Granzyme B May Act as an Effector Molecule to Control the Inflammatory Process in COPD

Abstract

Chronic obstructive pulmonary disease (COPD) is caused by smoking, but only a small proportion of smokers have disease severe enough to develop COPD. COPD is not always progressive. The question then arises as to what explains the different trajectories of COPD. The role of autoimmunity and regulatory T (Treg) cells in the pathogenesis of COPD is increasingly being recognized. Nine published studies on Treg cells in the lung tissue or bronchoalveolar lavage fluid have shown that smokers with COPD have fewer Treg cells than smokers without COPD or nonsmokers. Three studies showed a positive correlation between Treg cell count and FEV₁%, suggesting an important role for Treg cells in COPD progression. Treg cells can regulate immunological responses via the granzyme B (GzmB) pathway. Immunohistochemical staining for GzmB in surgically resected lungs with centrilobular emphysema showed that the relationship between the amount of GzmB+ cells and FEV₁% was comparable to that between Treg cell count and FEV₁% in the COPD lung, suggesting that GzmB could be a functional marker for Treg cells. The volume fraction of GzmB+  cells in the small airways, the number of alveolar GzmB+ cells, and GzmB expression measured by enzyme-linked immunosorbent assay in the lung tissue of smokers were significantly correlated with FEV₁%. These results suggest that the GzmB content in lung tissue may determine the progression of COPD by acting as an effector molecule to control inflammatory process. Interventions to augment GzmB-producing immunosuppressive cells in the early stages of COPD could help prevent or delay COPD progression.

Keywords:

• COPD
• autoimmunity
• regulatory T cell
• granzyme B
• centrilobular emphysema

Introduction

Chronic obstructive pulmonary disease (COPD) is generally triggered by tobacco smoking. The classic concept is that smoking induces an abnormal inflammatory response in susceptible individuals, damaging the airways (bronchitis and bronchiolitis) and alveoli (emphysema), accelerating the physiological decrease in lung function with age and leading to airflow obstruction and chronic respiratory symptoms [1]. Therefore, COPD is considered a progressive disease that occurs only in susceptible individuals.

However, this traditional concept does not always seem to apply in practice. Tobacco smoking causes histopathological abnormalities in the small airways, but only a small proportion of smokers have abnormalities that are severe enough to develop COPD [2]. This is consistent with epidemiological studies showing that only approximately 15% of smokers develop COPD [3]. Although tobacco smoking is a major risk factor of COPD, not all smokers develop COPD. However, pathologically, there are no qualitative differences between the lungs of smokers with and without airflow limitation [4]. Moreover, the observation posits that pathological abnormalities in the small airways do not deteriorate indefinitely but reach a morphological plateau [5]. In keeping with this observation, COPD is not always progressive in real life. In more than half of 2,163 patients with COPD, the rate of decrease in FEV₁ over three years was not greater than that in people without lung disease [6]. Therefore, the course of COPD varies, with some patients having a relatively stable course, whereas others experience an inexorable decline leading to severe shortness of breath. The question then arises as to what explains the different severities and courses of COPD among smokers. Although radiological and biological markers (biomarkers) can help predict disease progression [7], the bioactive molecules that are directly responsible for COPD progression are unknown.

We review the possible role of autoimmunity and specifically regulatory T (Treg) cells in the pathogenesis of COPD by evaluating the published literatures on the role of Treg cells in the blood and lungs of patients with COPD, as well as granzyme B (GzmB)-positive cells in surgically resected COPD lungs. Here, we found that the association between the amount of GzmB-positive cells and the degree of airflow limitation in the COPD lung is comparable to that between the number of Treg cells and the degree of airflow limitation in patients with COPD. We propose that granzyme B is an effector molecule in regulatory cells to control the inflammatory process during COPD progression. We also emphasize the therapeutic potential of promoting GzmB-producing immunosuppressive cells to prevent or delay disease progression in the early stages of COPD.

Pathogenesis of COPD: importance of airway count and attrition

As noted previously, remodeling and inflammatory changes in the airways of smokers with and without airflow limitation are morphologically subtle, although they become more exaggerated in GOLD 3 and 4 disease severity. However, even in very mild disease (GOLD 1), there are large differences in the number of small airways (consisting of terminal, transitional, and respiratory bronchioles) between those with GOLD 1 COPD and smokers without airflow limitation. Using micro-computed tomography of lung tissue sections from smokers with and without airflow limitation, Koo et al. [8] showed that lungs of patients with very mild GOLD 1 COPD (FEV₁ > 80%) had 56% fewer transitional bronchioles than smokers without any airflow limitation. When comparing GOLD 4 COPD lungs and control lungs, this value increased to 90%. These data suggest that airway loss or attrition plays an important role in the pathogenesis of COPD.

At the cellular level, destruction of airways begins with cell death in the form of apoptosis. Consistent with this notion, many studies have emphasized the existence of apoptosis in the lungs of patients with COPD. A growing body of data from both animal and human studies demonstrates the important role of apoptosis in the pathogenesis of COPD [9]. Although the pathways and molecules responsible for apoptosis are complex and beyond the scope of this review, granzymes constitute an important family of enzymes responsible for inducing apoptosis and cell death. At least five granzyme isoforms have been identified in humans. The most powerful enzyme is GzmB, produced by CD8+ T and natural killer cells [10]. When these cells encounter their target cells, they release secretory lysosomes containing perforin (which creates holes in the target cells to allow granzyme entry) and granzymes through exocytosis [11]. Once GzmB enters the cytoplasm of target cells, it binds to multiple protein substrates and triggers a cascade of events that leads to apoptotic cell death. Among these substrates, pro-caspase 3 is the most notable. Once the cascade is initiated, GzmB induces mitochondrial damage and the release of cytochrome C, leading to the activation of caspase 9 (also known as the apoptosome) and cell death [12].

Studies have shown increased numbers of CD8+ T cells in the small airways [13, 14] and lung parenchyma [15] of the surgically resected lungs of smokers with COPD. A significant association between the number of cytotoxic CD8+ T cells and decreased lung function in patients with COPD has also been demonstrated [13], indicating that these cells play a role in the pathogenesis of COPD. A recent systematic survey of 48 published articles on the function of CD8+ T cells in COPD found that not only are the numbers of CD8+ T cells increased in COPD, but these cells also have an increased ability to exert effector functions, which likely contribute to the pathogenesis of the disease [16]. Apoptosis triggered by CD8+ T cell-induced cytolysis has been proposed as a mechanism of lung destruction in COPD.

Autoimmunity in COPD

Autoimmunity is increasingly recognized to play an important role in the pathogenesis of COPD [17]. It has been suggested that, in the pathogenesis of COPD, CD8+ T cells are recruited and activated into GzmB-positive cells by tobacco-modified self-antigenic stimulation. The increasing awareness that COPD shares common features with autoimmune diseases has led to interest in the probable role of T cells, which are inextricably involved in the control of autoimmunity.

Regulatory T cell

Treg cells are a specialized subpopulation of T cells that govern unwanted immune responses. They play key roles in maintaining homeostasis within the immune system, including the innate and adaptive immune networks, and regulate inflammatory processes occurring in tissue injury, transplant rejection, and autoimmunity [18]. Since their discovery, FoxP3+ CD4+ cells have been identified as potent suppressors of T cell responses. Treg cells suppress various immune cells, particularly B cells, CD4+ T cells, CD8+ cytotoxic T cells, NK cells, NKT cells, macrophages, dendritic cells, neutrophils, and T cells [19]. In addition to their immunosuppressive action, Treg cells in nonlymphoid tissues promote tissue homeostasis, regeneration and repair [20].

The immune system is constantly confronted with antigens that arise not only from invading foreign pathogens but also from host tissues. Highly effective algorithms are in place to eradicate harmful pathogens through indiscriminate innate immunity and advanced antigen-specific adaptive immunity. Of the regulatory lymphocyte subtypes, FoxP3-expressing CD4+ T cells are the most potent and can be broadly divided into two main groups: natural Treg (nTreg) cells, which develop in the thymus, and induced or adaptive Treg (iTreg) cells, which develop in the periphery from naïve, conventional T cells. Both nTreg and iTreg cells play their own roles in the immune system, with nTreg cells mainly maintaining tolerance to self-structures, and iTreg cells developing in response to externally supplied antigens or commensal microbes. Furthermore, Treg cells acquire tissue-specific properties and adapt to function in the tissues in which they reside [21]. Upon encountering their cognate (self-)antigen, T cell receptor stimulation induces transcriptional modifications that cause naïve-like thymus-derived Treg cells circulating in secondary lymphoid organs to differentiate into effector Treg (eTreg) cells. These eTreg cells exhibit high proliferation rates and superior suppressive functions, and migrate to the peripheral tissues and control local immune and tissue homeostasis [22].

Immune safety in the lungs is attained through the innate and adaptive arms of the immune system. Innate immunity is exercised quickly by leukocytes such as macrophages and neutrophils, which indiscriminately eradicate pathogens and release detrimental substances. In contrast, adaptive immunity involving T and B lymphocytes and dendritic cells develops more slowly but is specific [23]. Airways are constantly exposed to inhaled antigens and pathogens; however, this delicate mucosal tissue must maintain a state of relative quiescence to maintain tissue design and enable gas exchange. In this context, overt immune responses to diseases, such as infection, must be strictly regulated. Treg cells induce multiple mechanisms that govern airway tissue homeostasis, by promoting effector immune responses, reducing inflammation, and enabling tissue maintenance and healing [24].

Appropriate migration and retention of Treg cells ensures that Treg cell suppression occurs in a microenvironment where regulation is necessary. Emerging data suggest that the compartmentalization and trafficking of Treg cells may exhibit tissue and/or organ specificity. The distinct expression patterns of chemokine receptors and integrins and their responsiveness contribute to the selective retention and trafficking of Treg cells [25]. In addition to Treg cells which are constitutively present in tissues, Treg cells can also be quickly recruited to sites of inflammation. The recruitment of T cells to inflamed tissues is a consequence of dramatic alterations in the expression of chemokines, adhesion molecules, and extracellular matrix components that occur during tissue inflammatory responses [26]. One of the main functions of Treg cells in nonlymphoid tissues is to regulate local inflammation [27].

Regulatory T cells in COPD

Many studies have examined the association between Treg cells and COPD by measuring the amount or function of Treg cells in the blood, bronchoalveolar lavage (BAL) fluid, and lung tissue of patients with COPD. Plumb et al. [28] reported increased numbers of Treg cells in the lymphocyte follicles in the parenchyma of surgically resected COPD lungs compared with those in normal smokers and nonsmokers. The concentration of Treg cells in the peripheral blood of patients with COPD has been reported to vary. Increased levels of blood Treg cells have been noted in patients with COPD compared to healthy subjects [29], and increased blood Treg cell subsets were present in COPD and normal smokers compared to healthy subjects [30]. However, other studies have shown that there are fewer blood Treg cells in patients with COPD compared to normal smokers and healthy controls [31–33]. There were more circulating Treg cells in patients during acute exacerbation of COPD (AECOPD) than in stable COPD or healthy controls [34, 35]. AECOPD patients had higher proportion of Treg cells in blood compared with normal smokers, but lower levels than stable COPD [36].

Beyond just numbers of Treg cells, there are reports of functional change in Treg cells in COPD. Roos-Engstrand et al. [37], for example, reported that smokers with COPD had an increased proportion of CD127⁺ helper T cells in BAL fluid, indicating the expansion of a T cell population without regulatory function. Tan et al. [38] reported impaired Treg cell-mediated suppression of CD4(+) T-cell activation in the blood of a subset of patients with COPD, all of whom had a high body mass index, suggesting that obesity and/or impaired homeostasis of Treg cell subsets may contribute to the increased inflammation observed in COPD. Based on miRNA microarray analysis, Chatila et al. [39] reported that COPD had a significant impact on miRNA expression profiles in blood Treg cells. They found that the miR-199a-5p response in COPD Treg cells was significantly blunted compared to Treg cells in healthy smokers, suggesting that the abnormal repression of miR-199a-5p in patients with COPD may be involved in modulating the adaptive immune balance.

Lee et al. [40] reported significantly fewer Treg cells (CD4⁺/CD25^hi/CD62L⁺) in the lungs of individuals with emphysema than in controls. In addition, the levels of elastin antibodies in the plasma of patients with emphysema were significantly higher than those in the control group. However, similar counts of Treg cells were found in the peripheral blood of controls and patients with emphysema, and their function was normal. The authors suggested that exposure to cigarette smoke induces secretion of proteolytic enzymes from innate immune cells leading to the release of lung elastin fragments, which, in turn, could induce T- and B-cell-mediated immunity to elastin in the lungs. Smyth et al. [41] reported incremental changes in Treg cells, defined by CD4CD25(bright) expression, in the BAL fluid of normal smokers and patients with COPD compared with healthy nonsmokers. They argued that the number of Treg cells increases in smokers to control cigarette smoke-induced inflammation. However, although smokers with COPD showed increased CD8+ cell counts, there was no further increase in the number of Treg cells. While there was no difference in the number of Treg cells between patients with COPD and normal smokers, the CD8 to Treg cell ratio in the BAL fluid was considerably higher in patients with COPD than in normal smokers. This suggests a relative paucity of Treg cells in the small airways of patients with COPD, which may contribute to uncontrolled lung inflammation. Treg cells are more abundant in the lungs than in the blood of patients with COPD and normal smokers, but not in healthy nonsmokers. There was no significant correlation between Treg cell count in BAL fluid and FEV₁% predicted in patients with COPD. Barcelo et al. [42] reported that patients with COPD have fewer CD4+ CD25+ Treg cells in the BAL fluid than normal smokers, and hypothesized that a Treg cell count insufficient to suppress cytolytic CD8+ T cells could contribute to COPD. They also observed that smokers with preserved lung function had a significant upregulation of Treg cells compared to nonsmokers, which was not observed in smokers with COPD. These observations demonstrate an attenuated Treg cell response to tobacco smoking in patients with COPD. This upregulation in smokers with normal lung function can be interpreted as an effort to regulate and minimize the inflammatory response induced by tobacco smoking, whereas the lack of this mechanism in smokers with COPD may contribute to the enhancement and/or dysregulation of the inflammatory response, which may contribute to the pathogenesis of the disease. However, no large differences were detected in peripheral blood Treg cells among the groups, whereas Treg cells in the BAL fluid were higher than those in the peripheral blood in the three groups. These data raise the possibility of physiological compartmentalization of T-lymphocyte trafficking. Isajevs et al. [43] showed elevated FoxP3 expression in the large airways of smokers with and without COPD, but reduced numbers of FoxP3-positive cells in the small airways of patients with COPD compared to normal smokers and nonsmokers in surgically resected lungs. There was a positive association between FoxP3⁺ cell counts in the small airways of COPD and FEV₁% predicted. Chu et al. [44] showed reduced FoxP3⁺ expression in the alveolar walls of individuals with COPD compared to that in normal smokers and nonsmokers in surgically resected lung tissues. The FoxP3/RORγt (a master regulator in the development of Th17 cells) ratio correlated positively with patients’ FEV₁% predicted. Hou et al. [45] reported that peripheral blood Treg cell counts in patients with COPD were similar to those in non-COPD smokers and nonsmokers using the conventional definition of CD4⁺CD25⁺FoxP3⁺; however, the proportions of resting Treg (rTreg) and activated Treg (aTreg) cells were reduced in patients with COPD compared to nonsmokers and normal smokers. In the BAL fluid, aTreg cells were also reduced in patients with COPD compared to those in normal smokers and nonsmokers. In addition, aTreg cells were upregulated in smokers without COPD compared to nonsmokers. There was a strong correlation between the ratio of (aTreg + rTreg):(Fr III, cytokine-secreting subset of Treg cells without suppressive activity) in BAL fluid and patients’ FEV₁% predicted. Sales et al. [46] demonstrated reduced FoxP3⁺ Treg cell density in the small airways of patients with COPD compared to that in nonsmokers or normal smokers in surgically resected lungs. Zheng et al. [47] reported that FoxP3⁺ Treg cell counts were considerably lower in patients with COPD than in nonsmokers and normal smokers in surgically resected lung tissue. In peripheral blood, the Treg/CD4+ T cell percentage was considerably lower in patients with COPD than in nonsmokers or normal smokers. Blood Treg/CD4+ T cell percentage was positively correlated with FEV₁%. Sileikiene et al. [48] reported lower levels of Treg (CD4+ CD25+) cells in the bronchial mucosa and Treg (CD3+ CD4+ CD25+^bright) cells in the blood of patients with severe/very severe COPD and nonsmokers compared to those with mild/moderate COPD and normal smokers. They suggested that higher Treg cell concentrations in the blood and mucosa of normal smokers may protect the lung from development of COPD. Lourenço et al. [49] reported that there is increased FOXP3 gene expression in the resected lungs of patients with GOLD I and II COPD and found decreased FOXP3 expression in the peripheral blood of patients with all stages of COPD compared to those of normal smokers. Hou et al. [50] recently showed that highly suppressive Treg cells, defined as the GARP⁺ subpopulation, which has been associated with a specific and functionally relevant marker expressed by activated nTreg cells, were decreased in the BAL fluid of patients with COPD and were related to the severity of the multi-organ loss of tissue phenotype of COPD. The frequency of FoxP3⁺GARP⁺ Treg cells was also reduced in the peripheral blood of patients with COPD. Ström et al. [51] reported a study that estimated the rate of pulmonary function decline over 5-year period. Patients with COPD with a fast decline had a substantially lower proportion of Treg cells (FoxP3⁺/CD4⁺CD25^bright) in the BAL fluid than patients with a non-fast decline.

In summary, five published studies have shown an increase in Treg cells in the peripheral blood of patients with COPD, including three studies in patients who were experiencing an acute exacerbation of COPD. Eight studies have shown decreased Treg cells or FOXP3 gene expression in the peripheral blood of patients with COPD. Three studies reported functional changes in blood Treg cells in patients with COPD. Nine studies on Treg cells in the small airways, alveoli, lung tissue, bronchial mucosa, or BAL fluid showed reduced Treg cell numbers in smokers with COPD compared to normal smokers or nonsmokers, although the studies used several different markers for Treg cells. Two studies showed increased Treg cells or FOXP3 gene expression in BAL fluid or resected lung tissue of COPD.

Regarding Treg cells in the peripheral blood, two studies showed no difference in the number of Treg cells between patients with COPD and normal smokers [40, 42], whereas four studies showed a reduction in Treg cells in the blood, with a concomitant reduction in Treg cells in the lung tissue, bronchial mucosa, or BAL fluid in patients with COPD [45, 47, 48, 50]. In contrast, two studies showed that Treg cells were higher in lung tissue or BAL fluid than in peripheral blood [41, 42]. Three studies showed upregulation of Treg cells in smokers with normal lung function in the BAL fluid or bronchial mucosa compared to that in nonsmokers [42, 45, 48]. Three studies showed a significant association between Treg cell counts in lung tissue or BAL fluid and FEV₁% predicted [43–45] and one study showed no association between them [41]. One study showed positive correlation between Treg/CD4+ T cell percentage in blood and FEV₁% predicted [47]. Overall, most published studies showed fewer Treg cells in peripheral blood, BAL fluid, or lung tissue of smokers with COPD compared to normal smokers and nonsmokers, and there is a positive association between the number of Treg cells in the lung and FEV₁% predicted.

The heterogeneity of Treg cells and the lack of exclusive marker for Treg cell

FoxP3-expressing CD4+ thymus-derived naturally occurring Treg cells play an essential role in maintaining self-tolerance and immune homeostasis. However, there is confusion regarding the identity, stability, and suppressive functions of human Treg cells.

Human Treg cells were first defined by their high CD25 expression, based on the finding that murine CD25+ CD4+ T cells demonstrated a strong suppressive effect. With the discovery of FOXP3 as the master control gene for CD4+ Treg cell development and function, these cells (human CD25^hiCD4⁺ T cells) were confirmed to be regulatory by the presence of FoxP3 expression [52]. However, human FoxP3+ cells may not be homogeneous. Various surface molecules have been proposed as Treg cell markers; however, none are expressed solely in Treg cells. The accurate identification of Treg cells is challenging because the expressed proteins are shared by activated classical effector T cells. In humans, low levels of FoxP3 expression in a significant number of activated, non-suppressive conventional T cells make it difficult to clearly identify a cell as a Treg cell based solely on FoxP3 positivity. Therefore, the use of FoxP3 as a specific indicator of human nTreg cells has been increasingly questioned [53].

FoxP3+ CD4+ T cells circulating in human blood can be divided into three main fractions using FoxP3, CD25, and CD45RA, indicators of naïve T cells: fraction I (Fr. I) CD45RA⁺FoxP3^lo/CD25^lo resting or naïve Treg cells, Fr. II CD45RA^-FoxP3^hi/CD25^hi effector Treg (eTreg) cells, and Fr. III CD45RA^-FoxP3^lo/CD25^lo cells, most of which are non-Treg cells with no or limited suppressive activity [52]. Therefore, FoxP3+ T cells are heterogeneous and encompass naïve and effector Treg cells, as well as non-Treg cells. Furthermore, assessing Treg cell counts alone can be misleading; for example, mice with severely disrupted Treg cell function may have an increased number of Treg cells. This highlights the need for a more thorough evaluation of both the phenotype and function of Treg cells to understand their role in autoimmunity [54]. Despite progress in Treg cell biology, a new foundation for reliably delineating human Treg cells remains necessary [55].

Granzyme B as an effector molecule of Treg cells

Suppressive mechanisms initiated by Treg cells can be divided into four basic modes of action: modulating the maturation or function of antigen-presenting cells, killing target cells, metabolic disruption, and inhibitory cytokines [56]. These processes are activated in different situations. Target cell killing is a mechanism of Treg cell-mediated suppression through cell contact-dependent cytolysis of target cells mediated by granzymes A (GzmA) and B (GzmB) in a perforin-dependent or -independent manner. The perforin-granzyme pathway is not only crucial for the function of NK and CD8+ T cells but can also be used by Treg cells to limit the activity of these cells. Thus, the granzyme pathway is one of the mechanisms through which Treg cells control immunological responses [57]. In addition to CD4+/CD8+ T cells and NK cells, active GzmB expression has been observed in B cells, dendritic cells, macrophages, mast cells, basophils, keratinocytes, and chondrocytes. To date, it is known that Treg cells, regulatory B cells and plasmacytoid dendritic cells secrete GzmB for immunosuppressive purposes.

The connection between GzmB and the immunosuppressive effects of Treg cells was initially demonstrated by the frequent presence of GzmB+ Treg cells in malignant tumors. An increasing number of studies have demonstrated that a high rate of tumor-infiltrating Treg cells in various solid tumors is significantly associated with poor prognosis. Furthermore, high numbers of GzmB+ Treg cells were negatively correlated with the occurrence of acute transplant-versus-host disease after hematopoietic stem cell transplantation, implying a regulatory influence on adaptive immunity [58].

Salti et al. [59] characterized the immune response in wild-type and GzmB-deficient mice infected with Sendai virus. GzmB deficiency resulted in a substantial increase in the number of CD8+ T cells in the lungs of virus-infected animals. Treg cells from wild-type mice expressed high levels of GzmB in response to infection, and the depletion of Treg cells from these mice resulted in an increase in the number of CD8+ T cells similar to that seen in mice with GzmB deficiency. Furthermore, GzmB-deficient Treg cells failed to suppress CD8+ T cell proliferation in vitro. These findings indicated that GzmB in the Treg cell compartment plays a crucial role in immune regulation during viral infections. Loebbermann et al. [60] found that selective depletion of Treg cells during acute respiratory syncytial virus (RSV) infection increased viral clearance and cell influx into the lungs of mice. Conversely, inflammation was reduced when Treg cell numbers and activity increased. Lung Treg cells from RSV-infected mice expressed GzmB, and bone marrow chimeric mice with a selective loss of GzmB in the Treg cell compartment showed a significant increase in cellular infiltration. They argued that GzmB-expressing Treg cells play a critical role in resolving acute infections. Interestingly, the pathologic findings in RSV infected mice with defective immune regulation closely resembled that observed in children with RSV-related bronchiolitis, suggesting an important role of Treg cells in the pathogenesis of bronchiolitis induced by RSV. Consistent with this observation, Bem et al. [61] reported increased GzmB expression in CD4+ T cells in the BAL fluid of children with severe RSV infections.

Granzyme B as a possible functional marker for regulatory cells in COPD

As mentioned earlier, the function of FoxP3 as a specific indicator of Treg cells has been questioned, and it remains unclear whether specific molecular markers are expressed only in Treg cells. Although the overall number of Treg cells in tissue is important, the functionality of these cells in vivo is of central importance for their regulatory effect [62]. There have been reports showing changes in the function of Treg cells in the blood of patients with COPD [37–39], suggesting that a functional assay is essential in determining the role of Treg cells in the pathogenesis and progression of COPD.

We [63] performed immunohistochemical staining with antibodies against CD8 and GzmB in surgically resected lungs with microscopically validated centrilobular emphysema (CLE) and small airway wall remodeling (Figure 1A). The aim was to investigate the role of GzmB in the cytolytic function of CD8+ cells in the pathogenesis of COPD. The results showed that the volume fraction (the volume of stained cells contained in the unit volume of tissue) of CD8+ and GzmB+ cells increased in the small airways of CLE lungs compared to those in nonsmoker lungs; however, the volume fraction of GzmB+ cells was unexpectedly higher than that of CD8+ cells (Figure 1B).

Figure 1. Small airway wall thickness (A) and volume fraction (Vv) of CD8 and granzyme B-positive cells (B) in small airway walls of nonsmoking control, panlobular emphysema (PLE) and centrilobular emphysema (CLE) lungs [63].

Furthermore, the localization of stained GzmB+ cells did not match that of CD8+ cells in serially sectioned slides from a small airway wall in the CLE group (Figure 2A,B). Separate CD8+ and GzmB+ cells were identified by double staining with CD8 and GzmB (Figure 2C). One double-stained cell (black arrow) and many CD8+ cells that were not positive for GzmB were present (Figure 2D).

Figure 2. CD8 (brown; A) and GzmB (red; B) staining in serially sectioned slides of the small airway wall of a subject with centrilobular emphysema (X 200). Double staining for CD8 (brown) and granzyme B (red) in the small airway wall of the same subject (C and D respectively; X 400) [63].

It has been previously reported that only recently activated CD8+ T cells are capable of GzmB production [64]. In chronic infection or cancer, CD8+ T cells are exposed to persistent antigenic and/or inflammatory signals, leading to T cell exhaustion [65] and progressive loss of effector function with reduced production of perforin and GzmB [66]. In addition to T cell exhaustion, senescent T cells with defects in proliferation and effector functions accumulate with age [67] and showed lower levels of perforin and GzmB after stimulation [68]. Although it is believed that CD8+ T cells may exert a GzmB-mediated cytolytic function in early stages of COPD, these reports support the notion that GzmB+ cells in the above old subjects with CLE (mean age: 63.8$\pm$7.7 years) are not necessarily CD8+ T cells. Consistent with this assumption, not all GzmB+ cells were CD8 positive and not all CD8+ cells were GzmB positive by double staining. As noted previously, Treg cells use GzmB to suppress cytotoxic CD8+ T cells [57]. Thus, there is a high probability that most of these GzmB+ cells are Treg cells.

In another of our studies [69], CD8 and GzmB staining of the alveolar walls of surgically resected lungs with CLE showed that the number of alveolar GzmB+ cells/mm was closely correlated with FEV₁% predicted (Figure 3A), although there was no significant relationship between the number of alveolar CD8+ cells and FEV₁% predicted (Figure 3B). This close correlation between the number of alveolar GzmB+ cells and FEV₁% predicted is similar to that observed between the number of Treg cells in lung tissue or BAL fluid and FEV₁% predicted in three previously published studies [43–45].

Figure 3. The relationship between the number of alveolar granzyme B+ cells and FEV₁% predicted (A) and that between CD8+ cells and FEV₁% predicted (B) in subjects with centrilobular emphysema [69].

This study also showed that the number of alveolar GzmB+ cells/mm was significantly greater in cases with FEV₁ 80% or more than in cases with FEV₁ less than 80% (Figure 4A), although there was no difference in the number of alveolar CD8+ cells between two groups (Figure 4B). This finding suggests that the roles of GzmB+ and CD8+ cells in the pathogenesis of COPD are different. Although the number of cytotoxic CD8+ T cells increased equally in both groups, the number of GzmB+ cells was higher in patients with relatively preserved lung function than in patients with reduced airflow, suggesting that the degree of lung damage and thus the degree of airflow limitation is determined by the number of GzmB+ cells. The higher GzmB+ cells in subjects with relatively preserved airflow than in those with low airflow are consistent with the results of previous reports showing upregulation of Treg cells in smokers without airflow limitation compared to nonsmokers.

Figure 4. Number of GzmB+ cells (A) and CD8+ cells (B) in the alveolar walls of subjects with centrilobular emphysema with FEV₁ below 80% predicted and those with FEV₁ 80% predicted or more [69].

Taken together, these two findings show that the relationship between GzmB+ cells and FEV₁ in CLE lungs is comparable to that between Treg cells in lung tissue or BAL fluid and FEV₁ in COPD. It suggests that GzmB may be a functional marker of Treg cells. Although the GzmB+ cells in this lung tissue have not been histochemically identified as Treg cells, there is some indirect evidence to support this assumption. It has already been shown that GzmB is an effector molecule of Treg cells in various tumors and viral lung infections [58–61]. A study that examined the gene expression of Foxp3 and IL-10 in lung tissues showed that although the FOXP3 gene was increased, IL10 gene expression was not increased in COPD compared to nonsmokers [49], suggesting that Treg cells likely effect an immunosuppressive process through a pathway other than the cytokine-mediated pathway. In addition, as mentioned above, active GzmB expression is observed in CD8+ T cells, NK cells, Treg cells, B cells, dendritic cells, macrophages, mast cells, basophils, keratinocytes and chondrocytes, and it has been reported that CD8+ T cells [13–16], NK cells [70], macrophages [71], mast cells [72], basophils [73] and B cells [74] were increased in the COPD lung. Therefore, the reduced GzmB+ cells in COPD lung tissue would correspond to Treg cells or dendritic cells.

On the other hand, it is known that in addition to Treg cells, dendritic cells also modulate T effector cells through a GzmB-dependent mechanism. COPD is associated with reduced numbers of (mature) CD83+ dendritic cells in the small airways [75] and alveoli [76] of surgically resected lungs. Therefore, GzmB may be a general functional marker of cells with regulatory function in COPD lung, including not only conventional Treg cells but also immunosuppressive dendritic cells.

Possible contribution of GzmB+ cells to COPD progression

As mentioned previously, there was a close correlation between the number of Treg cells in lung tissue or BAL fluid and FEV₁% predicted in patients with COPD [43–45]. This finding suggests that COPD progression may depend on the number of Treg cells in the lungs. However, these findings were based on cross-sectional studies. A follow-up study by Ström et al. [51] showed that the rate of lung function decline over a 5-year period in COPD was inversely dependent on the number of Treg cells in BAL fluid.

Consistent with these observations, the volume fraction of GzmB+ cells in the small airways (Figure 5A) and the number of alveolar GzmB+ cells/mm (Figure 3A) were closely correlated with FEV₁% predicted in subjects with microscopically validated CLE. In addition, there was a positive correlation between lung tissue expression of GzmB, measured directly by enzyme-linked immunosorbent assay and FEV₁% predicted in smokers with FEV₁ less than 80% (Figure 5B). These findings indicate that the fewer GzmB+ cells present in the lung tissue, the lower is the lung function. Conversely, the higher the number of GzmB+ cells in the lung tissue, the better is the lung function. It is believed that GzmB-containing cells in the lung tissue may protect against lung injury caused by tobacco smoking by suppressing cytolytic effector CD8+ T cells and promoting tissue repair. Therefore, it is conceivable that COPD progression depends on the tissue content of GzmB+ cells in the lungs of smokers, including the small airways and alveoli.

Figure 5. The relationship between the volume fraction (Vv) of granzyme B+ cells in small airways and FEV₁% predicted in subjects with centrilobular emphysema (CLE). P value for spearman’s rank correlation coefficient is applied to CLE subjects (A) and that between lung tissue expression of granzyme B, measured by enzyme-linked immunosorbent assay and FEV₁% predicted in smokers with FEV₁ less than 80% predicted (B) [63].

Peripheral blood Treg cells in patients with COPD

Two studies reported that Treg cells were present at equal frequencies in the peripheral blood of controls and subjects with COPD [40, 42]. Therefore, no difference was found in peripheral blood Treg cells between the groups, although Treg cells in the lung tissue or BAL fluid were decreased in those with COPD. However, four studies reported that the numbers of Treg cells in the lungs or BAL fluid and peripheral blood were simultaneously lower in smokers with COPD than in nonsmokers and normal smokers [45, 47, 48, 50]. If this finding is substantiated by further studies, it could explain why cardiovascular disease and diabetes mellitus are more common in COPD patients with smoking-induced thickening of small airway wall than in COPD without thickening of small airway wall [77]. Cardiovascular disease [78] and diabetes mellitus [79] have been reported to be associated with reduced circulating Treg cells. In fact, patients with COPD combined with type 2 diabetes mellitus had a significant decrease in blood Treg cells compared to patients with COPD, or type 2 diabetes mellitus, or in healthy controls [80]. Low blood Treg cells in patients with COPD may also explain why the incidence density of lung cancer decreases with increasing severity of baseline airflow obstruction [81], as low Treg cells would enhance the cytotoxic effect of CD8+ T cells to suppress cancer.

Conclusion

Studies have shown decreased numbers of Treg cells in the lung tissue or BAL fluid of patients with COPD. Studies have also shown a significant correlation between the number of Treg cells in lung tissue or BAL fluid and FEV₁% predicted in patients with COPD. However, there are no clear functional marker for Treg cells.

Our previous study suggests that GzmB+ cells in COPD lung tissues may be a general functional marker for regulatory cells, including Treg cells and dendritic cells. Staining of GzmB+ cells in small airways and alveoli of microscopically validated CLE and enzyme-linked immunosorbent assay of lung tissue from smokers demonstrated that amount of GzmB in lung tissue was positively correlated with FEV₁% predicted in study subjects. These results suggest that COPD progression may depend on GzmB content in the lung tissue. GzmB may act as an effector molecule to control the inflammatory process in COPD. Therefore, GzmB is thought to be an important bioactive small molecule involved in the COPD progression. However, further studies, including in vivo experiments, are needed to better understand the association between GzmB expression and COPD progression. Treg cells are candidate for the treatment of certain inflammatory and autoimmune diseases. Clinical trials have demonstrated the safety and effectiveness of Treg cell therapies for the treatment of inflammatory diseases [82]. We propose that granzyme B-producing immunosuppressive cells have therapeutic potential in the early stages of COPD to prevent or delay disease progression.

Disclosure statement

The authors report there are no competing interests to declare.

Additional information

Funding

The authors reported there is no funding associated with the work.