key: cord-0724305-5qtw90rr authors: Tao, K. P.; Chong, K. C.; Pun, J. C.; Tsun, J. G.; Chow, S. M.; Ng, C. S.; Wang, M. H.; Chen, Z.; Chan, P. K.; Li, A. M.; Chan, R. W. title: Suppression of influenza virus infection by rhinovirus interference at the population, individual and cellular levels date: 2021-08-10 journal: nan DOI: 10.1101/2021.08.09.21256656 sha: cc0bdb4948f189185f3470da2c14e455155ede48 doc_id: 724305 cord_uid: 5qtw90rr Background: Investigations of the natural viral interference effect between rhinovirus (RV) and influenza virus (IV) were conducted in temperate regions. We conducted an epidemiological study in Hong Kong, a major epicentre of influenza virus in the sub-tropical region. RV is the most prevalent respiratory virus year-round and causes asymptomatic to mild symptoms while IV infection exerts a great burden of public health. We aimed to examine the correlation of RV prevalence against IV activity. Methods: Nasopharyngeal aspirates (NPA) collected from patients hospitalized in the regional hospitals from 2015 to 2019 were examined for the presence of respiratory viruses. The correlation of the monthly prevalence between all pairs of virus infection, the co-infection rate and the temporal interference of RV and IV were tested. The viral interference was validated in vitro by conducting sequential RV and IV infection in the well-differentiated primary human airway epithelial cells. Findings: A total of 112,926 NPA were evaluated, and the Enterovirus/RV was the most prevalent respiratory virus detected. The negative correlation between EV/RV and IVs prevalence was independent of age and meteorological factors. Co-infection of EV/RV and IV was significantly less when compared with other virus pairs. Prior exposure to RV inhibited the replication of influenza A, B and oseltamivir-resistance stain in vitro and the inhibition is replication dependent. Interpretation: Epidemiological surveillance and the sequential infection in vitro suggested viral interference between EV/RV and IV operated at the population, individual and cellular levels. Influenza virus (IV) confers substantial morbidity and mortality worldwide every year. 62 The existence of natural reservoirs of IV makes it impossible to be eradicated in 63 Laboratory Centre, Centre for Health Protection, Department of Health, Hong Kong. 111 The respiratory virus panel, including influenza A virus (IAV, with subtyping of H1 112 and H3), influenza B virus (IBV), influenza C (ICV), parainfluenza viruses 1-4 (PIVs), 113 enterovirus/rhinovirus (EV/RV), respiratory syncytial viruses (RSV) and adenovirus 114 (ADV). The test-negative samples were retained as part of the essential denominator 115 to reflect the prevalence in the community to address the fluctuation in sample size 116 over the study period. This virological data covered nine episodes of IV peaks of four 117 consecutive years. Samples with missing or uncertain entries were excluded (<1%). 118 Meteorological data. We obtained the meteorological data including ambient 123 temperature (°C) and mean relative humidity (%) measured at the central monitoring 124 station run by the Hong Kong Observatory. The weekly averages of meteorological 125 record were matched with the prevalence over the study period. As absolute humidity 126 was showed to be associated with the respiratory infections, 13-15 we employed actual 127 vapour pressure (hPa) as a proxy of this humidity measure. The derivation of actual 128 vapour pressure was based on Teten's formula, 16 where e, TEMP, and RH denote actual vapour pressure, ambient temperature, and 131 relative humidity respectively. 132 133 Cell lines. H1-HeLa (CRL-1958) and Madin-Darby canine kidney cells (MDCK, were purchased from American Type Culture Collection. Cells were 135 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 cultured in minimal essential media with non-essential amino acids, 2mM L-136 glutamine supplemented with 1% penicillin and streptomycin, and 10% fetal bovine 137 serum. Both cell lines were maintained at 37°C and used for RV and IV virus 138 propagation and titration. Virus samples or culture supernatants were titrated in serial half-log 10 dilutions with 154 the corresponding culture medium before adding the diluted virus to the cell plates in 155 quadruplicate. The highest viral dilution leading to CPE was recorded and the 50% 156 tissue culture infectious dose (TCID 50 ) was calculated using the Karber method. The 157 infectivity of RV and IV was monitored by the infectious viral load in the supernatant, 158 as quantitated by viral titration in H1-HeLa or MDCK cells, respectively. 159 160 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 10, 2021. Primary HBECs and HNECs were washed with 125ul of PBS five times before 176 infection. Cells were exposed to infection regime 1) RV, 2) IV or 3) a prior RV 177 infection then a IV infection at 48 hpi of the initial RV inoculation at a multiplicity of 178 infection (MOI) of 0.01 ( Figure 6A ). In the infection step, 100ul of RV or sham 179 inoculum were added and allowed for virus attachment for 2 hours at 37°C. The 180 inoculum was discarded, and the cells were washed with PBS twice and the 181 basolateral compartment was replenished with 600ul of the medium. At 48 hours post 182 infection (hpi) of RV, cells were washed with PBS twice before secondary infection 183 of IAVs. The supernatant from the apical compartment of the transwell inserts were 184 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. The association of the monthly prevalence between all pairs of virus infections was 202 tested using Spearman's rank correlation coefficients. To assess whether IV 203 prevalence was statistically associated with the evolution of the future values EV/RV 204 prevalence, the Granger causality test was conducted and the significant lagged week 205 of IV prevalence was determined. 18 To examine the association between EV/RV and 206 IV prevalence independent of meteorological effects at different lagged times, a 207 quasi-Poisson generalized additive model (GAM) was used to control the total 208 number of weekly collected samples (i.e. model offset), long-term trend, and seasonal 209 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.21256656 doi: medRxiv preprint trend. The technical detail has been noted in the Supplementary File. The effect of 210 EV/RV on IV prevalence was quantified using adjusted relative risk (ARR) along 211 with its corresponding 95% confidence interval (CI). The reference value was set as 212 its median value. 213 The likelihood of viral co-infection was computed by Fisher's exact test and logistic 215 regression after adjustment to age and gender. Age group stratification with toddlers 216 (age <2), preschool (age 2-5), school-age (age 6-17), adult (age 18-64) and elderly 217 (age >65) were segregated for regression analysis. Differences in influenza titers and 218 viral gene expression was compared at respective time points with or without prior 219 EV exposure using two-way ANOVA followed by Bonferroni post-test for multiple 220 comparisons. One sample t-test was used to compare the changes (log 10 transformed) 221 in IV titer with prior RV infection, with null hypothesis assuming no difference with 222 sham treated control cells isolated from the same individuals. All statistical tests were 223 performed using Graphpad version 9.2.0 and IBM SPSS Statistics. Differences were 224 considered statistically significant at p < 0.05. 225 226 Opposing seasonality of EV/RV and IVs. A total of 112,926 NPA were included in 228 this study. EV/RV was the most prevalent viral infection ( Figure 1B) , and it 229 contributed a monthly positive rate of at least 8% throughout the study period ( Figure 230 2, red line). IVs were the second most dominant viral group being detected (Figure 2 , 231 navy line). Combining IAV, IBV and ICV, reached up to 35% positive rate during flu 232 season but remained low for the rest of the year. A strong seasonal pattern was 233 observed in both EV/RV and IVs, with robust biannual peaks of EV/RV occurred 234 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.21256656 doi: medRxiv preprint during spring and autumn, and one to two peaks of IV occurred in summers and 235 winters of Hong Kong, yet the onset, magnitude, duration and dominating subtypes of 236 the peaks varied extensively (Figure 2) . A staggered pattern between EV/RV and IVs 237 has been observed in which the intensity of flu peaks was often higher after a low 238 EV/RV season. During the spring of 2017, the shortest period of EV/RV peak was 239 followed by an early outburst of IAV of the H3N2 subtype. 240 241 Negative correlation between the prevalence of EV/RV with IVs. Interactions 242 between viruses may be confounded by other factors such as the age of the subjects 243 and meteorological factors during sampling. In the current study, we analyzed the 244 virological data by age stratification (Figure 1C ). EV/RV was the most prevalent in 245 those aged under 2 years old, while the median age of IAV and IBV positive cases 246 was significantly higher (11.19 and 7.84 years old, respectively) than that of EV/RV 247 (4.01 years old). Due to the variation of the influenza subtypes each year, IAV-H1, 248 IAV-H3, IBV and ICV were combined as IVs for regression analysis. Logistic 249 regression analysis revealed a significant negative correlation between the monthly 250 prevalence of EV/RV against IVs (-0.421, p<0.01, Figure 3C , navy) but not with 251 other virus pairs, such as PIVs, RSV and ADV. Spearman's bivariate analysis showed 252 a similar result in which a significant negative correlation was identified between 253 EV/RV against total IVs or IAV (Table 1 ). Significant negative correlations between 254 the monthly IVs prevalence and PIV2 and PIV4 were though their intensities were not 255 as strong as that between IVs and EV/RV. 256 257 Time series causality. IV prevalence was significantly associated with the evolution 258 of EV/RV (p<0.001) and the effect of IV was highly significant at lag zero (p<0.001), 259 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.21256656 doi: medRxiv preprint indicating a non-lagged interference between IV and EV/RV as assessed by the 260 Granger causality test. The disease-disease association at lag zero was further 261 examined via GAM analysis and a significant negative association between IV and 262 EV/RV was showed (Figure 4) . The ARR of EV/RV was 0.652 (95% CI: 0.571 to 263 0.745) when the prevalence of IV increased to 31.3% (i.e. 95 th percentile of IV), 264 whereas the ARR of EV/RV was 1.159 (95% CI: 1.079 to 1.244) when the prevalence 265 of IV decreased to 1.5% (i.e. 5 th percentile of IV), with a median reference value 266 (9.3%). samples were co-detected with two or more respiratory viruses, and 59.8% of these 279 co-infection cases were contributed by EV/RV (n=1,545) ( Figure 5A ). Co-infections 280 were more common in children, in which more than 80% of cases were found in 281 paediatric patients with age under 18 ( Figure 5B) . Interestingly, the odds to have both 282 EV/RV-IVs detected in the same specimen was exceptionally low (OR=0.15) when 283 compared with 0.75 for EV/RV-PIVs, 0.54 for EV/RV-RSV and 0.94 for EV/RV-284 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.21256656 doi: medRxiv preprint ADV co-detection using Fisher's Exact Test (Figure 5C) . A further reduction in odds 285 was observed between EV/RV and IVs after the adjustment of the confounding effect 286 due to age and gender by binary logistic regression ( Figure 5D) . inhibited. The viral load was significantly inhibited with a mean reduction of 1.44 299 log 10 and 1.22 log 10 in HBECs and 2.95 log 10 and 2.58 log 10 in HNECs at 24 h and 48 300 h post influenza virus infection, respectively, compared with those exposed to sham 301 treatment. (Figure 6B) . A significant reduction in the normalized IAV matrix gene 302 copies was also observed in both cell types (Supplementary Figure 2I) . 303 304 Active RV replication is essential to exert viral interference. Heterogenicity in 305 individual's susceptibility to RV was observed. HNECs derived from 4 out of 5 306 donors, and HBECs derived from 6 out of 9 donors supported productive replication 307 of RV (Supplementary Table 1 and Supplementary Figure 3) . Importantly, we 308 found that the interference effect depended on the active replication of RV. The 309 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.21256656 doi: medRxiv preprint inhibition effect to IAV infection was abolished in non RV-replicating cells. Using 310 ultraviolet-inactivated RV-A16 as the inoculum, we confirmed this observation again 311 and found that the UV-inactivated virus did not suppress the subsequent IVA and IVB 312 replication. 313 314 Suppression of IBV and oseltamivir-resistant IAV with prior exposure to RV-315 A16 and RV-1B in HBEC. The suppression by the prior exposure of RV-A16 was 316 not limit to IAV. A significant reduction of IBV (1.91 log 10 , and 2.25 log 10 at 24hpi 317 and 48hpi, p<0.05) and oseltamivir-resistant IAV strain (3.85 log 10 and 2.64 log 10 at 318 24 and 48hpi respectively, p<0.01, Figure 6C ) was observed in RV-A16 infected 319 cells compared with sham-exposed HBECs. Moreover, RV exposure protected HBEC 320 from IAV induced cell death. Extensive CPE was starting from 48hpi, while no CPE 321 was observed for at least 7 days if the cells were infected with RV-A16 before IVA 322 infection (Supplementary Figure 2C to 2H) . To rule out if the observation is a 323 specific effect of RV-A16, the same experiment setting was carried out using RV-1B 324 in HBECs. RV-1B suppressed the IAV replication by 2.59 log 10 at 48hpi but not at an The influenza virus exerts a great burden on the health system each year in terms of 331 frequent medical visits, hospitalization and flu-related death. We demonstrated the 332 viral interference between RV and IVs using epidemiological data and biological 333 experiments, suggesting a broad protective role of EV/RV in inhibiting subsequent 334 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.21256656 doi: medRxiv preprint IVs. We evaluated this interaction using the epidemiological data collected from 335 hospitalized patients from September 2015, when the EV/RV test was first introduced 336 as a routine test in the clinical settings, to December 2019, the last normal month 337 before the SARS-CoV-2 pandemic began. In this study, a total of 112,926 NPA 338 obtained from all ages were examined. The negative association between EV/RV and 339 IVs prevalence was independent of subject age and meteorological factors. Consistent 340 with studies performed in different climatic parameters, 10,11 an interference effect in 341 population-level is suggested. We also demonstrated the competitive effect between 342 the two could also operate at the individual level as the likelihood of getting co-343 detection between EV/RV and IVs was exceptionally low compared with other virus 344 pairs. During these nine flu seasons in Hong Kong, it is intriguing to see that EV/RV 345 prevalence oscillated in a counteracting manner. To convey the observation of viral interference from in vitro settings to the population 354 level, a transmission study using animal models will be a choice. It has been shown 355 that prior exposure to RV-1B can reduce the severity of mouse-adapted IAV PR8 in a 356 dose-dependent manner. 19 In the same study, mice with a double-stranded RNA 357 (dsRNA) mimic before IAV infection significantly reduce viral load. Aligned with 358 our findings in cells that did not support RV replication and UV-inactivated RV, it has 359 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.21256656 doi: medRxiv preprint been shown that UV-inactivated RV cannot induce antiviral cytokine responses. 20 The 360 lack of inhibitory effect to subsequent IAV infection suggested that active virus 361 replication within the host is required. Viral interference may be mediated by factors 362 such as IFNs, defective interfering particles, production of trans-acting proteases, 363 cellular factors, and nonspecific dsRNA. 21 Figure 4) . A recent finding suggested that the protective effect 378 conferred by prior EV exposure in respiratory cells is again due to the induction on 379 IFN-stimulated genes shared high consistency with our findings 26 . 380 In vitro infection using ALI differentiated cells and mathematical simulations also 382 agree that RV has an interference effect against SARS-CoV-2 at multiple levels. 27 383 These results all point to the fact that EV/RV infection, which usually causes mild or 384 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.21256656 doi: medRxiv preprint The cumulative evidence suggests the occurrence of viral interference at the 409 population, individual and cellular level. The understudied role of RV in providing 410 the baseline immunity to influenza virus replication warrants further attention. 411 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.21256656 doi: medRxiv preprint p IAV IBV ICV PIV1 PIV2 PIV3 PIV4 RSV ADV IVs EV/RV -0·421** 0·096 -0·223 0·117 0·114 0·240 0·147 0·052 0·138 -0·477*** AdV -0·199 Table 1 . Correlation of viral prevalence. Bivariate Spearman's cross-correlation coefficients between respiratory viruses using the monthly prevalence are shown. Asterisks indicate significance at p<0·05*, p<0·01** and p<0·001***. IVs in the last column indicates the sum of IAV, IBV and ICV, the cross-correlation within the IVs and IAV, IBV and ICV are therefore excluded from the analysis. Red and blue values indicate a negative and positive correlation, respectively. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.21256656 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.21256656 doi: medRxiv preprint The grey bar represents the number of samples tested in each group (right y-axis). (B) Prevalence of influenza subtypes (gradients of blue) compared to EV/RV (red) across the five-year study period. Typical influenza season (December to March, July to August) in Hong Kong is shaded in grey. (C) Logistic regression between the prevalence of EV/RV with other viruses with 95% CI marked in dotted lines. Significant negative correlation (-1·596 ± 0·3110, p<0·001*** with R 2 = 0·3123) was identified only between EV/RV and IVs. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.21256656 doi: medRxiv preprint Figure 4 . Adjusted relative risks (ARRs) with 95% confidence interval on EV/RV against IV prevalence. The estimated ARRs without controlling meteorological effects, with ambient temperature plus relative humidity controlled, and with actual vapor pressure adjusted are expressed as blue, red, and green colors respectively. Median IV prevalence was used as the reference value for comparison. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 10, 2021. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.21256656 doi: medRxiv preprint Rate of co-detecting more than one respiratory virus in different respiratory virus infection. The number of NPA samples detected with more than one pathogen detected was divided by the total number of the sample test-positive with the agent listed in each row. (B) Breakdown of co-infection cases according to age group. (C) The odd ratio of EV/RV and IVs co-infection with other respiratory pathogens using Fisher's exact test with null hypothesis assuming the likelihood of individual infection events were not interrelated (D) Logistic regression analysis of EV/RV and IV infection after adjustment to gender and age group with the same adjustments. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.21256656 doi: medRxiv preprint Suppression of IVA titers with prior RV-A16 infection in HBEC (n=9) and HNEC (n=5). Y-axis represents the difference in titer in RV exposed to the sham-treated cells. (C) Suppression of IVB and oseltamivir-resistance stain of IVA (R-IVA) in HBEC with prior RV-A16 exposure. (D) Suppression of seasonal IVA with prior RV-1B infection in HBEC. Error bars showing the SEM of means and asterisks indicating significance of p<0·05*, p<0·01** and p<0·001*** compared with sham treatment as examined by one-sample t-test. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.21256656 doi: medRxiv preprint Attenuation of Influenza A 475 Virus Disease Severity by Viral Coinfection in a Mouse Model Mouse models of rhinovirus-477 induced disease and exacerbation of allergic airway inflammation Immunological Outcomes of Coinfections Rhinovirus Reduces the 482 Severity of Subsequent Respiratory Viral Infections by Interferon-Dependent and -483 Independent Mechanisms Single treatment 485 with ethanol hand rub is ineffective against human rhinovirus--hand washing with 486 soap and water removes the virus efficiently Differentiating impacts of non-488 pharmaceutical interventions on non-coronavirus disease-2019 respiratory viral 489 infections: Hospital-based retrospective observational study in Taiwan. Influenza 490 Other Respir Viruses 2021. 491 25. Centre for Health Protection HKSAR. Statistics on Laboratory Surveillance Magnitude and timing of 494 the antiviral response determine SARS-CoV-2 replication early in infection Human rhinovirus infection blocks 497 SARS-CoV-2 replication within the respiratory epithelium: implications for COVID-498 19 epidemiology. The Journal of Infectious Diseases 2021 Interferon-Dependent and Respiratory Virus-Specific Interference in Dual Infections 504 of Airway Epithelia 0·92 (0·88 -0·95) *** Age Group 6 -17 0·66 (0·62 -0·70) *** 1·81 (1·69 -1·94) *** 18 -64 0·20 (0·19 -0·22) *** 0·65 (0·61 -0·70) *** > 65 0·16 (0·15 -0·17) *** 0·46 (0·43 -0·49) *** Co-detection PIVs 0·34 (0·31 -0·38) *** 0·20 (0·17 -0·23) *** RSV 0·22 (0·21 -0·24) *** 0·12 (0·10 -0·14) *** Adv 0·39 (0·35 -0·43) *** 0·19 (0·88 -0·95) ***