Mapping of sequences required for mouse neurovirulence of poliovirus type 2 Lansing.

Intracerebral inoculation of mice with poliovirus type 2 Lansing induces a fatal paralysis, while most other poliovirus strains are unable to cause disease in the mouse. To determine the molecular basis for Lansing virus neurovirulence, we determined the complete nucleotide sequence of the Lansing viral genome from cloned cDNA. The deduced amino acid sequence was compared with that of two mouse-avirulent strains. There are 83 amino acid differences between the Lansing and Sabin type 2 strain and 179 differences between the Lansing and Mahoney type 1 strain scattered throughout the genome. To further localize Lansing sequences important for mouse neurovirulence, four intertypic recombinants were isolated by exchanging DNA restriction fragments between the Lansing 2 and Mahoney 1 infectious poliovirus cDNA clones. Plasmids were transfected into HeLa cells, and infectious recombinant viruses were recovered. All four recombinant viruses, which contained the Lansing capsid region and different amounts of the Mahoney genome, were neurovirulent for 18- to 21-day-old Swiss-Webster mice by the intracerebral route. The genome of neurovirulent recombinant PRV5.1 contained only nucleotides 631 to 3413 from Lansing, encoding primarily the viral capsid proteins. Therefore, the ability of Lansing virus to cause paralysis in mice is due to the viral capsid. The Lansing capsid sequence differs from that of the mouse avirulent Sabin 2 strain at 32 of 879 amino acid positions: 1 in VP4, 5 in VP2, 4 in VP3, and 22 in VP1.     

Genetic basis of attenuation of the Sabin type 3 oral poliovirus vaccine.

The poliovirus type 3 Sabin oral poliovirus vaccine strain P3/Leon/12a1b differs in nucleotide sequence from its neurovirulent progenitor P3/Leon/37 by just 10 point mutations. The contribution of each mutation to the attenuation phenotype of the vaccine strain was determined by the construction of a series of recombinant viruses from infectious cDNA clones. The neurovirulence testing of recombinant viruses indicated that the attenuation phenotype is determined by just two point mutations: a C to U in the noncoding region at position 472 and a C to U at nucleotide 2034 which results in a serine-to-phenylalanine amino acid substitution in the structural protein VP3.     

Role of mutations G-480 and C-6203 in the attenuation phenotype of Sabin type 1 poliovirus.

Of the 55 point mutations which distinguish the type 1 poliovirus vaccine strain (Sabin 1) from its neurovirulent progenitor (P1/Mahoney), two have been strongly implicated by previous studies as determinants of the attenuation phenotype. A change of an A to a G at position 480, located within the 5' noncoding region, has been suggested to be the major attenuating mutation, analogous to the mutations at positions 481 and 472 in poliovirus types 2 and 3, respectively. In addition, the change of a U to a C at position 6203, resulting in an amino acid change in the polymerase protein 3D, has also been implicated as a determinant of attenuation, albeit to a lesser extent. To assess the contributions of these mutations to attenuation and temperature sensitivity, reciprocal changes were generated at these positions in infectious cDNA clones of Sabin 1 and P1/Mahoney. Assays in tissue culture and primates indicated that the two mutations make some contribution to the temperature sensitivity of the Sabin 1 strain but that neither is a strong determinant of attenuation.     

Changes in three of the four coat proteins of oral polio vaccine strain derived from type 1 poliovirus.

Little is yet known about the nature, or extent, of the changes involved in attenuation of neurovirulent poliovirus. The tryptic comparison reported here, of coat proteins from the Sabin type 1 polio vaccine and parental Mahoney virus, provides a useful approach and affords some insight into this question. The main obstacle, separation of the labile proteins VP1 and VP2 in an intact state from the vaccine strain, was overcome by incorporating 3.5 M urea into an otherwise standard preparative gel electrophoresis system. Tryptic maps revealed six altered leucine-containing peaks: two in VP1, none in VP2, three in VP3, and one in VP4. It is estimated, after correcting for leucine-free peptides, that the coat protein sequences may have undergone some 10 to 13 amino acid replacements, roughly 1.5% of the total, in the course of attenuation leading to the vaccine strain.     

Host range determinants located on the interior of the poliovirus capsid.

The inability of certain poliovirus strains to infect mice can be overcome by the expression of human poliovirus receptors in mice or by the presence of a particular amino acid sequence of the B-C loop of the viral capsid protein VP1. We have identified changes in an additional capsid structure that permit host-restricted poliovirus strains to infect mice. Variants of the mouse-virulent P2/Lansing strain were constructed containing amino acid changes, deletions and insertions in the B-C loop of VP1. These variants were attenuated in mice, demonstrating the importance of the B-C loop sequence in host range. Passage of two of the B-C loop variants in mice led to the selection of viruses that were substantially more virulent. The increased neurovirulence of these strains was mapped to two different suppressor mutations in the N-terminus of VP1. Whereas the B-C loop of VP1 is highly exposed on the surface of the capsid, near the five-fold axis of symmetry, the suppressor mutations are in the interior of the virion, near the three-fold axis. Introduction of the suppressor mutations into the genome of the mouse-avirulent P1/Mahoney strain resulted in neurovirulent viruses, demonstrating that the P2/Lansing B-C loop sequence is not required to infect mice. Because the internal host range determinants are in a structure known to be important in conformational transitions of the virion, the host range of poliovirus may be determined by the ability of virions to undergo transitions catalyzed by cell receptors.     

Genetic determinants of dengue type 4 virus neurovirulence for mice.

Mouse-adapted dengue type 4 virus (DEN4) strain H241 is highly neurovirulent for mice, whereas its non-mouse-adapted parent is rarely neurovirulent. The genetic basis for the neurovirulence of the mouse-adapted mutant was studied by comparing intratypic chimeric viruses that contained the three structural protein genes from the parental virus or the neurovirulent mutant in the background sequence of nonneurovirulent DEN4 strain 814669. The chimera that contained the three structural protein genes from mouse neurovirulent DEN4 strain H241 proved to be highly neurovirulent in mice, whereas the chimera that contained the corresponding genes from its non-mouse-adapted parent was not neurovirulent. This finding indicates that most of the genetic loci for the neurovirulence of the DEN4 mutant lie within the structural protein genes. A comparison of the amino acid sequences of the parent and its mouse neurovirulent mutant proteins revealed that there were only five amino acid differences in the structural protein region, and three of these were located in the envelope (E) glycoprotein. Analysis of chimeras which contained one or two of the variant amino acids of the mutant E sequence substituting for the corresponding sequence of the parental virus identified two of these amino acid changes as important determinants of mouse neurovirulence. First, the single substitution of Ile for Thr-155 which ablated one of the two conserved glycosylation sites in parental E yielded a virus that was almost as neurovirulent as the mouse-adapted mutant. Thus, the loss of an E glycosylation site appears to play a role in DEN4 neurovirulence. Second, the substitution of Leu for Phe-401 also yielded a neurovirulent virus, but it was less neurovirulent than the glycosylation mutant. These findings indicate that at least two of the genetic loci responsible for DEN4 mouse neurovirulence map within the structural protein genes.     

Genetic Determinants Responsible for Acquisition of Dengue Type 2 Virus Mouse Neurovirulence

Studies conducted some 50 years ago showed that serial intracerebral passage of dengue viruses in mice selected for neurovirulent mutants that also exhibited significant attenuation for humans. We investigated the genetic basis of mouse neurovirulence of dengue virus because it might be directly or indirectly associated with attenuation for humans. Analysis of the sequence in the C-PreM-E-NS1 region of the parental dengue type 2 virus (DEN2) New Guinea C (NGC) strain and its mouse-adapted, neurovirulent mutant revealed that 10 nucleotide changes occurred during serial passage in mice. Seven of these changes resulted in amino acid substitutions, i.e., Leu₅₅-Phe and Arg₅₇-Lys in PreM, Glu₇₁-Asp, Glu₁₂₆-Lys, Phe₄₀₂-Ile, and Thr₄₅₄-Ile in E, and Arg₁₀₅-Gln in NS1. The sequence of C was fully conserved between the parental and mutant DEN2. We constructed intertypic chimeric dengue viruses that contained the PreM-E genes or only the NS1 gene of neurovirulent DEN2 NGC substituting for the corresponding genes of DEN4. The DEN2 (PreM-E)/DEN4 chimera was neurovirulent for mice, whereas DEN2 (NS1)/DEN4 was not. The mutations present in the neurovirulent DEN2 PreM-E genes were then substituted singly or in combination into the sequence of the nonneurovirulent, parental DEN2. Intracerebral titration of the various mutant chimeras so produced identified two amino acid changes, namely, Glu₇₁-Asp and Glu₁₂₆-Lys, in DEN2 E as being responsible for mouse neurovirulence. The conservative amino acid change of Glu₇₁-Asp probably had a minor effect, if any. The Glu₁₂₆-Lys substitution in DEN2 E, representing a change from a negatively charged amino acid to a positively charged amino acid, most likely plays an important role in conferring mouse neurovirulence.     

Guanidine-selected mutants of poliovirus: mapping of point mutations to polypeptide 2C.

Sequence analysis of the genomic RNA of interstrain guanidine-resistant and antibody-resistant variant recombinants of poliovirus type 1 mapped the resistance of mutants capable of growth in 2.0 mM guanidine hydrochloride to a region located 3' of nucleotide 4444. This region of the viral genome specifies the nonstructural protein 2C. The sequence of genomic RNA encoding 2C from six independently isolated mutants resistant to 2.0 mM guanidine was determined. All six isolates contained a mutation in 2C at the same position in all cases, resulting in two types of amino acid changes. Dependent mutants were examined and found to contain two amino acid changes each within 2C. Mutants resistant to 0.53 mM guanidine were isolated and found to lack the mutations seen in variants resistant to 2.0 mM guanidine. A comparison of the amino acid sequences of the 2C proteins of poliovirus, foot-and-mouth disease virus, rhinovirus types 2 and 14, and encephalomyocarditis virus revealed a strong homology over regions totaling 115 residues. All of the mutations observed in guanidine-selected mutants were contained within this region. The amino acid region containing the mutations observed in poliovirus mutants resistant to 2.0 mM guanidine was compared with the homologous region in the other picornaviruses; a strong correlation was found between the amino acid present at this position and the sensitivity of the virus to 2.0 mM guanidine.     

Structural factors modulate the activity of antigenic poliovirus sequences expressed on hybrid hepatitis B surface antigen particles.

We have studied the functional expression of antigenic poliovirus fragments carried by various hybrid hepatitis B surface antigen (HBsAg) particles. Several constructions were made by using two different insertion sites in the HBsAg molecule (amino acid positions 50 and 113) and two different sequences, one derived from poliovirus type 1 (PV-1) and the other from PV-2. The inserted fragments each encompassed residues 93 to 103 of the capsid protein VP1, a segment which includes the linear part of the neutralization antigenic site 1 of the poliovirus. The antigenicity and immunogenicity of the hybrid particles were evaluated and compared in terms of poliovirus neutralization. A high level of antigenic and immunogenic activity of the PV-1 fragment was obtained by insertion at position 113 but not at position 50 of HBsAg. However, a cooperative effect was observed when two PV-1 fragments were inserted at both positions of the same HBsAg molecule. Antibodies elicited by the PV-2 fragment inserted at amino acid position 113 did not bind or neutralize the corresponding poliovirus strain. They did, however, bind a chimeric poliovirus in which the homologous antigenic fragment of PV-1 had been replaced by that of PV-2. The only virions that were neutralized by these antibodies were certain mutants carrying amino acid substitutions within the PV-2 fragment. These results show that position, constraints from the carrier protein, and nature of the inserted sequences are critically important in favoring or limiting the expression of antigenic fragments as viral neutralization immunogens.     

Correlation of virus polypeptide structure with attenuation of poliovirus type 1.

Several avirulent samples of poliovirus type 1 derived in the process attenuating the neurovirulent Mahoney strain show an altered virus capsid polypeptide, VP-1, on polyacrylamide gel electrophoresis of sodium dodecyl sulfate-disrupted virions.     

Characteristics of the genome of poliovirus type 3 isolated from patients with poliomyelitis in Moldavia

Twelve poliovirus isolates of serotype 3 from patients with paralytic poliomyelitis have been analyzed by oligonucleotide mapping of the viral genomes. All the studied strains were isolated from patients in different regions of the Moldavian SSR in 1982. The maps of all isolates are similar but they do not practically possess any large oligonucleotides characteristic of the vaccine strain of type 3 poliovirus. It is concluded that a wild neurovirulent strain of type 3 poliovirus, that circulated in 1982 in the Moldavian SSR was the cause of paralytic poliomyelitis cases. All the studied isolates are suggested to have been derived relatively recently from the common ancestor.     

Evolution of the Sabin type 1 poliovirus in humans: characterization of strains isolated from patients with vaccine-associated paralytic poliomyelitis.

Attenuated strains of the Sabin oral poliovirus vaccine replicate in the human gut and in rare cases cause vaccine-associated paralytic poliomyelitis (VAPP). Reversion of vaccine strains toward a pathogenic phenotype is probably one of the main causes of VAPP, a disease most frequently associated with type 3 and type 2 strains and more rarely with the type 1 (Sabin 1) strain. To identify the determinants and mechanisms of safety versus pathogenicity of the Sabin 1 strain, we characterized the genetic and phenotypic changes in six Sabin 1-derived viruses isolated from immunocompetent patients with VAPP. The genomes of these strains carried either few or numerous mutations from the original Sabin 1 genome. As assessed in transgenic mice carrying the human poliovirus receptor (PVR-Tg mice), all but one strain had lost the attenuated phenotype. Four strains presented only a moderate neurovirulent phenotype, probably due at least in part to reversions to the wild-type genotype, which were detected in the 5' noncoding region of the genome. The reversions found in most strains at nucleotide position 480, are known to be associated with an increase in neurovirulence. The construction and characterization of Sabin 1 mutants implicated a reversion at position 189, found in one strain, in the phenotypic change. The presence of 71 mutations in one neurovirulent strain suggests that a vaccine-derived strain can survive for a long time in humans. Surprisingly, none of the strains analyzed were as neurovirulent to PVR-Tg mice as was the wild-type parent of Sabin 1 (Mahoney) or a previously identified neurovirulent Sabin 1 mutant selected at a high temperature in cultured cells. Thus, in the human gut, the Sabin 1 strain does not necessarily evolve toward the genetic characteristics and high neuropathogenicity of its wild-type parent.     

Nucleotide sequence of a neurovirulent variant of the type 2 oral poliovirus vaccine.

Infectious cDNAs of the Sabin type 2 poliovirus vaccine virus and a vaccine-derived neurovirulent type 2 strain (P2/117) have been cloned in Escherichia coli. Nucleotide sequence analysis revealed that P2/117 differs from the vaccine strain by just 23 point mutations. Three occur in the 5' noncoding region. The remainder result in a total of 5 coding changes located in VP1, VP4, 2B, and 3D. The likely role of these mutations in the evolution to neurovirulence is discussed.     

Neurovirulence of type 1 polioviruses isolated from sewage in Japan.

Sixteen type 1 poliovirus strains were isolated from a sewage disposal plant located downstream of the Oyabe River in Japan between October 1993 and September 1995. The isolates were intratypically differentiated as vaccine-derived strains. Neutralizing antigenicity analysis with monoclonal antibodies and estimation of neurovirulence by mutant analysis by PCR and restriction enzyme cleavage (MAPREC) were performed for 13 type 1 strains of these isolates. The isolates were classified into three groups. Group I (five strains) had a variant type of antigenicity and neurovirulent phenotype. Group II (four strains) had the vaccine type of antigenicity and neurovirulent phenotype. Group III (four strains) had the vaccine type of antigenicity and an attenuated phenotype. Furthermore, it was demonstrated that the virulent isolates were neutralized by human sera obtained after oral poliomyelitis vaccine (OPV) administration, and the sera of rats immunized with inactivated poliovirus vaccine. Although vaccination was effective against virulent polioviruses, virulent viruses will continue to exist in the environment as long as OPV is in use.     

A mouse model for poliovirus neurovirulence identifies mutations that attenuate the virus for humans.

A mutation in the genome of poliovirus type 3 that is known to reduce neurovirulence in humans similarly reduces neurovirulence in mice when incorporated into a mouse-adapted-human poliovirus recombinant. Viral recombinants with a uracil at nucleotide position 472 in the 5'-noncoding regions of their genomes are unable to replicate in the mouse brain. Viral recombinants with a cytosine at this position are neurovirulent in mice. Neurovirulence of poliovirus in mice may therefore prove to be a useful indicator of the genetic stability of new attenuating mutations created by site-directed mutagenesis.     

In vitro phenotypic markers of a poliovirus recombinant constructed from infectious cDNA clones of the neurovirulent Mahoney strain and the attenuated Sabin 1 strain.

Infectious cDNA corresponding to the entire genome of the attenuated Sabin strain of type 1 poliovirus has been inserted into EcoRI site of bacterial plasmid pBR325. Two consecutive PstI fragments (nucleotide positions 1814 to 3421) of the infectious cDNA of the Sabin 1 strain were replaced by the corresponding DNA fragments prepared from an infectious DNA clone of the genome of the virulent Mahoney strain of poliovirus type 1. The exchanged segment encodes capsid protein VP1 and part of capsid protein VP3, a region in which a large number of amino acid differences between the attenuated Sabin and the parental, neurovirulent Mahoney strain cluster. The recombinant virus was obtained by DNA transfection of HeLa S3 cells, and several in vitro phenotypes of the virus were compared with those of the parental viruses. The recombinant virus was recognized by a neutralizing monoclonal antibody specific to the Mahoney strain. Growth of the Sabin strain of poliovirus has been shown to be quite dependent upon the bicarbonate concentration (d marker). The growth of the recombinant virus, however, was not highly dependent upon the concentration of bicarbonate in cell culture media, and thus resembled that of the Mahoney strain. On the other hand, the temperature-sensitive multiplication (rct marker) and the small-plaque morphology of the recombinant virus corresponded to the phenotype of the Sabin 1 strain. The in vitro recombination of infectious cDNA clones of genomic RNA and subsequent analysis of the growth properties of the recombinant virus have allowed us to correlate specific mutations in the genome of an RNA virus with certain biological characteristics of that virus.     

Sequence analysis of the downstream 5' nontranslated region of seven echoviruses with different neurovirulence phenotypes.

The downstream 5' nontranslated regions of seven echoviruses with different neurovirulent phenotypes were amplified and sequenced. Neurovirulent echovirus serotypes 4, 6, 9, 11, and 30 were identical to the putative poliovirus in 18S rRNA binding sequence and the flanking conserved sequences. Less neurovirulent echoviruses, serotypes 2 and 12, exhibited variations within these regions.     

Complete nucleotide sequences of all three poliovirus serotype genomes. Implication for genetic relationship, gene function and antigenic determinants

The complete nucleotide sequences of the genomes of the type 2 ( P712 , Ch, 2ab ) and type 3 (Leon 12a1b ) poliovirus vaccine strains were determined. Comparison of the sequences with the previously established genome sequence of type 1 (LS-c, 2ab ) poliovirus vaccine strain revealed that 71% of the nucleotides in the genome RNAs were common, that the 5' and 3' termini of the genomes were highly homologous, and that more than 80% of the nucleotide differences in the coding region occurred in the third letter position of in-phase codons, resulting in a low frequency of amino acid difference. These results strongly suggested that the serotypes of poliovirus derived from a common prototype. A comparison of the amino acid sequences predicted from the genome sequences showed highest variation in the capsid protein region, whereas non-structural proteins are highly conserved. Initiation of polyprotein synthesis occurs in all three strains more than 740 nucleotides downstream from the 5' end. An analysis of the non-coding region suggests that small peptides that could potentially originate from this region are conserved. The amino acid sequences immediately surrounding the cleavage signals, however, show a higher than average degree of variation. The analysis of the amino acid sequences of the capsid protein VP1 of all serotypes has led to the prediction of potential antigenic sites on the virion involved in neutralization.     

Restricted growth of attenuated poliovirus strains in cultured cells of a human neuroblastoma.

Cultured cells of a human neuroblastoma, SK-N-MC, were found to be highly resistant to Sabin attenuated poliovirus types 1 and 2 strains; no appreciable cytopathic effect was observed, and the total harvest was generally in the order of 1 PFU per cell or less. On the other hand, related neurovirulent strains of these antigenic types produced a relatively good (2 orders of magnitude higher) yield in a markedly protracted infectious cycle. The limited growth of the attenuated virus in the neuroblastoma cells appeared to be confined to a minor cell subpopulation. Experiments with intratypic (type 1) poliovirus recombinants suggested that the major genetic determinants limiting reproduction of the attenuated polioviruses in the neuroblastoma cells are located in the 5' half of the viral RNA, although the 3' half also appears to contribute somewhat to this phenotype. The possibility that neuroblastoma cells may represent an in vitro model for studying poliovirus neurovirulence is briefly discussed.     

Engineering a poliovirus type 2 antigenic site on a type 1 capsid results in a chimaeric virus which is neurovirulent for mice.

Poliovirus type 2 (PV-2) Lansing strain produces a fatal paralytic disease in mice after intracerebral injection, whereas poliovirus type 1 (PV-1) Mahoney strain causes disease only in primates. Atomic models derived from the three-dimensional crystal structure of the PV-1 Mahoney strain have been used to locate three antigenic sites on the surface of the virion. We report here the construction of type 1-type 2 chimaeric polioviruses in which antigenic site 1 from the PV-1 Mahoney strain was substituted by that of the PV-2 Lansing strain by nucleotide cassette exchange in a cloned PV-1 cDNA molecule. These chimaeras proved to have mosaic capsids with composite type 1 and type 2 antigenicity, and induced a neutralizing response against both PV-1 and PV-2 when injected into rabbits. Moreover, a six-amino-acid change in PV-1 antigenic site 1 was shown to be responsible for a remarkable host-range mutation in so far as one of the two type 1-type 2 chimaera was highly neurovirulent for mice.