Impact of fatty acid binding protein 5-deficiency on COPD exacerbations and cigarette smoke-induced inflammatory response to bacterial infection

Chronic obstructive pulmonary disease (COPD) is characterized by chronic airway inflammation, irreversible airflow limitation and emphysema [1]. The main risk factor of COPD is cigarette smoke (CS) [2]. COPD is presently the fourth leading cause of death worldwide, but it is predicted to become the third by 2030 according to the world health organization [3]. Bacterial and viral infections have been implicated as a major cause of COPD exacerbations. Pseudomonas aeruginosa (P. aeruginosa) is often disregarded as a COPD pathogen. However, it represents 5–10% of the pathogens that colonize COPD lungs [4, 5] and is not only involved in acute exacerbations, but also contributes to the chronic process of the disease [6]. In addition, several studies have shown that P. aeruginosa is the cause of more infections as severity of COPD increases [7‐9]. In hospitals, the organism contaminates moist/wet reservoirs such as respiratory equipment and indwelling catheters. Infections can occur in almost every body site but are particularly serious in the bloodstream.

Since the majority of COPD patients are current or former smokers, it is thought that CS has immunosuppressive effects that increase patients’ susceptibility to infections. To discover novel innate immune genes regulated by CS in humans, we took advantage of the non-biased C. elegans model. Using microarray and RNAi, we successfully identified lbp-7, a lipid binding protein, which was down-regulated after CS exposure and played a role in innate immunity [10]. Interestingly the human orthologue of this protein, FABP5, was up-regulated in response to bacterial infection, and was down-regulated in COPD patients as compared to healthy smokers [10]. We further showed that FABP5 exerts immunomodulatory functions in the airway epithelium against CS exposure and subsequent bacterial infection through its modulation of the nuclear receptor peroxisome proliferator-activated receptor (PPAR)-γ activity [11]. However, the anti-inflammatory function of FABP5 in COPD exacerbations is still unknown.

In this study we examined the effects of CS on inflammatory responses in vivo to determine the role of FABP5 in these processes. We also looked at the role of FABP5 in macrophages since these cells are at the center of the processes required for the resolution of inflammation, including engulfment of apoptotic cells, activation of PPARγ, and repair processes [12, 13]. We hypothesized that cigarette smoke decreases FABP5 expression, thus impairing its anti-inflammatory function and contributing to a more sustained bacterial-induced inflammation. To test this hypothesis, we measured FABP5 expression and PPARγ activation in peripheral blood mononuclear cells (PBMCs) of patients with or without COPD and characterized the relation between FABP5 suppression and disease exacerbations. In addition, we evaluated the effects of CS on FABP5 expression and PPARγ activity and defined the role of FABP5 in a murine model of COPD exacerbation and in macrophages.

We extracted the mRNA from freshly isolated PBMCs of 20 non-COPD patients and 36 COPD patients enrolled in the NIH-sponsored COPDGene study cohort [14]. Levels of FABP5 mRNA were significantly reduced in patients with COPD compared to non-COPD as measured by real-time RT-PCR (Fig. 1a). Furthermore, FABP5 mRNA expression was further decreased in COPD patients who reported flare-up and pneumonia episodes (Fig. 1b, c) within the last 12 months. We also showed a significant decrease in FABP5 protein levels among COPD patients who reported one or more exacerbation as compared to patients who did not report any exacerbation in the last 12 months (Fig. 1d). Taken together, these data indicate that FABP5 expression is decreased in patients with COPD and further decreased in patients reporting episodes of COPD exacerbation.

To define the role of FABP5 in CS- and exacerbation-induced responses, we measured the effects of CS and bacterial infection on FABP5 expression in wild type (WT) mice exposed to 5 days of CS or filtered air. As indicated in Fig. 2a, while P. aeruginosa infection increased FABP5 mRNA expression, CS exposure alone decreased FABP5 expression in the lung homogenates of WT mice. The combination of both CS exposure and P. aeruginosa infection resulted in similar FABP5 mRNA levels as observed for the infection alone in the lung homogenates of WT mice (Fig. 2a). Western blot analysis of FABP5 protein in lung homogenates further confirmed that FABP5 expression is increased after P. aeruginosa infection and suppressed after CS exposure (Fig. 2b). These results confirm that FABP5 expression is positively regulated during bacterial infection, but CS exposure inhibits FABP5 expression in murine lung tissues.

To assess the impact of FABP5-deficiency on CS exposure and bacterial-induced inflammation, WT and FABP5^−/− mice were exposed to CS 5 h a day for 5 days. Control mice were exposed to room air. Mice were then intranasally inoculated with 10⁷ CFU of viable P. aeruginosa. To address whether FABP5^−/− mice displayed increased inflammation, we performed histologic assessment of lung tissues. No overt inflammation was observed in air-exposed saline-treated WT and FABP5^−/− mice (Fig. 3A a, b). Sixteen hours post inoculation with P. aeruginosa, we observed similar perivascular and peribronchial inflammatory infiltrates in both WT and FABP5^−/− mice (Fig. 3A c, d) as illustrated by increased lung histopathology score compared to air-exposed saline-treated mice (Fig. 3B). CS exposure of saline-treated WT and FABP5^−/− mice showed minimal increase of inflammatory infiltrates, mainly mononuclear cells (Figs. 3A e, f). The combination of CS exposure and bacterial infection led to increased inflammation compared to infection alone (Figs. 3A g, h), with a significant worsening of lung histopathology score in FABP5^−/− mice compared to WT mice (Fig. 3B).

We next examined the cellular content (Fig. 3C) and inflammatory cytokine profiles (Fig. 3D) in the bronchoalveolar lavage (BAL) of P. aeruginosa-infected mice. As expected from the histologic assessment of lung tissues, we did not see a difference between WT or FABP5^−/− saline-treated air-exposed and saline-treated CS-exposed animals (data not shown). However, there was a significant increase in the numbers of neutrophils following P. aeruginosa in FABP5^−/− mice whether or not the mice were exposed to CS as compared to their WT counterpart (Fig. 3C). In addition, both inflammatory cytokines IL-6 and KC were increased in FABP5^−/− mice as compared to WT animals in all the different treatment groups (Fig. 3D). Taken together these data confirm that FABP5^−/− mice have increased inflammation following CS exposure and bacterial infection.

To determine whether FABP5 restoration in FABP5^−/− mice could protect them from CS and infection-induced inflammation, we performed in vivo alveolar macrophage FABP5 gene transfer in FABP5^−/− mice via intranasal inoculation of a lentiviral vector expressing mouse FABP5 (Lenti-FABP5) or the empty vector (Lenti-Control) (Fig. 4a). FABP5^−/− mice given Lenti-Control or Lenti-FABP5 were then exposed to CS for 5 days 5 h/day and infected with P. aeruginosa. FABP5^−/− mice reconstituted with Lenti-FABP5 show a significant decrease in lung tissue inflammation (Fig. 4b, c), inflammatory cells (Fig. 4d) and inflammatory cytokines (Fig. 4e) as compared to FABP5^−/− mice reconstituted with Lenti-Control. Taken together, these results suggest that FABP5 is critical to prevent CS and infection-induced pulmonary inflammation.

The studies described above demonstrate that CS exposure decreases FABP5 expression, which we have shown previously modulates PPARγ activity in airway epithelial cells [11]. Thus, we next hypothesized that reduced FABP5 expression leads to reduced PPARγ activity in macrophages and thus may lead to overt inflammation. We thus exposed THP-1 cells, a human monocytic cell line, to cigarette smoke extract (CSE) and used a lentiviral approach to knockdown FABP5. Interestingly, CSE resulted in FABP5 down regulation to a similar extend as FABP5 shRNA-mediated knockdown (Fig. 5a). We measured PPARγ binding to its consensus sequence in those same cells. While P. aeruginosa infection increased PPARγ activity, CSE greatly reduced PPARγ activity even in the presence of bacterial infection. PPARγ activity was reduced in cells treated with FABP5 shRNA-mediated knockdown (Fig. 5b). Similar PPARγ activity results were obtained in bone marrow-derived macrophages from WT and FABP5^−/− mice (Fig. 5c). Finally, we measured FABP5 protein levels and PPARγ activity in PBMCs of non-COPD and COPD patients and show a positive correlation between FABP5 protein levels and PPARγ activity (Fig. 5d). Taken together, these results suggest that FABP5 is required for induction of PPARγ activity, which in turn may regulate anti-inflammatory processes.

COPD exacerbations are triggered primarily by bacterial and viral infections and are associated with increased airway inflammation. This study provides new insights into the anti-inflammatory role of FABP5 in COPD exacerbations. Although FABP5 has been shown to be pro-inflammatory in several models, including high fat diet-induced skin inflammation [15], allergic asthma [16], LPS-induced liver damage [17] and neuropathic pain in mice [18], in this study we show that in the lungs of mice exposed to CS and infection, FABP5 expression, or re-expression, decreases inflammation. Ultimately, understanding how we may alter inflammatory responses in the lungs of CS-exposed mice will hopefully translate into promising pharmaceutical interventions that may be used in the treatment of bacterial-induced COPD exacerbations.

We have previously shown that FABP5 is decreased in COPD [10] and in airway epithelial cells exposed to CS [11, 19]. Here we report that FABP5 expression is decreased in PBMCs of patients with COPD and further decreased in patients reporting one or more episodes of COPD exacerbation. These findings confirm that COPD is not only an airway disease, but also a systemic disorder that involves systemic inflammatory manifestations [20, 21]. We also confirmed that CS decreases FABP5 mRNA and protein expression in WT mouse lung tissues.

COPD is a very complex disease that includes emphysema, bronchitis, and bronchiolitis. All aspects of the disease are impossible to recapitulate in one mouse model of COPD. A previous study has examined the concomitant effect of CS exposure and P aeruginosa infection, and showed that CS affects respiratory immune-inflammatory responses elicited by bacteria [22]. Using whole-body CS exposure in WT and FABP5^−/− mice to test the hypothesis that FABP5 is required for dampening CS- and infection-induced inflammation, we show that FABP5^−/− animals mount an exaggerated inflammatory response to P. aeruginosa infection following CS exposure. Furthermore, we were able to dampen the inflammatory response by re-expressing FABP5 in FABP5^−/− mice using a lentiviral approach.

Being at the interface between the host and the environment, the airway epithelium represents the first line of host defense against potential pathogens or irritants (such as cigarette smoke). This barrier not only provides the organism physiological function, but also protects it from damaging agents. The protective role of the epithelium includes recognition [23] and response to potentially dangerous particulates and microbes by producing various mediators such as proinflammatory cytokines (GM-CSF and IL-8), mucins (MUC5AC) and antimicrobial substances including peptides (β-defensins and LL-37) and proteins (lysozyme and lactoferrin) [24]. However, the persistent and repeated nature of infections within COPD patients indicates an abnormal epithelial cell function. We have previously shown that FABP5 is down regulated in primary airway epithelial cells of COPD patients as compared to patients without COPD [10]. Furthermore, primary human airway epithelial cells, which are knocked down for FABP5 and exposed to cigarette smoke, exhibit higher inflammatory cytokines secretion [11]. It is, thus, highly possible that the epithelial cells are also contributing to the increased inflammatory response seen in FABP5^−/− mice.

Interestingly, Perera et al. demonstrated that persistence of increased airway inflammation during exacerbation of COPD is associated with a prolonged symptom recovery time [25]. Thus, our FABP5^−/− mice exposed to CS and infected with P. aeruginosa may recapitulate an episode of exacerbation with heightened levels of inflammatory cells and cytokines. Further investigations are required to fully understand the implication of FABP5 in antimicrobial immunity during COPD exacerbations. In particular, studies looking at the effect of viral infection following the onset of bacterial infection are missing. Indeed, patients who develop recurrent exacerbations may be more susceptible to viral respiratory infections [26]. In addition, secondary bacterial infection is a common life-threatening event following influenza A virus infection. A thorough understanding of the events leading to susceptibility to this dual infection is lacking. We have shown that FABP5 plays a protective role in response to both influenza A virus (IAV) [27] and now P. aeruginosa infections, but the role of FABP5 in response to a dual IAV/P. aeruginosa infection has not been demonstrated, nor has the impact of CS been studied in this model.

Fatty acids, in addition to providing cell energy and structure, participate in a variety of cellular processes by modulating the activity of PPARs [28, 29]. PPARs are ligand-dependent transcription factors and members of the nuclear receptor superfamily [30]. Since fatty acids are hydrophobic, they rely on FABPs to exert their signaling effects [31]. FABP5 has been shown to stimulate PPARβ/δ transactivation in a breast cancer cell line (MCF-7 cells) [32]. Here, we investigated the functional cooperation between FABP5 and PPARγ in a monocytic-derived cell line and in human PBMCs in response to CS exposure and infection. PPARγ has been shown to have anti-inflammatory properties [33, 34], and as such has been targeted for therapeutic interventions in COPD [35]. In this study, we are showing an interesting association between FABP5 expression and PPARγ activity. Decreased FABP5 expression, either by CS exposure or by shRNA-mediated downregulation, lead to decreased PPARγ activity. We also show a positive correlation between FABP5 protein levels and PPARγ activity in human PBMCs. PPARγ has long been studied for its anti-inflammatory role, since PPARγ agonists were shown to dampen macrophage activation in vitro [36, 37]. Other studies have linked PPARγ with the development of pro-resolving macrophages [38] and resolution of inflammation [34]. It is, thus, tempting to speculate that FABP5 may exert its anti-inflammatory or pro-resolving properties through PPARγ activation. Further studies are required to confirm this hypothesis.

At this time, we do not know if this CS-induced FABP5 inhibition is reversible. Our human data would suggest that it is not, since COPD patients that quit smoking still have lower levels of FABP5. Our data also suggest that the level of FABP5 suppression may be an important factor of COPD exacerbations, and polymorphisms or other alterations that contribute to FABP5 decrease may contribute to disease susceptibility. Whether decreased FABP5 is a risk or a consequence of COPD exacerbation still remains to be elucidated.

In summary, our study shows that FABP5 promotes the resolution of CS and infection-induced inflammation and that it may do so through PPARγ activation. Understanding the role of FABP5 during infections in the presence of CS may open up new therapies for the treatment and/or prediction of recurrent exacerbations of COPD.