Dideoxy sequencing of RNA using reverse transcriptase

Sub-3-Å cryo-EM structure of RNA enabled by engineered homomeric self-assembly

High-resolution structural studies are essential for understanding the folding and function of diverse RNAs. Herein, we present a nanoarchitectural engineering strategy for efficient structural determination of RNA-only structures using single-particle cryogenic electron microscopy (cryo-EM). This strategy-ROCK (RNA oligomerization-enabled cryo-EM via installing kissing loops)-involves installing kissing-loop sequences onto the functionally nonessential stems of RNAs for homomeric self-assembly into closed rings with multiplied molecular weights and mitigated structural flexibility. ROCK enables cryo-EM reconstruction of the Tetrahymena group I intron at 2.98-Å resolution overall (2.85 Å for the core), allowing de novo model building of the complete RNA, including the previously unknown peripheral domains. ROCK is further applied to two smaller RNAs-the Azoarcus group I intron and the FMN riboswitch, revealing the conformational change of the former and the bound ligand in the latter. ROCK holds promise to greatly facilitate the use of cryo-EM in RNA structural studies.

Analysis of RNA by Primer Extension

For mapping the 5' termini of mRNA molecules, primer extension is the method of choice. A purified oligonucleotide is end-labeled using polynucleotide kinase. The probe and a population of mRNA are allowed to hybridize, and the primers and template are used to carry out reverse transcription using an enzyme cloned from the Moloney murine leukemia virus. The primer extension products are separated on a denaturing polyacrylamide gel and analyzed by radiography.

Top-down tandem mass spectrometry of tRNA via ion trap collision-induced dissociation

Transfer RNA is a class of highly modified and structured non-coding RNA molecules generally comprised of 74-95 nucleotides. In this study, tandem mass spectrometry of intact multiply charged tRNA anions of roughly 25 kDa in mass has been demonstrated using a quadrupole/time-of-flight tandem mass spectrometer adapted for ion/ion reaction studies. The sample proved to be a mixture of tRNA molecules. The mass of the most abundant component of the mixture was not consistent with that of the nominal identity of the tRNA from the supplier, viz., tRNA(Phe); rather, the mass was consistent with tRNA(Phe) bearing an incomplete 3'-terminus. Multiply-charged anions from the major components were isolated in the gas phase and subjected to ion trap collision-induced dissociation without subsequent ion/ion reactions. Abundant fragments from the 5'- and 3'-termini of the molecule could be used to identify the major component as tRNA(Phe)-3'adenosine (without 3'-phosphorylation). Roughly 15% of the primary sequence of the intact tRNA was unambiguously reflected in the product ion spectrum. The existence of a possible tRNA(Phe) variant and the intact tRNA(Phe) was also supported by ion trap CID data. The multiply-charged fragment ions derived from tRNA(Phe)-3'adenosine were further charge-reduced to mostly singly- and doubly-charged species via proton transfer ion/ion reactions with benzoquinoline cations. The resulting reduction in spectral overlap and charge state ambiguity simplified interpretation of the product ion spectrum and allowed for the identification of product ions from roughly 60% of the sequence.

Ion trap collision-induced dissociation of multiply deprotonated RNA: c/y-ions versus (a-B)/w-ions

The dissociation of model RNA anions has been studied as a function of anion charge state and excitation amplitude using ion trap collisional activation. Similar to DNA anions, the precursor ion charge state of an RNA anion plays an important role in directing the preferred dissociation channels. Generally, the complementary c/y-ions from 5' P-O bond cleavage dominate at low to intermediate charge states, while other backbone cleavages appear to a limited extent but increase in number and relative abundance at higher excitation energies. The competition between base loss, either as a neutral or as an anion, as well as the preference for the identity of the lost base are also observed to be charge-state dependent. To gain further insight into the partitioning of the dissociation products among the various possible channels, model dinucleotide anions have been subjected to a systematic study. In comparison to DNA, the 2'-OH group on RNA significantly facilitates the dissociation of the 5' P-O bond. However, the degree of excitation required for a 5' base loss and the subsequent 3' C-O bond cleavage are similar for the analogous RNA and DNA dinucleotides. Data collected for protonated dinucleotides, however, suggest that the 2'-OH group in RNA can stabilize the glycosidic bond of a protonated base. Therefore, base loss from low charge state oligonucleotide anions, in which protonation of one or more bases via intramolecular proton transfer can occur, may also be stabilized in RNA anions relative to corresponding DNA anions.

Sequence confirmation of modified oligonucleotides using chemical degradation, electrospray ionization, time-of-flight, and tandem mass spectrometry

We report the sequencing of highly modified oligonucleotides containing a mixture of 2'-deoxy, 2'-fluoro, 2'-O-methyl, abasic, and ribonucleotides. The passenger and guide strands each containing 48% and 86% of modified nucleotides, respectively, are representative sequences of synthetic short interfering RNAs (siRNAs). We describe herein the sequence confirmation of both strands using a series of robust chemical reactions, followed by analysis via ESI-TOF and ion trap mass spectrometry (ITMS). The following method enables the rapid sequence confirmation of highly modified oligonucleotides.

Identification and characterization of archaeal and fungal tRNA methyltransferases

All organisms modify their tRNAs by use of evolutionarily conserved enzymes. Members of the Archaea contain an extensive set of modified nucleotides that were early evidence of the fundamental evolutionary divergence of the Archaea from Bacteria and Eucarya. However, the enzymes responsible for these posttranscriptional modifications were largely unknown before the advent of genome sequencing. This chapter explains methods to identify tRNA methyltransferases in genome sequences, emphasizing the identification and characterization of six enzymes from the hyperthermophilic archaeon Methanocaldococcus jannaschii. We describe methods to express these proteins, purify or synthesize tRNA substrates, measure methyltransferase activity, and map tRNA modifications. Comparison of the archaeal methyltransferases with their yeast homologs suggests that the common ancestor of the archaeal and eucaryal organismal lineages already had extensive tRNA modifications.

Mass spectrometry of RNA

A complex population of non-coding RNAs is present in higher organisms. These RNAs have a multitude of functions and execute control over gene expression through various, often poorly understood, mechanisms. At present, the identification and analysis of functional regulatory RNAs and disparate ribonucleoprotein complexes remain an experimental challenge for biologists. They require specially designed approaches and techniques in genomics and RNA biochemistry. Developments in technologies based on mass spectrometry could offer sensitive and efficient solutions to analysis of the sequence, structure, modification and composition of RNA.

Identification of the mass-silent post-transcriptionally modified nucleoside pseudouridine in RNA by matrix-assisted laser desorption/ionization mass spectrometry

A new method using matrix-assisted laser desorption/ionization (MALDI) mass spectrometry for the direct analysis of the mass-silent post-transcriptionally modified nucleoside pseudouridine in nucleic acids has been developed. This method utilizes 1-cyclohexyl-3-(2-morpholinoethyl)carbodiimide to derivatize pseudouridine residues. After chemical derivatization all pseudouridine residues will contain a 252 Da 'mass tag' that allows the presence of pseudouridine to be identified using mass spectrometry. Pseudouridine residues can be identified in intact nucleic acids by obtaining a mass spectrum of the nucleic acid before and after derivatization. The mass difference (in units of 252 Da) will denote the number of pseudouridine residues present. To determine the sequence location of pseudouridine, a combination of enzymatic hydrolysis and mass spectrometric steps are used. Here, MALDI analysis of RNase T1 digestion products before and after modification are used to narrow the sequence location of pseudouridine to specific T1 fragments in the gene sequence. Further mass spectrometric monitoring of exonuclease digestion products from isolated T1 fragments is then used for exact sequence placement. This approach to pseudouridine identification is demonstrated using Escherichia coli tRNAS: This new method allows for the direct determination of pseudouridine in nucleic acids, can be used to identify modified pseudouridine residues and can be used with general modification mapping approaches to completely characterize the post-transcriptional modifications present in RNAs.

Sequencing RNA by a combination of exonuclease digestion and uridine specific chemical cleavage using MALDI-TOF

The determination of DNA sequences by partial exonuclease digestion followed by Matrix-Assisted Laser Desorption Time of Flight Mass Spectrometry (MALDI-TOF) is a well established method. When the same procedure is applied to RNA, difficulties arise due to the small (1 Da) mass difference between the nucleotides U and C, which makes unambiguous assignment difficult using a MALDI-TOF instrument. Here we report our experiences with sequence specific endonucleases and chemical methods followed by MALDI-TOF to resolve these sequence ambiguities. We have found chemical methods superior to endonucleases both in terms of correct specificity and extent of sequence coverage. This methodology can be used in combination with exonuclease digestion to rapidly assign RNA sequences.

A sequencing method for RNA oligonucleotides based on mass spectrometry

Synthetic oligoribonucleotides up to 22 bases have been sequenced by observing different kinetics in exonuclease-induced phosphodiester bond hydrolysis and detecting the fragments by matrix-assisted laser desorption/ionization time of flight mass spectrometry (MALDI-TOF-MS). Common mass spectrometric sequencing methods have disadvantages concerning read length, cost, and specialist instrumentation using RNA as the target molecule because uridine and cytidine have similar masses. Now we are in the position to distinguish U and C by different peak intensities. The method proposed has been verified using specific endonucleases and 13C-labeled nucleotides. This new nongel-based and nonlabeling sequencing strategy offers first RNA sequencing data using the advantages of fast and accurate MALDI-TOF-MS. Preparation steps and measurements are performed in less than 1 h.

Matrix-assisted laser desorption/ionization mass spectrometry (MALDI) of endonuclease digests of RNA

The determination of RNA sequences using base- specific enzymatic cleavages is a well established method. Different synthetic RNA molecules were analyzed for uniformity of degradation by RNase T1, U2, A and PhyM under reaction conditions compatible with Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry (MALDI-MS), to identify the positions of G, A and pyrimidine residues. In order to get a complete set of fragments derived from cleavage at every phosphodiester bond, the samples were also subjected to a limited alkaline hydrolysis. Additionally, the 5'-terminus fragments of a 49mer RNA transcript were isolated by way of 5'-biotinylation and streptavidin-coated magnetic beads (Dynal), followed by a RNase U2digestion. MALDI-MS of the generated fragments is presented as an efficient technique for a direct read out of the nucleotide sequence.

Genetic determinants of Sindbis virus neuroinvasiveness

After peripheral inoculation of mice, Sindbis virus replicates in a variety of tissues, leading to viremia. In some cases, the virus can enter the central nervous system (CNS) and cause lethal encephalitis. The outcome of infection is age and virus strain dependent. Recently, two pairs of Sindbis virus variants differing in neurovirulence and neuroinvasiveness were derived by limited serial passaging in mouse brain. Two early passage isolates (SVA and SVB) were neurotropic but did not cause lethal encephalitis. SVB, but not SVA, was neuroinvasive. A second independent pair of isolates (SVN and SVNI), which had undergone more extensive mouse brain passaging, were highly neurotropic and caused lethal encephalitis. Only SVNI could reach the brain after peripheral inoculation. From these isolates, virion RNAs were obtained and used to construct full-length cDNA clones from which infectious RNA transcripts could be recovered. The strains recovered from these clones were shown to retain the appropriate phenotypes in weanling mice. Construction and analysis of recombinant viruses were used to define the genetic loci determining neuroinvasion. For SVB, neuroinvasiveness was determined by a single residue in the E2 glycoprotein (Gln-55). For SVNI, neuroinvasive loci were identified in both the 5' noncoding region (position 8) and the E2 glycoprotein (Met-190). Either of these changes on the SVN background was sufficient to confer a neuroinvasive phenotype, although these recombinants were less virulent. To completely mimic the SVNI phenotype, three SVNI-specific substitutions on the SVN background were required: G at position 8, E2 Met-190, and Lys-260, which by itself had no effect on neuroinvasion. These genetically defined strains should be useful for dissecting the molecular mechanisms leading to Sindbis virus invasion of the CNS.

Isolation and genomic analysis of the rat polymeric immunoglobulin receptor gene terminal domain and transcriptional control region

The polymeric immunoglobulin receptor (pIgR) transports IgA and IgM across secretory epithelial cells and is essential in external immunity maintenance. We report here the structural characterization of the single-copy rat gene distributed over 30 kb of chromosomal DNA and analysis of its transcriptional control region. RNA sequencing and genomic analysis show a 5' terminal region originates at a major (+1) and a minor site producing an unusual 124-bp nontranslated exon I separated from a small 96-bp initiator ATG coding exon II by a 7.5-kb intron. The pIgR 5' region comprises a structured promoter with abundant helix-loop-helix (bHLH) cis elements positioned within an equivalent internal -70, -290, -528, and three centered at -745. The three latter bHLH elements each occur within 30-bp repeats at -690 to -780. Transient expression assays show a 1.3-kb 5' region is sufficient to drive expression in rat primary hepatocyte monolayer cultures, transformed human hepatic (HepG2) cells, and a mammary epithelial tumor cell line MCF-7, but is inactive in the rodent fibroblast 3T3 cell line. A minimal transcriptional promoter domain was deduced from sequentially deleted vectors revealing a +40 to -922 sequence to be sufficient for full activity. Further deletions within this region yield incremental losses in cis activity, indicating that multiple subregions comprise an extended transcriptional control region.

Mutational analysis of the 3'-terminal extra-cistronic region of poliovirus RNA: secondary structure is not the only requirement for minus strand RNA replication

A series of mutations were introduced in the 3'-terminal untranslated region (3'-UTR) of full-length infectious poliovirus cDNA clones, and following transfection of COS-1 cells the ability of these constructs to generate viable viral particles and/or to support viral RNA synthesis was assayed. Substitution of the 3'-UTR of poliovirus RNA with the equivalent sequences of HAV RNA abrogated viral RNA replication, whereas the introduction of extended 'foreign' sequences between the open reading frame and the 3'-UTR was well tolerated. Point mutation that either destabilized the stem-and-loop structure or altered the sequence of the loop in domain 'Y' (nomenclature as per Pilipenko et al., [Nuclei Acids Res. 20 (1992) 1739-1745]) abolished both the infectivity and viral RNA synthesis. These were not restored by compensatory mutation that reconstructed the native secondary structure of this domain, suggesting that the secondary/tertiary folding of the 3'-UTR is not the only determinant for template recognition at initiation of RNA synthesis, but rather that a specific primary sequence is indeed required.

Maturation of pre-tRNA(fMet) by Escherichia coli RNase P is specified by a guanosine of the 5'-flanking sequence

The C+1/A+72 base pair at the top of the acceptor stem of Escherichia coli tRNA(fMet) accounts for several of the specialized roles of this tRNA in translation initiation. According to the rules of RNA substrate recognition by RNase P, the C+1/A+72 pair is likely to disfavor the 5'-maturation of pre-tRNA(fMet). Indeed, in contrast to other E. coli tRNA species, tRNA(fMet) was not properly matured when overproduced from a multicopy expression vector. Half of the recovered tRNA(fMet) retained an extension at the 5' side. Such a defect of tRNA(fMet) processing could be cured by changing bases C+1 and A+72 by a Watson-Crick base pair or by non-paired bases, provided one of them was a G. It could also be compensated by either (i) over-expression of RNase P or (ii) introduction within the plasmid of one out of the three 5'-flanking sequences naturally occurring in the four E. coli tRNA(fMet) genes. The effect of these flanking sequences on the maturation of tRNA(fMet) could be accounted for by the presence of a G located 2 bases upstream from C+1. Notably, this G is the only residue that is conserved in the 5'-flanking sequences of all four E. coli tRNA(fMet) genes.

Sequences at four termini of the giardiavirus double-stranded RNA

The linear double-stranded RNA (dsRNA) genome of giardiavirus (GLV) was estimated to be 6100 nucleotides (nt) [Wang et al., Proc. Natl. Acad. Sci. USA 90 (1993) 8585-8599]. As the 5'- and 3'-untranslated regions (UTR) of viral genomes are known to contain critical information for viral replication, we reexamine the sequences at all four termini of GLV dsRNA by (i) direct RNA sequencing with RT, (ii) tailing GLV dsRNA with UTP or CTP in addition to ATP in 3' rapid amplification of cDNA ends (RACE) and (iii) adding poly(dG) to the products of primer extension in 5'-RACE. The results confirmed the reported sequence for the 5'-terminus of the GLV sense strand RNA, but uncovered an additional 177 nt at the 3'-terminus. The new study also showed conclusively that there are no protruding overhangs at either terminus of GLV dsRNA.

Self-coded 3'-extension of run-off transcripts produces aberrant products during in vitro transcription with T7 RNA polymerase

More than 70% of the RNA synthesized by T7 RNA polymerase during run-off transcription in vitro can be incorrect products, up to twice as long as the expected transcripts. Transcriptions with model templates indicate that false transcription is mainly observed when the correct product cannot form stable secondary structures at the 3'-end. Therefore, the following hypothesis is tested: after leaving the DNA template, the polymerase can bind a transcript to the template site and the 3'-end of the transcript to the product site and extend it, if the 3'-end is not part of a stable secondary structure. Indeed, incubation of purified transcripts with the polymerase in transcription conditions triggers a 3'-end prolongation of the RNA. When two RNAs of different lengths are added to the transcription mix, both generate distinct and specific patterns of prolonged RNA products without any interference, demonstrating the self-coding nature of the prolongation process. Furthermore, sequencing of the high molecular weight transcripts demonstrates that their 5'-ends are precisely defined in sequence, whereas the 3'-ends contain size-variable extensions which show complementarity to the correct transcript. Surprisingly, a reduction of the UTP concentration to 0.2-1.0 mM in the presence of 3.5-4.0 mM of the other NTPs leads to faithful transcription and good yields, irrespective of the nucleotide composition of the template.

Preferential selection of adenosines for modification by double-stranded RNA adenosine deaminase

Double-stranded RNA adenosine deaminase (dsRAD), previously called the double-stranded RNA (dsRNA) unwinding/modifying activity, modifies adenosines to inosines within dsRNA. We used ribonuclease U2 and a mutant of ribonuclease T1 to map the sites of modification in several RNA duplexes. We found that dsRAD had a 5' neighbor preference (A = U > C > G) but no apparent 3' neighbor preference. Further, the proximity of the strand termini affected whether an adenosine was modified. Most importantly, dsRAD exhibited selectivity, modifying a minimal number of adenosines in short dsRNAs. Our results suggest that the specific editing of glutamate receptor subunit B mRNA could be performed in vivo by dsRAD without the aid of specificity factors, and support the hypothesis that dsRAD is responsible for hypermutations in certain RNA viruses.

Beta-tubulin genes of Trichomonas vaginalis

Microtubules, formed by polymerization of alpha and beta-tubulins, are major structural components of the mitotic spindle, cytoskeleton, and flagella, and are also an important target for the antiparasitic benzimidazole drugs. Trichomonas vaginalis, a flagellated protozoan responsible for urogenital tract infections in humans, is highly sensitive to certain benzimidazoles in vitro. As a first step towards defining the roles of microtubules in this organism, the regulation of their expression, and the basis for their benzimidazole sensitivity, we have characterized the genes encoding T. vaginalis beta-tubulin. A combination of genomic DNA cloning using bacteriophage lambda and PCR amplification using conserved beta-tubulin gene primers was employed. Southern blots of DNA from two different T. vaginalis strains suggest there are 6-7 beta-tubulin gene copies. Sequencing identified three distinct genes: btub1, btub2, and btub3. Amplification of cDNA with gene-specific primers indicated that the relative expression of RNA transcripts was btub1 > btub2 >> btub3. The promoter region from btub1 includes a 15-bp repeat also found (with 1-bp difference) upstream of the T. vaginalis ferredoxin gene. Primer extension suggests the 5' leader of the mRNA transcribed from btub1 is only 10 nucleotides long, similar to the lengths found in other anaerobic protozoa. In 152 residues examined by PCR, btub2 and btub3 differed by 1 and 12 amino acids, respectively, from btub1. All three sequences, however, have diverged considerably (20-24%) from beta-tubulins of other protozoa. T. vaginalis beta-tubulins include residues Tyr167 and Phe200, previously implicated in resistance and sensitivity, respectively, to the benzimidazole derivative benomyl.

Nucleotide sequence of sweet clover necrotic mosaic dianthovirus RNA-1

The complete nucleotide sequence of sweet clover necrotic mosaic dianthovirus (SCNMV) RNA-1 has been determined. RNA-1 consists of 3876 nucleotides in length, containing three large open reading frames (ORFs). The 5'-proximal, internal and 3'-terminal ORFs potentially encode 27-kDa, 57-kDa and 37-kDa proteins, respectively. The frameshift event between the C-terminus of the 27-kDa protein and extension of the N-terminus of the 57-kDa protein may result in the formation of a 88-kDa protein which is presumed to be a replicase. The 37-kDa coat protein ORF is located immediately downstream of the 57-kDa ORF. The same genome organization and high similarity (80-92%) of both the nucleotide sequences and the deduced amino acid sequences between red clover necrotic mosaic dianthovirus and SCNMV suggest that they originate from a common progenitor, but have divergent evolution later. Striking similarity was detected between the putative RNA-dependent RNA polymerase of SCNMV and that of the tombus-, carmo-, necro-, machlomo- and luteoviruses, supporting a proposal that they belong to the same virus supergroup although there is a relatively low degree of coat protein sequence similarity in these viruses.

The 5' nontranslated region of hepatitis A virus RNA: secondary structure and elements required for translation in vitro

Although the lengthy 5' nontranslated regions (5'NTRs) of other picornaviral RNAs form highly ordered structures with important functions in viral translation, little is known about the 5'NTR of hepatitis A virus (HAV). We determined the nearly complete 5'NTR nucleotide sequences of two genetically divergent HAV strains (PA21 and CF53) and included these data in a comparative phylogenetic analysis of the HAV 5'NTR. We identified covariant nucleotide substitutions predictive of conserved secondary structures and used this information to develop a model of the 5'NTR secondary structure, which was further refined by thermodynamic predictions and nuclease digestion experiments. According to this model, the 5'NTR comprises six major structural domains. Domains I and II (bases 1 to 95) contain a 5'-terminal hairpin and two stem-loops followed by a single-stranded and highly variable pyrimidine-rich tract (bases 96 to 154). The remainder of the 5'NTR (domains III to VI, bases 155 to 734) contains several complex stem-loops, one of which may form a pseudoknot, and terminates in a highly conserved region containing an oligopyrimidine tract preceding the putative start codon by 13 bases. To determine which structural elements might function as an internal ribosome entry site, RNA transcripts representing the HAV 5'NTR with progressive 5' deletions were translated in rabbit reticulocyte lysates. The translation product was truncated, unprocessed P1 polyprotein. Removal of the 5'-terminal 354 bases of the 5'NTR had little effect on translation. However, deletion to base 447 slightly decreased translation, while deletion to base 533 almost completely abolished it. These data indicate that sequences 3' of base 355 play an important role in the translation mechanism utilized by genomic-length HAV RNA. Significantly, this region shares several conserved structural features with the internal ribosome entry site element of murine encephalomyocarditis virus.

Bovine lactoferrin mRNA: sequence, analysis, and expression in the mammary gland

The mRNA sequence for bovine lactoferrin expressed in the mammary gland was determined by sequencing three over lapping cDNA clones and by direct sequencing of the mRNA. The mRNA (2351 bases) codes for a 708 amino acid protein with a 19 amino acid signal peptide immediately preceding a sequence identical to the N-terminal 40 amino acids reported for bovine lactoferrin. A putative destabilizing sequence (AUUUA) was identified in the 3'-untranslated region. The nucleic acid sequence and deduced amino acid sequence are highly homologous with other transferrin family members. Lactoferrin mRNA concentrations in bovine mammary tissue were quite low two days before parturition and during lactation but were high three days after the cessation of milking, a sharp contrast from the pattern of regulation of the other milk proteins.

Identification of antigenically important domains in the glycoproteins of Sindbis virus by analysis of antibody escape variants

To study important epitopes on glycoprotein E2 of Sindbis virus, eight variants selected to be singly or multiply resistant to six neutralizing monoclonal antibodies reactive against E2, as well as four revertants which had regained sensitivity to neutralization, were sequenced throughout the E2 region. To study antigenic determinants in glycoprotein E1, four variants selected for resistance to a neutralizing monoclonal antibody reactive with E1 were sequenced throughout the E2 and E1 regions. All of the salient changes in E2 occurred within a relatively small region between amino acids 181 and 216, a domain that encompasses a glycosylation site at residue 196 and that is rich in charged amino acids. Almost all variants had a change in charge, suggesting that the charged nature of this domain is important for interaction with antibodies. Variants independently isolated for resistance to the same antibody were usually altered in the same amino acid, and reversion to sensitivity occurred at the sites of the original mutations, but did not always restore the parental amino acid. The characteristics of this region suggest that this domain is found on the surface of E2 and constitutes a prominent antigenic domain that interacts directly with neutralizing antibodies. Previous studies have shown that this domain is also important for penetration of cells and for virulence of the virus. Resistance to the single E1-specific neutralizing monoclonal antibody resulted from changes of Gly-132 of E1 to either Arg or Glu. Analogous to the findings with E2, these changes result in a change in charge and are found near a glycosylation site at residue 139. This domain of E1 may therefore be found near the 181 to 216 domain of E2 on the surface of the E1-E2 heterodimer; together, they could form a domain important in virus penetration and neutralization.

Mutagenesis of the 3' nontranslated region of Sindbis virus RNA

A cDNA clone from which infectious RNA can be transcribed was used to construct 42 site-specific mutations in the 3' nontranslated region of the Sindbis virus genome. The majority of these mutations were made in the 3'-terminal 19-nucleotide conserved sequence element and consisted of single nucleotide substitutions or of small (1 to 8) nucleotide deletions. An attempt was made to recover mutant viruses after transfection of SP6-transcribed RNA into chicken cells. In most cases, viable virus was recovered, but almost all mutants grew more poorly than wild-type virus when tested under a number of culture conditions. In the case of mutations having only a moderate effect, the virus grew as well as the wild type but was slightly delayed in growth. Mutations having a more severe effect led to lower virus yields. In many cases, virus growth was more severely impaired in mosquito cells than in chicken cells, but the opposite phenotype was also seen, in which the mutant grew as well as or better than the wild type in mosquito cells but more poorly in chicken cells. One substitution mutant, 3NT7C, was temperature sensitive for growth in chicken cells and severely crippled for growth in mosquito cells. Insertion mutations were also constructed which displaced the 19-nucleotide element by a few nucleotides relative to the poly(A) tail. These mutations had little effect on virus growth. Deletion of large regions (31 to 293 nucleotides long) of the 3' nontranslated region outside of the 19-nucleotide element resulted in viruses which were more severely crippled in mosquito cells than in chicken cells. From these results, the following principles emerge. (i) The entire 3' nontranslated region is important for efficient virus replication, although there is considerable plasticity in this region in that most nucleotide substitutions or deletions made resulted in viable virus and, in some cases, in virus that grew quite efficiently. Replication competence was particularly sensitive to changes involving the C at position 1, the A at position 7, and a stretch of 9 U residues punctuated by a G at position 14. (ii) The panel of mutants examined collectively deleted the entire 3' nontranslated region. Only mutants in which 8 nucleotides in the 3' terminal 19 nucleotides had been deleted or in which the 3' terminal C was deleted were nonviable. Although the 3' terminal C was essential for replication, it could be displaced by at least 7 nucleotides from its 3' terminal position adjacent to the poly(A) tract.(ABSTRACT TRUNCATED AT 400 WORDS)