Genetics, recombination and clinical features of human rhinovirus species C (HRV-C) infections; interactions of HRV-C with other respiratory viruses

Abstract

To estimate the frequency, molecular epidemiological and clinical associations of infection with the newly described species C variants of human rhinoviruses (HRV), 3243 diagnostic respiratory samples referred for diagnostic testing in Edinburgh were screened using a VP4-encoding region-based selective polymerase chain reaction (PCR) for HRV-C along with parallel PCR testing for 13 other respiratory viruses. HRV-C was the third most frequently detected behind respiratory syncytial virus (RSV) and adenovirus, with 141 infection episodes detected among 1885 subjects over 13 months (7.5%). Infections predominantly targeted the very young (median age 6-12 months; 80% of infections in those <2 years), occurred throughout the year but with peak incidence in early winter months. HRV-C was detected significantly more frequently among subjects with lower (LRT) and upper respiratory tract (URT) disease than controls without respiratory symptoms; HRV-C mono-infections were the second most frequently detected virus (behind RSV) in both disease presentations (6.9% and 7.8% of all cases respectively). HRV variants were classified by VP4/VP2 sequencing into 39 genotypically defined types, increasing the current total worldwide to 60. Through sequence comparisons of the 5'untranslated region (5'UTR), the majority grouped with species A (n = 96; 68%, described as HRV-Ca), the remainder forming a phylogenetically distinct 5'UTR group (HRV-Cc). Multiple and bidirectional recombination events between HRV-Ca and HRV-Cc variants and with HRV species A represents the most parsimonious explanation for their interspersed phylogeny relationships in the VP4/VP2-encoding region. No difference in age distribution, seasonality or disease associations was identified between HRV-Ca and HRV-Cc variants. HRV-C-infected subjects showed markedly reduced detection frequencies of RSV and other respiratory viruses, providing evidence for a major interfering effect of HRV-C on susceptibility to other respiratory virus infections. HRV-C's disease associations, its prevalence and evidence for interfering effects on other respiratory viruses mandates incorporation of rhinoviruses into future diagnostic virology screening.

Figure 1. Phylogeny of VP4/2 gene sequences amplified from study subjects and comparison with corresponding full length sequences of HRV-C and other samples where 5′UTR groups had been previously assigned .

Symbols identify HRV-Cc (ie. with species C-like 5′UTR sequences) or HRV-Ca (species A-like) as described in the key. The neighbour-joining tree was constructed using maximum composite likelihood distances estimated between sequences in the amplified region (positions 629–1063 numbered according to the HRV-B14 reference sequence (NC_001490). Data was bootstrap re-sampled 100 times to assess robustness of branches; values of 70% or greater shown. Sequences showing incompatibilities between 5′UTR and phylogenetic grouping in VP4/VP2 suggestive of recombination are arrowed.

Figure 2. Seasonal distribution of (A) HRV-C positive samples and (B) those positive for other common respiratory viruses.

The y-axis depicts the proportion of samples positive from the total in each category (shown above each bar in part A). Frequencies of HRV-C detection have been subdivided into HRV-Ca and HRV-Cc subgroups.

Figure 3. Age distribution of (A) HRV-C infected subjects and (B) those infected with other common respiratory viruses.

The y-axis depicts the proportion of subjects positive from the total in each category. Frequencies of HRV-C detection have been subdivided into HRV-Ca and HRV-Cc subgroups.

Figure 4. Detection frequency of HRV-C and other respiratory viruses among subjects presenting with symptoms or diagnoses of LRTIs, URTIs and those with no relevant (NR) respiratory symptoms.

For the purposes of this analysis, cough was considered a symptom of URTI.

Figure 5. Analysis of potential virus interference through comparison of mono- and co-infection frequencies.

Each of the three most frequently detected respiratory viruses (HRV-C, RSV and AdV) and a combined group of all others (influenza A and B, PIV1-3, hMPV, coronaviruses, HBoV, HPeV) were separately analysed (x-axis, top). For each, mono-infection frequencies (unfilled bars) were compared with those where HRV-C, RSV, AdV and other respiratory viruses were co-detected (indicated on bottom x-axis, grey filled boxes). The existence of statistically significant differences in frequencies was determined by chi-squared test; NS: not significant.