Expression of intercellular adhesion molecule-1 (ICAM-1) in nasal epithelial cells of atopic subjects: a mechanism for increased rhinovirus infection?


  • A. Bianco,

    1. Lung Injury and Inflammation Research Group, Centre for Cell and Molecular Medicine, Keele University, and Department of Respiratory Medicine, North Staffordshire Hospital Trust, Stoke-on-Trent, UK
    Search for more papers by this author
  • S. C. Whiteman,

    1. Lung Injury and Inflammation Research Group, Centre for Cell and Molecular Medicine, Keele University, and Department of Respiratory Medicine, North Staffordshire Hospital Trust, Stoke-on-Trent, UK
    Search for more papers by this author
  • S. K. Sethi,

    1. Lung Injury and Inflammation Research Group, Centre for Cell and Molecular Medicine, Keele University, and Department of Respiratory Medicine, North Staffordshire Hospital Trust, Stoke-on-Trent, UK
    Search for more papers by this author
  • J. T. Allen,

    1. Lung Injury and Inflammation Research Group, Centre for Cell and Molecular Medicine, Keele University, and Department of Respiratory Medicine, North Staffordshire Hospital Trust, Stoke-on-Trent, UK
    Search for more papers by this author
  • R. A. Knight,

    1. Lung Injury and Inflammation Research Group, Centre for Cell and Molecular Medicine, Keele University, and Department of Respiratory Medicine, North Staffordshire Hospital Trust, Stoke-on-Trent, UK
    Search for more papers by this author
  • M. A. Spiteri

    1. Lung Injury and Inflammation Research Group, Centre for Cell and Molecular Medicine, Keele University, and Department of Respiratory Medicine, North Staffordshire Hospital Trust, Stoke-on-Trent, UK
    Search for more papers by this author

Correspondence: Dr Monica A. Spiteri, Department of Respiratory Medicine, North Staffordshire Hospital Trust, Newcastle Road, Stoke-on-Trent, Staffordshire ST4 6QG, UK.E-mail:


Since clinical experimental studies indicate that upper respiratory tract viral infections may exacerbate acute asthma symptoms in atopic/asthmatic individuals, we have investigated the expression and modulation of ICAM-1 on human nasal epithelial cells (HNEC) from normal and atopic subjects. ICAM-1 is the attachment molecule for the majority of serotypes of human rhinovirus (HRV), including HRV-14, and is also critical for the migration and activation of immune effector cells. Basal ICAM-1 expression was significantly higher in HNEC obtained by brushings from atopic compared with non-atopic subjects (P = 0·031), and was also significantly increased on atopic HNEC harvested in season compared with out of season (P < 0·05). Atopic HNEC showed further up-regulation in ICAM-1 expression when cultured with clinically relevant allergen (P = 0·032). ICAM-1 levels on normal HNEC were also increased by infection with HRV-14 (P < 0·05). Basal expression of ICAM-1 on atopic nasal polyp epithelial cells (EC) was significantly higher than on both normal and atopic nasal HNEC. This elevated nasal polyp ICAM-1 level was not increased further by allergen, although HRV infection resulted in a small significant increase. Recovered viral titres from HRV-infected nasal polyp EC were 1·5-fold higher than from infected normal nasal HNEC. The data are consistent with the hypothesis that allergen, by enhancing expression of the HRV attachment target on host cells, facilitates viral infection in atopic subjects; simultaneously HRV-induced increases in ICAM-1 levels would favour migration and activation of immune effector cells to the airway, resulting in enhanced atopic inflammation.


Asthma is a common and sometimes life-threatening respiratory disease affecting approximately 30% of the UK population and 20% of USA [1]. Several studies show a marked relationship between acute asthma exacerbations and upper respiratory viral infections [2–4]. Chanarin et al. [5] reported a close temporal association between the peak in seasonal asthma deaths and subjective/objective evidence of respiratory viral infection. Separate observations indicated that patients with atopic/asthma appear to have a higher incidence of viral respiratory infections than non-atopic individuals [6–9]. Whilst the virus–asthma relationship is complicated by other factors such as allergen load, improved understanding of the cellular mechanisms by which the atopic asthmatic airway is both susceptible and more reactive to viral infection will provide ways of targeting antiviral therapy to the host.

In vitro studies utilizing HRV-16 have demonstrated that the virus enhances allergen-induced responses in the lower airway, including histamine release, eosinophil recruitment, and the induction of late-phase airway response [7,8]. Furthermore, by inducing HRV-16 infection in normal subjects and atopic asthma patients, Fraenkel et al. [6] found that whilst all subjects developed increased numbers of T cells and eosinophils within the airway mucosa during the acute infection phase, only the atopic asthmatics had eosinophilia persisting 6–10 weeks later.

ICAM-1 is the key cell receptor for viral attachment and cell entry for 90% of human rhinovirus (HRV) serotypes. Both this major group of HRV, and the 10% of HRV which use an alternative receptor for cell attachment, enhance ICAM-1 expression [10–12]. Furthermore, inflammatory cytokines, particularly Th-2 cytokines such as IL-13, whose expression is increased in the atopic airway, also increase ICAM-1 expression [13]. HRV-mediated enhancement of ICAM-1 expression is dependent on NF-κB [11], and antigen-primed NF-κB-deficient mice show markedly reduced eosinophil infiltration following antigen challenge.

In addition to its role as a docking molecule for the majority of HRV serotypes, ICAM-1, with its cognate ligand LFA-1 (CD18/CD11a), is critical for the migration of inflammatory cells into an area of atopic inflammation [14]. In murine models of asthma, ICAM-1 has been shown to regulate lymphocyte and eosinophil infiltration of the lower airway [15,16]. Human bronchial biopsy studies have also shown that epithelial and endothelial ICAM-1 expression is increased in atopic, but not non-atopic, subjects [17].

We have previously shown, using cultured human bronchial H292 epithelial cell lines, that both HRV infection and the Th2-associated cytokines characteristic of atopic inflammation [9,18–20] increase cell surface ICAM-1 expression [10,11,13,21]. To validate these in vitro findings, the objectives of the present study were, first, to compare basal ICAM-1 expression on human nasal epithelial cells (HNEC) obtained by nasal brushings from normal and atopic subjects, and to assess whether there are differences in expression on cells recovered from atopic donors in and out of season. Second, using primary HNEC from normal subjects, we wished to confirm that HRV infection indeed increased surface ICAM-1 levels. Finally, we made use of the availability of the relatively large yield of primary EC from nasal polyps removed surgically from atopic donors, to assess the recovered titres of HRV after infection, and co-effects of allergen exposure.



The study population consisted of: (i) 13 atopic/asthmatic subjects (six males, seven females with a mean age of 31 ± 4·3 years) with a positive skin prick test reaction for grass allergen (cutaneous wheal response of > 3 mm diameter compared with saline control); and negative to all other tested common aeroallergens; and positive specific IgE antibody to grass (≥ grade 2; 0·7 kU/l), total serum IgE level > 100 U/ml. None of the subjects had received any nasal immunosuppressant medication for at least 2 weeks prior to the study; (ii) 11 non-atopic normal volunteers (four males and seven females with a mean age of 33 ± 3·6 years) with a negative history of asthma or atopy, negative skin prick tests; and mean IgE level 45·0 ± 2·3 U/ml. All above subjects were lifelong non-smokers and had not suffered a viral infection within at least the previous 8 weeks; (iii) separate experiments utilized epithelial cells (EC) isolated from nasal polyps surgically removed from three atopic/asthmatic patients (one male, two females, mean age 34 ± 3·2 years) with mean total IgE level of 450 ± 90 U/ml, and skin test positivity (≥ grade 2) to grass pollen. Treatment history of these patients indicated that they had received intermittent nasal steroid and/or anti-histamine preparations prior to surgery. Control nasal brushing EC were obtained from a further three non-atopic healthy volunteers (one male, two females, mean age 29 ± 3 years).

The study received prior approval by the local Research Ethics and Scientific Committee. Formal written consent was obtained from all subjects

Nasal epithelial cell sampling

HNEC were obtained at the height of the hay fever season and, out of season when patients were symptom-free, by direct endoscopic brushing of the nasal epithelium using a sterile nylon 2-mm brush, previously dipped in Hams F-12 (Sigma, Poole, UK). Samples were taken from between the inferior nasal turbinate and lateral nasal wall of the cavity, the nasal septum at the level of the inferior turbinate and the inferior turbinate itself. Retrieved samples were gently agitated into Hams F-12 medium supplemented with anti-microbials (20 000 U/ml penicillin, 20 mg/ml streptomycin and 50 μg/ml amphotericin B (all from Sigma)). Viability was assessed using the trypan blue exclusion test. Cytospins were prepared for staining with May–Grünwald–Giemsa (Sigma) to identify cell types present in each sample and analysed by immunocytochemistry for cytokeratin staining. HNEC brushings yielded a total of 1·5–2·0 × 106 cells, of which > 90% were epithelial cells.

Tissue culture flasks (Corning, Corning, NY), of a size appropriate for the number of recovered HNEC, were coated with a 1:10 dilution of bovine dermal purified collagen solution (Sigma); left for 5 h at room temperature and then fixed with a 2·5% solution of glutaraldehyde (Sigma) in PBS for 10 min. The fixed collagen layer was washed thoroughly with PBS (Sigma) for 3 min, three times. Hams F-12 supplemented with anti-microbials was then added to the flask and placed at 37°C in humidified air containing 5% CO2 until required.

Propagating HNEC

HNEC suspensions in Hams F-12 medium were centrifuged at 500 g for 10 min at room temperature. The cell pellet was gently resuspended in Hams F-12 serum-free culture medium supplemented with anti-microbials, l-glutamine, 5 μg/ml transferrin, 5 μg/ml insulin, 25 ng/ml epidermal growth factor, 15 μg/ml endothelial growth supplement, 200 p m triiodothryonin and 100 n m hydrocortisone (all from Sigma). This supplemented Hams F-12 culture medium will now be referred to as Hams F-12 6X medium. The EC suspension (105 cells/ml) was then introduced into the collagen-coated flasks and cultured in Hams F-12 6X at 34°C in humidified air containing 5% CO2. The culture medium was replaced with freshly made Hams F-12 6X at day 1 and then every 2 days after this, until a non-overlapping semiconfluent monolayer was achieved.

Extraction of epithelial cells from human nasal polyps

As experiments identifying the combined influence of allergen and HRV infection over a time period required a large number of EC, we utilized nasal polyp (NP) specimens from three known atopic asthmatics undergoing polypectomies to provide a greater cell yield. These nasal polyp specimens are considered a different model of chronic airway inflammation, although the samples were isolated from atopic subjects. As obvious background differences occur between EC obtained from nasal polyp and those from atopic nasal brushings, data analyses were performed separately.

Nasal EC were isolated from NP by established protease digestion methods [22,23]. Nasal specimens were washed with Hams F-12 supplemented with anti-microbials and incubated in 0·1% protease type XIV (Sigma) in Hams F-12 at 4°C overnight. After incubation, 10% fetal calf serum (FCS; Sigma) was added to neutralize the protease activity. Nasal specimens were then gently pounded to release epithelial cells. The cell suspension was then filtered through a sterile 60 mesh cell dissociation sieve. The cells were then processed in the same manner as HNEC.

In these experiments, control cultures of HNEC were obtained by nasal brushings from three normal healthy volunteers and pooled together to ensure a sufficient number of cells were obtained to perform these experiments.


A semiquantitative three-step indirect immunoenzymatic labelling protocol [19] for analysis of surface ICAM-1 expression [24–28] was used as described in previous studies [10,12,13]. Briefly, EC were incubated with a purified ICAM-1 MoAb (RR1/1.1; Boehringer Ingelheim, Ridgefield, CT ) for 30 min, after which rabbit anti-mouse IgG was added for 30 min followed by swine anti-rabbit IgG (Dako, Glostrup, Denmark) to amplify the staining intensity. Both secondary and tertiary antibodies are conjugated to the same enzyme (peroxidase). Cells were then incubated with benzidine (Sigma, Poole, UK) and stained with Mayer's Hemalum solution (Merck Ltd., UK). Three hundred cells were counted per field. Surface ICAM-1 expression on each of these EC was scored according to a five-point rating scale based on intensity of staining and appearance of nucleus: 0 = grey/brown, 1 = light brown, 2 = medium brown, 3 = medium/dark brown, 4 = dark brown; in grades 0–2 the nucleus appears well defined, in 3–4 the nucleus is partially or fully obliterated. The number of cells thus counted in each grade was then multiplied with the respective grade index and the resulting values summed. Immunocytochemistry was performed at the start of the experiment at then at 1, 4, 6 and 8 days. The final result for surface ICAM-1 on EC was expressed as the POX score as defined as the difference between the sum of the specific and background staining: {((a × 0) + (b × 1) + (c × 2) + (d × 3) + (e × 4)) – value for the control slide = POX score}, where each letter represents number of cells in the respective grade [19]. All slides were scored blind and by a second independent observer.

Rhinovirus-14 stock generation

The main seed of rhinovirus (HRV-14) was kindly provided by J. Kent (University of Leicester, UK). A stock solution of HRV-14 was generated by infecting confluent monolayers of HeLa Ohio cells (European Collection of Animal Cell Culture (ECACC), Salisbury, UK). The HeLa Ohio cell line was cultured as previously described [19]. Briefly, HRV-14 was propagated using confluent non-overlapping monolayers of HeLa Ohio cells. The medium was first removed and the monolayer was then inoculated with a known titre of rhinovirus and incubated for 90 min on roller cultures at 33°C in a humidified atmosphere containing 5% CO2. Residual inoculum was decanted and replaced by maintenance medium (Eagle's minimum essential medium (EMEM) supplemented with 5% FCS, 4% sodium bicarbonate, 10 000 U/ml penicillin, 10 mg/ml streptomycin and 25 mg/ml amphotericin B (Sigma) and incubated as roller cultures at 33°C in humidified air containing 5% CO2. Flasks were checked daily for cytopathic effect (cpe); once cpe was > 80%, flasks were frozen and thawed three times in order to release virus from cells. Medium containing virus was centrifuged at 600 g for 10 min, after which the viral suspension was stored at −70°C until required. Supernatants from uninfected HeLa were also frozen in aliquots, to be used as a negative control.

Viral purification

Prior to use in experiments, HRV-14 was purified by an established sucrose gradient method to remove ribosomes and soluble factors of HeLa cell origin [28]. In brief, 20 μg/ml RNase A (Sigma) were added to the above viral suspension and incubated at 35°C for 20 min. Sodium Sarkosyl (35 μm; 1% solution; Sigma) and 1 μl/ml 2-mercaptoethanol (Sigma) were then added to the viral suspension. A solution containing; 20 m m Tris acetate (Sigma), 1 m NaCl, 30% (w/v) sucrose (Sigma) was prepared, 3 ml of which were added to a 50-ml centrifuge tube. Viral suspension (27 ml) was then poured onto this solution. The resulting layered solutions were centrifuged at 200 000 g at 16°C for 5 h. The supernatant was discarded and the resulting viral pellet resuspended in medium and stored at −70°C until required.

Viral titre assay

The TCID50 method was used to calculate the concentration of virus (viral titres) in supernatants. Serial dilutions of virus were incubated with EC using roller cultures as described above, to assess the dilution of virus that was not able to cause cpe in cells and so quantify the amount of infectious virus. Serial 10-fold dilutions of HRV-14 suspension were prepared with medium (EMEM) supplemented with 5% FCS. The TCID50 was then calculated using the Karber formula [28]: TCID50 = L − d (S/n− 0·5), where L = log10 of lowest dilution, d = log10 of difference between the highest and lowest dilutions, S= number of positive tests, n= number of duplicate tests per dilution.

Study cultures

All cultures described below were carried out in triplicate.

Expression of ICAM-1 on HNEC In order to analyse baseline surface ICAM-1 expression, cultures of HNEC from 13 atopic subjects in and out of season and 11 non-atopic normal subjects were set up in collagen-precoated 24-well plates (ICN Flow Labs, Basingstoke, UK) using Hams F-12 6X medium. Cells were seeded at a density of 1 × 105 cells/well and allowed to reach confluence. Once confluent, HNEC were retrieved gently using a rubber policeman for analysis of surface ICAM-1 expression.

Effect of allergen on HNEC Cultures, set up as described above, were incubated with clinically relevant and irrelevant allergen (6 grass mix 10 μg/ml and house dust mite 15 μg/ml, respectively; ALK, Reading, UK), for 24 h to assess influence of aeroallergens on surface ICAM-1 expression of HNEC in the above in and out of season atopic subjects and normal controls. The concentration of allergen used represented the respective optimum determined on ex vivo EC. After 24 h the cells were retrieved and cytospins prepared for ICAM-1 evaluation.

Effect of rhinovirus infection on HNEC and nasal polyp EC Previous studies [10,13] using H292 epithelial cell lines have shown that HRV infection itself enhances cell surface ICAM-1, and that maximum HRV replication occurs between days 0 and 8. Thus, the influence of HRV-14 infection on surface ICAM-1 expression of primary epithelial cells was evaluated over 8 days, utilizing nasal polyp EC from three known atopic asthmatics; for normal control samples we pooled HNEC from three healthy non-atopic subjects. Medium was removed from confluent monolayers of epithelial cells which were then inoculated with 1 ml of HRV-14 (102·5 TCID50/ml). After a 90-min incubation at 34°C/5% CO2 in air the inoculum was removed and cells washed. Viral titres of the supernatant retrieved at the end of the period of infection (90 min) were assessed to calculate the viral uptake by the EC during the period of infection. The viral titre of inocula after incubation in EC-free wells acted as a control to assess non-specific adherence of the virus to the wells. The medium was replaced by 2 ml of serum-free Hams F-12 6X medium and incubation continued at 34°C. For each time point separate replicate wells were set up. Medium was not replaced further until supernatants were recovered at the previously listed time points after infection and stored at −70°C for viral titre analysis. Cells were recovered, counted, and then viability assessed by trypan blue exclusion. Cytospins for immunocytochemistry were made using a cell suspension at 1 × 104 cells/ml.

Effect of rhinovirus infection on allergen-conditioned nasal polyp EC We investigated the influence of HRV-14 infection on surface ICAM-1 expression in allergen prestimulated cells.

Nasal polyp EC were prestimulated with optimal concentrations of grass pollen (10 μg/ml) for 24 h; medium was removed and the cell monolayer washed three times. The cells were then infected with HRV-14 for 90 min. Control HNEC cultures were prestimulated with the same allergen alone or infected with HRV-14 alone. Cells and supernatants were retrieved at days 0, 1, 4, 6 and 8 in order to assess surface ICAM-1 expression and viral titres performed to assess viral replication.

Statistical analysis

Non-parametric statistical methods were used for analysing the data in this study for reasons of small sample size. The Mann–Whitney test was used when comparing the data of two independent samples. Statistical significance was quoted for P < 0·05 (the 5% probability level) and two-tailed hypothesis testing was applied. The statistical software package used for calculation was SPSS version 7 for Windows 95 [29].


Analysis of EC

Cell viability in each HNEC and nasal polyp sample was always > 95%. Immunostaining of cells immediately after sampling confirmed that > 90% of cells stained positive for cytokeratin.

Experiments using HNEC

Basal ICAM-1 expression HNEC from atopic subjects both in and out of season showed significantly increased baseline ICAM-1 expression compared with normal non-atopic controls (P < 0·04 and P < 0·05, respectively; Fig. 1). ICAM-1 expression on atopic HNEC recovered out of season was significantly lower than on atopic HNEC recovered in season (P < 0·05). In each case, the change in ICAM-1 expression titrated with the concentration of grass allergen used. The data in Fig. 1, and subsequently, show the maximum responses obtained.

Figure 1.

Comparison of basal ICAM-1 expression (▪) and ICAM-1 expression in response to relevant grass allergen (hatched) and irrelevant mite allergen (□) in normal human nasal epithelial cells (HNEC) and in/out of season atopic HNEC. Atopic HNEC both in and out of season had higher basal ICAM-1 levels compared with normal controls (*P < 0·04 and **P < 0·05, respectively). ICAM-1 expression on atopic HNEC recovered out of season was significantly lower than atopic HNEC recovered in season (†P < 0·05). Data are expressed as ± s.e.m.

Allergen-induced ICAM-1 expression No increase in basal expression was seen with HNEC from normal subjects cultured with grass pollen or mite allergens for 24 h. HNEC from atopic subjects recovered in season and out of season showed significant increases in ICAM-1 expression after culture with clinically relevant allergen (grass) (P = 0·04, P = 0·032, respectively), but not with clinically irrelevant (mite) allergens ( Fig. 1).

HRV-14-induced ICAM-1 expression These experiments were only performed with HNEC from normal subjects. There was a steep four-fold increase in ICAM-1 expression from day 0 to day 4 post-infection, with a further gradual rise between days 4 and 8 ( Fig. 2). The level of ICAM-1 expression on day 8 was significantly higher than on day 4 (P < 0·05).

Figure 2.

The effect of HRV-14 infection on ICAM-1 expression on human nasal epithelial cells (HNEC) from normal subjects. □, ICAM-1 expression in uninfected HNEC; ▪, virus-infected HNEC. A steep four-fold up-regulation of ICAM-1 expression is observed from day 0 to day 4 post-infection, with a further gradual rise from day 4 to day 8 (*P < 0·05). Data are expressed as ± s.e.m.

Experiments using nasal polyp EC

Basal ICAM-1 expression on atopic nasal polyp EC was much higher than on control HNEC, with POX scores reaching values of 800 ( Table 1). Culture with clinically relevant grass allergen for 24 h produced a small non-significant increase. Infection with HRV produced a small significant increase (P < 0·05; Table 2). Allergen pretreatment for 24 h prior to HRV infection did not produce any further increases in ICAM-1 expression over that produced by HRV alone ( Table 2).

Table 1.  The effect of clinically relevant grass and irrelevant house dust mite allergens on surface ICAM-1 expression of nasal polyp epithelial cells (EC)
  1.   Basal expression of ICAM-1 in nasal polyp EC was 16-fold greater than basal expression of control human nasal epithelial cells (HNEC) (POX score 42·9 ± 5·6). There was a small increase approaching significance in nasal polyp EC ICAM-1 expression in response to stimulation with grass (P = 0·06), but no effect with house dust mite. Data are expressed as ± s.e.m.

POX score800 ± 6·54 850 ± 7·22800 ± 8·6
Table 2.  The effect of HRV-14 infection on ICAM-1 expression in unstimulated and grass allergen-conditioned nasal polyp epithelial cells (EC)
 BaselineHRV-14Grass + HRV-14
  1.   HRV-14 alone induces a small increase in ICAM-1 expression on day 1, reaching maximal up-regulation on day 8 (*P < 0·05). No additional increase in ICAM-1 levels was seen in allergen-stimulated cell cultures. Data are expressed as ± s.e.m.

Day 0800 ± 2·9790 ± 4·7800 ± 8·2
Day 1800 ± 4·7820 ± 5·7820 ± 13·1
Day 4800 ± 5·2840 ± 2·1840 ± 7·8
Day 8800 ± 4860 ± 3·1 *860 ± 8

Recovered viral titres were assayed after infection of atopic nasal polyp EC and a pool of three normal HNEC. Although nasal polyp and HNEC represent different biological sources, a comparison is considered relevant since recovered viral titres from the two sources on day 0 were similar. From 1 to 8 days after HRV infection, viral titres were consistently higher in supernatants from the polyps, although this only reached significance on day 8 ( Fig. 3). Pretreatment of nasal polyp EC with grass allergen prior to viral infection had no significant effects on recovered viral titres (data not shown).

Figure 3.

Time course study of recovered viral titres (TCID50/ml) from HRV-14-infected human nasal epithelial cells (HNEC) from normal subjects (□) and nasal polyp epithelial cells (EC; ▪). A steep increase in viral titres from day 0 to day 1 was seen in both cell types (†P < 0·01). From day 1, viral titres recovered from normal HNEC increased more gradually, reaching significance on days 6 and 8 (*P < 0·05); viral titres recovered from nasal polyp EC continued to rise steeply from day 1 to day 8. At day 8, nasal polyp EC viral titres were significantly greater than those from normal HNEC (**P < 0·05). Data are expressed as ± s.e.m.


The present study shows that expression of ICAM-1 is elevated on HNEC from atopic subjects compared with controls, and is higher on in-season than on out-of-season samples. The interpretation that this chronic elevation of ICAM-1 on atopic samples is allergen-driven is supported by the data showing that culture of atopic HNEC with clinically relevant allergen in vitro further enhances ICAM-1 levels. ICAM-1 expression can also be enhanced by HRV infection [10,11,13], suggesting that allergen and rhinoviruses may have synergistic effects in the atopic host. Thus allergen exposure, by enhancing expression of the principal rhinoviral cell attachment, could influence susceptibility of atopic subjects to HRV infection. In addition, chronic elevation of ICAM-1 by the combined effects of allergen and HRV infection would favour migration of immune effector cells into the airway. Since many of these cells in the atopic airway would be committed Th2 and B cells [30,31], the enhanced local production of Th2-associated cytokines and IgE may accentuate the atopic state [32,33].

Increased ICAM-1 expression following exposure to relevant but not irrelevant allergens is surprising, given that 90% of cells in the HNEC samples were cytokeratin-positive. However, there are likely to be residual T cells within the samples, which may release cytokines when re-stimulated with relevant allergen. Indeed, previous studies have shown that Th2-associated cytokines, such as IL-13 [10,13,20], are particularly potent stimulators of ICAM-1 expression on epithelial cells.

In contrast, only relatively small increases in ICAM-1 expression on atopic nasal polyp EC were observed following HRV infection, and less so with allergen challenge. We are unable immediately to account for this disparity between results obtained with HNEC and polyp EC. However, the baseline POX scores of polyp EC were among the highest we have recorded, even with cytokine-stimulated epithelial cell lines in vitro[10,13], and it may be that the baseline ICAM-1 levels on polyp EC represent the maximum expression that can be achieved.

The increased ICAM-1 expression following HRV infection of primary HNEC also confirms our findings with bronchial epithelial cell lines. In experiments utilizing cell lines, cytokine-induced increases in ICAM-1 were associated with increased recovered viral titres following HRV infection compared with non-cytokine-treated cells [10,13]. In the present study we attempted to assess whether this was also true using ex vivo material. Nasal brushing provides too few cells to perform these viral recovery experiments, so we took advantage of the availability of the larger numbers of cells obtainable from nasal polyps surgically removed from atopic subjects. Atopic polyp EC express very high levels of ICAM-1, which may be due to their increased content of Th2 cells and eosinophils expressing IL-4, IL-5 and IL-13, together with RANTES [19,23,30,34]. Indeed, nasal polyp ICAM-1 expression, largely on ciliated and basal epithelial and submucosal cells, has been correlated with their eosinophil content [32,33]. While pooled HNEC are not an ideal control, non-atopic polyp EC were not available for study; however, the starting recovered HRV titres on day 0 from polyps and HNEC were similar. With this proviso, HRV titres from infected nasal polyp EC were significantly higher 8 days after infection than from HNEC.

We hypothesize that the increased basal levels of ICAM-1 on atopic HNEC, which can be further enhanced by exogenous allergen in vitro and by environmental allergen in vivo, provide an increased array of attachment sites for HRV and thus enhance the likelihood of host cell infection. Further augmentation of ICAM-1 expression by HRV itself could exacerbate and prolong viral infection in atopic subjects. In addition, by increasing sites for interaction with inflammatory cell LFA-1, up-regulated ICAM-1 levels promote recruitment and activation of inflammatory cells within the virally infected airway [16]. Since there is a pre-existing predominance of Th2-like cells in atopic airway disease [20], the HRV-induced ICAM-1 expression and the consequent ICAM-1/LFA-1 interactions are likely to recruit and activate increased proportions of this cell type [16]. Increased release of Th2 cytokines by the recruited cells will both still further increase adhesion molecule expression, reinforcing HRV attachment and Th2 cell migration, culminating in clinical exacerbation of atopic/asthmatic symptoms. Whilst this hypothesis is plausible, and supported by the epidemiological observations of close association between HRV infection and asthma exacerbations, the scenario in vivo may also be influenced by the severity of the host's atopic/asthmatic status and the magnitude of immune response to infection. Fleming et al. [35] recently showed that the airway inflammatory responses to nasal inoculation with HRV16 were similar in very mild asthmatics and healthy subjects; in addition, only 5% of the infected asthmatic subjects suffered increased symptoms that were clinically relevant. These data suggest that viral infection alone would not be sufficient to provoke worsening asthma symptoms. Under such circumstances we postulate that the host response to invading viral infection may result in an over-riding Th2→Th1 switch, with enhanced local production of interferon-gamma (IFN-γ). Indeed, we have shown that whilst IFN-γ increases ICAM-1 expression, in the presence of HRV infection this Th-1 mediator down-regulates ICAM-1 levels on infected cells, thereby decreasing available cellular binding sites for viral attachment and limiting host infectivity [10,13]. Clearly, more detailed studies are now required to define these phenomena in vivo.


Mr Tim Little (ENT Department, North Staffordshire Hospital Trust), Mr Richard Skinner (Ear Eye Throat Hospital, Shrewsbury) and Dr Salvatore Rotondetto (Department of Respiratory Medicine North Staffordshire Hospital Trust). Supported by the British Medical Association TV James Fellowship (1994) to M.A.S. and the European Respiratory Society Scientific Research Fellowship (1996) to A.B.