protein complexes with synthetic polynucleotides.

TMV-like RNP complexes were reconstituted from TMV protein and synthetic polynucleotides. Analysis of the pH stability of RNP with polynucleotides containing U, G, or their analogues reveals a correlation between the stability of their structure and the pK values of the bases, and indicates that the -NH-CO-groups of U and G are involved in hydrogen bonding with protein. It is suggested that TMV protein has two U- and one G-specific binding sites which, according to the phase position of the protein subunits relative to the origin of TMV assembly (D. Zimmern (1977), Cell 11, 463) are likely to be organized as UGU. The binding of the A and C residues of RNA with TMV protein is nonspecific. TMV protein groups with pK 6.3, 7.5 and 9.7 were found to be essential in the protein-protein interactions in RNP. A group of the protein with pK 8.2 is also involved in RNP stabilization. Both protein-protein interactions and interactions of protein with RNA phosphate groups were shown to be mediated by a conformational change in the protein induced by base binding. The effect of bases on both types of interactions changes in the order G approximately equal to much greater than A, and incorporation of C in RNP proceeds in a compulsory way at the expense of interaction of the neighbouring nucleotide residues in polynucleotides with protein. The data obtained are used to discuss the principles of the cooperativity of the interactions between TMV components and the mechanism of initiation and elongation in TMV self-assembly.     

Protein-RNA interactions during TMV assembly

A review of the structural studies of tobacco mosaic virus (TMV) is given. TMV is essentially a flat helical microcrystal with 16 1/3 subunits per turn. A single strand of RNA runs along the helix and is deeply embedded in the protein. The virus particles form oriented gels from which high-resolution X-ray fiber diffraction data can be obtained. This may be interpreted by the use of six heavy-atom derivatives to give an electron density map at 0.4 nm resolution from which the RNA configuration and the form of the inner part of the protein subunit may be determined. In addition, the protein subunits form a stable 17-fold two-layered disk which is involved in virus assembly and which crystallizes. By the use of noncrystallographic symmetry and a single heavy-atom derivative, it has been possible to solve the structure of the double disk to 0.28 nm resolution. In this structure one sees that an important structural role is played by four alpha-helices, one of which (the LR helix) appears to form the main binding site for the RNA. The main components of the binding site appear to be hydrophobic interactions with the bases, hydrogen bonds between aspartate groups and the sugars, and arginine salt bridges to the phosphate groups. The binding site is between two turns of the virus helix or between the turns of the double disk. In the disk, the region proximal to the RNA binding site is in a random coil until the RNA binds, whereupon the 24 residues involved build a well-defined structure, thereby encapsulating the RNA.     

An extended secondary structure model for the TMV assembly origin, and its correlation with protection studies and an assembly defective mutant

Recognition of the unique internal assembly origin on tobacco mosaic virus (TMV) RNA by the disk aggregate of the viral coat protein probably involves an extended region of the RNA (larger than that coated by a single disk) folded into a specific conformation. A secondary structure model is proposed for the RNA preferentially coated by limiting amounts of coat protein disks on the basis of partial nuclease digestion data. Part of this sequence can form three symmetrically spaced hairpins with marginally stable base paired sequences at the tips of the stems. The pattern of progressive protection of the RNA from nuclease attack during assembly suggests that these three hairpins are successively coated by the first three disks to add. The spacing of these hairpins is identical to that of three hairpins in the pseudo assembly origin (part of the coat protein gene homologous to the assembly origin). In Ni 2519, a TMV mutant whose assembly is defective at high temperature because it can no longer discriminate between the true and pseudo assembly origins, a point mutation has occurred near the tip of the third metastably base paired stem of the true assembly origin which would disrupt its structure and alter one copy of a repeated heptanucleotide. This suggests an important role for the ordered and cooperative recognition of successive loops in determining the specificity of assembly.     

Studies of tobacco mosaic virus reassembly with an RNA tail blocked by a hybridised and cross-linked probe

Segments of cloned cDNA to tobacco mosaic virus RNA, 150--300-bases long, have been hybridised and cross-linked to the RNA, which has then been used for reassembly experiments. This enables the elongation reaction, which does not encapsidate the double-stranded region generated, to be stopped at specific regions along the RNA and the resulting particles to be characterised, by measuring the lengths of the rods in the electron microscope. With hybridisation to the 3'-tail the entire RNA contiguous to the nucleation region is encapsidated, from the 5'-terminus up to the modified region. When the double-stranded region is on the 5'-side of the nucleation region, the mean length of the particles corresponds to a situation in which the double-stranded region is unable to enter the central hole of the growing rod, but the 3'-tail of the RNA is completely encapsidated. The longest particles hybridised on the 5'-tail (i.e. in a class longer than the mean length) show an effect complementary to those with a 3'-block, and have lengths which correspond to encapsidation from the modified region to the 3'-terminus, despite the continued presence of the 5'-tail up the rod. In all cases where there is a remaining 5'-tail the lengths observed can only be explained if elongation has occurred substantially, or probably completely, along the 3'-tail. Hence elongation must have occurred simultaneously along both the 5' and 3'-tails of the tobacco mosaic virus RNA after initiation on the internal nucleation region.     

Structure and function of disk aggregates of the coat protein of tobacco mosaic virus

Experiments have been carried out on the coat protein of tobacco mosaic virus (TMVP) to test for the occurrence of the previously postulated RNA-induced direct switching, during in vitro assembly of tobacco mosaic virus (TMV), of the subunit packing from the cylindrical bilayer disk to the virus helical arrangement. No evidence was found for such RNA-induced switching and no evidence for the direct participation of the bilayer disk in either the nucleation or elongation phases of the in vitro virus assembly. Instead, virus assembly proceeds by an initiation step involving the binding of the RNA to the previously characterized two-plus turn helical aggregate that is formed from small oligomers of subunits. However, a bilayer disk, which has been characterized in high ionic strength crystals, has been observed in low ionic strength virus assembly solutions only as a transient species upon depolymerization of dimers of bilayer disks formed in solution at high ionic strength, and not as an equilibrium species of TMVP.     

Sequence of a specifically encapsidated RNA fragment originating from the tobacco-mosaic-virus coat-protein cistron.

When 25-S tobacco mosaic virus (TMV) protein aggregate and TMV RNA, which has been partially digested by T1 RNase, are mixed under conditions suitable for reconstitution, only a few RNA fragments are encapsidated. These fragments were isolated and purified by polyacrylamide gel electrophoresis. The sequence of the three main fragments, the longest of which (fragment 1) was estimated to contain 103 nucleotides, has been determined. The two smaller fragments are portions of the longer chain produced by an additional specific scission. Because of the great affinity of 25-S TMV protein for this nucleotide sequence, it will be referred to as the "specifically encapsidated RNA fragment". The occurrence of a "hidden break" in the sequence has been demonstrated: fragment 1, purified by electrophoresis on a polyacrylamide gel without 8 M urea, gives rise upon further electroporesis in the presence of urea to two new bands corresponding to the two halves of the molecule. A stable hair-pin secondary structure has been derived from the base sequence which can account for the specificity of action of the enzyme. Because of its properties, we have suggested elsewhere that the sequence of fragment 1 might correspond to the disk recognition site for reconstitution, which is known to be located at the 5' end of the intact RNA. But experiments with TMV RNA whose 5'-OH end has been radioactively phosphorylated with polynucleotide kinase show that this is not the case. Analysis of the amino acid coding capacity of the fragment has instead revealed that fragment 1 is a portion of the TMV coat protein cistron.     

The interaction between tylophorine B and TMV RNA

Tylophorine B exhibits 60% inhibition against tobacco mosaic virus (TMV) at a concentration of 1.0 x 10(-6) g/ml. In our study, high affinity for TMV RNA and assembly origin of TMV RNA (oriRNA) was revealed, accompanied by the conformational change of RNA. Considering that TMV assembly begins with the specific recognition by the coat protein aggregate of oriRNA, and that tylophorine B has favorable interaction with oriRNA, we speculate that tylophorine B likely exerts its virus inhibition by binding to oriRNA and interfering with virus assembly initiation. This work may shed light on the possible molecular inhibition mechanism against TMV by tylophorine B, and provide clues in rational design of sequence-specific RNA binding antivirus drugs.     

The isolation of tobacco mosaic virus RNA fragments containing the origin for viral assembly

Upon mixing purified TMV RNA with limited amounts of viral coat protein in the form of the disk aggregate, a unique region of the whole RNA becomes protected from nuclease digestion. The protected RNA consists of fragments up to 500 nucleotides long in varying yields, which are found in nucleoprotein particles having a protein-nucleic acid ratio similar to the mature virus. The protected RNA, when reextracted, is able to rebind to coat protein disks rapidly, quantitatively and with high affinity, becoming once more RNAase-resistant in the process. Small aggregates of TMV protein (A protein) are inactive in formation of the nuclease-resistant complexes. On the basis of this evidence, we identify the isolated RNA fragments as portions of TMV RNA containing the origin or initiation site for in vitro reassembly, which have been protected from digestion by incorporation into assembly nucleation complexes.The yield, but not the length distribution, of the protected RNA pieces is found to double upon increasing the protein added from 1–2 disk-equivalents of protein per RNA molecule. This implies that the formation of the nucleation complexes may involve a highly cooperative initial addition of protein.     

Specific encapsidation of fragments of TMV RNA.

The in vitro reconstitution of tobacco mosaic virus (TMV) is initiated by the binding of a disk of TMV protein to the 'disk recognition site', a region of the RNA chain at or near the 5'-terminus for which the disk has special affinity. In order to gain insight into the recognition process, we have studied the ability of disks to encapsidate short RNA fragments produced by partial pancreatic or T1 RNase digestion of TMV RNA. The disk is capable of dicriminating among such fragments, encapsidating only a few of the many present in the digest. The products of encapsidation are short nucleoprotein rods of the same diameter as TMV and of length proportional to that of the encapsidated RNA fragment. The particles differ from TMV, however, in one significant aspect (apart from their length): they possess rings of RNA-free protein at one or both extremities of the rod. In the case of T1 RNase digestion the principal encapsidated fragments were fragments T1 (105 nucleotides) and a family of smaller fragments containing elements of the same sequence. Partial digestion with pancreatic RNase generated only one major fragment (fragment P1; 150 nucleotides) with affinity for the disk. Fragment T1 has been sequenced and shown to represent a portion of the coat protein cistron. Fragment P1 has been partially sequenced but its function is not yet known. Several lines of evidence indicate that fragment T1 is not the disk recognition site. The portion of the TMV RNA chain from which fragment P1 is derived, on the other hand, is encapsidated early in the reconstitution process; thus fragment P1 may contain the disk recognition site. Fragment T1 and fragment P1 both have purine-rich and cytosine-poor sequences near their termini. In addition, fragment T1, and possibly fragment P1, possess a periodicity of order three in purine residues. It seems likely that one or both of the aforesaid properties are largely responsible for the affinity of these fragments for the disk.     

Structures and roles of the polymorphic forms of tobacco mosaic virus protein. 8. Initial stages of assembly of nucleoprotein rods from virus RNA and the protein disks

The initial stages of the assembly of tobacco mosaic virus have been investigated by reassembling the RNA with a radioactively labelled protein disk preparation and then completing the reaction by the addition of a large excess of an unlabelled disk preparation. This gives measurements of the numbers of subunits incorporated at early times and the growth curves have been plotted.These curves have been analysed in terms of a bimolecular nucleation reaction, which is first order in the disk concentration, with a rate constant of 1.3 × 10³ mol^?1 s^?1, and then an elongation which saturates at high protein concentrations to a maximum rate of 7.6 subunits s^?1, with a K_m of 0.84 mg/ml for the disk preparation.These kinetic parameters, and the predicted overall assembly curves, have been compared with data previously determined by other methods and agree closely, showing that the different experimental techniques give consistent results. The measurements are fully compatible with our earlier hypotheses Butler &; Klug 1971 that the nucleation with virus RNA and protein disks is rapid compared with the subsequent rod elongation and that this elongation can occur most rapidly directly from the protein disks. They are not compatible with the contention of some other workers that elongation cannot occur directly from disks, but only from the smaller A-protein.