key: cord-0720462-f5zicj80
authors: Arden, Katherine E.; Mackay, Ian M.
title: Newly identified human rhinoviruses: molecular methods heat up the cold viruses
date: 2010-02-01
journal: Rev Med Virol
DOI: 10.1002/rmv.644
sha: a03f338dc963c74bd7fe4b8ac0f4929ad1957c8e
doc_id: 720462
cord_uid: f5zicj80

Human rhinovirus (HRV) infections cause at least 70% of virus‐related wheezing exacerbations and cold and flu‐like illnesses. They are associated with otitis media, sinusitis and pneumonia. Annually, the economic impact of HRV infections costs billions in healthcare and lost productivity. Since 1987, 100 officially recognised HRV serotypes reside in two genetically distinct species; HRV A and HRV B, within the genus Enterovirus, family Picornaviridae. Sequencing of their ∼7kb genomes was finalised in 2009. Since 1999, many globally circulating, molecularly‐defined ‘strains’, perhaps equivalent to novel serotypes, have been discovered but remain uncharacterised. Many of these currently unculturable strains have been assigned to a proposed new species, HRV C although confusion exists over the membership of the species. There has not been sufficient sampling to ensure the identification of all strains and no consensus criteria exist to define whether clinical HRV detections are best described as a distinct strain or a closely related variant of a previously identified strain (or serotype). We cannot yet robustly identify patterns in the circulation of newly identified HRVs (niHRVs) or the full range of associated illnesses and more data are required. Many questions arise from this new found diversity: what drives the development of so many distinct viruses compared to other species of RNA viruses? What role does recombination play in generating this diversity? Are there species‐ or strain‐specific circulation patterns and clinical outcomes? Are divergent strains sensitive to existing capsid‐binding antivirals? This update reviews the findings that trigger these and other questions arising during the current cycle of intense rhinovirus discovery. Copyright © 2010 John Wiley & Sons, Ltd.

We use the term 'strain' to indicate the molecular equivalent of a classical human rhinovirus (HRV) serotype or distinct newly identified HRV (niHRV) detected from a patient specimen. A 'variant' is a clinical detection of the same strain from a different individual, usually sharing 95-100% nucleotide or amino acid identity, depending on the analysed region. To simplify strain identification, we add the species identifier to the strain names e.g. HRVB-14 or HRVC-QPM.

When relevant, we refer to nucleotide-based strain typing targets using their encoding region name and viral protein (VP) name is used for the encoded product only. Thus, the 1A/1B region is the RT-PCR target which encodes VP4 and VP2.

An HRV C strain is defined in this review as a member of the proposed species consisting of HRV strains that cluster most closely with the first polyprotein sequence described, HRVC-QPM [1] . These include HRVC-NAT001, HRV-HRVC-NAT045, HRVC-C024, HRVC-C025, HRVC-C026, HRVC-NY074, HRVC-QCE, HRVC-N4 and HRV-N10. If the 5 0 NTR is used for strain typing instead of the capsid sequence, the 'HRV C' membership

In the 1950s and 1960s a flurry of research into colds and flu-like illnesses (CFLIs; [2] ) and identification of 'common cold viruses' resulted in the first reports [3] of what became known as rhinoviruses [4, 5] . By 1987, 100 serotypes had been amassed during three stages of submission and neutralising antibody studies [5] . HRVs are currently organised into two International Committee on Taxonomy of Viruses (ICTV)-approved species, Human rhinovirus A and Human rhinovirus B with a third, Human rhinovirus C, receiving initial approval by the Executive Committee and awaiting ratification by the ICTV [6] (Table 1) .

HRVs infect the upper and lower respiratory tract (URT and LRT) [7, 8] (Figure 1 ) and they cause more asthma [9] [10] [11] [12] and chronic obstructive pulmonary disease (COPD) [13] exacerbations than any other factor identified to date, in addition to the majority of CFLIs [14, 15] . The knowledge of viral involvement in asthma and wheeze has been percolating through the literature since at least 1957 [16, 17] , but the important role of HRVs in wheeze was made much more obvious once PCR became available [18, 19] . Up to a quarter of children worldwide experience asthma symptoms, with prevalence plateauing in some countries while rising in other parts of the world [20] . In adults, COPD exacerbations are predicted to soon become the world's third leading cause of death [21] and HRVs play a central role in illness among adult transplant recipients [22] . HRVs have been the most common reason for prescribing antibiotics [23] and are associated with pneumonia [24] , otitis media [25] and sinusitis [26] . It is interesting to note that, in contrast to other well-known respiratory viruses, the clinical symptoms of HRV infection are primarily caused by the host's immune response to infection rather than by viral cytopathicity [27] [28] [29] [30] . Intriguingly, whether in the midst of an influenza pandemic or during an outbreak of an emergent virus such as the SARS-CoV, HRVs are the most common virus detected in patients meeting appropriate clinical criteria for presentation to hospitals or clinics [31, 32] . The HRVs, therefore, create an enormous direct and indirect socio-economic burden across the developed and developing world [33, 34] .

For decades HRV testing data have correlated poorly with clinical outcomes or yielded untypeable strains, undermining HRV epidemiology [35, 36] and reinforcing the 'common cold virus' M and H indicate early cell tropism-based classification (monkey, human) abandoned in favour of a sequential numbering system [145] . HRV strains were later divided into the major and minor groups defined by receptor tropism [146, 147] . Receptor-designated minor group HRV strains are underlined, major group are shown in bold. Antiviral groups (A and B) are labelled [148, 149] . HRV-87 is not included as it has been previously defined as a variant of HEV-68 [150] . HRV-Hanks (not listed) and HRV-21, HRV-8 and HRV-95 are most likely the same serotype [119] .

The nature and impact of newly identied rhinoviruses The nature and impact of newly identied rhinoviruses label that resulted in the underestimation of HRV impact. With the advent of diagnostic PCR, it became apparent that HRV infections are the most frequent of viral respiratory infections, even in hospitalised young children [23, [37] [38] [39] . While the first classical HRV strain sequences were lodged onto GenBank in 1994 [40] , it was not until 1999, using specimens collected in 1992-1995 [41] , that the first hint of new HRV diversity appeared ( < 98% identity with classical strains in the 5 0 nontranslated region (NTR)). In 2006, using phylogeny of 1A/1B sequences from HRV positive specimens collected in 2003, we announced a distinct and previously undefined clade of HRVs [42] which was quickly followed by a report of related HRV strains among some cases of a 2004/2005 CFLI outbreak in New York [43] . Global identification of clade members ensued [15, 36, 37, [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] and it now seems that the much overlooked HRVs have been brought in from the cold. In our opinion, the current HRV discovery phase is ongoing, not, as was once suggested, largely complete [61] ; there will be descriptions of many more novel strains in the years to come. The niHRV strains do not appear to be emerging viruses in that they have likely not arisen from a recent zoonotic event [62] but have instead been circulating for at least a decade [49, 52] and probably much longer [51] , without previous detection. Historically, HRV culture was fraught with unreliability and continues to fail in all attempts to isolate niHRVs [1, 57] . Even when using a strain-typing PCR, HRV Cs have been missed if the HRV-positive patient specimens had first Cell lines used to attempt HRV C isolations are listed. *Recurrent attacks of shortness of breath and wheezing caused by spasmodic contraction of the bronchi, attributed to infection. Adapted from Reference [165] been passaged in culture [63] . By overcoming the insensitivity of culture, PCR with subsequent sequencing produced a large number of diverse viral discoveries from within the human respiratory tract [1, [64] [65] [66] [67] [68] [69] . Unfortunately, the inclusion of HRV screening has not yet become as accepted as it has become for other respiratory viruses including human respiratory syncytial virus (HRSV), influenza viruses (IFVs), parainfluenza viruses (HPIVs) and metapneumovirus (HMPV) [70] [71] [72] .

HRVs were one of the clinical virology laboratory's earliest targets for conventional PCR [73] . We have previously reviewed in detail the failings of culture-based HRV detection methods and listed many PCR-based methods. A review of the many recent papers reporting niHRVs reveals some general trends. HRV strains, both known and new, form a significant, although variable, proportion of respiratory virus detections ( Table  2) . It is not especially difficult to assign detections to one of the three species using sequence obtained from a variety of coding regions. Strain typing, however, is more challenging, as we discuss below. A combination of HRV As and Cs have the highest prevalence within the HRVs, although in different studies either the As or Cs may be much more frequently detected [37, 47, 50, 51] . HRV Bs are consistently rare, constituting around 7% of HRV detections overall [50, 54] , thus smaller studies may fail to detect any [47, 48, 56] . Many different HRV strains may be detected within single studies, covering short periods of time [15, 45] , emphasizing the great diversity of this group.

Since 1988 [74] the transcriptionally important [75, 76] 5 0 NTR ( Figure 2 ) has been the pre-eminent target for contemporary screening and is popular for strain typing (Table 3 ). Many subtly different assays target this area. It is relatively easy to amplify all HRV and human enterovirus (HEV) strains due to multiple, small, broadly conserved regions within its sequence [41, 74, 77] . It is more sensitive than the longer and more variable 1A/ 1B region [46, 54, 78] which is increasingly favoured for strain typing.

Today, several well-considered rtPCR or PCRderived methods are popular for detecting HRVs in clinical specimens [51, 57, 79, 80] . In particular, Lu et al. described a meticulously validated novel rtPCR assay which is capable of detecting HRV Cs (recently enhanced by a modified forward primer CPXGCCZGCGTGGY [Lu et al., personal communication] ) in addition to the classical strains [79] . The assay detected 5 Â 10 1 synthetic RNA copies per reaction and was 5 10 5 -fold less sensitive to HEVs [79] , although HRV/HEV dual infections were identified [79] . A more recent rtPCR designed specifically to include the HRV Cs detected 10 synthetic RNA copies per reaction [80] . At present, viral loads of respiratory viruses in specimens, including HRV Cs, typically range between 10 4 -10 7 copies/ml [46, [81] [82] [83] [84] [85] . Other assays do not comprehensively detect niHRVs, possibly underestimating positive patients by more than two-fold [80] . Our experience with a conventional assay [86] uncovered very specific (equivalent amplicon size as from an HRV target) cross-amplification of human genomic DNA (accession number AL139807) which may result in over-estimation of HRV positives if strain typing is not used.

The niHRVs include an as-yet unquantified number of strains that seem to have arisen from, or contributed to, genetic recombination [45] . Recombination can thwart phylogeny based on the 5 0 NTR by producing tree structures whose branching patterns cannot be compared with those from other regions because the evolution of the 5 0 NTR may not always be congruent with that of the capsid coding regions [87] (Figure 3 ). Nucleotide sequences are particularly useful for recombination and evolution studies. However, because the virus immunogenically 'shows' the protein to its host, the translated capsid region may be a more clinically relevant sequence to examine for strain typing, antiviral design and for monitoring treatment efficacy, something not possible with the 5 0 NTR. Among the capsid regions, VP4 shows the least variation in protein sequence among strains. A clinical detection may differ by 5 15% from the nearest GenBank nucleotide match in 1A/1B but still be 5 98% identical in amino acid sequence to a previously identified strain. This is likely due to the sequestration of VP4 and the amino terminus of VP2 within the capsid of HRVs, which affords

The nature and impact of newly identied rhinoviruses The nature and impact of newly identied rhinoviruses 159 159 (38) some protection from immune pressure. Thus, examining only nucleotide sequences may overestimate the number of novel strains present but sole reliance on VP4/VP2 may underestimate strain diversity [63] . The 3 0 portion of VP2, VP3 or VP1 sequences may provide better strain typing targets, although they have not been thoroughly evaluated. While sequencing other subgenomic regions of HRVs improves phylogenetic power [45, 63] , the deduction of complete polyprotein sequences yields the ultimate phylogenetic information [88] .

In 2007, Lee et al. nominated a distinct clade of HRV strains be named HRV C based on pairwise nucleotide identity thresholds derived only from the classical strains' 5 0 NTR region [57] (Madison clade; Figure 3 ). In the same month a separate proposal of a novel HRV C species was made by Lau et al. for the HRV-QPM-like strains. The latter was based on complete coding sequence analyses [48] which we subsequently confirmed with additional analyses and predictive modelling [89] . The Madison clade was subsequently noted by Kiang (23) 1 Not all clinical HRV detections were/could be genotyped; 2 Specimens had previously tested negative by culture and direct immunofluorescence for IFV A and B, HRSV, HadVs; 3 83 specimens had been previously found positive for HboV, 120 were negative [151] ; 4 Culture or PCR negative for other viruses including, but not limited to, HRSV and IFV [58, 152] ; 5 Small, mixed populations with a range of diagnostic investigations, screened for HRVs and collected from Africa, Asia, Australia, Europe and North America; 6 Not known as HRV C at the time but we believe this is currently the best label for these strains; 7 Specimens combined from categories including clinic-derived, hospitalbased and asymptomatic; 8 We have used the authors numbers in this table however some of their 'A' strains align best with the International Committee on Taxonomy of Viruses proposed HRV C strains; 9 These strains were originally described as a sublineage of HRV A called HRV-A2 but have subsequently been renamed to HRV C [ [80] nPCR-nested PCR; snPCR -semi-nested PCR; 1A and 1B -regions of the genome encoding the VP4 and VP2 polyprotein cleavage products, respectively; A -array-based detection of randomly amplified RNA. [45] and strains from the Wisconsin clade [57] . HRVC-QCE, HRVC-QPM, HRVC-NAT001 and HRVC-NAT045 were excluded from the alignments because of incomplete 5 0 NTR sequences. This figure is available in colour online at www.interscience.wiley.com/journal/rmv et al. [90] , called HRV C' by Tapparel et al [80, 91] and Cc by Huang et al [45] all using the 5 0 NTR. However, examination of the 1A/1B region of a HRVC-N10 [45] , the first complete polyprotein sequence of this clade, shows limited distinctiveness. The polyprotein of HRVC-N10 does not meet the ICTV sequence identity criteria within P1 and 2C+3CD ( 5 70% amino acid identity) exactly, however its falls within only a few percent of doing so for the average identities of the HRV C species. The significant point here is that reliable phylogeny depends on an appropriate choice of sequences. Use of too short a region, an inappropriate region (non-coding, too variable, too conserved) or insufficient reference sequences will stymie the resultant tree, no matter what program is used or how many bootstraps are performed.

An example of the problems experienced when using 5 0 NTR phylogenetic trees that contain clinical strains for typing is the co-localisation of some HRV A sequences (usually HRV A-51, -65, -71 -12, -45 and -78) with HRVC-QPM-like sequences. Using this region makes rendering a clear definition of the two species almost impossible. For simplicity's sake, we call this newly defined clade HRV AC; the pattern is similar to the merging of HEV species when typing using the 5 0 NTR. Kiang et al. [90] followed by Piralla et al. [46] have divided the HRV A group into nucleotide-based clades HRV A2, GAC-1 and GAC-2. Simmonds et al. defined additional clades too, this time using the 1A/1B region [60] . In doing so, they developed the first coding-region based objective threshold for defining inter-( > 10% pairwise nucleotide difference) and intraspecies ( < 10% pairwise nucleotide distance) variation which is key to aiding the identification of new clinical detections as known or novel strains [60] . The clinical value of these genomically defined clades and cut-offs remains unknown. These clades are not apparent when the coding region is used for typing. We support the use of nucleotide-based coding region sequences for improved strain typing results [87] .

It was once expected that HRVs were involved in frequent recombination events due to their similarity to the HEVs, some of which are well known for participation in genomic recombinations [63, 92] . When Huang et al. compared clinical HRV strain typing results using the 5 0 NTR with those from 1A/1B they noted two branching patterns which were the HRVC-QPM-like clade (which they called HRV Ca) and the Madisonlike clade (HRV Cc) [45] . Further in silico investigation predicted recombination [45] . While these data suggest that the 5 0 NTR shows HEV-like potential as a recombination 'hotspot' [45, 88, 91, 93] , data for frequent recombination elsewhere in the genome are rare [91] . In addition, empirical demonstration of recombination among the HRVs is lacking. Data describing recombination in HRVs are derived only from in silico predictions of the likelihood that shared regions of sequence similarity between strains, identified through phylogeny, originate from a mixing between parental strains. Recombination seems to involve the classical strains [88] but our ability to detect potential recombination has been enhanced by the discovery of niHRVs [45] . At present we lack sufficient numbers of full HRV C genome sequences to perform reliable recombination analysis.

Previously co-detection was seen as a sign of weakness indicating that the respiratory virus, often an HRV, could not cause significant illness [94, 95] . Some have noted more severe clinical outcomes among multiply positive patients [96] [97] [98] ; others have not [99, 100] . Strong correlation between detection of a specific virus and the cooccurrence of certain clinical signs or symptoms are often made when testing for the known respiratory viruses is incomplete [15, 50, 101, 102] . The value of clinical data from patients lacking comprehensive testing is questionable although the caveat must be added that any association risks challenge while novel endemic respiratory viruses remain unidentified. Co-detections are influenced by the quality of the assays used to obtain the data. Previously the HRVs were often the forgotten diagnostic target, now, ironically they have become the focus of investigations in which other viruses are sometimes forgotten [36, 50, 55, 56, 103] .

In HRV C studies so far, no clear clinical difference has been noted between patients with different numbers of pathogens detected in single specimens [48, 49, 56] . Of particular interest is that despite differences in study years and assays used, most HRV co-detections are with HRSV [15, 44, 45, 49, 52, 59] . We recently reported that in a

The nature and impact of newly identied rhinoviruses The nature and impact of newly identied rhinoviruses 163 163 study of 1247 specimens, including 660 positive for at least one virus, HRVs were statistically the least likely virus of 17 examined to be associated with co-detections [104] . These co-detection data further prompt a rethink of the role of HRVs in driving respiratory illness rather than being passengers. Several studies describing HRV Cs have identified multiple HRV strains from different HRV species in the same specimen [45, 50, 57] although not in a study of acute otitis media [36] . HRV/HRV co-detection suggests that strain typing studies need an intermediate amplicon cloning and analysis method to accurately represent every HRV strain present. It is interesting to note that at this point, only HRVs from different species have been co-detected. Further work is required to determine whether we are missing same-species co-detections because the sequences are too similar, or if there is some immune-mediated mechanism preventing same-species co-detections.

Why do we need to differentiate strains?

The amino acid similarity at key genomic regions (P1, 2C+3CD; Figure 2 ) aids assignation of an HRV strain to a species within the genus Enterovirus, but no equivalent molecular criteria exist to define an HRV detection as a known or novel HRV strain [89] . The viral genome is directly or indirectly responsible for a virus' antigenic and immunogenic potential, seasonal and epidemic circulation patterns, response to antiviral and vaccine interventions and clinical effects; factors which contribute to the impact of the virus. Genome deduction and characterisation and the development of strain-defining criteria are the first requirements for robust analysis and definition of respiratory viruses without which we cannot reduce the burden of infection. Strain-specific molecular epidemiology studies are rare for the HRV super-group [1] despite these being commonplace for the less populated respiratory virus species such as HMPV and HRSV. When Kistler et al. described the HRVs as being under 'purifying' selective pressure [105] (environmental forces that encourage each strain to remain intact, once formed) they described a very important feature of some, if not all members [106] , of the heavily populated HRV group; their unexpected ability to remain largely unchanged over time. This can be seen during molecular epidemiology studies as high levels of conservation ( > 96% nucleotide identity) among variants of the same strain detected over the same time period. We saw this when studying the complete 1A, 1B and 1D regions of 17 variants of HRV-QPM detected throughout 2003 in Queensland, Australia in which we defined a 96% nucleotide identity threshold to identify HRVC-QPM variants. It should be noted that we used highly specific rtPCR assays which may have limited the diversity we saw within this strain. Conservation can also be seen among temporally disparate classical strains that we described in 2006 from the same specimen population; the close phylogenetic relationship between, for example, QPID03-007 (from 2003) and HRVA-80 (from 1967, [107] ), which share 92% pairwise nucleotide and 100% amino acid identity in 1A (VP4) and QPID03-0027 and HRVB-48 (also from 1967, [107] ) shared 88% and 100% amino acid identity. It is worth noting that despite low nucleotide identity, something found by others [49] , they all shared 100% amino acid identity.

A useful way to define what makes a strain different is to look at it from the host's perspectivewhat degree of sequence difference allows penetration of the host defenses by subsequent HRV challenges? In 1977 five co-infections were noted using culture; none contained the same HRV species in the pairing [108] . We can now show that each strain pairing shared < 51% amino acid identity across the polyprotein sequence. In a more recent report two cases of recurrent CFLI with wheeze were reported, each yielding two HRV C strains, demonstrating that the same patient could be infected by genetically distinct strains (sharing 85% amino acid identity in VP4) within a period of 7 months [103] . In another report, two different (sharing 72% amino acid identify in VP4) HRV C strains were identified from recurrent human bocavirus (HBoV)-positive specimens collected two weeks apart [48] . These data may provide a practical threshold which are useful in molecularly defining the antigenic distinctiveness of strains. Future studies identifying and expanding upon the characteristics of consecutive HRV infections could significantly define this area of clinical rhinovirology and could contribute to understanding a role for nucleotide-based thresholds.

The HRV Cs are quite distinct from the As and Bs and they exhibit considerable intra-species diversity. At the time of writing there were 10 HRV C complete polyprotein coding sequences available for public analysis, not all with a complete 5 0 NTR. HRVC-QCE and HRV-NY074 have the shortest polyproteins (2140aa), HRVC-C025 the longest (2152aa). Ten protease cleavage sites, common to all enteroviruses, are predicted to cleave each HRV C polyprotein into typical picornavirus structural (VP1-4) and non-structural proteins (2A-C and 3A-D, [109] [110] [111] ). The VP1/2A junction of HRVC-N4 and NY074 VP1/2A junction consists of the more HRV A-like V/G scissile bond whereas the same HRVC-N10 junction consists of an L/D cleavage site, unique among the HRVs. HRVC-N10 also has an unusual Asp insertion at position 2 in VP1. A translation initiation site (MGAQVS) shared with all other HRVs and motifs (YGDD, YGL, TFLKR and SIRWT) including those crucial for RNA polymerase binding shared by HRV-As, HRV-Bs and HEVs [112, 113] are present within most HRV C strains; HRVC-N4 and -N10 have SFLKR. HRV Cs differed in the amino acid sequence of two of three exposed VP1 motifs previously found to be conserved among HRV polyproteins [114] . Phylogeny inferred from a discontinuous alignment of 10 amino acids in VP3 and VP1, which comprise the predicted intercellular adhesion molecule (ICAM)-1 footprint [115] on major group HRV strains (Figure 4) , placed the HRV Cs on a branch with HEV strains and, unexpectedly, HRV-1, but distinct from the known major (employing the ICAM-1 molecule as receptor) or minor (employing the very low density lipoprotein receptor, VLDLR) group HRV strains. Key residues on the VP1 BC loop involved in receptor contact with the VLDLR [116] are missing in strains from HRV C and there are differences in the HI loop compared to the classical strains. Antigenic site A [117] is absent in HRV C strains due to deletions within VP1 while the sequence of Site B is uniquely variable among the HRV C species. As with other HRV C strains identified to date [118] , a conserved loop sequence motif constituting a cis-acting replication element (cre), R 1 NNNA 1 A 2 R 2 NNNNNNR 3 was identified in the 1B region of HRV C strains described by Piralla et al. in the context of an appropriate predicted RNA stemloop structure. The secondary structure of the cre is essential for viral genome replication and its location differentiates all the species within the genus Enterovirus [118] .

To date, no HRV C detection has been isolated in culture. Traditional methods have been attempted using cell types shown in Figure 1 [1, 48, 49, 55] and even allowing for the variability of HRV isolation, it appears that different cell types, organ tissue or reverse engineered viruses will be required for success. We sought a molecular surrogate to replace some of the data important for taxonomic description and traditionally generated by culture [89] .

Without an isolate or infectious clone to test, it is only possible to predict the likely interaction with The nature and impact of newly identied rhinoviruses The nature and impact of newly identied rhinoviruses 165 165 capsid-binding anti-picornaviral drugs in silico. Their site of action has been mapped to a hydrophobic pocket beneath a canyon in the viral capsid [109] which encircles the union of five monomeric capsid units. Among the 25 key discontinuous contact residues of the 'footprint' of pleconaril (an anti-picornaviral drug), there are 12 HRV C-specific differences (INFQFLXIIMFP-SVFITYMINMLAF) compared to the HRV A (INLQFSILYMYASVFLLYMTNMLHH) or HRV B (INLSFSYIAMYPSVFVVYCINMIHG) consensus sequences [119] . There are 11 differences between HRV A and B strains. Changes to two sites important for identifying naturally resistant strains were identified as Tyr 152 (HRV B residue numbering from Reference [119] ) to Phe 152 in some HRV C strains and Val 191 to Thr 191 in all strains. The His 245 position conserved in most other classical HRV strains (excluding HRV-8, -45 and -95 which contain Leu) was highly variable among the HRV Cs (Gly, Ala, Val or Ile instead). Other highly conserved HRV A and HRV B residues including Ile/ Leu 106 (Phe in HRV C) and Tyr/Phe/Ala 150 (Ile or Val in HRV C), differed in the HRV C sequences. The empirical impact on HRV C sensitivity to pleconaril remains unknown.

The HRV Cs are found in patients with the same clinically indistinct, broad range of clinical outcomes as the HRV As and Bs and other respiratory viruses ( Table 4) . Symptoms of infection with HRVCs include CFLIs [1, 48, 102] , pharyngitis [50] , croup-like cough [89] , wheeze [15, 48, 103] , acute otitis media [36] , febrile convulsion [89] , bronchiolitis [1, 15, 102] and pneumonia [15, 48, 120] among otherwise healthy children and adults as well as those with underlying conditions including asthma [1, 53, 55, 103] , immunocompromise [47, 120] , cystic fibrosis [89] or multiple sclerosis [121] .

In our strain-specific study we identified that most sole detections of HRV-QPM were in patients with signs of mainly LRT illness related to wheeze [89] . Jin et al. reported that children with HRV C sole detections did not have wheezing, but were diagnosed with bronchopneumonia [44] . These diagnoses were not dissimilar to those linked to other respiratory virus detections and in fact HRVs, including Cs, are found in similar numbers to HRSV in children hospitalised due to CFLI [37] .

Many recent studies do include data on the HRV As, Bs and Cs making comparisons between species possible. Species-specific illness is not always observed (2879} [37] but when found, the HRV Cs, more so than As or Bs, are seen as the major contributors to febrile wheeze in infants and toddlers, to asthma exacerbations in older children (who are able to be accurately diagnosed as asthmatic) and to illness in hospitalised children with asthma [48, 49, 53] . Not only did more children with an HRV C have a cough than those with an HRV A (who had more fever than the HRV C positive children) but HRV Cs were associated with more wheezing and supplemental oxygen than those positive for HRV A strains [49, 54] . First time and recurrent wheezers with HRV C detected all responded well to bronchodilators (with or without steroids) [15] . HRV C strains accounted for half of all detections from middle ear fluids of children with acute otitis media [36] .

The strain-specific severity data may be related to the higher average 'viral loads' (relative quantification based on comparative threshold cycles in the absence of normalisation) reported in HRV C positive specimens relative to positives with other HRV species [46] . Higher levels of RNA can equate with more severe illnesses, usually categorised with LRT rather than URT symptoms [46] although there may not be a definite correlation for asthmatics; the mere presence of HRV can induce illness that, while more severe in asthmatics, does not significantly differ either in virus load or duration of viral RNA detection [122] . HRV C-positive children have more pre-existing conditions and exhibit more LRT than URT illness than adults. HRV C strains are often detected in more serious clinical outcomes than HRV A or B [46] although hospitalisations may be fewer for HRV Cs than the other species [44] .

To date, most studies have sought HRV C strains in samples collected from the respiratory tract (mostly the nasopharnyx) although extrarespiratory detections have been reported in the blood and pericardium [120, [123] [124] [125] [126] . HRV C studies have mostly examined patients from asthma or hospital-based populations often with a focus on inpatients (Table 4) , which may characterise the HRV Cs as more severe pathogens than they really are. While differences between study popu- The nature and impact of newly identied rhinoviruses The nature and impact of newly identied rhinoviruses 167 167 Cough (82), runny nose (53), [44] in/outpatient, CFLI crackle (53), fever (37) . URTI (53) Asthma, bronchiolitis.

[ 164] retrospective (Singapore)

Children-includes neonates (0-1 month), infants (1-12 months), toddler (12-24 months) and children (2-14 years); CFLI-cold and flu-like illness; FEVforced expiratory volume; ILI-influenza-like illness (nonspecific ARTI with fever > 38 C, cough and/or pharyngitis [43] ); LRT-lower respiratory tract;

MEF-middle ear fluid; NA-not available; NL-nasal lavage; NPA-nasopharyngeal aspirate; NPS-nasopharyngeal swab; NPW-nasopharyngeal wash; NTSnose/throat swabs; SRI-symptomatic respiratory illness; URTI-upper respiratory tract illness; ARTI-acute respiratory tract infection; 1 visiting health care provider; 2 some hospitals admit patients, particularly children, from the majority of the surrounding populated area and so some of these are more accurately described as 'population-based' studies; 3 strain-specific study; 4 unless specific, the period was inclusive of the months described in the reference; 5 IFV and HRSV PCR-negative [152] ; 6 Patient numbers not specimen numbers; 7 only examines HRV-positive specimens.

lations can be a problem, it is also important to investigate a wide range of illness groups, as they give an indication of the scope of clinical outcomes possible. Well considered, large community-based studies have not yet included HRV C investigations but are important to present the full clinical spectrum of clinical outcomes. Much longer study timeframes will also be required if strain diversity and circulation patterns are to be measured usefully and if HRV species-specific clinical outcomes are to be robustly investigated.

While the HRV Cs, like many other respiratory viruses, are found among some patients with none of the study-defined symptoms at the time of sampling, they are more often detected among sick children than asymptomatic controls [37] . Viruses detected in asymptomatic people can be perceived as playing a minor role in illness. However, for the definition of a symptom to be most useful, it should include any departure from normal function or feeling, not just the presence of one or more signs such as cough, fever above 38 C, rhinorrhoea or vomiting. Mild headaches or mood changes may not be as noticeable or describable but nonetheless probably reflect a minor, immune-managed symptom that could be correlated to PCR-related HRV detections. Negative mood and reduced alertness and reaction times have previously been identified in those suffering HRV infections [127] but do not register among typical clinical criteria.

Reports of unusually long periods of HRV positivity ( > 2-3 weeks [128, 129] ) have increased in frequency since more sensitive PCR methods replaced cell culture for HRV detection. Identification of the same HRV serotype over a four-week period has been reported [130] . HRV RNA has been detected days before symptoms commence through to five or more weeks after they cease [129, 131, 132] although commonly strain typing is not done. Epidemiology that incorporates strain typing usually does not find chronic shedding [133] ; strain typing indicates that HRV shedding normally ceases within 11-21 days [79, 134] . Thus, the perception of persistence is more than likely the result of serial or overlapping infections by multiple untyped strains [130, 135, 136] and caution is required when describing persistent HRV infections [137] in the absence of strain typing investigations if the implication is that persistence is a feature of individual HRV strains. Few data are available [37] to address persistence since preand post-sampling clinical data are rarely described [138, 139] and the definition of 'well' subjects prior to or at the time of sampling or inoculation is sometimes not clear especially for young children who cannot reliably report symptoms [129, 133, 140] . It is likely that any plan to link long-term HRV positivity in otherwise healthy individuals will require strain typing, regular and frequent sampling and, as with the previous section, identification of a truly asymptomatic, rather than adequately asymptomatic, state.

To date, true persistence, seen as ongoing detection of a single confirmed HRV strain, has been limited to individuals with underlying immunosuppression or immune dysfunction [141] . Piralla et al. reported that HRV C strains were detected more than three-times longer in immunocompromised young patients than in immunocompetent children, with a mean of 16 versus 53 days [46] . Multiple detection of the same strains (100% identical 1A sequence in each patient over time) extended to 4 months in haematopoietic stem cell transplant recipients.

During the 1960s, large epidemiology studies of HRVs in temperate climes identified the major epidemic peak during autumn with a smaller spring peak [61, [142] [143] [144] . It can be assumed that the HRV C strains did not contribute to these findings which employed culture methods. HRV Cs and other HRV species have been identified across all months in tropical, subtropical and semi-arid regions [15, 45, 47] . In our initial study (subtropical [42] ), HRV Cs were mainly detected in spring and in our single strain study we noted HRV-QPM variants had a bi-modal peak in which detection frequencies were 5.0% of all virus positives in August (late winter) and 4.2% in February (summer) [1] . A bimodal trend was also apparent when HRV Cs were studied in other subtropical and temperate areas [37, 48, 49] while single HRV C epidemic peaks were described in two studies in temperate climates [58, 59] . No strong conclusions can be drawn from these data, beyond the need to obtain more data, since sample size and length of sampling differed between studies. It is clear that many HRV C strains circulate in a single year and while some studies indicate that the same strains can be detected in neighbouring years [63] , we did not find this to be the case for the HRV-QPM strain in Brisbane [1] .

With properly deployed PCR-based testing, it is no longer possible to overlook the central role HRVs play in human morbidity. The finding of the niHRVs has reinvigorated HRV research but raised many questions that need data on a scale that does justice to the size of the HRV supergroup.

Future respiratory virus research must include more powerful studies to confirm the trend towards HRV C dominance in serious HRV illness and will face a daunting array of questions needing attention. What makes an HRV distinct to the host and how do nucleotide clades inform our understanding of HRV diversity? What will community-based studies teach us about the clinical impact of niHRV? What is needed to isolate the niHRVs in the laboratory and what cellular receptor(s) do they use? Can we rely on in silico-derived structural predictions or do we need crystal structures?

Apart from the basic virology aspects of identifying and characterising the rhinovirome, we have some fascinating questions to answer about why there are so many stable strains of HRV compared to other respiratory viruses and what their role is in 'training' our naïve immune systems to better defend against viral insult. If their role is central, what would be the impact on the population of introducing a vaccine? What is the immunobiology of the niHRVs and do the few strains used in studies to date accurately reflect the pathogenesis of infection by any HRV strain?

The discovery of this newfound diversity has surprised some authors and perhaps re-defined the assumptions made by others. Hopefully the HRV Cs can also teach the importance of an open mind towards rhinoviruses and their role in human illness.

Characterisation of a newly identified human rhinovirus, HRV-QPM, discovered in infants with bronchiolitis

Upper respiratory virus detection without parentreported illness in children is virus-specific

Propagation of commoncold virus in tissue cultures

The taxonomic position of common cold viruses and some others

Rhinoviruses: a numbering system

Localization of human rhinovirus replication in the upper respiratory tract by in situ hybridization

Basal cells of differentiated bronchial epithelium are more susceptible to rhinovirus infection

Personal exposure to nitrogen dioxide (NO2) and the severity of virus-induced asthma in children

Community study of role of viral infections in exacerbations of asthma in 9-11 year old children

Rhinovirus and respiratory syncytial virus in wheezing children requiring emergency care

The clinical importance of rhinovirus-associated early wheezing

Overview of virus-induced airway disease

Viral etiology of common cold in children

High prevalence of human rhinovirus C infection in Thai children with acute lower respiratory tract disease

Viruses as precipitants of asthma symptoms. I. epidemiology

A collaborative study of the aetiology of acute respiratory infections in Britain 1961-4. A report of the medical research council working party on acute respiratory virus infections

Viruses as precipitants of asthmatic attacks in children

Respiratory viruses and exacerbations of asthma in adults

Worldwide time trends in the prevalence of symptoms of asthma, allergic rhinoconjunctivitis, and eczema in childhood: ISAAC Phases One and Three repeat multicountry cross-sectional surveys

Alternative projections of mortality and disability by cause 1990-2020: Global Burden of Disease Study

Respiratory viruses in bronchoalveolar lavage: a hospital-based cohort study in adults

Picornavirus infections: a primer for the practitioner

Viral pneumonia in the first month of life

Ruuskanen O. Rhinovirus in otitis media with effusion

Computed tomographic study of the common cold

Pathology of naturally occurring colds

The cold case: Are rhinoviruses perfectly adapted pathogens?

The host response, not the virus, causes the symptoms of the common cold

Association between interleukin-8 concentration in nasal secretions and severity of experimental rhinovirus colds

A variety of respiratory viruses found in symptomatic travellers returning from countries with ongoing spread of the new influenza A(H1N1)v virus strain

SARS surveillance during emergency public health response

The economic burden of non-influenza-related viral respiratory tract infection in the United States

Cost burden of viral respiratory infections: Issues for formulary decision makers

The complex epidemiology of respiratory virus infections

Novel species of human rhinoviruses in acute otitis media

Rhinoviruses are a major cause of wheezing and hospitalization in children less than 2 years of age

A collaborative report: Rhinoviruses -extension of the numbering system from 89 to 100

Rhinovirusassociated hospitalizations in young children

Polymerase chain reaction and sequencing for typing rhinovirus RNA

Improved detection of rhinoviruses in clinical samples by using a newly developed nested reverse transcription-PCR assay

Frequent detection of human rhinoviruses, paramyxoviruses, coronaviruses, and bocavirus during acute respiratory tract infections

MassTag Polymerase-Chain-Reaction Detection of Respiratory Pathogens, Including a New Rhinovirus Genotype, That Caused Influenza-Like Illness in New York State during

Prevalence and clinical characterization of a newly identified human rhinovirus C species in children with acute respiratory tract infection

Evidence of recombination and genetic diversity in human rhinoviruses in children with acute respiratory infection

Clinical severity and molecular typing of human rhinovirus C strains during a fall outbreak affecting hospitalized patients

Multiplex MassTag-PCR for respiratory pathogens in pediatric nasopharyngeal washes negative by conventional diagnostic testing shows a high prevalence of viruses belonging to a newly recognized rhinovirus clade

Clinical features and complete genome characterization of a distinct human rhinovirus genetic cluster, probably representing a previously undetected HRV species, HRV-C, associated with acute respiratory illness in children

A novel group of rhinoviruses is associated with asthma hospitalizations

A recently identified rhinovirus genotype is associated with severe respiratory-tract infection in children in Germany

Global distribution of novel rhinovirus genotype

Detection of rhinoviruses by tissue culture and two independent amplification techniques, nucleic acid sequence-based amplification and reverse transcription-PCR, in children with acute respiratory infections during a winter season

Novel human rhinoviruses and exacerbation of asthma in children

Human rhinovirus C associated with wheezing in hospitalised children in the Middle East

Pan-viral screening of respiratory tract infections in adults with and without asthma reveals unexpected human coronavirus and human rhinovirus diversity

Rhinovirus associated with severe lower respiratory tract infections in children

A diverse group of previously unrecognized human rhinoviruses are common causes of respiratory illness in infants

Detection of human rhinovirus C in children with acute lower respiratory tract infections in South Korea

Human rhinovirus group C infection in children with lower respiratory tract infection

Screening respiratory samples for detection of human rhinoviruses (HRVs) and enteroviruses: comprehensive VP4-VP2 typing reveals high incidence and genetic diversity of HRV species C

Rhinovirus infections in Tecumseh, Michigan: Frequency of illness and number of serotypes

Host range and emerging and reemerging pathogens

Phylogenetic analysis of rhinovirus isolates collected during successive epidemic seasons

A newly discovered human pneumovirus isolated from young children with respiratory tract disease

Identification of a new human coronavirus

Characterization and complete genome sequence of a novel coronavirus, coronavirus HKU1, from patients with pneumonia

Cloning of a human parvovirus by molecular screening of respiratory tract samples

Identification of a third human polyomavirus

Identification of a Novel Human Polyomavirus from Patients with Acute Respiratory Tract Infections

Human metapneumovirus in children with acute respiratory tract infections in Suzhou

Respiratory viral infections detected by multiplex PCR among pediatric patients with lower respiratory tract infections seen at an urban hospital in Delhi from

Human coronavirus NL63 infections in infants hospitalised with acute respiratory tract infections in South Africa

Human rhinoviruses: The cold wars resume

Polymerase chain reaction amplification of rhinovirus nucleic acids from clinical material

A picornaviral loop-to-loop replication complex

The 5 0 -untranslated regions of picornavirus RNAs contain independent functional domains essential for RNA replication and translation

Use of polymerase chain reaction for diagnosis of picornavirus infection in subjects with and without respiratory symptoms

Recovery of new viruses (Coryzavirus) from cases of common cold in human adults

Real-time reverse transcription-PCR assay for comprehensive detection of human rhinoviruses

New molecular detection tools adapted to emerging rhinoviruses and enteroviruses

Development of one-step, real-time, quantitative reverse transcriptase PCR assays for absolute quantitation of human coronaviruses OC43 and 229E

Croup is associated with the novel coronavirus NL63

Comparison of real-time PCR assays with fluorescentantibody assays for diagnosis of respiratory virus infections in children

Human respiratory syncytial virus (hRSV) RNA quantification in nasopharyngeal secretions identifies the hRSV etiologic role in acute respiratory tract infections of hospitalized infants

Rhinovirus infections in children: a retrospective and prospective hospital-based study

Amplification of rhinovirus specific nucleic acids from clinical samples using the polymerase chain reaction

5 0 noncoding region alone does not unequivocally determine genetic type of human rhinovirus strains

Sequencing and Analyses of All Known Human Rhinovirus Genomes Reveals Structure and Evolution

Distinguishing molecular features and clinical characteristics of a putative new rhinovirus species, human rhinovirus C (HRV C)

Assay for 5 0 noncoding region analysis of all human rhinovirus prototype strains

New respiratory enterovirus and recombinant rhinoviruses among circulating strains

Frequency and dynamics of recombination within different species of human enteroviruses

Rhinovirus-induced interferon-gamma and airway responsiveness in asthma

Development and implementation of a molecular diagnostic platform for daily rapid detection of 15 respiratory viruses

Rhinovirus infections of children in hospital; Isolation of three possibly new rhinovirus serotypes

Respiratory pathogens in children with and without respiratory symptoms

Single versus dual respiratory virus infections in hospitalized infants: impact on clinical course of disease and interferongamma response

Human metapneumovirus in severe respiratory syncytial virus bronchiolitis

Prevalence and clinical characteristics of human metapneumovirus infections in hospitalized infants in Spain

HMPV infections are frequently accompanied by co-infections

Simultaneous detection of human bocavirus and adenovirus by multiplex real-time PCR in a Belgian paediatric population

Role of rhinovirus C in apparently lifethreatening events in infants

Recurrent human rhinovirus infections in infants with refractory wheezing

Do rhinoviruses reduce the probability of viral co-detection during acute respiratory tract infections?

Genomewide diversity and selective pressure in the human rhinovirus

Phylogenetic relationships and molecular adaptation dynamics of human rhinoviruses

A collaborative report: rhinoviruses-extension of the numbering system

Demonstration of dual rhinovirus infection in humans by isolation of different serotypes in human heteroploid (HeLa) and human diploid fibroblast cell cultures

Cleavage site analysis in picornaviral polyproteins: Discovering cellular targets by neural networks

Human rhinovirus 2: Complete nucleotide sequence and proteolytic processing signals in the capsid protein region

The complete nucleotide sequence of a common cold virus: human rhinovirus 14

Sequence analysis of human rhinoviruses in the RNA-dependent RNA polymerase coding region reveals within-species variation

Complete genomic sequencing shows that polioviruses and members of human enterovirus species C are closely related in the noncapsid coding region

Alignment of capsid protein VP1 sequences of all human rhinovirus prototype strains: conserved motifs and functional domains

Molecular characterization of human rhinovirus field strains isolated during surveillance of enteroviruses

X-ray structure of a minor group human rhinovirus bound to a fragment of its cellular receptor protein

Structure of human rhinovirus serotype 2 (HRV2)

The cis-acting replication elements define human enterovirus and rhinovirus species

VP1 sequencing of all human rhinovirus serotypes: Insights into genus phylogeny and susceptibility to antiviral capsid-binding compounds

Pneumonia and pericarditis in a child with HRV-C infection: a case report

Sequence analysis of human rhinovirus aspirated from the nasopharynx of patients with relapsingremitting MS

Enhanced severity of virus associated lower respiratory tract disease in asthma patients may not be associated with delayed viral clearance and increased viral load in the upper respiratory tract

Virological studies of sudden, unexplained infant deaths in Glasgow 1967-1970

Rhinovirus viremia in children with respiratory infections

Studies with rhinoviruses in volunteers: production of illness, effect of naturally acquired antibody, and demonstration of a protective effect not associated with serum antibody

Effects of the common cold on mood and performance

Persistence of rhinovirus and enterovirus RNA after acute respiratory illness in children

Picornavirus infections in children diagnosed by RT-PCR during longitudinal surveillance with weekly sampling: association with symptomatic illness and effect of season

Epidemiology of the common cold in military recruits with emphasis on infections by rhinovirus types 1A, 2, and two unclassified rhinoviruses

Persistence of rhinovirus and enterovirus RNA after acute respiratory illness in children

Presence of viral and bacterial pathogens in the nasopharynx of otitis-prone children. A prospective study

Rhinovirus transmission within families with children: incidence of symptomatic and asymptomatic infections

Relationship of upper and lower airway cytokines to outcome of experimental rhinovirus infection

Epidemiology of infections with rhinovirus types 43 and 55 in a group of university of Wisconsin student families

Human rhinoviruses: coming in from the cold

Persistence of rhinovirus RNA after asthma exacerbation in children

Rhinovirus/enterovirus RNA in tonsillar tissue of children with tonsillar disease

Detection of viruses in human adenoid tissues by use of multiplex PCR

Comparison of results of detection of rhinovirus by PCR and viral culture in human nasal wash specimens from subjects with and without clinical symptoms of respiratory illness

Chronic rhinoviral infection in lung transplant recipients

Rhinovirus infections in an industrial population I. The occurrence of illness

Virologic studies of acute respiratory disease in young adults. IV Virus isolations during four years of surveillance

Epidemiology of viral respiratory infections

Serotypes of viruses (rhinoviruses) isolated from common colds

Many rhinovirus serotypes share the same cellular receptor

Isolation of a monoclonal antibody that blocks attachment of the major group of human rhinoviruses

Two groups of rhinoviruses revealed by a panel of antiviral compounds present sequence divergence and differential pathogenicity

Phylogenetic analysis of human rhinovirus capsid protein VP1 and 2A protease coding sequences confirms shared genus-like relationships with human enteroviruses

Human rhinovirus 87 identified as human enterovirus 68 by VP4-based molecular diagnosis

Clinical and molecular epidemiology of human bocavirus in respiratory and fecal samples from children in Hong Kong

Application of a fluorogenic PCR assay for typing and subtyping influenza viruses in respiratory samples

Development of three multiplex RT-PCR assays for the detection of 12 respiratory RNA viruses

Diagnostic system for rapid and sensitive differential detection of pathogens

Simultaneous detection of influenza A, B and C viruses, respiratory syncytial virus, and adenoviruses in clinical samples by multiplex reverse transcriptase nested-PCR assay

Improved detection of rhinoviruses by nucleic acid sequence-based amplification after nucleotide sequence determination of the 5 0 noncoding regions of additional rhinovirus strains

Active surveillance for respiratory virus infections in adults who have undergone bone marrow and peripheral blood stem cell transplantation

Rapid detection of human rhinoviruses in nasopharyngeal aspirates by a microwell reverse transcription-PCR-hybridization assay

Sequence analysis of human rhinovirus aspirated from the nasopharynx of patients with relapsingremitting MS

Microarraybased detection and genotyping of viral pathogens

Real-time PCR for rapid diagnosis of entero-and rhinovirus infections using LightCycler

High-throughput, sensitive, and accurate multiplex PCR-microsphere flow cytometry system for large-scale comprehensive detection of respiratory viruses

Identification of enteroviruses in clinical specimens by competitive PCR followed by genetic typing using sequence analysis

Human rhinovirus group C in hospitalized children

Challenges facing real-time PCR characterization of acute respiratory tract infections

The authors would like to thank many people for graciously providing additional data for this review including: Pedro Avila, Zhao-jun Duan, Dean Erdman, James Gern, Giuseppe Gerna,