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)     

The deletion of 41 proximal nucleotides reverts a poliovirus mutant containing a temperature-sensitive lesion in the 5'' noncoding region of genomic RNA.

We generated a number of small deletions and insertions in the 5' noncoding region of an infectious cDNA copy of the poliovirus RNA genome. Transfection of these mutated cDNAs into COS-1 cells produced the following phenotypic categories: (i) wild-type mutations, (ii) lethal mutations, (iii) mutations exhibiting slow growth or low-titer properties, and (iv) temperature-sensitive (ts) mutations. The deletion of nucleotides 221 to 224 produced a ts virus, 220D1. Mutant 220D1 was found to have a dramatic reduction in growth, virus-specific protein and RNA synthesis, and the shutoff of host cell protein synthesis at 37 or 39 degrees C compared with 33 degrees C. Temperature shift experiments showed that the mutant viral RNA is not an effective template for protein or RNA synthesis at 39 degrees C and suggested a decreased stability of the 220D1 RNA at 39 degrees C. Selection for a non-ts revertant of 220D1 yielded the virus R2, which was no longer ts for growth or viral protein and RNA synthesis. Sequencing the 5' noncoding region of the genomic RNA from R2 revealed the deletion of 41 proximal nucleotides for an overall deletion of nucleotides 184 to 228. These data suggest that the deleted sequences are nonessential to the poliovirus life cycle during growth in HeLa cells. According to computer-predicted RNA secondary structures of the 5' noncoding region of poliovirus RNA, the R2 revertant virus has deleted an entire predicted stem-loop structure.     

A Predicted Secondary Structural Domain within the Internal Ribosome Entry Site of Echovirus 12 Mediates a Cell-Type-Specific Block to Viral Replication

The enterovirus 5' nontranslated region (NTR) contains an internal ribosome entry site (IRES), which facilitates translation initiation of the viral open reading frame in a 5' (m(7)GpppN) cap-independent manner, and cis-acting signals for positive-strand RNA replication. For several enteroviruses, the 5' NTR has been shown to determine the virulence phenotype. We have constructed a chimera consisting of the putative IRES element from the Travis strain of echovirus 12 (ECV12), a wild-type, relatively nonvirulent human enterovirus, exchanged with the homologous region of a full-length infectious clone of coxsackievirus B3 (CBV3). The resulting chimera, known as ECV12(5'NTR)CBV3, replicates similarly to CBV3 in human and simian cell lines yet, unlike CBV3, is completely restricted for growth on two primary murine cell lines at 37 degrees C. By utilizing a reverse-genetics approach, the growth restriction phenotype was localized to the predicted stem-loop II within the IRES of ECV12. In addition, a revertant of ECV12(5'NTR)CBV3 was isolated which possessed three transition mutations and had restored capability for replication in the utilized murine cell lines. Assays for cardiovirulence indicated that the ECV12 IRES is responsible for a noncardiovirulent phenotype in a murine model for acute myocarditis. The results indicate that the 5' NTRs of ECV12 and CBV3 exhibit variable intracellular requirements for function and serve as secondary determinants of tissue or species tropism.     

Mutations responsible for adaptation of hepatitis A virus to efficient growth in cell culture.

Chimeric genomes of hepatitis A virus strain HM-175 were constructed from cDNA clones of the wild-type virus and its cell culture-adapted variant. RNA transcribed in vitro from each construct was assayed for infectivity by transfection of cultured cells. RNA transcribed from the wild-type cDNA clone was minimally infectious and produced virus that grew inefficiently in vitro, whereas that transcribed from certain chimeric genomes consistently produced virus that grew efficiently in cultured cells. Mutations in the P2 region were found to be necessary for efficient virus growth in vitro, while mutations in the 5' noncoding region imparted a conditional enhancement of growth in vitro.     

Sequence of the region coding for virion proteins C and E2 and the carboxy terminus of the nonstructural proteins of rubella virus: comparison with alphaviruses

The sequence of the 3' 4508 nucleotides (nt) of the genomic RNA of the Therien strain of rubella virus (RV) was determined for cDNA clones. The sequence contains a 3189-nt open reading frame (ORF) which codes for the structural proteins C, E2 and E1. C is predicted to have a length of 300 amino acids (aa). The N-terminal half of the C protein is highly basic and hydrophilic in nature, and is putatively the region of the protein which interacts with the virion RNA. At the C terminus of the C protein is a stretch of 20 hydrophobic aa which also serves as the signal sequence for E2, indicating that the cleavage of C from the polyprotein precursor may be catalyzed by signalase in the lumen of the endoplasmic reticulum. E2 is 282 aa in length and contains four potential N-linked glycosylation sites and a putative transmembrane domain near its C terminus. The sequence of E1 has been previously described [Frey et al., Virology 154 (1986) 228-232]. No homology could be detected between the amino acid sequence of the RV structural proteins and the amino acid sequence of the alphavirus structural proteins. From the position of a region of 30 nt in the RV genomic sequence which exhibited significant homology with the sequence in the alphavirus genome at which subgenomic RNA synthesis is initiated, the RV subgenomic RNA is predicted to be 3346 nt in length and the nontranslated region from the 5' end of the subgenomic RNA to the structural protein ORF is predicted to be 98 nt. In a different translation frame beginning at the 5' end of the RV nt sequence reported here is a 1407 nt ORF which is the C terminal region of the nonstructural protein ORF. This ORF overlaps the structural protein ORF by 149 nt. A low level of homology could be detected between the predicted amino acid sequence of the C-terminus of the RV nonstructural protein ORF and the replicase proteins of several positive RNA viruses of animals and plants, including nsp4 of the alphaviruses, the protein encoded by the C-terminal region of the alphavirus nonstructural ORF. However, the overall homology between RV and the alphaviruses in this region of the genome was only 18%, indicating that these two genera of the Togavirus family are only distantly related. Intriguingly, there is a 2844-nt ORF present in the negative polarity orientation of the RV sequence which could encode a 928-aa polyprotein.     

Mutations in the 5' nontranslated region of bovine viral diarrhea virus result in altered growth characteristics

The 5' nontranslated region (NTR) of pestiviruses functions as an internal ribosome entry site (IRES) that mediates cap-independent translation of the viral polyprotein and probably contains additional cis-acting RNA signals involved in crucial processes of the viral life cycle. Computer modeling suggests that the 5'-terminal 75 nucleotides preceding the IRES element form two stable hairpins, Ia and Ib. Spontaneous and engineered mutations located in the genomic region comprising Ia and Ib were characterized by using infectious cDNA clones of bovine viral diarrhea virus. Spontaneous 5' NTR mutations carrying between 9 and 26 A residues within the loop region of Ib had no detectable influence on specific infectivity and virus growth properties. After tissue culture passages, multiple insertions and deletions of A residues occurred rapidly. In contrast, an engineered mutant carrying 5 A residues within the Ib loop was genetically stable during 10 tissue culture passages. This virus was used as starting material to generate a number of additional mutants. The analyses show that (i) deletion of the entire Ib loop region resulted in almost complete loss of infectivity that was rapidly restored during passages in cell culture by insertions of variable numbers of A residues; (ii) mutations within the 5'-terminal 4 nucleotides of the genomic RNA severely impaired virus replication; passaging of the supernatants obtained after transfection resulted in the emergence of efficiently replicating mutants that had regained the conserved 5'-terminal sequence; (iii) provided the conserved sequence motif 5'-GUAU was retained at the 5' end of the genomic RNA, substitutions and deletions of various parts of hairpin Ia or deletion of all of Ia and part of Ib were found to support replication, but to a lower degree than the parent virus. Restriction of specific infectivity and virus growth of the 5' NTR mutants correlated with reduced amounts of accumulated viral RNAs.     

Changes in both gp120 and gp41 can account for increased growth potential and expanded host range of human immunodeficiency virus type 1.

Virus derived from an infectious molecular clone of the ELI strain of human immunodeficiency virus type 1 (HIV-1) replicates well in peripheral blood mononuclear cells and in some CD4-positive cell lines but exhibits a delayed time course of infection in CEM and H9 cells and fails to infect SupT1 and U937 cells. If the virus that emerges from infected H9 cells is used to infect CEM and H9 cells, the time course of infection is accelerated and the virus is able to infect U937 and SupT1 cells. In this study, we used the technique of polymerase chain reaction-single-strand conformation polymorphism to localize changes in both the extracellular gp120 and the transmembrane gp41 components of the envelope gene associated with adaptation to growth in tissue culture cell lines. Specifically, mutations were identified both in a region of gp120 implicated in CD4 binding and in the amino-terminal portion of gp41 adjacent to the region involved in fusion. No changes were found in the V3 loop of gp120, a region previously shown to be involved in viral tropism. When these mutations were introduced into the original molecular clone, they conferred an enhanced replicative capacity on ELI. These results demonstrate that two additional determinants in the HIV-1 envelope protein influence viral tropism and growth in vitro. They also may have important implications for the generation of viruses with increased growth potential and expanded host range seen in the late stages of HIV disease.     

Mapping of attenuating sequences of an avirulent poliovirus type 2 strain.

A mouse model for poliomyelitis was used to identify genomic sequences that attenuate neurovirulence of poliovirus strain P2/P712. This type 2 strain is avirulent in primates and mice yet grows as well as virulent strains in cell culture. The approach used was to exchange portions of the genome of the mouse-virulent P2/Lansing strain with the corresponding region from P2/P712 to identify sequences that could attenuate Lansing neurovirulence in mice. A full-length infectious cDNA of P2/P712 was assembled and used to construct recombinants between P2/P712 and P2/Lansing. The results of neurovirulence testing of 11 recombinants indicated that strong attenuating determinants are located in the 5' noncoding region of P2/P712 and a region encoding capsid protein VP1 and 2Apro, 2B, and part of 2C. An attenuating determinant was further localized to between nucleotides 456 and 628 of P2/P712. A third sequence from P2/P712, nucleotides 752 to 2268, encoding VP4, VP2, and part of VP3, was weakly attenuating. The sequence from nucleotide 4454, approximately halfway through the 2C-coding region, to the end of the P2/P712 genome did not contain attenuating determinants. Nucleotide sequence analysis revealed that P2/P712 differs from the type 2 Sabin vaccine strain by only 22 nucleotides. Six differences lead to amino acid changes in the coding region, and four differences are in the 5' noncoding region. These studies show that, like the type 1 and type 3 Sabin vaccine strains, the attenuated type 2 strain P712 contains multiple attenuating sequences, including strongly attenuating sequences in the 5' noncoding region of the genome.     

A mutation in integrase can compensate for mutations in the simian immunodeficiency virus att site.

Sequences at the left terminus of U3 in the left long terminal repeat (LTR) and at the right terminus of U5 in the right LTR are important for integration of retroviral DNA. In the infectious pathogenic molecular clone of simian immunodeficiency virus strain mac239 (SIVmac239), 10 of the 12 terminal base pairs form an imperfect inverted repeat structure (5' TGGAAGGGATTT 3' [nucleotides 1 to 12] and 3' ACGATCCCTAAA 5' [nucleotides 10279 to 10268]). Nineteen different mutant forms of SIVmac239 proviral DNA with changes at one or more of the positions in each of the 12-terminal-base-pair regions were constructed. Viral replication was severely or completely compromised with nine of these mutants. Revertants appeared 40 to 50 days after transfection in two independent experiments with mutant 7, which contained changes of AGG to TAC at positions 5 to 7 in U3 and TCC to GAA at positions 10275 to 10273 in U5. Virus produced at these times from mutant 7 transfection replicated upon reinfection with only a slight delay when compared to the wild type. Sequence analysis of the LTR and integrase regions from infected cultures revealed two predominant changes: G to A at position 10275 in U5 and Glu to Lys at position 136 in integrase. Derivatives of clone 7 in which these changes were introduced individually and together were constructed by site-specific mutagenesis. Each change individually restored replication capacity only partially. However, the combination of both mutations restored replicative capacity to that of the original revertants. These results indicate that changes in integrase can compensate for mutations in the terminal nucleotides of the SIV LTR. The results further indicate that resistance to integrase inhibitors may include both integrase and LTR mutations.     

Spontaneous substitutions in the vicinity of the V3 analog affect cell tropism and pathogenicity of simian immunodeficiency virus.

Simian immunodeficiency virus (SIV) exists within tissues of infected macaques as a mixture of diverse genotypes. The goal of this study was to investigate the biologic significance of this variation in terms of cellular tropism and pathogenicity. PCR was used to amplify and clone 3'-half genomes from the spleen of an immunodeficiency SIV-infected pig-tailed macaque (Macaca nemestrina). Eight infectious clones were generated by ligation of respective 3' clones into a related SIVsm 5' clone, and virus stocks were generated by transient transfection. Four of these viruses were infectious for macaque peripheral blood mononuclear cells (PBMC) or monocyte-derived macrophages (MDM). Three viruses with distinct tropism for macaque PBMC or MDM were tested for in vivo infectivity and pathogenicity. The ability of these three viruses to infect PBMC and macrophages correlated with differences in infectivity and pathogenicity. Thus, a virus that was infectious for both PBMC and MDM was highly infectious for macaques and induced AIDS in half of the inoculated animals. In contrast, virus that was less infectious for PBMC and not infectious for MDM induced only transient viremia. Finally, a virus that was not infectious for either primary cell type did not infect macaques. Chimeric clones exchanging portions of the envelope gene of the 62A and smH4 molecular clones and a series of point mutants were used to map the determinant of tropism to a 60-amino-acid region of gp120 encompassing the V3 analog of SIV. Naturally occurring mutations within this region were critical for determining tropism and, as a result, pathogenicity of these SIVsm clones.     

Transcriptional control of the mouse prealbumin (transthyretin) gene: both promoter sequences and a distinct enhancer are cell specific.

The mouse genomic clone for the prealbumin (transthyretin) gene was cloned, and its upstream regulatory regions were analyzed. The 200 nucleotides 5' to the cap site when placed within a recombinant plasmid were sufficient to direct transient expression in HepG2 (human hepatoma) cells, but this DNA region did not support expression in HeLa cells. The sequence of the 200-nucleotide region is highly conserved between mouse and human DNA and can be considered a cell-specific promoter. Deletions of this promoter region identified a crucial element for cell-specific expression between 151 and 110 nucleotides 5' to the RNA start site. A region situated at about 1.6 to 2.15 kilobases upstream of the RNA start site was found to stimulate expression 10-fold in HepG2 cells but not in HeLa cells. This far upstream element was invertible and increased expression from the beta-globin promoter in HepG2 cells. Unlike the simian virus 40 enhancer, the prealbumin enhancer would not stimulate beta-globin synthesis in HeLa cells, and even the simian virus 40 enhancer did not stimulate the prealbumin promoter in HeLa cells. Thus, we identified in the prealbumin gene two DNA elements that respond in a cell-specific manner: a proximal promoter including a crucial sequence between -108 and -151 nucleotides and a distant enhancer element located between 1.6 and 2.15 kilobases upstream.     

The mismatched nucleotides in the 5'-terminal hairpin of minute virus of mice are required for efficient viral DNA replication.

The 5'-terminal sequence in the DNA of the parvovirus minute virus of mice (MVM) is a palindrome. It can form a hairpin, the stem of which is entirely base-paired except for three consecutive unpaired nucleotides which form a bubble. Since this structure is well conserved among different parvoviruses, we examined its importance for viral replication by generating MVM mutants with alterations in this region. A clone of MVMp DNA which contained the entire 3' end and more than half of the 5' palindrome was made. Although it lacked the sequence information to form a wild-type bubble, this DNA was infectious. On transfection into A9 fibroblasts, it gave rise to a virus (MVMs) which had a bubble in its 5' palindrome. The bubble consisted of four mismatched nucleotides in the same location as the unpaired nucleotides of the wild-type palindrome. Apparently, neighboring plasmid sequences were incorporated into the viral DNA, enabling formation of the mismatch. This observation suggested that a bubble is critical for growth of MVM but that its sequence is not. To find out whether MVM lacking a bubble in the 5' palindrome is viable, we made a second clone in which the plasmid sequences incorporated in MVMs were removed. Transfection of this DNA gave rise to a virus (MVMx) in which the nucleotides unpaired in the wild-type hairpin are now fully base-paired. Although MVMx can be propagated, it is defective in comparison with wild-type MVMp; it exhibited about a 50-fold-lower ratio of plaque-forming units to DNA content. In mixed infections, MVMp consistently outgrew the bubbleless MVMx. The rate of accumulation of DNA replication intermediates was lower for MVMx than for the wild-type virus. Quantitative analysis of the 5' termini of replicative form DNA suggested that the ability of MVMx to convert hairpin 5' termini to extended termini is impaired. In contrast, the virus with the altered bubble, MVMs, behaved like the wild-type MVMp in all the assays. We conclude that MVM lacking a bubble in its 5'-terminal DNA hairpin is less infectious than and has a selective disadvantage compared with wild-type MVM. The nucleotide sequence of the bubble is not critical. We provide evidence that the presence of a bubble is necessary for efficient viral DNA replication.     

Mouse primase p49 subunit molecular cloning indicates conserved and divergent regions

Primase is a specialized RNA polymerase that synthesizes RNA primers for initiation of DNA synthesis. A full cDNA clone of the p49 subunit of mouse primase, a heterodimeric enzyme, has been isolated using a primase p49-specific polyclonal antibody to screen a lambda gt11 mouse cDNA expression library. The cDNA indicated the subunit is a 417-amino acid polypeptide with a calculated molecular mass of 49,295 daltons. The p49 mRNA is approximately 1500 nucleotides long with a 5'-untranslated region of 74 nucleotides and a 3'-untranslated region of 200 nucleotides. Comparison with a similar sized primase subunit from yeast showed highly conserved amino acid sequences in the N-terminal halves of the polypeptides and included a potential metal-binding domain suggesting the functional importance of this region for DNA binding. In contrast, the 3' portion of the cDNA has rapidly diverged in nucleotide sequence, as primase mRNA can be detected in mouse and rat cells with a 3' probe (including coding and noncoding) but not in RNA from hamster or human cells. A full-length cDNA probe detected mRNA from hamster and human cell lines, indicating a conserved 5' portion and divergent 3' region of the expressed gene. The rapid divergence may be related to the species-specific protein interactions found for the DNA polymerase alpha-primase complex. The mRNA is detected in proliferating but not in quiescent cells consistent with its function in DNA replication.     

Monoclonal antibody-defined epitope map of expressed rubella virus protein domains.

An expanded library of murine monoclonal antibodies (MAbs) was generated by infecting BALB/C mice with the Therien strain of rubella virus (RV) and selecting secreting hybrids by enzyme-linked immunosorbent assay (ELISA) using purified virion targets. A panel of plasmids containing specified RV cDNA fragments was also constructed by using a variety of strategies with pGE374- and pGE374-derived expression vectors. Hybrid RecA-RV-beta-galactosidase (LacZ)- or RecA-RV-truncated LacZ-containing proteins collectively representing the entire open reading frame of the structural proteins of RV were overexpressed in Escherichia coli. Bacterial lysates were then probed by ELISA with selected MAbs and by immunoblot following separation by electrophoresis under denaturing conditions. With this approach, MAbs that appeared to react with linear determinants defined epitopes localized within the following domains: MAbs C-1, C-2, and C-8 bind epitopes within the predicted amino-terminal 21 amino acids of the capsid region C9 to C29; MAb C-9 binds to a domain bounded by C64 and C97; MAbs E2-1 through E2-6 bind to the E2 glycoprotein backbone region from E2(1) to E2(115); MAbs E1-18 and E1-20 bind to the E1 glycoprotein region from E1(202) to E1(283). MAb E1-18 neutralizes RV infectivity; MAb E1-20 neutralizes infectivity and modestly inhibits hemagglutination. Analyses with selected synthetic peptides have confirmed several of the molecular domains deduced with the expressed proteins. These plasmid constructions and peptides have proven useful in beginning to unravel the molecular organization of several antigenic sites of this human pathogen.     

A mouse immunoglobulin heavy chain deletion mutant: isolation of a cDNA clone and sequence analysis of the mRNA

The mouse cell line IF2 secretes an immunoglobulin heavy chain lacking the CH1 domain. We have isolated and characterised a recombinant plasmid containing cDNA copies of the IF2 mutant mRNA. The cloned sequence extends from the nucleotides coding for amino acid 96 in the variable region through 100 nucleotides of untranslated region at the 3' end. The sequence of the cDNA insert reveals no discontinuity at the variable-hinge region junction, the site of the CH1 deletion. Experiments employing direct priming on the poly(A) tail of the IF2 heavy chain mRNA suggest that the 3' end of the cDNA clone (sequence C-C-C-T-G-C) is also the 3' end of the mRNA.     

Isolation of the chicken middle-molecular weight neurofilament (NF-M) gene and characterization of its promoter.

We have isolated and sequenced genomic DNA clones covering the coding region of the chicken mid-size neurofilament (NF-M) gene and greater than 1 kb of its 5' upstream region. The NF-M gene contains two introns which both are located within the highly conserved C-terminal region of the rod domain. The 5' end of the corresponding mRNA was assigned to a G residue 40 nucleotides upstream of the translation start site and in appropriate distance from a potential TATA box. To functionally analyze the NF-M promoter, constructs carrying 112, 222, and 1026 nucleotides of the 5' upstream region in front of a luciferase reporter gene were tested for their capability to direct luciferase expression after transient transfection into various cell lines. Significant luciferase activity was recorded both in rat phaeochromocytoma (PC12) cells and murine fibroblasts. In PC12 cells, in which neurite outgrowth is induced by nerve growth factor (NGF), expression was stimulated up to 13-fold within 3 days of NGF treatment. This closely resembles expression of the endogenous NF-M gene in response to this hormone.     

Construction of chimeric plasmids containing histone H5 cDNA from hen erythrocyte. DNA sequence of a fragment derived from the 5' region of H5 mRNA.

We report that construction and characterization of chicken erythrocyte histone H5 cDNA recombinant plasmids. cDNA was synthesized from poly(A)+ polysomal RNA enriched in H5 mRNA and inserted into the PstI site of pBR322. Several clones containing H5 cDNA sequences were obtained and one of them (p541), expressing H5 antigenic determinants, was sequenced. The DNA insert of p541 contains 118 nucleotides from the 5' non-translated region of H5 mRNA and sequences coding for up to residue 46 of the N-terminus of the arginine (position 15) H5 variant. There is a strikingly high number of repeated sequences both in the leader and coding region; among these, the octanucleotide 5' GCG GCG GC 3' is found five times along the sequence. Although the H5 mRNA 5' leader is GC-rich (66%), there is an AT-rich region, about 16 nucleotides long, which shares strong homology with the leaders of sea urchin histone H1 mRNAs.     

Cell proteins bind specifically to West Nile virus minus-strand 3' stem-loop RNA.

The first 96 nucleotides of the 5'noncoding region (NCR) of West Nile virus (WNV) genomic RNA were previously reported to form thermodynamically predicted stem-loop (SL) structures that are conserved among flaviviruses. The complementary minus-strand 3' NCR RNA, which is thought to function as a promoter for the synthesis of plus-strand RNA, forms a corresponding predicted SL structure. RNase probing of the WNV 3' minus-strand stem-loop RNA [WNV (-)3' SL RNA] confirmed the existence of a terminal secondary structure. RNA-protein binding studies were performed with BHK S100 cytoplasmic extracts and in vitro-synthesized WNV (-)3' SL RNA as the probe. Three RNA-protein complexes (complexes 1,2, and 3) were detected by a gel mobility shift assay, and the specificity of the RNA-protein interactions was confirmed by gel mobility shift and UV-induced cross-linking competition assays. Four BHK cell proteins with molecular masses of 108, 60, 50, and 42 kDa were detected by UV-induced cross-linking to the WNV (-)3' SL RNA. A preliminary mapping study indicated that all four proteins bound to the first 75 nucleotides of the WNV 3' minus-strand RNA, the region that contains the terminal SL. A flavivirus resistance phenotype was previously shown to be inherited in mice as a single, autosomal dominant allele. The efficiencies of infection of resistant cells and susceptible cells are similar, but resistant cells (C3H/RV) produce less genomic RNA than congenic, susceptible cells (C3H/He). Three RNA-protein complexes and four UV-induced cross-linked cell proteins with mobilities identical to those detected in BHK cell extracts with the WNV (-)3' SL RNA were found in both C3H/RV and C3H/He cell extracts. However, the half-life of the C3H/RV complex 1 was three times longer than that of the C3H/He complex 1. It is possible that the increased binding activity of one of the resistant cell proteins for the flavivirus minus-strand RNA could result in a reduced synthesis of plus-strand RNA as observed with the flavivirus resistance phenotype.