Anti-Inflammatory Effects of Lactobacillus Rahmnosus and Bifidobacterium Breve on Cigarette Smoke Activated Human Macrophages

Background Chronic obstructive pulmonary disease (COPD) is a major global health problem with cigarette smoke (CS) as the main risk factor for its development. Airway inflammation in COPD involves the increased expression of inflammatory mediators such as CXCL-8 and IL-1β which are important mediators for neutrophil recruitment. Macrophages are an important source of these mediators in COPD. Lactobacillus rhamnosus (L. rhamnosus) and Befidobacterium breve (B. breve) attenuate the development of ‘allergic asthma’ in animals but their effects in COPD are unknown. Objective To determine the anti-inflammatory effects of L. rhamnosus and B. breve on CS and Toll-like receptor (TLR) activation. Design We stimulated the human macrophage cell line THP-1 with CS extract in the presence and absence of L. rhamnosus and B. breve and measured the expression and release of inflammatory mediators by RT-qPCR and ELISA respectively. An activity assay and Western blotting were used to examine NF-κB activation. Results Both L. rhamnosus and B. breve were efficiently phagocytized by human macrophages. L. rhamnosus and B. breve significantly suppressed the ability of CS to induce the expression of IL-1β, IL-6, IL-10, IL-23, TNFα, CXCL-8 and HMGB1 release (all p<0.05) in human THP-1 macrophages. Similar suppression of TLR4- and TLR9-induced CXCL8 expression was also observed (p<0.05). The effect of L. rhamnosus and B. breve on inflammatory mediator release was associated with the suppression of CS-induced NF-κB activation (p<0.05). Conclusions This data indicate that these probiotics may be useful anti-inflammatory agents in CS-associated disease such as COPD.


Introduction
COPD is currently the 4th biggest killer worldwide and is expected to be the third leading cause of death over the next 10 years [1]. Smoking is the most important lifestyle risk factor for pathogenesis of COPD [2], the presence of which also enhances the risk of developing lung cancer [3]. The level of inflammation in the airways of COPD patients correlates strongly with disease severity and is critical to the development and progression of disease [4][5][6]. Attention has traditionally centered on the role of macrophages and neutrophils in disease development [7]. Macrophages synthesize and secrete many mediators which play a role in COPD inflammation and innate immune responses to potentially pathogenic organisms [8].
The chronic inflammation in COPD shares many epidemiological, inflammatory and immune characteristics with other chronic diseases such as inflammatory bowel disease (IBD) [9]. The crosstalk between the pulmonary and intestinal mucosal in COPD and IBD has recently been extensively reviewed [10][11][12][13][14]. The concept that probiotics confer health benefits has gained much attention [15]. Most probiotics contain virulent lactic acid-producing bacteria (Lactobacillus, Streptococcus, Bifidobacterium, and Enterococcus) or non-pathogenic yeasts such as Saccharomyces boulardii [16,17], and have been advocated for the prevention and treatment of various conditions, including gastroenteritis, clostridium-associated diarrhea, inflammatory bowel disease, food allergies, and dental cavities [18][19][20][21]. With the exception of treatment of infectious diarrhea in both adults and children [19 22], the evidence of their effectiveness remains inconclusive [23,24].
Administration of lactobacilli orally may not only modulate local inflammation [25], but might also act systemically and have effects on other organs and tissues [26]. The probiotic hypothesis proposes that perturbations in the gut microbiota resulting from antibiotic use and dietary differences can disrupt the normal mechanisms of immunologic tolerance in the gut mucosa leading to an increase in the incidence of allergic disease, including asthma [27] and viral infections [28,29]. Sagar et al demonstrated that the combination of Bifidobacterium breve (B. breve) with non-digestible oligosaccharides suppresses airway inflammation in a murine model for chronic asthma [30], which could explain the link of gut and respiratory system.
Macrophages within the airway are derived from blood monocytes that track into the lung following migration signals and are a major cell type in the pathogenesis of COPD [5,7,8]. Peripheral blood monocytes and monocyte-derived-macrophages from COPD patients have defective cellular functions and this may reflect signals derived from other organs including the gut as these cells traffic around the body [5,7,8]. It is possible, therefore, that probiotic signals may directly target blood monocytes as they traffic through the gut.
Studies have also shown that probiotics such as L. reuteri inhibit TNF-αinduced CXCL-8 expression in intestinal epithelial cells [31] and the release of proinflammatory cytokines by human macrophages via inhibition of c-Jun pathways [32]. Although the mechanism of action of these beneficial bacteria still needs to be elucidated, probiotics have been shown to contain Toll-like receptor (TLR) ligands and can thereby attenuate TLR-driven Th1 responses [33]. TLRs and NODlike receptors (NLRs) are key pattern recognition receptor (PRR) families in the innate immune response, which are also involved in the activation and shaping of adaptive immunity [34].
There is no data examining the effect of probiotics on cigarette smoke-induced inflammation in macrophages despite these being key cells in the pathogenesis of COPD and are activated by TLRs. In the current study we investigated the mechanisms by which Lactobacillus rhamnosus (L. rhamnosus) and B. breve modulate cigarette smoke-induced inflammatory mediator expression in human macrophages.
Preparation of cigarette smoke extracts (CSE) CSE was prepared as described previously [36]. Briefly, CSE was generated by the burning of commercially available Lucky Strike cigarettes without filter (British-American Tobacco, Groningen, The Netherlands), using the TE-10z smoking machine (Teague Enterprises, Davis, CA, USA) which is programmed to smoke cigarettes according to the Federal Trade Commission protocol (35ml puff volume drawn for 2s/min). The machine produced main and side stream smoke from one cigarette through 5ml RPMI without phenol red. The absorbance was then measured using a spectophotometer and the media were standardized to a standard curve of CSM concentration against absorbance at 320nm. The pH of the resultant extract was titrated to pH 7.4 and diluted with RPMI medium. This concentration (OD = 4.0) was serially diluted with untreated media to 0.03, 0.06 and 0.12 OD and applied to the cells.
were grown in MRS (Oxoid, Basingstoke, UK), supplemented with 0.5g/L L-cysteine for Bifidobacteria, at pH 6.5 and under anaerobic conditions. Bacteria were harvested in the early stationary phase, washed with phosphate buffered saline (PBS, Lonza Leusden, The Netherlands) and stored in 20% glycerol in aliquots at-80°C as described earlier [37,38].
Bacterial cells were harvested from exponential phase cultures, washed three times in PBS and resuspended in appropriate media at 2×10 8 bacteria/ml and then labeled with FITC as described previously [39]. Briefly, bacteria were heat killed for 1h at 70°C, centrifuged (16000 xg for 5 min), washed three times in carbonate buffer, pH 9.4 and re-suspended in the same buffer. FITC was added to a final concentration of 0.15mg/ml and the bacteria were incubated for 30min at room temperature on a rotary mixer. Cells were then washed three times in carbonate buffer to remove all traces of free FITC, adjusted to 2×10 8 /ml in PBS and stored at −20°C until used. For phagocytosis assays THP-1 cells were incubated with FITC-conjugated L. rhamnosus and B. breve (MOI 1:6) at 37°C or at 4°C for 2h and before washing with 4x cold PBS and then centrifuged at 100 xg. The percentage of cells that bound FITC-conjugated bacteria was measured by flow cytometry (40,000 events). FITC-Dextran (Sigma-Aldrich, USA) was used as a positive control. Phagocytosis was determined by comparing the intensity of green fluorescence (FITC) before and after trypan blue quenching of membrane-bound, labeled bacteria.

Real-time quantitative PCR
Total RNA was extracted using High Pure RNA Isolation Kit (Roche Applied Science, USA) according to the manufacturer's instructions. The quantity and purity of the RNA was determined by Nanodrop (Nanodrop Tec, Wilmington, DE, USA) and the ratio of 260/280 nm of all the samples was higher than 2. Equal amounts of total RNA was reverse-transcribed using Transcriptor first strand cDNA synthesis kit (Roche) using oligo(dT). Real-time PCR was performed using SYBR Green PCR Master-Mix (ABGene) in 20μl reactions with 0.5μM primers for 40 cycles on an ABI Prism 7000 sequence detector (Applied Biosystems). PCR conditions were 50°C for 2min and 95°C for 15min, followed by 40 cycles of 95°C for 15s and 60°C for 1min. Primers were designed using the Primer beb3 software which are as follows: The raw CTs from the reactions were analyzed by a modified delta-C t method with efficiency correction using a PCR data analysis program, qBase to obtain relative quantification values.

Preparation of whole cell extracts and Western blot analysis
Cells were washed twice with PBS and total protein extracts were prepared using radioimmunoprecipitation assay (RIPA) buffer supplemented with protease inhibitor cocktail (MiniTM protease inhibitors, Roche Diagnostics). Samples were stored at -70°C until analysed. Protein concentrations were determined by BCA protein assay kit (Pierce) and samples (25μg) were subjected to SDS/PAGE [10% (w/v) gel] as previously described [40]. The separated proteins were electro blotted on PVDF membranes (Bio-Rad). Membranes were then washed once with Tris/HCI, pH 7.4, containing 159mM NaCI and 1% Tween 20 (TBS-T), and then blocked in super-blocking buffer (Pierce) for 1h. After washing the membranes with TBS-T, antibodies against IL-1β (Cleaved IL-1β (Asp116) Antibody #2021, Bioke, Leiden, The Netherlands) was added for 24h at 4°C. Bands were detected using 1:20,000 HRP-conjugated goat anti-rabbit IgG for IL-1β (Dako B.V. Heverlee, Belgium) and visualized with an enhanced chemiluminescence Western blot analysis system (Amersham Pharmacia Biotech). Films were scanned and analyzed on a GS7-10 Calibrated Imaging Densitometer equipped with Quantity One v. 4.0.3 software (Bio-Rad, USA). For detection of equal loaded protein on the gel, the membranes were stripped with stripping buffer (Pierce), incubated with antibody to histone H1 (Santa Cruz Biotechnology, Santa Cruz, CA), as a housekeeping protein and visualized by ECL.

Measurement of NF-κB activity and p65 translocation
NF-κB activity was determined using the secreted embryonic alkaline phosphatase (SEAP) activity assay. Aliquots of culture medium were clarified by centrifugation at 14,000 xg for 2min, heated at 65°C for 5 min to inhibit endogenous phosphatase activities, adjusted to 1x SEAP assay buffer (0.5M carbonate, pH 9.8, 0.5mM MgCl 2 ), and then incubated at 37°C for 10min in a 96-well culture dish. Fifty microliters of 6M p-nitrophenylphosphate (Sigma-Aldrich, USA) dissolved in SEAP assay buffer (pre warmed to 37°C) was added to the mixture (to a final volume of 200μl). The absorbance at 405nm (A 405 ) of the reaction mixture was read in a Wallac 1420 plate reader (PerkinElmer, USA). SEAP activity is given in milliunits (mU) per ml. One milliunit is defined as the amount of phosphatase that hydrolyzes 1.0pmol of pnitrophenylphosphate per min, and this corresponds to an increase of 0.04mU per min.
Western blot analysis was used to evaluate p65 nuclear translocation. Nuclear and cytoplasmic cell extracts were prepared using an NE-PER kit (Thermo Fisher Scientific, USA) according to the manufacturer's instructions. Subcellular localization of p65 was detected using a rabbit antihuman p65 antibody (Santa Cruz Biotechnology, USA) and visualized as described above.

Viability assay
Viability of cells was determined by staining cells with Annexin-V or 7-ADD by flow cytometry analysis (BD, USA).

Statistics
Results are expressed as percent of unstimulated and-stimulated controls and are shown as means ±standard error of the mean (S.E.M.). For statistical evaluation of the data the program SPSS 18.0 was used and normally distributed data compared using ANOVA and student's Ttest. Some of the data sets were not normally distributed and group comparisons were performed using non parametric tests (Friedman and Mann-Whitney U-tests) and data presented as median (95% confidence intervals); p<0.05 was considered to indicate significance.  (Table 1 and Fig 1). This was significant for B. breve (p<0.05). Inhibition of phagocytosis by cytochalasine D (10mg/ml) prevented uptake of FITC labelled bacteria (data not shown) and no phagocytosis occurred in cells cultured at 4°C ( Table 1). Incubation of cells with 1.5% CSE did not significantly suppress the phagocytosis of FITC-dextran (Fig 1) or further reduced the ability of these cells to phagocytose bacteria (Fig 1).

L. rhamnosus and B. breve enhance basal CXCL-8 release but suppress CSE-induced CXCL-8 production
There was a dose-dependent induction of CXCL8 release from transformed THP-1 cells by L. rhamnosus and B. breve up to 20 bacteria/cell but this is not further enhanced at 50 bacteria/ cell (Fig 2A). LR and BB The induction of CXCL-8 by B. breve was greater than that observed with L. rhamnosus. In all subsequent experiments, a sub-maximal ratio of 1:10 was selected.
CSE caused a concentration-dependent induction of CXCL-8 release from transformed THP-1 cells (data not shown). 1.5% CSE significantly increased the release of CXCL-8 (p<0.001, Fig 2A) and this concentration was used in all subsequent experiments.
In contrast to the effect of L. rhamnosus and B. breve on basal expression of CXCL-8, pretreatment of cells with L. rhamnosus and B. breve attenuated CSE-induced CXCL-8 release (Fig  2A). There was no difference in the ability of L. rhamnosus and B. breve to suppress CSEinduced CXCL-8 release and no obvious dose-response was observed (Fig 2A). A similar, pattern was seen when examining the effect of L. rhamnosus and B. breve on baseline and CSEinduced CXCL-8 mRNA expression (Fig 2B). CSE and the bacteria had no effect on cell viability ( Fig 2C) at a dose of 10 bacteria per cell.
A similar ability of L. rhamnosus and B. breve to attenuate stimulated CXCL-8 release was also seen when cells were stimulated with the TLR4 agonist LPS (1000ng/ml) (Fig 2D) and the TLR9 agonist CPG ODN (3μM) (Fig 2E). L. rhamnosus attenuated CXCL-8 release to a similar extent in LPS-and CpG-stimulated cells (55 and 50% suppression for LPS and CPG  stimulation, p0.05 and P0.01 respectively) and a similar effect was seen with B. breve (50 and 65% suppression for BB with LPS and CPG stimulation, p0.05 and P0.01 respectively).
To show that probiotics but not pathogens inhibit the pro-inflammatory effects of CSE, THP-1 cells were incubated with E.coli bacteria and then stimulated with CSE for 16 h and the levels of CXCL-8 was evaluated in supernatants. In contrast to probiotics, E. coli enhanced the ability CSE to stimulate CXCL-8 release (Fig 2F).
In addition, to determine whether the effects seen were cell-type dependent, we used a second macrophage cell type U937 and conducted similar experiment. U937 macrophages were stimulated with LR and BB in the presence and absence of 1.5% CSE and CXCL-8 release determined after 16hrs. A similar response to that observed in THP-1 cells was observed (Fig 2G).
Effect of L. rhamnosus and B. breve on IL-10, TNF-α, IL-1β, IL-6, and IL-23 release in the presence and absence of CSE L. rhamnosus (1:10 bacteria:cells) significantly enhanced the expression of basal IL-23 ( Fig 3A) but had no effect on the basal release of IL-10 (Fig 3B), IL-1β (Fig 3C), TNF-α (Fig 3D) or IL-6 ( Fig 3E). B. breve, in contrast, at the same dilution significantly stimulated basal production of both IL-23 and IL-10 ( Fig 3A & 3B) but again had no effect on basal release of IL-1β, TNFα or IL-6. Both L. rhamnosus and B. breve had a similar significant ability to suppress CSE-induced expression of IL-10, TNF-α, IL-1β, IL-6, and IL-23 by 40-50% in all cells (Fig 3A-3E). We have subsequently examined the expression of the mRNA for these cytokines in this study and show similar effects (data not shown).

L. rhamnosus and B. breve suppress CSE-induced NF-κB activity
Since CXCL-8 and many of the other inflammatory mediators up-regulated by CSE are controlled by the transcription factor NF-κB, we next analyzed whether L. rhamnosus (1:10 bacteria:cells) and B. breve (1:10 bacteria:cells) had any effect on basal and CSE-stimulated NF-κB activity as determined by SEAP assay. Pre-incubation of transformed THP-1 cells with B. breve, but not L. rhamnosus, significantly increased basal NF-κB activity (Fig 4A). CSE exposure also resulted in a similar significant induction of NF-κB activity. However, pre-incubation of cells with both L. rhamnosus and B. breve resulted in a significant reduction in CSE-stimulated NF-κB activation back to baseline levels ( Fig 4A).
Next we determined nuclear p65 translocation by Western blotting as an alternative measure of NF-κB activation (Fig 4B). Neither L. rahmanosus nor B. breve had any effect on basal p65 nuclear translocation at the single 15min time point examined whereas CSE significantly increased p65 nuclear import. Consistent with the effects of L. rahmanosus and B. breve pre- incubation on NF-κB activity, CSE-enhanced p65 nuclear translocation was markedly suppressed by these bacteria (Fig 4B).

Effects of L. rhamnosus and B. breve on IL-1β activation
We have previously reported that CSE activation of cells involves IL-1β activation [36]. In Fig  2B we report that CSE-enhanced release of mature IL-1β is suppressed by both L. rhamnosus and B. breve. Western blot analysis of cellular IL-1β expression demonstrated that CSE exposure greatly enhanced the formation of the mature form of IL-1β after 6h (Fig 5, compare lanes  1 and 6). As predicted from the ELISA data (Fig 3B), pre-treatment of cells with L. rhamnosus and B. breve for 2h resulted in a reduction in CSE-induced formation of mature IL-1β. In the absence of CSE-exposure both L. rhamnosus and B. breve increased mature IL-1β expression (Fig 5). However, this was not reflected in the ELISA data (Fig 3B) for L. rhamnosus at the single time point studied suggesting a failure to release IL-1β from the cells in response to L. rhamnosus.
Effects of L. rhamnosus and B. breve on the HMGB1 expression HMGB1 is also released in response to CSE [41]. We, therefore, determined the effect of L rhmanosus and B. breve on the expression of HMGB1 within cells at 5h and the release of HMGB1 from cells at 16h at baseline and in response to CSE exposure. There was a small but Effects of L. rhamnosus (LR) and B. breve (BB) on cigarette smoke extract (CSE)-induced NF-κB activation. Cells 5x10 4 /well were preincubated for 2h with LR or BB (both at 10 bacteria/cell) and then incubated for 24h in the presence or absence of 1.5% CSE. SEAP activity was measured in cell culture supernatant (A). *p<0.05 compared to the control and # p< 0.05 compared to CSE alone. The effect of RL and BB on NF-κB p65 nuclear import was also assessed by Western blotting with histone H1 (H1) as a nuclear loading control (B). A representative blot of 3 independent experiments is shown (upper panel) with the mean ±S.E.M. data presented graphically (lower panel). *p < 0.05; **p < 0.01 versus control and # p<0.05 versus CSE-stimulated p65 nuclear import.
doi:10.1371/journal.pone.0136455.g004 significant increase in basal HMGB1 cell expression ( Fig 6A) and release (Fig 6B) following L. rhamnosus exposure which was not seen with B. breve. CSE exposure resulted in a highly significant induction of HMGB1 expression and release (Fig 6A & 6B). This was significantly inhibited by 2h pre-treatment with both L. rhamnosus and B. breve. The effect of L. rhamnosus on CSE-induced HMGB1 expression and release was greater than that seen with B. breve. This suggests that distinct mechanisms are involved in IL-1β activation by L. rhamnosus and B. breve.

Discussion
Our data demonstrates that L. rhmanosus and B. breve are able to suppress the expression and release of a number of important pro-inflammatory mediators from a CSE-exposed human macrophage-like cell. A similar ability to suppress TLR4 and TLR-9-induced CXCL-8 was also seen. The ability to suppress CS-induced mediators was associated with the attenuation of the key regulatory NF-κB pathway. In contrast, these bacteria enhanced the basal expression of IL-23, IL-10 and CXCL-8 release.
Probiotics are widely used as supplements in dietary products and are defined as live microorganisms that may have the ability to provide a health benefit to the host. Desirable features of a probiotic strain are protective colonization in the intestinal tract and increasing the efficiency of digestion and nutrient absorption and reduced capacity to cause immune responses to deviate towards Th1 and/or Th17 immune responses [42]. Several studies have advocated the potential of probiotics as a therapeutic moiety in asthma and airway infections [38,43]. However, the rational use of probiotics in airways disease requires a greater knowledge of the effects of these agents on innate immune and inflammatory pathways in the lung.
We used a human macrophage cell line exposed to cigarette smoke as a model to determine the actions of two different beneficial bacterial strains (L. rahmonsus and B. breve) on the inflammatory response of human macrophages in COPD. These two bacterial strains were selected as they belong to genera that are widely used in dietary products. Our studies focused on CXCL-8 as this is considered an important chemokine which contributes to COPD pathogenesis by stimulating neutrophil recruitment and promoting their activation and release of proteases [44][45][46].
Although these strains of bacteria caused increased release of the pro-inflammatory mediators CXCL-8 and IL-23, they also enhanced the release of the anti-inflammatory cytokine IL-10. It is unclear which of these two opposing traits will have a dominating effect in overall macrophage and down-stream T-cell function. Furthermore, there are likely to be distinct mechanisms involved in the pro-inflammatory effects of L. rahmonsus and B. breve since the suppression of mediator release was not dose-dependent and occurred equally with both bacteria. Future studies will require the use of gene and/or protein arrays to obtain a better picture of the actions of probiotics on basal macrophage function and testing in a good animal model of COPD. In COPD, the oxidative stress seen after the initial cigarette smoking challenge persists even in subjects who have ceased smoking [47,48]. This suggests that macrophages from COPD patients will remain in an 'activated' state and that modulation of this state is more important than the effects seen in basal unstimulated cells. Ideally, experiments should be performed to examine the effect of probiotics on inflammatory parameters in primary airway or lung macrophages from COPD patients.
We did not find any significant detrimental effects of probiotics on macrophage phagocytosis either on basal or CSE stimulated cells. Defective phagocytosis is an important defect contributing to the pathobiology of COPD [49,50]. It is important, therefore, that these bacteria did not affect this process as this may affect the ability to control infection.
In contrast, we were able to demonstrate a significant inhibition of both L rhmanosus and B. breve on CSE-activated NF-κB activity. The degree of inhibition of this pathway correlated well with the suppression of inflammatory mediator release. However, since we did not observe complete suppression of mediator release, the involvement of other pathways such as the JNK pathway, which has been previously implicated in probiotic actions [51], cannot be ruled out.
There was a difference in NF-κB activation dependent upon the assay used. SEAP measures functional NF-κB activity after overnight incubation of cells whereas p65 nuclear translocation by Western blot analysis was performed 30 minutes after cell stimulation. NF-κB activity can be enhanced not just by nuclear import but also by post-translational modifications which can markedly affect transcriptional activity without significantly affecting the nuclear levels of p65 [3]. The precise molecular mechanism(s) by which these bacteria are able to activate basal NF-κB activity in macrophages but repress CSE-induced activity is unknown. All bacteria, not just probiotics contain TLR ligands and we hypothesize that it is the panoply of these ligands in probiotics combined with the effect of cigarette smoke on the expression profile of TLRs that give rise to the specific anti-inflammatory effects seen here. Further research in the area should investigate whether these probiotic bacteria may act as partial agonists at TLRs and compete with CSE-induced activation or whether there is cross-talk between cell surface receptors at the level of intracellular co-factors.
The precise mechanism by which IL-1β induced is regulated by CSE and by probiotics unclear. Previous data indicates that nonpathogenic L. rhamnosus activates the inflammasome leading to caspase-1 activation and IL-1β production [52]. However we were unable to see any induction of IL-1β by L. rhamnosus or B. breve from baseline in this study and these probiotics reduced attenuated CSE-induced IL-1β expression. The mechanism by which probiotics to inhibit IL-1β, IL-18 and HMGB1 expression and the possible involvement of inflammasome in the response to TLR3, TLR4 or TLR9 activation and the subsequent effect on IL-1β, may be an area of future research. It is likely that these processes may be more important in exacerbations since there is little evidence for IL-1β and inflammasome activation in stable COPD [53].
Our study here confirms previous work showing that CSE induces the release of HMGB1 [54][55][56][57]. HMGB1 expression is increased in COPD patients [58] but the association between HMGB1 expression and DAMPs such as S100 proteins, defensins which are also increased in COPD patients [59], is unclear. We found that the basal release of HMGB1 was increased by CSE treatment and by L. rahmanosus but that both L rhmanosus and B. breve suppress CSEinduced expression and release of HMGB1. The effect of these probiotics on the expression of other innate immune mediators and the mechanism by which they reduce CSE-induced HMGB1 expression needs to be elucidated. The induction of TLR and nucleotide-binding oligomerization domain (NOD)-mediated cross-tolerance may explain these effects of probiotics shown here and this may be a novel therapeutic option in COPD.
Despite the limitations of study, which include the restricted range of probiotics used, the restricted time course for NF-κB analysis and the use of cultured macrophages rather than primary human airway or lung macrophages, we have shown that L. rhamnosus and B. breve inhibit the expression and release of several pro-inflammatory mediators induced by exposure of macrophages to CSE. The precise characterization of the active factor(s) that mediate this effect and its mechanism of action will facilitate the development of new therapeutic strategies for smoking-related airway diseases. Further studies are required to confirm whether L. rhamnosus and B. breve could act as a beneficial probiotics in COPD.