Published ahead of print on October 5, 2006, doi:10.1164/rccm.200604-561OC
© 2006 American Thoracic Society doi: 10.1164/rccm.200604-561OC
Cigarette Smoke Impacts Immune Inflammatory Responses to Influenza in MiceDepartment of Pathology and Molecular Medicine, Centre for Gene Therapeutics, Department of Biology, and Department of Medicine, McMaster University, Hamilton; and Department of Biochemistry, Microbiology, and Immunology, University of Ottawa, Ottawa, Ontario, Canada Correspondence and requests for reprints should be addressed to Martin R. Stämpfli, Ph.D., McMaster University, Department of Pathology and Molecular Medicine, MDCL, Room 4011, 1200 Main Street West, Hamilton, ON, L8N 3Z5 Canada. E-mail: stampfli{at}mcmaster.ca
Rationale: Studies have shown that cigarette smoke impacts respiratory host defense mechanisms; however, it is poorly understood how these smoke-induced changes impact the overall ability of the host to deal with pathogenic agents. Objective: The objective of this study was to investigate the impact of mainstream cigarette smoke exposure on immune inflammatory responses and viral burden after respiratory infection with influenza A. Methods: C57BL/6 mice were sham- or smoke-exposed for 3 to 5 mo and infected with either 2.5 x 103 pfu (low dose) or 2.5 x 105 pfu (high dose) influenza virus.
Measurements and Main Results: Although smoke exposure attenuated the airway's inflammatory response to low-dose infection, we observed increased inflammation in smoke-exposed compared with sham-exposed mice after infection with high-dose influenza, despite a similar rate of viral clearance. The heightened inflammatory response was associated with increased expression of tumor necrosis factor- Conclusion: Our study suggests that, in mice, cigarette smoke affects primary antiviral immune-inflammatory responses, whereas secondary immune protection remains intact.
Key Words: chronic obstructive pulmonary disease • inflammation • immunity • neutrophils
Epidemiologic studies clearly show that smoking is associated with an increased incidence of both upper and lower respiratory tract infections (1). For example, Aronson and colleagues reported that young smokers were at significantly greater risk of acquiring lower respiratory tract illness, with a longer duration of cough than nonsmokers (2). Furthermore, chronic obstructive pulmonary disease (COPD), a condition largely associated with cigarette smoking, is associated with periods of acute exacerbation of symptoms largely due to bacterial and viral infections (3–7). Together, these findings suggest that cigarette smoking may alter the way respiratory pathogenic microorganisms are handled. Multiple host defense mechanisms are involved in protecting the lung from the potentially harmful effects of infection. These include physical barriers as well as a complex network of innate and adaptive immune mechanisms (8–10). That exposure to tobacco smoke impacts a number of these processes is suggested by studies demonstrating that cigarette smoke impairs mucociliary clearance and damages epithelial cell tight junctions (11, 12) (reviewed in Reference 13). Other studies have shown that cigarette smoke impacts numerous cell types of the immune system, including bronchial epithelial cells (14), alveolar macrophages (15–18), natural killer cells (19–21), dendritic cells (22, 23), and B and T lymphocytes (24–26). It is poorly understood, however, how these smoke-induced changes impact the overall ability of the host to deal with pathogenic agents. In the present study, we assessed the impact of mainstream tobacco smoke (MTS) on the course of viral infection with a replication-competent, mouse-adapted influenza A virus. Influenza infection is a significant concern for smoking populations. Nicholson and colleagues reported influenza infection in 23% of smokers compared with 6% in nonsmokers in a nonimmunized population between the ages of 60 and 90 yr (27). In another 3-yr study, 20% of reported hospital admissions for acute exacerbation of COPD presented with positive serology for influenza A (28). In addition to smoking cessation, annual influenza vaccination is an important measure for preventing COPD exacerbation. We demonstrate that MTS differentially affects airway inflammatory responses depending on the initial infectious dose. Specifically, we observed decreased inflammation in the lungs of mice infected with 2.5 x 103 pfu of influenza virus. In contrast, MTS exposure exacerbated airways inflammation and induced death after infection with a 2-log higher infectious dose (2.5 x 105 pfu), despite a similar lung viral burden as in sham-exposed mice. Importantly, MTS exposure did not interfere with the induction of protective antiviral immunity. Our study suggests that inappropriate inflammatory responses to viral agents may in part contribute to the inflammation observed in smokers and the decline in health status associated with COPD exacerbations. Some of the results in this study were previously reported in abstract form (29).
Animals Female C57BL/6 mice (6–8-wk-old) were purchased from Charles River Laboratories (Montreal, PQ, Canada). All experiments described in this study were approved by the McMaster University (Hamilton, ON, Canada) Animal Research Ethics Board.
Cigarette Smoke Exposure
Influenza Infection
Collection and Measurement of Specimens
Myeloperoxidase Activity
Measurements of Cytokines and Immunoglobulins
Plaque Reduction Assay
Measurement of Viral Titers in the Lung
Flow Cytometry
Innate Immunostimulation with Polyinosine-Polycytidylic Acid
Statistical Analysis
Clinical Outcome To assess the impact of MTS on the clinical course of influenza infection, sham- and MTS-exposed mice were infected with either low- (2.5 x 103 pfu) or high- (2.5 x 105 pfu) dose influenza A virus. Neither sham- nor MTS-exposed mice displayed overt clinical symptoms after infection with low-dose virus. On the other hand, infection with high-dose influenza resulted in a progressive worsening of clinical symptoms in both sham- and MTS-exposed animals (data not shown). Animals became ruffled, hypomobile, and displayed evidence of labored breathing. Importantly, three of eight MTS-exposed animals died between 6 and 7 d after infection (Table 1). None of the sham-exposed mice died.
BAL and Peripheral Blood Inflammatory Profile Infection of sham-exposed mice with either low- or high-dose influenza virus resulted in the induction of significant airways inflammation (Figure 1). Unlike the lower infectious dose, infection with 2.5 x 105 pfu influenza was associated with a more pronounced BAL neutrophilia. Whereas MTS attenuated the inflammatory response after low-dose viral challenge, we observed increased inflammation in MTS- compared with sham-exposed mice after high-dose infection. Specifically, 7 d after low-dose infection, we detected significantly fewer mononuclear cells and a trend toward decreased numbers of neutrophils in MTS- compared with sham-exposed mice. In contrast, infection with high-dose influenza A virus resulted in increased numbers of inflammatory cells in the BAL of MTS- compared with sham-exposed animals; we observed more mononuclear cells and neutrophils in the airways of MTS-exposed mice compared with those of sham-exposed animals at 3 and 5 d after infection.
Compared with uninfected animals, we observed significantly fewer lymphocytes in the peripheral blood of both sham- and MTS-exposed mice after high-dose influenza infection (Figure 2). Furthermore, MTS exposure was associated with elevated numbers of peripheral blood neutrophils compared with sham exposure at 3 and 5 d after infection. In contrast, we did not observe any significant differences in the number of leukocytes isolated from the peripheral blood of sham- and MTS-exposed animals after infection with low-dose influenza virus (data not shown).
Histologic Assessment of the Lung Next, we investigated the impact of cigarette smoke exposure on lung pathology. We demonstrate that 3–5 mo of cigarette smoke exposure alone does not lead to overt tissue pathology (Figure 3B). Low-dose influenza infection was mainly associated with peribronchiolar and perivascular inflammation in both sham- and MTS-exposed mice (Figures 3C and 3D). In addition to peribronchiolar and perivascular inflammation, high-dose influenza infection resulted in the accumulation of inflammatory cells within the lung parenchyma of both sham- and MTS-exposed mice, suggestive of greater alveolar involvement (Figures 3E–3H). Consistent with these observations, we isolated fewer cells from the lungs of sham- and MTS-exposed animals infected with low-dose (9.07 ± 2.50 x 106 and 9.16 ± 2.94 x 106 cells/lung, respectively; mean ± SEM; n = 10) compared with high-dose (20.24 ± 3.65 x 106 and 22.79 ± 4.93 x 106 cells/lung, respectively; mean ± SEM; n = 9) influenza virus.
To determine whether MTS exposure was associated with increased tissue inflammation in animals infected with high-dose influenza, we assessed MPO activity in lung tissue homogenates. Although MPO activity was increased in the lungs of animals infected with influenza compared with uninfected mice (Table 2), we did not observe any significant differences in MPO levels between sham- and MTS-exposed, influenza-infected animals.
Cytokine Levels in the Airway The result that MTS exposure altered inflammatory cell responses associated with influenza infection led us to investigate whether the expression of inflammatory cytokines were similarly affected. Levels of MIP-2, TNF-  , IL-6, and IFN- were assessed in the BAL fluid of MTS- and sham-exposed mice infected with either low- or high-dose influenza A virus. We observed increased expression of MIP-2, TNF-, and IL-6 after infection with either low- or high-dose influenza A compared with uninfected animals (Table 3). We detected similar levels of MIP-2, TNF-, and IL-6 in sham- and MTS-exposed animals 3 and 7 d after infection with low-dose influenza. Similarly, infection with high-dose influenza was associated with comparable levels of MIP-2 in the airways of sham- and MTS-exposed animals. However, MTS exposure led to significantly increased levels of IL-6 and TNF- compared with sham-exposed animals 3 and 5 d after infection. Plaque reduction assays further demonstrated that MTS exposure induced heightened type-1 IFN production compared with sham exposure after infection with high-dose influenza (Figure 4).
Compared with uninfected mice, IFN- levels in the BAL fluid were elevated to a similar extent in both sham- and MTS-exposed animals 1 wk after low-dose infection (data not shown). We did not observe increased levels of IFN- in either sham- or MTS-exposed animals after high-dose infection (measurements were taken 5 d after infection, which is likely too early a time point to detect this cytokine).
Viral Burden in the Lung
T-Cell Profile of the Lung The involvement of T-cell responses in the resolution of influenza infection is well established. To determine if MTS impacted the expansion and activation of CD4 and CD8 T cells after infection with influenza, sham- and MTS-exposed mice were infected with either low- or high-dose influenza virus. We assessed the percentage of CD4 (CD3+/CD4+) and CD8 (CD3+/CD8+) T cells in the lung, as well as expression of the early activation marker CD69. Infection with both infectious doses of influenza led to significant and comparable increases in the percentage of activated CD4 and CD8 T cells in the lungs of sham- and MTS-exposed mice compared with uninfected animals (Table 4). MTS exposure resulted in a modest reduction in the percentage of CD69-expressing CD4 T cells in animals infected with high-dose influenza. We also determined the impact of MTS on intracellular expression of granzyme B expression in CD8 T cells, a previously described surrogate for antigen-specific T-cell activation (32). Table 4 demonstrates that, compared with sham exposure, MTS attenuated the percentage of granzyme B+ CD8 T cells in mice infected with low-dose influenza. Unexpectedly, we did not observe an increase in the percentage of granzyme B+ CD8 T cells in high-dose–infected animals compared with uninfected mice.
Secondary Responses to Influenza A Because MTS exposure was associated with an attenuated airways inflammatory response and accelerated clearance of influenza from the lung after infection with low-dose influenza, we investigated whether MTS affected the establishment of protective antiviral immunity. To this end, serum (IgG1 and IgG2a) and BAL fluid (IgA) levels of influenza-specific immunoglobulins were assessed 6 wk after infection. Similar levels of influenza-specific IgG1, IgG2a, and IgA were detected in sham- and MTS-exposed animals (Figure 6A). We did not detect influenza-specific immunoglobulins in uninfected mice (data not shown).
To assess whether MTS impacted immune protection, sham- and MTS-exposed mice infected with 2.5 x 103 pfu influenza virus were rechallenged with high-dose influenza 12 wk after the initial infection; 5 d later, we assessed the BAL inflammatory profile in these animals. That we did not observe airways inflammation in either sham- or MTS-exposed mice suggests that both groups of animals were completely protected after rechallenge (Figure 6B). This was corroborated by a complete absence of virus in the lungs of rechallenged mice (Figure 6C). Moreover, neither sham- nor MTS-exposed mice displayed any clinical symptoms after rechallenge with high-dose influenza (data not shown).
The ability to mobilize and activate antigen-specific memory T cells is likely important in protecting against infection with heterologous influenza strains, where previously established immunoglobulin responses may be ineffective. To determine the effect of MTS on the establishment of influenza-specific memory T-cell responses, the percentage of influenza-specific memory T cells was assessed in the lung and spleen by tetrameric flow cytometric staining. We detected similar percentages of nucleoprotein (NP) 366–374+ CD62Lhi and
Innate Immune Intervention Our data show that smoking affects inflammatory processes elicited by an initial encounter with influenza virus. Therefore, recently described antiviral intervention strategies that target innate immune mechanisms may be influenced by smoking. Specifically, poly(I:C), a synthetic mimic of viral double-stranded RNA, has been shown to confer protection against subsequent viral challenge in a nonimmunized host (34). To assess whether intranasal administration of poly(I:C) elicits protection against influenza infection, sham- and MTS-exposed animals were inoculated intranasally with poly(I:C) 24 h before viral infection. Poly(I:C) administration led to a significant decrease in total cells and the number of neutrophils in both sham- and MTS-exposed animals after infection with low-dose influenza (Figure 7). Although poly(I:C) administration did not affect the total number of inflammatory cells in the BAL of either sham- or MTS-exposed animals after high-dose influenza infection, we observed significantly fewer mononuclear cells in the airways of poly(I:C)-treated, MTS-exposed animals compared with untreated, MTS-exposed mice (Figure 7).
The objective of this study was to investigate the impact of active smoking on the course of respiratory viral infection in mice. Animals were exposed to MTS using a smoke exposure system for small rodents that was initially developed for guinea pigs (35) and has since been adapted for mice (30). Animals were infected with one of two infectious doses of a replicating, mouse-adapted influenza A virus: 2.5 x 103 pfu (low dose) or 2.5 x 105 pfu (high dose). The low infectious dose led to an airways inflammatory response, but did not induce clinical symptoms, whereas infection with the high dose induced significant clinical symptoms and was associated with a more pronounced airway inflammation. Unlike administration of low-dose influenza, the higher infectious dose resulted in significant morbidity. Animals became ruffled, hypomobile, and displayed evidence of labored breathing. Importantly, whereas all of the sham-exposed mice survived through 1 wk of infection, MTS exposure resulted in death for three of eight animals. Our data are consistent with clinical studies in large military cohorts demonstrating a worsening of clinical symptoms and an increased incidence of influenza-related deaths among smokers compared with nonsmokers (36–38). MTS exposure attenuated the BAL inflammatory response to low-dose influenza infection. Although the underlying mechanism behind this observation remains unclear, the cellular toxicity of cigarette smoke has previously been demonstrated (39–41). Therefore, MTS may limit the extent of viral replication, either through its effects on airway epithelial cells, the primary infectious target of the virus, or by directly impacting viral replication. Consequently, the obligation to respond to a lower viral burden may result in the recruitment of fewer inflammatory cells into the airways. Alternatively, cigarette smoke may be toxic to immune inflammatory cells, and the reduced inflammation in the BAL is a reflection of this cytotoxicity. For example, it has recently been shown that activated granulocytes are susceptible to nitric oxide, a component of cigarette smoke (42). MTS exposure led to significantly increased numbers of mononuclear cells and neutrophils in the airways of mice infected with a 2-log higher infectious dose of influenza (2.5 x 105 pfu) when compared with sham-exposed animals. The exacerbating effect of MTS on lung inflammation was specific to BAL specimens, as we isolated similar numbers of cells and detected comparable MPO activity in the lung tissue of sham- and MTS-exposed, influenza-infected mice. Therefore, in our model, it is the extent of inflammation in the BAL rather than the lung tissue that is the best indicator of clinical status. Similar viral titers in the lungs of sham- and MTS-exposed animals indicated that factors other than antigen load likely contributed to the MTS-induced increase in BAL inflammation. Specifically, MTS exposure may have stimulated the bone marrow, as indicated by the increased numbers of neutrophils observed in the peripheral blood of MTS-exposed, influenza-infected animals. It has been demonstrated, in both animal and clinical studies, that cigarette smoking stimulates the bone marrow to release neutrophils into the circulation (43, 44). Cigarette smoke may also suppress local defense mechanisms in the lung, necessitating increased recruitment of monocytes and neutrophils from the periphery to compensate for local deficiencies. This hypothesis is supported by in vitro studies showing that bronchial epithelial cells produce fewer cytokines after stimulation with viral and bacterial agents in the context of cigarette smoke and its components (14).
Inflammatory cytokines are a consequence of and help to drive inflammatory processes. The elevated levels of TNF- T cells play an important role in the resolution of influenza infection in the lung. Taking the total cell number into consideration, we detected similar numbers of activated (CD69+) CD4 and CD8 T cells in the draining lymph nodes and lungs of sham- and MTS-exposed mice after infection with low- and high-dose influenza virus (data not shown, and Table 5). We have previously reported that MTS decreased the number of activated CD4 and CD8 T cells after infection with a replication-deficient adenovirus (22), suggesting that cigarette smoke differentially impacts immune responsiveness depending on the nature of the virus. CD8 T cells kill virus-infected cells through the release of cytotoxic granules, including perforin and granzyme B. Consequently, intracellular granzyme B expression has been demonstrated to be a functional marker of antigen specificity (32). In our studies, MTS exposure resulted in a statistically lower percentage of granzyme B–expressing CD8 T cells in the lungs of animals infected with low-dose influenza compared with sham exposure. Interestingly, this was not associated with an increased viral burden in the lungs of these animals. Unexpectedly, sham- and MTS-exposed mice infected with high-dose influenza exhibited decreased percentages of CD4, CD8, and granzyme B–expressing CD8 cells compared with low-dose infection. This observation is consistent with a recent report by Legge and colleagues demonstrating that high doses of influenza induce apoptosis of virus-specific CD8 T cells, a process mediated by lymph node–resident dendritic cells in an IL-12–regulated, FasL-dependent manner (50). Alternatively, the observed differences in T-cell percentages between the low- and high-dose influenza–infected groups may be a reflection of the timing of the measurements, 7 versus 5 d after infection for low- and high-dose infection, respectively. Influenza virus is largely cleared from the airway lumen and from within airway epithelial cells by mucosal IgA (51, 52). Serum-derived IgG isotypes are instrumental in clearing virus from the lung parenchyma and in protecting against reinfection (53–55). We demonstrate that low-dose influenza infection induced similar levels of influenza-specific mucosal IgA and serum IgG1 and IgG2a in sham- and MTS-exposed animals. Furthermore, animals were completely protected on rechallenge (Figures 5B and 5C). Mackenzie and colleagues previously showed that short- and long-term smoke exposure either enhanced or depressed humoral responses, respectively (56). It is difficult to directly compare this study with our own, because the smoking apparatus and protocol, as well as the infectious strain of influenza, differed markedly. The methodology for assessing humoral responses was also dissimilar between the two studies. However, our data are consistent with those from subsequent studies by Mackenzie and colleagues and others, demonstrating that MTS does not impair secondary responses to influenza in experimental animals, and may account for the efficacy of annual influenza vaccination strategies in lowering the incidence of influenza infection in both asymptomatic smokers and patients with COPD (27, 57–59). Razani-Boroujerdi and colleagues recently demonstrated that in vivo administration of nicotine, a constituent in tobacco smoke with known immunosuppressive properties, significantly inhibited the tissue inflammatory response to a virulent mouse-adapted influenza virus (PR8) in rats and mice that was associated with increased viral titers in the lung (60). Although this study assessed the effect of a single tobacco constituent on viral infection, tobacco smoke contains more than 4,500 compounds in the particulate and vapor phases. Our study assessed the collective effect of all these compounds on host defenses to influenza, and suggests a differential effect for individual tobacco components and whole-smoke exposure on antiviral responses. The need for annual reformulation of influenza vaccines arises from the continuous mutation of influenza coat proteins that render previously acquired humoral immunity ineffective. It has been suggested that novel vaccination strategies will need to induce memory T-cell responses to more conserved peptides if they are to protect against variant viruses that have undergone significant antigenic drift and antigenic shift. In addition to being highly conserved, the core NP is the dominant influenza T-cell epitope (61). Using major histocompatibility complex class I–restricted, NP-specific tetramer staining, we assessed influenza-specific memory T cells in the lungs and spleens of animals infected with low-dose influenza. We detected antigen-specific, central (CD62Lhi) and effector (CD62Llo) memory T-cell populations in both the lungs and spleens of influenza-infected mice. Furthermore, MTS exposure did not interfere with the development of influenza-specific memory T-cell responses after either initial or secondary exposure to virus. To our knowledge, this is the first study to incorporate tetramer technology to assess the impact of MTS on pathogen-specific T-cell responses, and suggests that smokers would benefit from novel vaccine strategies aimed at amplifying memory T-cell populations. Interaction between pathogens and the host is mediated by pattern recognition receptors, including Toll-like receptors (TLRs). Recent studies have demonstrated that TLR2 expression is altered in cigarette smoke–exposed mice, as well as in smokers and patients with COPD, suggesting that smoking impacts innate immune responsiveness (62, 63). The importance of TLR3 in the innate recognition of many viruses, including influenza, is well documented (64). We show that MTS exposure did not impact cell surface/intracellular expression of TLR3 in the lung or interfere with the ability of animals to induce an antiviral state after influenza infection (data not shown and Figure 4). Moreover, Figure 7 demonstrates that poly(I:C) administration attenuated airway inflammation after influenza infection in both sham- and MTS-exposed animals, suggesting that TLR3 signaling was not compromised by cigarette smoke exposure. An increasing number of clinical studies have demonstrated the potential benefit of using innate bacterial immunostimulants to reduce both the number and severity of acute exacerbations (65, 66). Our findings suggest that similar consideration may be given to intervention strategies aimed at inducing innate antiviral immunity. The Global Initiative for Chronic Obstructive Lung Disease states that patients with COPD mount inappropriate inflammatory responses to noxious particles or gases within cigarette smoke (67). In our study, we show that cigarette smoke exacerbates inflammatory processes elicited by viral infection. We have previously demonstrated that MTS exacerbates inflammatory responses to bacterial agents and is associated with a decline in health status (68). We postulate that this altered responsiveness to respiratory pathogens contributes to the chronic inflammation observed in smokers and the decline in health status associated with COPD exacerbation. Furthermore, the fact that MTS exposure did not lead to increases in viral burden in the lung suggests that normal pathogen clearance does not preclude the development of exaggerated and potentially harmful inflammatory responses. In summary, we show that the effect of MTS on host responses to respiratory infection with influenza A is complex and at least partly dependent on the magnitude of the infectious dose. Although MTS exposure attenuated airways inflammation in mice infected with low-dose influenza, it exacerbated inflammation, increased the production of inflammatory cytokines, and led to death in animals infected with a 2-log higher infectious dose. We believe a comprehensive understanding of the impact of cigarette smoke on host responses to respiratory infections will yield novel insight into the underlying mechanisms connecting cigarette smoking, inflammatory processes, and COPD disease progression.
The authors thank Joanna Kasinska, Sussan Kianpour, and Suzanna Gonchorova for their expert technical support. They also thank Mary Kiriakopoulos for her secretarial assistance.
Supported by the Canadian Institutes of Health Research. M.R.S. is a holder of a Canadian Institutes of Health Research New Investigator award. This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org Originally Published in Press as DOI: 10.1164/rccm.200604-561OC on October 5, 2006 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form April 24, 2006; accepted in final form October 5, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
statistically significant compared with sham-exposed, influenza-infected mice.
CD62Llo T cells in the lungs and spleens of sham- and MTS-exposed animals 12 wk after infection (
statistically significant compared with MTS-exposed, uninfected mice; 
statistically significant compared with MTS-exposed, influenza-infected mice.
, IL-2, IFN-
, and NF-
B signaling pathways. J Immunol 2003;171:1035–1041.