OUP user menu

Infection with respiratory syncytial virus and water-soluble components of cigarette smoke alter production of tumour necrosis factor α and nitric oxide by human blood monocytes

Muhammad W. Raza, Steve D. Essery, Donald M. Weir, Marie M. Ogilvie, Robert A. Elton, C. Caroline Blackwell
DOI: http://dx.doi.org/10.1111/j.1574-695X.1999.tb01310.x 387-394 First published online: 1 July 1999


Cigarette smoke and virus infections contribute to the pathogenesis and exacerbation of chronic obstructive pulmonary disease and asthma. The objective of this study was to examine the effects of a water-soluble cigarette smoke extract (CSE) and/or respiratory syncytial virus (RSV) infection on release from monocytes of the blood from donors of tumour necrosis factor a (α) and nitric oxide (NO). Both RSV infection and CSE stimulated α release from monocytes and there was an additive effect if both the agents were present. There was a decrease in NO release, but the effect was significant only with CSE or a combination of CSE and RSV infection. Interferon γ significantly increased α release and cotinine significantly increased NO release. Nicotine decreased both α and NO responses. The general pattern observed for individual donors was increased α and decreased NO. The proportion of extreme responses with very high α and very low NO in the presence of both RSV and CSE increased to 20% compared with 5% observed with CSE or RSV alone. The results show that RSV infection and components of cigarette smoke elicit inflammatory responses that could contribute to damage to the respiratory tract and these individual factors could be more harmful in combination.

  • Cigarette smoke
  • Respiratory syncytial virus
  • Tumor necrosis factor
  • Monocyte
  • Inflammation

1 Introduction

Bacterial and viral infections and non-infectious air pollutants such as cigarette smoke are important determinants in the pathogenesis of respiratory disease. Their influence on the inflammatory and immune responses underlies the pathological processes in the respiratory tissues. The factors that determine whether the disease is mild and short-lived or severe and chronic are not clear. A number of reports have examined the release of inflammatory mediators from alveolar macrophages and there have been clinical and experimental studies on smoking and virus infection as contributory factors to chronic obstructive pulmonary disease (COPD) [1,2]. Both tumour necrosis factor a (α) and nitric oxide (NO) are important inflammatory mediators in COPD and asthma [35]. Infection with respiratory syncytial virus (RSV) induces release of α and NO from human alveolar macrophages, bovine peripheral blood mononuclear cells and a murine monocyte cell line [68]. There are, however, conflicting reports on the effect of smoking on α release [5,9,10]. Both episodic and habitual smoking reduced NO exhalation [11]; but, in pigs challenged with cigarette smoke, a vasodilator response due to NO release was recorded [12].

Both blood monocytes and alveolar macrophages can be infected with respiratory syncytial virus (RSV) [13,14], and both cell types are expected to be exposed to water-soluble components of cigarette smoke absorbed across mucosal surfaces. The objective of the present study was to evaluate release of α and NO from human blood monocytes challenged with either RSV, a water-soluble cigarette smoke extract (CSE) or both. Since many virus infections stimulate release of interferon γ (γ) [15] that might in turn mediate other secretory functions, the effect of γ on release of α and NO in this system was also analysed. Nicotine is metabolised in the liver to cotinine which is secreted in body fluids including those of the respiratory tract [16]; therefore, the effects of nicotine and cotinine on α and NO release were also examined.

2 Materials and methods

2.1 Preparation of cigarette smoke extract

A water-soluble extract of cigarette smoke (CSE) was prepared by the use of a vacuum pump to pass the smoke of 10 filter-tipped cigarettes (Benson and Hedges) through 100 ml of Dulbecco′s modified Eagle′s medium (DMEM; Gibco) containing 0.45% (w/v) d-glucose and 4 mg l pyridoxin [17]. To reduce lipopolysaccharide contamination, the glass bottles were heated at 134°C for 1 h. The CSE was sterilised by filtration and aliquots were stored at −20°C.

2.2 Preparation of RSV stock

The Edinburgh strain of RSV (subgroup A) was grown in HEp-2 cells (an epithelial cell line) as described previously [18] except that the virus was harvested in growth medium (GM) containing DMEM supplemented with 10% (v/v) foetal calf serum (Gibco), 2 mm ll-glutamine (Gibco), 200 µg ml streptomycin (Gibco) and 100 IU ml penicillin (Gibco). The virus suspension was adjusted to 2×106 plaque-forming units ml.

2.3 Separation and stimulation of monocytes from blood

One-day-old buffy coats from the blood of group O, Rh+ donors were obtained from The Scottish National Blood Transfusion Service (SNBTS), Royal Infirmary, Edinburgh. Buffy coats were diluted 1 in 4 with sterile phosphate-buffered saline (PBS). The diluted preparations (40 ml) were layered over 12 ml of Histopaque 1077 (Sigma) in 50-ml polypropylene conical tubes (Greiner, Gloucestershire, UK). The tubes were centrifuged at 400×g for 30 min at 25°C. Mononuclear cells, in the opaque band formed at the interface of serum and Histopaque, were aspirated carefully. The cells were washed twice in sterile PBS at 150×g for 10 min and the supernatant fluid was discarded. The cells were resuspended in 20 ml of GM, transferred to a 75-cm2 tissue culture flask (Greiner) and incubated at 37°C for 30 min to separate monocytes from non-adherent cells. The medium containing non-adherent cells was poured off and the monocytes harvested by gentle scraping with a cell scraper in 20 ml of fresh GM. A viable count was performed by the trypan blue dye exclusion method and the concentration of monocytes adjusted to 1×108 ml in GM. The cells (1 ml) were distributed in 24-well tissue culture plates (Costar) with 1×106 cells in each well. Viability of cells at the end of each experiment was also assessed by microscopic examination of cells in the wells for exclusion of trypan blue.

The cells were cultured at 37°C in 5% CO2 in 1 ml of GM, GM with CSE at various dilutions, RSV at a multiplicity of infection (MOI) of 2.0 [6], or with a combination of RSV and CSE. Some cells were also exposed to γ (Sigma), nicotine (Sigma) or cotinine (Sigma) ranging from 25 to 400 ng ml. Samples were collected from each well after 48 h for determination of α (100 µl) and after 72 h for determination of NO (400 µl). Negative control samples to which no cells were added included culture medium alone, medium with CSE and/or RSV, or with γ, nicotine or cotinine. The samples were stored at −20°C until analysed.

The proportion of monocytes infected with RSV in each sample 24 h post-infection was determined by an indirect immunofluorescence technique with monoclonal antibody to glycoprotein G of RSV [18].

2.4 TNF bioassay and detection of NO

L-929 cells (a mouse fibroblast cell line) were used in a bioassay for α activity [19] and the results expressed as per cent cytotoxicity [20]. NO was detected as nitrites by the spectrophotometric assay described by Zhang et al. [21]. The samples were clarified by centrifugation at 12 600×g with a microcentrifuge (Sorval MC 12C, Dupont) for 5 min. Supernatant fluids (400 µl) were reacted with equal volumes of Greiss reagent which contained 0.3% (w/v) naphthylethylenediamine dihydrochloride (Sigma), and 1% (w/v) sulfanilamide (Sigma) in 5% (v/v) orthophosphoric acid (BDH), mixed 1:1 immediately before use. After incubation for 10 min at room temperature, the absorbance at 540 nm was determined in a spectrophotometer (Jeway 6100). Concentrations of nitrites in the samples were derived from a standard curve for sodium nitrite prepared for each experiment.

2.5 Statistical methods

The results obtained with buffy coats from 24 donors are presented here. The results from some samples for some treatments could not be presented because of contamination in individual wells; therefore, the mean control values corresponding to different experiments were not the same. The data from monocytes treated with different agents were compared with those from the monocytes incubated with medium only. Comparisons were made also between different treatment groups. The significance values obtained with paired t-tests of the data were similar to values obtained with a non-parametric test (Wilcoxon′s). The values obtained from t-tests are presented here.

3 Results

3.1 RSV infection of cells

On average, more than 40% of monocytes from each individual tested were infected with RSV at MOI of 2.0 in these assays. One-way — analysis of variance—indicated no significant differences in the proportion of RSV-infected cells among the donors.

3.2 Standardisation of the assay

Ten-fold dilutions of CSE ranging from smoke of 0.1 to 0.0001 cigarette ml were tested with monocytes from four donors, and a dilution of 0.001 cigarette ml was selected for the assays on the basis of maximum effects on the production of α and NO without killing the monocytes (Fig. 1). Doubling dilutions of γ, nicotine and cotinine ranging from 400 ng ml to 6.25 ng ml were tested [22]. A dose of 25 ng ml for these reagents was used for further study (data not shown). More than 90% of monocytes survived until the end of the experiments under the conditions selected for the assays.

Figure 1

Per cent viability of monocytes (▲), per cent cytotoxicity of L-929 cells due to α (■) and levels of sodium nitrite (solid line) produced by monocytes incubated for 24 h with various dilutions of cigarette smoke extract (CSE).

Time course experiments with monocytes from four donors (6–72 h) found the maximal α bioactivity occurred at 48 h and nitrite accumulation at 72 h in response to CSE or RSV (data not shown). α bioactivity was not detected in control samples without cells containing culture medium or medium with CSE, RSV, CSE and RSV, γ, nicotine or cotinine. For the detection of nitrites, the spectrophotometer was blanked on these individual controls for assessment of their respective test samples. No effect due to presence of these agents was recorded at the absorbance used to detect sodium nitrite.

3.3 The effect of CSE and RSV infection on TNF-α bioactivity

The α bioactivities expressed as per cent cytotoxicity of L-929 cells observed in experiments with monocytes from a total of 24 donors were compared. Fig. 2 represents paired differences in L-929 cytotoxicity due to α bioactivities caused by different treatments of monocytes compared with monocytes incubated with medium only. Compared with cell culture fluids from cells (24 donors) incubated with medium alone (mean 48%, S.E.M. 5.2), cell culture fluids from cells incubated with CSE (mean 60%, S.E.M. 5.1) had significantly increased α bioactivities (95% CI of paired differences 3.7, 19.7, t=3.05, P=0.006) as did the cell culture fluids from RSV-infected cells (mean 68%, S.E.M. 4.6) (95% CI of paired differences 16.4, 34.15, t=5.9, P=0.000). There was no correlation between the per cent cytotoxicity for L-929 cells and proportions of RSV-infected cells in RSV-infected samples. Compared with α bioactivities detected in cell culture fluids from cells (15 donors) exposed to medium alone (mean 38%, S.E.M. 5.4), a significant increase was observed in cell culture fluids from cells incubated with both CSE and RSV (mean 71%, S.E.M. 6.6) (95% CI of paired differences 15.4, 49.5, t=4.08, P=0.002).

Figure 2

Paired difference in α bioactivity expressed as % cytotoxicity for L-929 cells from monocytes of individual donors exposed to medium only or to CSE and/or RSV. Error bars represent S.E.M.

Compared with α bioactivities of monocytes from 15 donors exposed to CSE alone, the bioactivities observed with the combination of CSE and RSV infection were significantly higher (paired differences 26%, 95% CI 10.4, 41.3, t=3.58, P=0.003). Compared with α bioactivity found for cells incubated only with RSV, addition of CSE did not significantly increase α bioactivities (95% CI −12.9, 21.5). This indicates that the main contribution to increased levels of α was due to the virus infection.

Compared with α bioactivity of monocytes from six donors exposed to medium only (mean 32%, S.E.M. 7.5), there was no significant increase observed in cell culture fluids of cells incubated with nicotine (mean 34%, S.E.M. 8.7) (95% CI of paired differences −14.2, 15.4) or cotinine (mean 39%, S.E.M. 7.2) (95% CI of paired differences −1.6, 15.2). Incubation with γ did, however, result in significantly increased α bioactivity (mean 51%, S.E.M. 3.0) (95% CI of paired differences 6.2, 31.7, t=3.83, P=0.012).

3.4 The effect of CSE and RSV infection on NO release from monocytes

The supernatant fluids from cells in the same experiments were examined for nitrite levels. Paired differences between nitrite levels resulting from different treatments of monocytes compared with those incubated with medium alone are given in Fig. 3. Compared with supernatant fluids from cells incubated with medium only (mean 0.41 nM, S.E.M. 0.09), supernatant fluids from cells incubated with CSE (mean 0.34 nM, S.E.M. 0.08) had significantly lower levels of nitrite (95% CI of paired differences −0.14,-0.008, t=2.3, P=0.031) as did RSV-infected cells (mean 0.35 nM, S.E.M. 0.05), but the results were not significant (95% CI of paired differences −0.023, 0.099). There was no correlation between the levels of nitrite detected and ratios of RSV-infected cells in the samples. Compared with cells incubated with medium only, cells incubated with both CSE and RSV showed a significant decrease in nitrite production (mean 0.25 nM, S.E.M. 0.06) (95% CI of paired differences −0.4, −0.01, t=2.26, P=0.04).

Figure 3

Paired difference in sodium nitrite production from monocytes of individual donors exposed to medium only or to CSE and/or RSV. Error bars represent S.E.M.

In experiments with monocytes from 15 donors, there was no significant difference between nitrite levels found for cells incubated with CSE alone compared with those incubated with both CSE and RSV. In comparison with nitrite detected in supernatants from cells incubated with RSV only, supernatants from RSV-infected cultures containing CSE had lower levels of nitrite with marginal significance (paired difference −0.098 nM, S.E.M. 0.048, 95% CI −0.006, −0.2, t=2.02, P=0.063). This indicated that the main decrease in nitrite levels was caused by CSE.

Experiments with cells from six individual donors showed that, compared with nitrite levels observed with cells incubated with medium only (mean 0.10 nM, S.E.M. 0.03), addition of INF-γ (mean 0.15 nM, S.E.M. 0.03) (95% CI 0, 0.057) or nicotine (mean 0.13 nM, S.E.M. 0.03) (95% CI 0, 0.03) to cells did not result in significant changes in nitrite release, but addition of cotinine significantly increased release of nitrite from cells (mean 0.26 nM, S.E.M. 0.03) (95% CI of paired differences 0.03, 0.15, t=3.48, P=0.018).

3.5 Variability of TNF-α and NO response of individual donors

Each set of experiments was carried out with cells from different donors. Individual α and NO responses to CSE, RSV infection and combinations of both are summarised in Table 1. α and NO responses of the test samples were arbitrarily classified as very high if the levels of killing of L-929 cells or levels of nitrite were more than twice the value of the controls in which monocytes were cultured with medium alone. Responses were classified as very low if the levels of test samples were less than half the control. The most common pattern observed was increased α bioactivity and decreased NO production compared with controls incubated with medium only. In the presence of CSE or RSV, 4% of donors exhibited very high α bioactivity and very low nitrite levels. In the presence of both CSE and RSV, this rose to 20%.

View this table:
Table 1

Individual responses to CSE and/or RSV

Response toPer cent of donorsPer cent of donors with extreme response
  • High ≥2×control; low ≤1/2 control.

4 Discussion

RSV is a common pathogen affecting infants and the elderly [23]. Complete immunity does not follow exposure to RSV, hence reinfections are not uncommon [24]. Significant numbers of patients with COPD or bronchial asthma suffer bouts of exacerbation and possible residual effects due to RSV infection in the course of their disease [25 (abstract)]. Cigarette smoking is a major cause of COPD [1]. Both viral infection and cigarette smoking enhance bacterial binding to epithelial cells in model systems [18,26,27] and cause immunosuppression [28,29]. Since α and NO are important mediators of inflammation in the respiratory tract, α and NO responses of blood monocytes to RSV infection and CSE were assessed.

Peripheral blood monocytes were used in the study for four reasons: (1) their ready availability in sufficient numbers; (2) alveolar macrophages in the lungs are derived from monocytes; (3) they were less likely to have been exposed to respiratory viruses and air pollutants; (4) they were more likely to be in an unstimulated state. Some stimulation of monocytes due to the use of Histopaque for isolation of monocytes could not be ruled out. There is evidence that genetic factors contribute to levels of α in response to endotoxins [30]. The variable spontaneous α and NO release from monocytes could be attributed to the individual donor′s genetic make up or condition at the time of blood donation, their smoking habits, or variable responses of monocytes to Histopaque. The assays examined the responses to various agents in relation to background levels of each individual donor.

The dose of CSE (0.001 cigarette ml) used in the assays was similar to dilutions of smoke extract used in experiments with alveolar macrophages by Higashimoto et al. [17]. It was based on the range of numbers of cigarettes an average person can smoke and the water-soluble components of the inhaled smoke that cross the mucosal lining and are diluted in the body fluids.

Bioactivity of α in samples can differ from total α detected by ELISA because of the soluble α receptors produced by monocytes that block the functions of the cytokine. There was no correlation between the data from the bioassay and the ELISA with 200 samples (data not shown). This could partly be due to variable levels of α receptors in the cell culture fluid and partly to variable rate of degradation of α by the monocytes. The bioassay was selected for this study because it measures the levels of α activity in the solution at a given time.

Midulla et al. found RSV infection elicited variable α responses from alveolar macrophages from individual donors [31]. Cells from most of the donors in the present study showed increased α responses to RSV infection or exposure to CSE. A minority exhibited very high responses to either of the agents. α is thought to play a protective role in RSV infection. Prior incubation with α reduced the replication of RSV in alveolar macrophages by about half [32]. Cells from 8% of donors in this study exhibited a decreased α response to RSV infection. Individuals with this response might be particularly susceptible to severe infection RSV.

The beneficial pulmonary vasodilatory, possible bronchodilatory, and bactericidal effects of NO in patients with COPD or asthma might be offset by its induction of exudate formation, inflammation, DNA toxicity and cytotoxicity. It is generally agreed that mild NO induction is protective in the respiratory tract while higher levels might be associated with deleterious consequences [3,33]. Two distinct populations of donors based on the level of NO production have been recognised [34]. The assay for nitrite used in this study was sensitive down to NaNO2 concentrations of approximately 1 pM. Results presented here indicate that increased responses of α are not, in most donors, accompanied by increased NO responses to RSV or CSE. α reduced the half-life of mRNA encoding NO synthase in human umbilical vein endothelial cells [35]. The results indicate that α activity might also affect the production of this enzyme in monocytes.

Some of the effects of virus infections are mediated through release of γ from monocytes [36]. The present data indicated that stimulatory effects of RSV infection on α bioactivity could be due in part to γ. Significant increases in α and significant decreases in NO response in the presence of CSE did not match with the observed slight increase in both responses mediated by nicotine alone. This suggests other factors in CSE are responsible for the effects observed in these assays. The significant increase of NO from cells treated with cotinine indicated that some of the effects of cigarette smoking on inflammatory mediators in vivo might be mediated by this metabolite of nicotine.

The results presented here examined some of the effects of two environmental factors that exacerbate COPD and asthma. Smoking or passive exposure to cigarette smoke and virus infections of the respiratory tract do not always lead to similar degrees of acute or chronic illness. This could reflect the individual differences in responses observed in this study. In addition to enhancing bacterial colonisation of the respiratory mucosa and immunosuppression, these agents appear generally to enhance α response and reduce NO levels. The extreme responses noted with cells from a minority of subjects might contribute to increased susceptibility to chronic inflammatory disease of the respiratory tract or exacerbations. Comparison of monocytes from healthy donors with those from patients with these conditions for production of inflammatory cytokines is needed to obtain further evidence for this hypothesis.


This work was supported by the Chest, Heart and Stroke Association, Scotland.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
  36. [36].
View Abstract