The family Picornaviridae
contains two large and important genera of common human pathogens, Enterovirus
. In structural and genetic properties, these two genera are very much alike. However, rhinoviruses, which multiply mainly in the nasal epithelium, differ from enteroviruses, infectious agents of the alimentary tract, by their sensitivity to acid and by their growth at a lower optimal temperature. The genus Enterovirus
contains 64 serotypes pathogenic to humans, which have been distinguished by the neutralizing antibodies against them (17
). There may still be uncharacterized serotypes, as some clinical enterovirus isolates are not typeable by existing antisera and show genetic segregation indicative of an independent serotype (22
). Nucleotide analysis of the RNA genomes of different human enterovirus (HEV) serotypes has provided new insight into the classification of enteroviruses (23
), resulting in the division of these viruses into four main genetic clusters, designated HEV species A to D. Poliovirus serotypes 1 to 3 are genetically related to HEV-C but are classified as a species of their own (15
). The genus Rhinovirus
contains 102 serotypes, which are numbered from 1 to 100 (8
). Serotype 1 contains two subtypes, 1A and 1B. More recently, a strain referred to as the Hanks strain has been proposed to represent a new serotype (2
). We have generated partial capsid sequences of all human rhinovirus (HRV) prototypes, and with the exception of HRV87, all prototypes segregated into two previously established genetic clusters, HRV-A and HRV-B. HRV87 was found to cluster together with a representative of HEV-D, enterovirus 70 (EV70) (26
). Through further analysis, we found that HRV87 showed striking nucleotide identity with the partial sequence (obtained from GenBank) of the other member of HEV-D, EV68. This prompted further investigations on the relationship of HRV87 to the viruses of the HEV-D cluster. In this study, we examined (i) the nucleotide sequences of the 5′ untranslated regions (UTRs), two separate capsid regions, and the 3D RNA polymerase genes of HRV87 and two lines of EV68; (ii) the antigenic characteristics of both HRV87 and EV68; (iii) their acid sensitivities; and (iv) their receptor usage in HeLa cells.
In this work, we have demonstrated the close genetic similarity in four genomic regions between HRV87 and EV68, which is classified in the HEV-D species together with EV70 (15
). It has been shown recently that the nucleotide sequence encoding VP1 correlates well with enterovirus serotype (20
). Our VP1 gene sequences did not cover the entire VP1 region, but compared with the intraserotypic divergence limits of 25% for the nucleotide sequence and 12% for the amino acid sequence, the observed identities of 95.2 and 98.1%, respectively, are clearly indicative of the same serotype. EV70 is of a different serotype but is a member of the same genetic cluster. Antigenic characterization of the strains confirmed this conclusion. The one-way antigenic cross-reactivity between HRV87 and EV68 is typical of the prototype strain-prime strain relationship previously described for both enteroviruses (17
) and rhinoviruses (5
). Analysis of individual human sera suggested antigenic differences between strains now proposed to belong to a single serotype. However, this kind of variability of response is not new and has been described, for instance, for poliovirus type 3 (10
The clustering of HRV87 and EV68 with EV70 was evident for the coding regions as well for the 5′ UTR. The apparent divergence between the first two viruses and EV70 in the 5′ UTR was definitely greater than the corresponding interserotypic differences in the HEV-B species. The divergence of HRV87 from other HRVs with regard to the 5′ UTR has been described previously (1
The initial characterization of the EV68 Fermon strain showed that it was resistant to low pH, so it was classified as an enterovirus (27
). However, according to the data presented here, the titer of the acid-treated virus aliquot was somewhat lower than that of the untreated virus. We performed the acid sensitivity test by use of a standard method (4
) and also repeated the test by the protocol used by Schieble et al. (27
). The acid sensitivities of HRV87 and EV68 were obvious in both tests, but the differences between the titers of the acid-treated and untreated viruses were greater when the current standard method was used.
Originally, EV68 strain Fermon was recovered from a throat swab of a 10-month-old female with pneumonia. Despite a careful search for rhinovirus and enterovirus isolation data, no references were found in the literature to the isolation of either EV68 or HRV87 from clinical specimens after their first appearance and characterization. Serotyping of rhinovirus isolates is laborious and has been carried out in only a few studies (6
). However, the typing of enteroviruses has been widespread and has resulted in lengthy reports of clinical enterovirus findings. Even so, no isolation of EV68 has previously been reported (7
). In this study, neutralizing antibodies to these virus strains were found in all tested sera. Although the number of sera was small, these results indicate that infections caused by these viruses are not uncommon. The absence of clinical isolates of HRV87 or EV68 in the literature may be due to the acid-sensitive nature of these viruses and the infrequent serotyping of acid-sensitive viruses in routine virus diagnostics.
Approximately 90% of rhinovirus serotypes bind to intercellular adhesion molecule 1, while all but one of the remaining serotypes use the low-density lipoprotein receptor (9
). The attachment of HRV87 to the cell surface is known to require sialic acid (29
), which has also been reported for EV70 (30
). Our present results suggest that the receptor might be the DAF and might thus be shared by EV70 and several echovirus serotypes (13
In conclusion, we have shown in this study that HRV87 and EV68 are genetically and antigenically highly similar and that they also share homology with EV70. Therefore, we propose that this virus pair, HRV87 and EV68, should be reclassified as a single HEV-D serotype. Our results also demonstrate a conflict between the classical and modern phylogenetic classification criteria. HRV87/EV68 and EV70 definitely belong to the same genetic cluster and appear to share receptor specificity in HeLa cells. However, according to acid sensitivity assays, HRV87 and EV68 should be classified as rhinoviruses whereas EV70 is a typical enterovirus. The dividing line between these two genera is even more blurred than previously thought.
This work was partly supported by Wyeth-Lederle Vaccines and by grants from the Academy of Finland and the Päivikki and Sakari Sohlberg Foundation, Helsinki, Finland.
We thank Douglas Lublin (Washington University School of Medicine, St. Louis, Mo.) for the DAF SCR-3-specific murine monoclonal antibody 1H4 used in the study. The kind advice of Mirja Stenvik and expert technical assistance by Kristiina Aitkoski, Mervi Eskelinen, and Päivi Hirttiö are gratefully acknowledged.