Tobacco or cigarette smoke regulates airway gene expression via two main mechanisms, by altering the status of (i) chromatin remodeling, and (ii) DNA methylation of the target genes (Figure 1) (52–54).

Chromatin remodeling is a result of a disrupted balance in histone acetylation/deacetylation (55). Excessive activation of more than 20 transcription factors including NF-κB, and lipoprotein peroxidation products (peroxinitrite, acrolein, and 4HNE from tobacco smoking) contributes to such anomaly. NF-κB is a key inflammatory and redox-sensitive transcription factor that plays a direct role in cigarette smoke-induced airway inflammation. NF-κB has been described as a “smoke-sensor” due to its sensitive activation by tobacco residues (56). Stimulation of multiple signaling cascades [p38 mitogen-activated protein (MAPK) kinases, mitogen and stress-activated kinase 1 (MSK1), protein kinase C zeta (PKCζ), and IκB kinase (IKK) complex (IKKα, IKKβ, and NEMO)] by tobacco residues promotes the activation and nuclear translocation of transcription factor NF-κB RelA/p65 (54, 57–64). This is followed by a complex formation of NF-κB/CBP-p300 [coactivator, CREB-binding protein (CBP) or CBP/p300] at target DNA sequences. It should be noted that CBP/p300 also has intrinsic histone acetyltransferase (HAT) activity. Subsequent acetylation and phosphorylation of the subunit p65 in the NF-κB/CBP-p300 complex by the activated MSK1/PKCζ-signaling pathways (and other 11 different kinases), and CBP/p300, respectively, are required for the full activation of NF-κB (57, 60, 63). This enhances the DNA binding affinity of the complex. Histones H3 and H4 in the chromatin complex of target sequences are then being acetylated (histone H3 at Lys9; H4 at Lys8 and Lys12) and phosphorylated (histone H3 at Ser10) by the subunit CBP of the NF-κB/CBP-p300 complex, and the activated MSK1 and PKCζ, respectively. The hyperacetylated core histones, however, fail to be neutralized or deacetylated by a dysfunctional histone deacetylase (HDAC2). Peroxinitrite nitrates the tyrosine residues of the HDAC2 and causes inhibition of activation and reduced expression of the protein. Of note, peroxinitrite is a by-product generated from the immune cell-derived nitrite oxide (NO) and reactive oxygen species (ROS) of cigarette smoke (65, 66).

Cigarette or tobacco smoke disturbs the DNA methylation status of target genes through several mechanisms. Firstly, DNA damage caused by cigarette smoke stimulates the DNA methyltransferase 1 (DNMT) to actively induce CpGs methylation at the damaged site (67). The hypermethylation is prone to introduce error of methylation in some target genes, resulting in reduced gene expression. Secondly, activation of nicotine signaling pathway by tobacco smoke causes CaMKII/IV and ERK/MAPK pathway activation that subsequently induces the activity of CBP to suppress the expression of DNMT1. This may result in reduced DNA methylation and thus altered level of gene repression by DNMT (68–70). Finally, enhanced activities of transcription factors such as hypoxia inducible factor 1 due to the high level of carbon monoxide and hypoxia have also been reported to influence airway gene expression (71).

Consequently, the combinatorial effect from both aberrant acetylation of histone and DNA methylation promotes the transformation of chromatin from a condensed structure to an activated open conformation. This facilitates irregular accessibility of DNA for transcription machineries, hence irregular gene expression by various cell types in the airway. The mechanisms reported are responsible for increased expression of NF-κB-dependent proinflammatory gene products [i.e., IL-1β, IL-6, IL-8, CCL-5 cyclooxygenase (COX)-2, and MIP-2/CXCL2] in both pulmonary structural cells (bronchial, small airway, and alveolar epithelial cells) and immune cells (alveolar macrophages), increased VEGF and iNOS in nasal fibroblasts and lymphocytes (Jurkat T cells), respectively, and decreased activity of antioxidant transcription factor Nrf2 and α1-antitrypsin in bronchial epithelial cells (54, 56, 57, 59, 62–64, 72–79). These may contribute to the anatomical anomalies in the airway and excessive inflammatory responses among smokers during the course of COPD.

COPD is associated with chronic inflammation in the peripheral airways orchestrated by both innate and adaptive immune responses that are interconnected via dendritic cells (80). Increasing numbers of inflammatory cells (neutrophils, macrophages, T and B lymphocytes, mast cell, eosinophils, and dendritic cells) and inflammatory mediators are accumulated in the airway lumen/wall in the lung parenchyma (19, 81). These immune cells and inflammatory mediators can hence be detected in the sputum and BAL fluid of COPD patients. The level of accumulation is positively correlated with disease severity. An increasing number of studies using animal models and clinical tissues have reported the nature of excessive airway inflammatory responses in COPD. Despite this, the heterogeneity in symptoms progression among COPD patients remain unexplained. The overall mechanism of COPD inflammatory immune response is depicted in Figure 2.

Respiratory viral infections are often the primary cause in the infection-dependent AECOPD, and virus was identified as single or multiple infecting strains from up to 64% of COPD patients with exacerbations recorded between years 2001–2017 (137–145). The most common infecting viruses are, by far, human rhinovirus, influenza virus A, and respiratory syncytial virus, whereas parainfluenza virus, coronavirus, echovirus, human metapneumovirus, and adenovirus are considerably rare.

Bacterial infections contribute to an average of 50% of infective acute exacerbations with a prevalence being reported ranging from 26 to 81% (132, 135, 146–148). The most commonly pathogenic bacterial species isolated from the lower airway of COPD patients during AECOPD are NTHi, Moraxella catarrhalis, Streptococcus pneumoniae, Staphylococcus aureus, Pseudomonas aeruginosa, and Klebsiella pneumoniae (11, 129, 133, 136, 149–152). It has been suggested that infection with new strains of the infecting species, rather than a new species, is highly associated with an increased risk of exacerbation (11, 153, 154). Atypical bacteria that cause exacerbations are Chlamydia spp., Legionella pneumophilia, and Mycoplasma spp.

In contrast to viral infections that are diagnosed in 5–45% of COPD patients with stable disease and increase to 39.3–64% during COPD exacerbations, bacterial colonization in the airways are more common with the same species during both stable disease (25–86%) and exacerbations (58.8–81%) (11, 132, 136, 137, 142, 155–158). Hence the precise or direct role of bacterial infection as the primary cause in triggering AECOPD remains controversial although a significantly increased bacterial load is observed during exacerbation in several patients. This further suggests that bacteria might be more involved as secondary invaders after an initial viral infection.

Viral infections have been reported to cause several physiological changes in the lung that in turn facilitates secondary bacterial invasion. The mechanism of bacterial superinfection has been described for H. influenzae, S. pneumoniae, S. aureus, and many other airway pathogens (159–161). Firstly, viral infections destroy the tight junctions of the airway epithelial barrier while inducing epithelium apoptosis. This results in the onset of airway epithelium lining repair whereby the sloughed off dead cells would become a rich nutrient source for growth of infecting bacteria. The damaged epithelium lining also enables bacterial adherence to the exposed basement membrane and ECM. Secondly, the demolished ciliated clearance as a result of the virus-damaged airway epithelium lining further promotes bacterial colonization and subsequent epithelial transmigration into deeper tissues (162–164). Lastly, viral infections are also detrimental to the airway immune defense by causing degradation of antimicrobial peptides (AMP), and by triggering IFN-γ secretion by immune cells. This results in suppressed macrophage and neutrophil responses to infecting bacteria, and thus enables bacterial evasion of the airway immune defense (165–168). Nevertheless, viral and bacterial coinfection have greater impact in the AECOPD airway inflammatory responses than bacteria or virus infection alone (168, 169). This is in parallel with the co-isolation of both respiratory viruses and bacteria from 6 to 30% of AECOPD patients (129, 133, 136, 170–174).

Infective AECOPD is also attributed to impaired functions of AMP, macrophages, and neutrophils triggered by inhaled irritants such as tobacco smoke. Expression of microbial-induced AMP (human β-defensin 2) is suppressed in airway epithelial cells when exposed to cigarette smoke (175, 176). Both the alveolar and monocyte-derived macrophages in patients with COPD are defective in phagocytosis of bacteria such as H. influenzae and S. pneumoniae (177, 178), and in efferocytosis of apoptotic neutrophils and epithelial cells. In addition, neutrophils from COPD patients are aberrant in chemotactic response with defective accuracy (179). All these factors contribute to the failure to resolve inflammation in COPD leading to facilitated chronic microbial colonization, also during exacerbations.

The low number of cultivable bacteria found in healthy individuals previously led to the conclusion that healthy and normal lungs are virtually sterile. This hypothesis is currently being revised, since the introduction of 16S rDNA based molecular diagnostics has shown that even healthy lungs have a distinct microbial community, different from that seen in the upper respiratory tract (180, 181). This has led to the concept of a core human lung microbiome which can be altered in COPD stable disease and during exacerbations (182). The role of the lung microbiome in the pathogenesis of COPD by influencing host immune response has also been suggested (151, 183–188).

The stability of the lung microbiome has profound impact on maintaining local immune homeostasis (189). According to the “vicious circle” hypothesis, airway inflammation and impaired immune defenses caused by either viral infections or irritant inhalation have ecological influence on the airway microenvironment and growth conditions that would eventually lead to dysbiosis of the lung microbiota (182, 190). The changed lung microbiome would then cause a maladaptive immunological response resulting in further inflammation and damage of the lung immune defenses, and additional alteration of the lung microbiome. The chain of events thus generates a vicious circle that contributes to COPD progression and exacerbation.

Several studies have documented that COPD progression from stable state to an exacerbation could induce microbiota shift in the lower airway (bronchioles), sputum, and throat (151, 190–196). Alteration in the microbiome complexity or richness is associated with the inflammatory process and changes in ECM protein expression in the lung, as observed in COPD (185, 197). Declined diversity in the lung microbiome has been reported to be related to disease severity, inflammation and decreased lung functions in COPD. This includes the increased emphysematous destruction, bronchial tissue remodeling, lymphoid follicle formation, elevated autoantibodies, and IL-17A production, and finally increased neutrophil extracellular traps (NET) formation in the airway of animal models or AECOPD patients (198–201). It has recently been reported that lung microbiome diversity is also associated with genetic factors. Mannose-binding lectin (MBL) deficiency has also been associated with disease severity and exacerbations in patients with cystic fibrosis and bronchiectasis (202). However, COPD patients with a genetic deficiency in MBL are less susceptible to Haemophilus spp. colonization, lowering the risk of exacerbations while their lung microbiota is more diverse than normal COPD patients (203).

The epithelium and alveolar macrophages are predominant cell types in the airway compartment. They comprise the first line of defense in the cellular immune response against potential inhaled pathogens and antigens. The sensing of bacteria, and particularly NTHi in the lower airways is initiated via PRRs expressed on innate immune cells and endothelium in addition to epithelial cells (238–240). TLRs are PRRs that sense stimulation by NTHi-derived pathogen-associated molecular patterns (PAMPs), and play a primary role in initiating effector cellular responses and intracellular signaling for NF-κB activation (238). Among the different TLRs, most of the studies on NTHi infection have by far been focused on TLR2 and 4. Lipoproteins including NTHi P6, and LOS are potent immunomodulators for activation of TLR2 and TLR4, respectively, and has been described in several studies on airway epithelial cells and alveolar macrophages (241–244).

Interaction of NTHi lipoprotein P6 with TLR2 on human epithelial cells [type II alveolar A549 and human middle ear epithelial cells (HMEE)] causes NF-κB-dependent activation via two distinct TLR-signaling pathways, that is, the NF-κB translocation-dependent, and -independent pathways (242). The NF-κB nuclear translocation-dependent pathway requires activation of NF-κB-inducing kinase IKK complex. In the second pathway, the MKK3/6-p38 MAPK signaling cascade is recruited for direct nuclear phosphorylation, and thus activation of NF-κB. The branching of both pathways may occur at the TGF-β activated kinase 1 (TAK1) signaling junction. NTHi stimulation via TLR2 and downstream activation of p38 MAPK/NF-κB-dependent pathways result in expression of COX-2 and prostaglandin (E2) (PGE2) that promote inflammatory responses (245).

TLR4 stimulation by NTHi LOS also contributes to the activation of NF-κB via two signaling pathways, the primary activating pathway of MyD88 cascade and the alternative pathway of Toll/IL-1R domain-containing adapter-inducing interferon-β (TRIF). Both pathways activate NF-κB through phosphorylation and degradation of inhibitor IκBα (243, 246). NTHi-TLR4 signaling mediates an effective innate immune response that leads to upregulation of TNF-α, IL-1β, IL-6, macrophage-inflammatory protein (MIP)-1α, MIP-2, and neutrophil infiltration in the airways of mice. The TLR4 response promotes efficient pulmonary clearance of bacteria in TLR4-expressing animals compared to CD14/TLR4 knockout mice (243, 244). A recent study by Jungnickel et al. revealed that, in parallel with the infection-induced pulmonary neutrophilic inflammation, NTHi-dependent stimulation of both TLR2 and TLR4 in a transgenic mouse [(Kras^LA1) with oncogenic Kras allele in the lung epithelium] additionally promotes the proliferation of Kras-induced early adenomatous lesion in the lung in an TLR-dependent manner (247). The association or role of NTHi-induced airway inflammation in lung cancer progression, however, is not supported by another recent cohort study showing the lack of differences in NTHi specific-antibodies between cancer- and non-cancer COPD patients (12).

Lastly, Dectin-1 and the epidermal growth factor receptor (EFGR) pathway also have proinflammtory effects upon interaction with NTHi (248, 249). Activation of the Dectin-dependent proinflammatory response requires NTHi-induced phosphorylation of the Dectin-1 hem-immunoreceptor tyrosine-based activation motif (hemITAM) (248). Direct activation of EFGR in alveolar cells and HMEE by NTHi-derived EGF-like factor has been shown to contribute to NF-κB activation. The EFGR-dependent NF-κB activation is mediated via an NF-κB nuclear translocation-independent pathway, which involves both MKK3/6-p38 and PI3K/Akt signaling pathways (249). Surprisingly, the interaction of EFGR and NTHi also results in negative regulation and suppression of the induction of TLR2 via the Src-MKK3/6-p38 α/β MAP kinase-dependent signaling cascade, and this in turn may facilitate NTHi infection (250). The actual components of NTHi that exhibit the EGF-like factor activity have, however, yet to be defined. The EFGR-dependent negative regulation of TLR2 may thus suggest a novel mechanism targeted by NTHi for immune evasion by attenuating the responses of host PRR, despite the contradicted role of EFGR in proinflammatory and innate immune responses of the airway epithelium (251). NTHi infection also upregulates the NRLP3-inflammasome during NTHi-induced inflammation in the airway epithelium and alveolar macrophages, leading to increased secretion of IL-1β and IL-8, and thus neutrophilic influx to the lung (252).

Some of the endogenous inflammatory mediators that are produced in response to NTHi infection, including TNF-α, IL-1α, and TGF-β1, may act synergetically with NTHi on the airway epithelial and immune cells. The synergetic interaction drives a positive feedback loop to amplify the NF-κB transcriptional activity on proinflammatory genes and further augments airway inflammation.

The synergetic activation of NF-κB by NTHi and TNF-α in HMEE and normal human bronchial epithelial (NHBE) cells occurs via NF-κB nuclear translocation-dependent and independent pathways. The latter pathway involves MAPK/extracellular signal regulated kinase kinase kinase 1 (MEKK1)-dependent activation of MAPK kinase 3/6–p38 MAPK pathway (253). However, the synergetic action of NTHi with TGF-β1 is mediated by another mechanism which involves Smad3/4-protein kinase A (PKA)-p300-dependent signaling cascade. The pathway components, PKA and p300, phosphorylates residue Ser276 and acetylates Lys221 of the NF-κB subunit p65, respectively. This results in enhanced DNA-binding activity of NF-κB (254).

The synergetic action of NTHi with both TNF-α and TGF-β1 enhances the production of TNF-α, IL-1β, and IL-8 from airway epithelial cells and interstitial polymorphonuclear infiltrates. Recently, it has been reported that co-infection of human rhinovirus and NTHi on the airway epithelial cells (NHBE cells and the BEAS-2B cell line) also results in synergetic induction of CCL20 and IL-8, albeit the exact mechanism remains to be elucidated (255). Of note, activated macrophages also release increased concentrations of TNF-α and IL-1α (256), further enhancing the inflammatory synergetic effect of surrounding immune cells.

Finally, IL-1α acts synergetically with NTHi to upregulate the expression of AMP β-defensin 2 (DEFB-4) via the p38/MAPK pathway (257). Of note, IL-1α could also act individually to upregulate the expression of DEFB-4 via the Src-dependent MEK1/2-ERK1/2 signaling pathway (258). Taken together, the synergetic action may aid in the expansion of the inflammatory response and in some cases worsen the clinical outcome.

Alveolar macrophages located in the air-parenchyma interface are the primary professional phagocytes in the lung (259, 260). These cells are responsible for infection eradication through its phagolysosomal machinery while releasing a plethora of inflammatory cytokines and chemokines for promoting a local inflammatory response and recruitment of neutrophils. Neutrophils are the first responder cells recruited from circulation to the airway for efficient killing of pathogens through an array of microbicidal strategies (261, 262). During NTHi lung infection, both alveolar macrophages and neutrophils are the main innate immune cells involved in the pulmonary bacterial clearance through phagocytosis. They are also an important source of cytokine secretion required for induction of other immune cells and enhanced bacterial killing. Eradication of NTHi by alveolar macrophages involves adhesion or contact, phagocytosis and phagolysosomal processing of bacteria, in addition to secretion of TNF-α. Phagocytic clearance of NTHi by alveolar macrophages is orchestrated through actin polymerization, plasma membrane lipid rafts, and phosphatidylinositol 3-kinase (PI3K) signaling cascade upon induction of macrophage PRRs by NTHi (256).

Interestingly, in response to NTHi infection, human alveolar macrophages, and blood neutrophils produce extensive amount of intracellular and extracellular ROS as a component of the antimicrobial defense. This leads to the formation of macrophage and neutrophil extracellular traps (METs and NETs, respectively), with co-expression of MMP-12 for enhanced bacterial killing (263, 264). Nevertheless, the overexpression of MMPs may adversely result in a protease imbalance and contribute to alveolar emphysematous destruction and bronchiectasis in COPD (265). Moreover, excessive endogenous ROS production could also introduce airway oxidative stress that is detrimental by causing chronic inflammation and tissue damage in the lung, and thus contributing to the COPD exacerbation (266, 267). The NET formation is elicited mainly by NTHi LOS in addition to other Haemophilus PAMPs (264).

Several studies by King et al. have revealed that T cell-mediated adaptive immune responses against NTHi airway infection in patients with idiopathic bronchiectasis and COPD has been predominated by a Th2/Tc2 response (268–270). The activated T cells produce reduced level of the CD40 ligand and IFN-γ, and increased levels of TNF-α, IL-13, and IL-17, as well as altered IgG subclass production by plasma cells. It is to be noted that the Th2/Tc2-mediated immune response is less effective in suppressing NTHi infection. Redirecting the Th2/Tc2-mediated immune response to Th1/Tc1 dominant (which is more protective) by adding the Th1/Tc1 mediators (CD40 ligand and IFN-γ) has helped to restore the T cell-mediated immune killing of NTHi (269). However, a separate study in a COPD mouse model by Lu et al. reported that NTHi infection causes increased production of airway type 1 interferon (1-IFN) (271). It was further reported that DNA of NTHi acts as a PAMP in stimulating the STING/TBK1/IRF3 pathway, and thus the production of 1-IFN. The impact of the bacterial DNA-induced 1-IFN in host immune/inflammatory response, which may potentially induce a Th1/Tc1 response requires further investigations.

COPD patients also have abnormally higher number of Treg cells, myeloid-derived suppressor cells (MDSC), and exhausted effector T cells (PD-1⁺) than healthy individuals (272, 273). Cigarette smoke-induced anti-inflammatory activity of Tregs in a COPD model is further suppressed by NTHi infection. The pathogen causes downregulation of Foxp3 (biomarker of Tregs), and thus impairs the anti-inflammatory/pro-inflammatory balance of Tregs (273, 274). This may lead to the extensive immunosuppressive activity by Tregs on the proliferation of NTHi P6-specific effector T cells, causing a diminished response of effector T cells to sputum IL-6 and IL-8 induction, and increased levels of IL-10 and TGF-β1 (272, 275). Recently, it has been reported that mucosal-associated invariant T cells (MAIT) from COPD patients are more effective in response to NTHi stimulation and thus produce increased levels of IFN-γ, 3-, to 10-fold more than the COPD Th (CD4⁺) and Tc (CD8⁺) cells (276). However, the pulmonary MAIT cell immune responses are compromised in the presence of corticosteroids that are commonly used for the treatment of COPD. This may potentially prone the T cell-mediated immunity to a Th2/Tc2 response in COPD patients treated with corticosteroids (277). Interestingly, antigen-specific Th17 cells from NTHi-immunized non-COPD mice model recognize both homologous and heterologous strains of NTHi, and are able to confer protection upon adoptive transfer (278). However, it is unclear whether the Th17 cell which is prone to the inflammatory response could be “trained” to counteract the NTHi infection in COPD patients, particularly during exacerbations.

During the systemic humoral immune response in NTHi-infected COPD patients, greater concentrations of NTHi-specific IgG, IgA, IgM, and IgE serum antibodies are produced compared to non-infected controls (12, 279–281). Some of the NTHi-specific serum immunoglobulins are specific to P2, P5, and P6 (12, 282, 283). However, decreased mucosal antibodies associated with sIgA deficiency, or decreased total IgG in the small airways have been reported in COPD patients, and might be associated with disease severity (283, 284). Importantly, NTHi-specific mucosal sIgA has been found to be lower in the airways of NTHi-infected COPD patients than the non-colonized patients (285, 286).

The epithelial polymeric immunoglobulin receptor (pIgR) is essential for the generation of mucosal sIgA. It is, however, downregulated in COPD patients with a positive correlation to disease severity and increased level of TGF-β (287). The combinatorial effects of downregulated plgR and elevated TGF-β1 contribute to an impaired mucosal IgA immunity in COPD patients. A mouse model lacking the pIgR (^−/−) is therefore devoid of sIgA and are susceptible to airway stimulation by an NTHi lysate resulting in increased inflammation and airway neutrophilia. Interestingly, introduction of exogenously added sIgA mitigated the airway inflammation (288). NTHi-infected COPD patients with greater airway inflammation have also decreased NTHi-specific mucosal IgG1 in the BAL fluid compared to the non-colonized patients (283). Interestingly, the phenomenon with decreased NTHi-specific antibodies seems to be restricted to the airways, since the specific serum antibodies are not affected. Therefore, the reduced mucosal IgG is unlikely to be associated with hypogammaglobulinemia (IgG deficiency), despite the latter was reported as a contributing factor in NTHi infection (289). Decreased airway IgA might be attributed to the expression of IgA proteases by NTHi. The bacterial IgA protease degrades the local airway IgA during airway colonization to avoid immune exclusion by sIgA (226, 228). Reduced mucosal antibodies might promote host immune evasion and resistance to complement-mediated killing of NTHi, thus enable persistent colonization of NTHi in the airways of COPD patients, in addition to a plethora of various other virulence mechanisms (4, 5, 207).

Whilst the role of macrophage extracellular traps (MET) for killing of NTHi remains unknown, it has been reported that blood neutrophils and NET from COPD patients are defective in the killing of planktonic or biofilm/NET-entrapped NTHi, respectively (263, 264, 290). A series of studies by Berenson et al. revealed that alveolar macrophages derived from COPD patients are basically dysfunctional in eradication of NTHi (177, 291–293). Intriguingly, TLR2 and TLR4 expressed on alveolar macrophages from COPD patients are intrinsically unresponsive to the potent immunomodulatory lipoprotein P6 and LOS, respectively. This causes decreased LOS/P6-induced expression of TLRs, reduced NF-κB nuclear activation and consequently diminished IL-8, TNF-α, and IL-1β responses by alveolar macrophages from COPD patients. The compromised TLR expression and signaling potentially contribute to the defective complement-dependent and independent phagocytosis of NTHi. The defective phagocytosis is greater for NTHi than for M. catarrhalis, and correlates with disease severity. Interestingly, the phagocytosis disability was not detected in monocyte-derived macrophages in COPD. In contrast, however, Taylor et al. reported that monocyte-derived macrophages from COPD patients are also defective in phagocytosis of NTHi and S. pneumoniae. The author also suggested that the defective monocyte-derived macrophages are not attributed to the alteration in cell surface TLR2 or TLR4 expression, macrophage receptor with collagenous structure (MARCO), CD163, CD36 or the mannose receptor (178). The unresponsive TLR2 and TLR4 in COPD alveolar macrophages to NTHi lipoprotein and LOS might be explained by the recently reported phenomenon of TLR tolerance (294). Repetitive stimulation of COPD alveolar macrophages with the same TLR ligands, Pam3CSK4 and LPS desensitizes the TLR2 and TLR4, respectively, and generates TLR tolerance. Moreover, the repetitive TLR stimulation further reduced the production of TNF-α, CCL5, and IL-10 without affecting the constantly augmented level of IL-6 and IL-8 in alveolar macrophages. This may provide alternative explanations for diminished immune responses against the recurrent/repetitive infection by NTHi.

The intrinsically reduced expression of TLRs in COPD patients may also contribute to the impaired pulmonary immune response thus facilitating NTHi persistent colonization. Expression of TLR2 or TLR4 are found to be lower on sputum neutrophils, alveolar macrophages, nasal epithelium, and T cells in COPD patients despite high concentrations of IL-8 and MMP-9 (295–298).

The lack of the more protective Th1/Tc1 immune response in COPD patients against NTHi infection might be attributed to upregulated antagonists (A20, IRAK-M, and MyD88s) of the MyD88/IRAK/MAPK signaling pathway in COPD T cells (295). It should be noted that the MyD88/IRAK/MAPK pathway is required for expression of TLR4 in Th1, whereas production of IFN-γ in Th1/Tc1 is TLR4-dependent via the TLR4/TRIF/IKKe/TBK1 signaling pathway. The antagonists prevent the NTHi LOS-induced TLR4 expression in Th1 and Tc1 and thus a reduced secretion of IFN-γ. In addition, unusual high numbers of Tregs in COPD patients have also contributed to effector T cell dysfunction or a Th2/Tc2 predominant immune response (272). However, Freeman et al. reported that Tc (CD8⁺) cells from COPD patients have increased expression of TLR1, TLR2, TLR4, TLR6, and TLR2/1 as well as Tc1 cytokines (IFN-γ and TNF-α) compared to healthy individuals that may imply the auto-aggressive response of lung Tc cells in COPD lung inflammation (299). However, the COPD Tc cells can only be stimulated by ligands for TLR2/1 (Pam3CSK4) yet tolerant to other agonists, indicating the dysfunctional TLRs or TLR tolerance on T cells despite their high level of receptor expression.

Inversely, peripheral blood neutrophils isolated from COPD patients have increased expression of TLR2, TLR4, and NLRP3 (298, 300). Nevertheless, the increased TLRs expression might not improve the microbicidal ability of COPD peripheral neutrophils probably due to the inaccurate responses to cytokines (179). In addition, certain types of SNPs (SNPs) in TLR2 and TLR4 have also been associated with decreased lung function, enhanced inflammatory responses and increased immune cell infiltration in COPD (301). Interestingly, the diminished IL-8 responsiveness of COPD alveolar macrophage to NTHi infection has a strong association with the carriage of TLR9 (T1237C) polymorphism instead of TLR2 (Arg753Gln), TLR4 (Thr399Ile; Asp299Gly), and TLR9 (T1486C) (302). The carriage of TLR9 (T1237C) is also positively correlated with diminished lung function. Of note, the activation of TLR9-signaling cascade in pro-inflammatory cytokine response requires stimulation from microbial DNA (303).

The microbicidal malfunction in both innate and adaptive immune cells is also potentially linked to the deleterious effect of tobacco smoke, the major risk factor for COPD. It has been reported that, exposure of tobacco or cigarette smoke can impair phagocytosis/engulfment of NTHi by alveolar macrophages isolated from COPD patients (256, 304). Moreover, the chemical exposure also suppressed the TLR-induced TNF-α, IL-6, and IL-10 production in COPD alveolar macrophages that have been pre-stimulated with TLR2, 4, or 5 ligands (Pam3CSK4, LPS, or phase I flagellin, respectively), or whole NTHi bacteria (305). This may potentially delay the macrophage-dependent bacterial clearance. The suppressive effect of cigarette smoke in macrophage-dependent phagocytosis is due to the suppression of the PI3K signaling cascade which is required for optimal phagocytic activity and movement (256). Meanwhile, the cigarette smoke also inhibits the activation of the p38-ERK signaling pathway and p65/NF-κB, thus dampens the NTHi LOS-induced cytokine production of COPD alveolar macrophages (305). The diminished alveolar macrophage responsiveness could also be related to anticholinergic agents used by COPD patients that results in lower concentrations of NTHi-induced TNF-α (306). Nevertheless, the impaired phagocytosis of NTHi by COPD alveolar macrophages could be improved in the presence of nuclear erythroid related factor 2 and microRNA MiR-328 (307, 308). Interestingly, in addition to the constant exacerbated inflammatory effect observed in different murine model studies, Gaschler et al. observed a rapid pulmonary clearance of NTHi in mice upon exposure to cigarette smoke, and this was positively correlated with an increased neutrophilia in the animal BAL fluid (236, 309–311). However, in other COPD animal studies, cigarette smoke also impaired the IL-22 production that has a potential anti-bacterial activity while delaying the airway clearance of NTHi (311–313). Interestingly, IL-22 might play a protective role in COPD exacerbation as supplementation of IL-22 manages to restore the homeostasis of airway immune response and improve NTHi clearance (313).

The increased airway neutrophilia might be due to the enhanced production of pulmonary IL-17 triggered by cigarette smoke (152, 236, 311, 314). This may imply the important microbicidal role of neutrophils (neutrophilia) in compensating the COPD- or cigarette smoke-associated dysfunctional alveolar macrophages (96, 315). However, such compensation may not be adequate to provide optimal immune defense to eradicate persistent NTHi lower airway colonization, since the cigarette smoke also has profound suppressive effect on the host adaptive immunity, thus constantly risking the COPD patients to episodes of exacerbation and relapsed infection. In adaptive immunity, cigarette smoke impairs the antigen-specific B and T cells responses to NTHi infection. It suppresses the secretion of IFN-γ and IL-4 by NTHi-specific T cells. Antibody production by B cells has also been attenuated, with lower levels of specific anti-P6 antibodies and compromised IgG1, IgG2a, and IgA class switching (311, 312).

A recent and some previous cohort studies revealed that the level of airway antimicrobial cathelicidin (hCAP18/LL-37) in COPD patients increase gradually from the stable disease to exacerbation states (176, 316). Moreover, higher levels of cathelicidin are positively associated with NTHi airway colonization, sputum neutrophilia, and higher concentrations of IL-8, particularly in the NTHi-infected COPD patients. Of note, cathelicidin and other AMPs play important roles in the innate immune defense against different pathogens and persist immunomodulatory properties (317–319). Ironically, it is plausible that the increased level of cathelicidin could diminish or alter the balance in lung microbiota, and the immune/inflammatory response. This might contribute to the “vicious circle,” thus considerably increasing the risk for NTHi infection during COPD exacerbations (214, 320). Moreover, the microbicidal property of cathelicidin could be compromised by the inflammatory conditions in the airway, such as low pH, or the effect of cigarettes that causes peptide citrulination and modification (321, 322). Finally, expression of AMPs (human beta defensin 2 and S100A7) by COPD airway epithelium in response to NTHi infection, is also disturbed by cigarette smoke. The insulted airway cells have also a reduced expression of TLR4 and IL-8, and impaired NTHi-induced NF-κB activation (175, 296). Thus, a large body of evidence exists on the deleterious effects of tobacco smoke.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.