|Home | About | Journals | Submit | Contact Us | Français|
Rhinoviruses cause serious morbidity and mortality as the major etiological agents of asthma exacerbations and the common cold. A major obstacle to understanding disease pathogenesis and to the development of effective therapies has been the lack of a small-animal model for rhinovirus infection. Of the 100 known rhinovirus serotypes, 90% (the major group) use human intercellular adhesion molecule-1 (ICAM-1) as their cellular receptor and do not bind mouse ICAM-1; the remaining 10% (the minor group) use a member of the low-density lipoprotein receptor family and can bind the mouse counterpart. Here we describe three novel mouse models of rhinovirus infection: minor-group rhinovirus infection of BALB/c mice, major-group rhinovirus infection of transgenic BALB/c mice expressing a mouse-human ICAM-1 chimera and rhinovirus-induced exacerbation of allergic airway inflammation. These models have features similar to those observed in rhinovirus infection in humans, including augmentation of allergic airway inflammation, and will be useful in the development of future therapies for colds and asthma exacerbations.
Rhinoviruses cause the common cold, hospitalization of infants1, pneumonia in the immunosuppressed2 and the majority of acute exacerbations of asthma3–6 and chronic obstructive pulmonary disease (COPD)7. These diseases generate hundreds of billions of dollars in healthcare costs8, and the morbidity and mortality attributable to rhinovirus infections is enormous. No effective treatment is available.
Attempts to develop small-animal models of rhinovirus infection in many species, from mice to monkeys, have failed, and an intracellular block to rhinovirus replication in mouse cells was proposed to be the cause9. The lack of small-animal models of rhinovirus infection has thus severely hampered mechanistic insight into and therapeutic development for rhinovirus infections over the past 50 years.
Ten percent of the ~100 rhinovirus serotypes make up the minor receptor group of viruses, which can use both the human and mouse forms of the receptor to enter cells of either species10. The remaining 90% comprise the major receptor group and use human ICAM-1 for cell attachment and entry11, but these viruses do not bind mouse ICAM-1 (refs. 12,13). Here we report the development of in vivo mouse models of major- and minor-group rhinovirus infection and of rhinovirus-induced asthma exacerbation.
We inoculated BALB/c mice with 5 × 106 50% tissue-culture infective dose (TCID50) of minor-group rhinovirus-1B, UV-inactivated rhinovirus-1B or major-group rhinovirus-16 (Fig. 1). Only rhinovirus-1B induced rapid neutrophilic inflammation, detectable at 8 h, peaking 1–2 d after infection (with neutrophils representing 60–70% of total bronchoalveolar lavage (BAL) cells) and returning to baseline levels by day 4 (Fig. 1a). A prolonged increase in lymphocytes detectable at day 2 and peaking 4–7 d after infection (~20% of BAL cells) was also observed only in rhinovirus-1B–infected mice (Fig. 1a). Macrophage numbers remained unchanged in all groups (Supplementary Fig. 1a online). Examination of stained lung sections revealed areas of extensive peribronchial and perivascular cellular infiltration (Fig. 1b) that were not detected in the lungs of control mice. Induction of mucin-5, subtypes A and C (Muc5AC), and Muc5B mRNA in lung (Fig. 1c,e), Muc5AC and Muc5B protein in BAL (Fig. 1d,f) and staining for Muc5AC and Muc5B in bronchial epithelium (Supplementary Fig. 1b) were observed only in rhinovirus-1B–infected mice, with Muc5AC protein abundance peaking at day 1 and remaining elevated at day 7 and Muc5B elevated throughout days 1–14 after infection. Rhinovirus-1B induced the early production of the neutrophil chemokines macrophage inflammatory protein-2 (MIP-2) and KC (Fig. 2a,b) concurrently with the onset of neutrophilia (Fig. 1a) and also induced the dendritic and T cell chemokine MIP-3α (CCL20, refs. 14,15); Fig. 2c), whereas the T cell chemokines IP-10 (CXCL10), RANTES (CCL5) and I-TAC (CXCL11) were induced later, peaking at 1–2 d (Fig. 2d–f) and preceding the lymphocytosis on days 2–7 (Fig. 1a). Rhinovirus-1B induced the proinflammatory cytokines tumor necrosis factor-α (Supplementary Fig. 1c), interleukin-6 (IL-6) and IL-1β between 8 and 48 h after infection (Fig. 2g,h). Thymic stromal lymphopoietin, IL-4, IL-13 and IL-17 were all undetectable (data not shown).
The levels of rhinovirus-1B genomic RNA in lung and of infectious virus particles in BAL from 8 to 24 h after infection were significantly increased compared to those in mice treated with UV-inactivated virus (Fig. 3a,b). Viral replication was further shown by in situ hybridization using an antisense RNA probe to detect genomic viral RNA and a sense RNA probe to detect replicative viral RNA (Fig. 3c). Strong staining for replicative RNA was detectable in the airway epithelium 8 h after infection (data not shown), peaking at 12 h (Fig. 3c) and becoming undetectable by 48 h (data not shown), whereas genomic RNA peaked at 24 h (Fig. 3c) and was observable until 72 h after infection (data not shown). Rhinovirus-1B induced interferon-α (IFN-α), IFN-β and IFN-λ proteins in BAL fluid (Fig. 3d); IFN-β was produced earliest at 8 h. All IFNs peaked at 24 h, and IFN-α and IFN-β had returned to baseline by 48 h (Fig. 3d), whereas IFN-λ remained significantly elevated at this time (Fig. 3d).
IFN-γ production was increased in lung leukocytes from rhinovirus-1B–infected mice compared to those from rhinovirus-16–infected mice (Fig. 3e). IFN-γ production at day 4 was due to nonspecific activation of leukocytes, as only leukocytes obtained at day 7 produced significantly more IFN-γ when re-stimulated ex vivo with rhinovirus-1B (Fig. 3e), which is consistent with induction of a virus-specific acquired immune response. Rhinovirus-1B–specific IgG titers were significantly higher at day 7 than at day 0 and were further increased at day 10 in rhinovirus-1B–infected mice compared to those in mice infected with UV-inactivated virus (Fig. 3f).
We generated a BALB/c mouse strain transgenic for the rhinovirus-binding extracellular domains 1 and 2 of human ICAM-1 (Supplementary Methods and Supplementary Fig. 2a online). Rhinovirus-16 infection of transgenic mice, but not rhinovirus-16 infection of nontransgenic BALB/c controls or UV-inactivated rhinovirus-16 infection of transgenic mice, resulted in outcomes similar to those observed with the rhinovirus-1B infection model, including BAL neutrophilia and lymphocytosis (Fig. 4a); increased Muc5B protein in BAL (Fig. 4b); increased viral RNA levels (Fig. 4c); and induction of IFNs (Fig. 4d), chemokines, IL-1β (Fig. 4e–g) and virus-specific antibodies (Supplementary Fig. 2b).
Finally, to develop a mouse model of rhinovirus-induced asthma exacerbation, we challenged BALB/c mice sensitized to ovalbumin (OVA) with OVA or PBS control. During the challenge, mice were also inoculated with either rhinovirus-1B or UV-inactivated rhinovirus-1B. The combination of OVA challenge and rhinovirus-1B infection (RV-OVA) resulted in a significant increase in the BAL of neutrophils on day 1 (Fig. 5a) and eosinophils on day 7 (Supplementary Fig. 3a online), and a prolonged increase in lymphocytes in BAL from days 3 to 14 (Fig. 5a), compared with mice infected with virus but not challenged (RV-PBS) and allergen-challenged mice dosed with UV-inactivated virus (UV-OVA). In addition to increased airway inflammation, RV-OVA mice also showed significantly enhanced airway hyper-responsiveness on day 1 compared with that observed in mice infected with virus alone or mice challenged with allergen alone, whether the hyper-responsiveness was measured non-invasively (Fig. 5b, left) or invasively (Fig. 5b, right). RV-OVA mice showed significantly increased Muc5AC protein abundance in BAL at day 7 and Muc5B expression at day 14 compared with UV-OVA–treated and RV-PBS–treated mice (Fig. 5c), as well as significant induction of IL-4 (day 1) and IL-13 (days 1 and 2) expression compared with UV-OVA controls (Fig. 5d,e), whereas IL-5 abundance (Supplementary Fig. 3b) was not significantly increased. Significant induction of virus-specific IFN-γ was also observed in RV-OVA mice compared to those receiving either viral or allergen challenge alone (Fig. 5f).
Rhinoviruses, the cause of common colds, are the most frequent precipitants of acute exacerbations of asthma and COPD3–7, as well as causes of other serious respiratory diseases1,2. The mouse models of minor- and major-group rhinovirus infection reported herein are characterized by several relevant disease-related outcomes, including airway neutrophilic and lymphocytic inflammation16,17, mucin secretion (a cardinal symptom of common colds that is also induced by rhinovirus infection of human airway epithelial cells in vitro18) and the induction of various chemokines and proinflammatory cytokines induced in human rhinovirus infections.
Our recent work has highlighted deficient induction of IFN-β and IFN-λ in lung cells from asthmatic donors in response to rhinovirus infection ex vivo19,20. However, the relative importance of the IFN subtypes in rhinovirus infections is unknown, and although IFN-λ1 has been shown to have antiviral effects21,22, it has not been known whether IFN-λs are produced by rhinovirus infections in vivo. Our data suggest that IFN-λ is the most sustained and abundant IFN subtype induced in the lung by rhinovirus infections.
The absence of pathologic, physiologic or antiviral responses in mice inoculated with UV-inactivated virus and in nontransgenic mice inoculated with rhinovirus-16 indicates that these responses were replication dependent. Replication was studied further with in situ hybridization, which showed early production of negative-sense replicative RNA and subsequent synthesis of viral genomic RNA in epithelial cells, which is consistent with the rhinovirus replication cycle. We observed further evidence of viral replication in the increased levels of rhinovirus-16 RNA in the lungs of transgenic mice compared to those in control mice and in the induction of type 1 IFN-α and IFN-β and type 3 IFN-λs. The use of viruses better adapted for replication in mice23,24 or different mouse strains, including those with deficient IFN responses19,20, may allow for better virus replication in mice.
Because rhinoviruses are the major precipitants of acute exacerbations of asthma3–6, we studied the interactions between rhinovirus infection and allergic airway inflammation. In allergen-sensitized and allergen-challenged mice, rhinovirus infection exacerbated neutrophilic, eosinophilic and lymphocytic airway inflammation, airway hyper-responsiveness, mucus secretion and production of both T helper type 1 (TH1) and TH2 cytokines. These responses have been associated with rhinovirus-induced exacerbations of asthma25, including fatal asthma26, and are consistent with either rhinovirus infection substantially exacerbating allergic airway inflammation or, conversely, allergic airway inflammation substantially exacerbating rhinovirus-induced inflammation.
The mechanisms involved in the synergistic interaction between virus infection and allergen exposure to increase the risk of asthma exacerbations are unknown27,28. TH2-mediated inflammation is clearly implicated in allergic airway inflammation; however, it is not known whether rhinovirus infection can augment TH2 responses to allergens. Conversely, virus infection is strongly associated with TH1 responses, but it is not known whether allergic airway inflammation can augment rhinovirus-induced TH1 responses. We show here that rhinovirus infection increases production of both IL-4 and IL-13 in response to allergen challenge, indicating that rhinovirus infection can exacerbate TH2 responses to allergens. Similarly, allergen challenge increased rhinovirus induction of IFN-γ, confirming that allergen exposure can enhance TH1 responses to rhinoviruses. Further studies will be required to increase our understanding of these possible mechanisms of disease to aid the development of new therapeutic approaches.
In conclusion, these models of rhinovirus infection should be useful in investigating the pathogenesis and treatment of the common cold and acute exacerbations of asthma. With further development, they will probably provide a similar boost to research into exacerbations of COPD7 and other diseases in which rhinovirus infection is implicated2.
We purchased 6-week-old female BALB/c mice from Harlan. We generated human-mouse ICAM transgenic mice in house (Supplementary Fig. 2 and Supplementary Methods online). We grew major-group rhinovirus serotype 16 and minor-group rhinovirus serotype 1B in Ohio HeLa cells (European Collection of Cell Cultures). We obtained viruses from the American Type Culture Collection and passaged them five times in HeLa cells before purification. We titrated the viruses on HeLa cells by standard methods and inactivated them by exposing them to UV light at 1,200 mJ/cm2 for 30 min. We lightly anesthetized the mice with isofluorane and infected them intranasally with 50 μl of rhinovirus.
We excised the left upper lobe (~15% of total lung) from each mouse, purified the RNA (RNeasy miniprep kit, Qiagen) and reverse-transcribed 5 μg using random hexamers (Omniscript RT kit, Qiagen) as primers. We normalized real-time PCR assays to 18S rRNA with Quantitect Probe PCR master mix (Qiagen) and primers and probe to human 18S rRNA29. The TaqMan assays for measuring rhinovirus RNA have been described previously20.
We sensitized the mice with 50 μg OVA in 2 mg aluminum hydroxide intraperitoneally in a volume of 200 μl on day 1. We challenged lightly anesthetized (isofluorane) mice with 50 μg OVA (Calbiochem) in 30 μl of PBS (controls received PBS alone) on three consecutive days (days 10–12) to induce allergic airway inflammation. While they were under anesthesia during the third OVA challenge, we infected the mice intranasally with 50 μl rhinovirus-1B (2.5 × 106 TCID50) or UV-inactivated rhinovirus-1B. We assessed lung function at 24 h after infection and killed the mice on the indicated days (Fig. 5). for end-point analyses.
We cannulated mouse lungs through the trachea and then lavaged the lungs. We processed cellular fractions for differential staining by cytospinning. We analyzed BAL fluid for interferons, cytokines, chemokines and mucins by ELISA (R&D Systems) according to manufacturers’ specifications.
We carried out ELISpot analysis of IFN-γ production by purified lung leukocytes (harvested on day 4 or day 7 after infection). We cultured the cells for 3 d with or without ex vivo rhinovirus-1B stimulation. We read the plates with Leica Q Windows image software, quantifying the spot number per well.
We coated plates (Maxisorp, Nunc) overnight with purified rhinovirus stock (2 × 108 TCID50/ml). We incubated diluted serum samples for 1 h at 37 °C before detection with peroxidase-conjugated antibody to mouse IgG (Sigma).
All data were normally distributed and are expressed as means ± s.e.m. We analyzed data by ANOVA and pinpointed differences between groups using Bonferroni’s post-test, with 95% confidence levels (Prism 4, GraphPad) to determine significant differences. The numbers of mice used are indicated in the figure legends, and we repeated all experiments at least three times.
Further details on these and other methods are available in the Supplementary Methods.
This work was supported by Medical Research Council UK grant number G9824522, GlaxoSmithKline, Sanofi Pasteur and Asthma UK grant numbers 03/073, 04/052, 05/067 and 06/050.
Note: Supplementary information is available on the Nature Medicine website.
AUTHOR CONTRIBUTIONS S.L.J. conceived the studies and was principle investigator; all authors contributed to the design and execution of the experiments, helped draft the manuscript and approved the final version for publication.
Reprints and permissions information is available online at http://npg.nature.com/reprintsandpermissions