Introduction

Particulate matters (PMs) with an aerodynamic diameter of less than 2.5 μm (PM2.5) are the most critical and objectionable group of pollutants in the ambient air (Wang et al. 2017). PM2.5 includes several components such as heavy metals, polycyclic aromatic hydrocarbons (PAHs), unburned or incompletely burned fuel, sulfur compounds, and nitric compounds (Prieto-Parra et al. 2017; Zhang et al. 2018). They can penetrate the alveolar region of the lung effectively and either deposit there or enter the bloodstream considering their size (Kulkarni et al. 2006). These characteristics allow PM2.5 to introduce various pollutants to respiratory cells for an extended period compared to the gaseous pollutants. Smoking is the most critical risk factor in incidence or exacerbation of respiratory diseases due to induction of permanent inflammation (Cosio et al. 2002). However, according to the epidemiological studies, various lung diseases and respiratory symptoms are attributed to the exposure to the environmental PM2.5 (Cohen et al. 2017; Dominici et al. 2006). Respiratory pathologies most commonly reported in the studies include inflammation in lung parenchyma (functional tissue of lung) or airways, oxidative stress, metabolic interruptions, and mutagenicity (Feng et al. 2016). Moreover, the relationship between exposure to particulate matters and some respiratory diseases such as exacerbation of asthma and COPD, and lung cancer has been confirmed (Raaschou-Nielsen et al. 2016). Inflammation is the most critical event in all of these diseases. Inflammation is a complicated response to exposure to pathogens or tissue injury. In normal condition, it has several steps including secretion of several cytokines as inflammatory mediators for recruitment of other immune cells, phagocytosis of damaged cells’ debris or pathogens, and finally, anti-inflammatory and tissue repairing step (Medzhitov 2008). Inflammation in the respiratory system could involve different cell types and can be chronic. Depending on the cell type involved in the inflammation in the lung, induced chronic inflammation can activate different immunological pathways and lead to individual severe diseases (Abbas et al. 2019a).

Lung diseases are generally categorized as obstructive or restrictive. Interstitial lung diseases (ILD) are the dominant group of restrictive diseases. ILD are chronic diseases with the insidious but progressive onset of dyspnea and cough. Lung fibrosis is one of the main ILD (Chilosi et al. 2012). Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive, and usually lethal pulmonary disease; moreover, it is more prevalent among the elders (King Jr et al. 2011). IPF involves abnormal and continuous production and deposition of extracellular matrix (ECM), including collagen/elastin fibers (Wilson and Wynn 2009). Figure 1 illustrates the histopathological appearance of IPF. The fibrosis can be diagnosed by increasing the thickness of ECM because of abnormal deposition of collagen in the tissue (King Jr et al. 2011). The regions with normal density of pneumocytes and fibrotic area are shown in Fig. 1.

Fig. 1
figure 1

Fibrosis in lung tissue (1: normal area; 2: fibrosis area)

Full size image

Several pathogeneses have been proposed for the stimulation of fibrosis; however, the most accepted hypothesis is the chronic inflammation in epithelial cells in the lung parenchyma. Various inflammatory reaction pathways could cause fibrosis. They start from cell damage due to contact with irritant agents such as pathogens or particulates and finally lead to abnormal collagen secretion by fibroblasts/myofibroblasts. Several inflammatory mediators are involved in these pathways. Some of them are unique to a specific pathway, and the others are common between pathways (Kumar et al. 2014).

In a different scenario, inflammatory reaction cascade can activate different pathways and lead to obstructive lung diseases such as emphysema. The clinical definition of emphysema is the air space enlargement due to the alveolar wall distraction. The pathophysiological feature of emphysema is shown in Fig. 2. The healthy and enlarged alveoli are indicated by numbers 1 and 2 in this figure, respectively. Following chronic exposure to antigens, alveolar macrophages and airway epithelial cells can secrete various cytokines such as interleukin-8 (IL-8) and granulocyte-macrophage colony-stimulating factor (GM-CSF). These proinflammatory mediators can signal an increase in the recruitment of leukocytes in the lung (Ruwanpura et al. 2017). The leukocytes, especially neutrophils, secrete serine and cysteine proteinases. These enzymes are elastases and can destruct connective tissue components such as collagen and elastin (Taraseviciene-Stewart and Voelkel 2008). Consequently, alveolar attachment to the bronchioles would be decreased, and the alveoli area would be increased, followed by long-term secretion from them. Anatomically, it can lead to emphysema following hyperinflation, a significant reduction of forced vital capacity (FVC%), and other clinical symptoms (Cottin et al. 2017).

Fig. 2
figure 2

Emphysema in human lung (1: normal alveoli; 2: enlarged alveoli following alveolus wall destruction)

Full size image

Therefore, chronic inflammation can induce both restrictive and obstructive lung diseases. Thus, numerous studies have been tried to determine the inflammatory effect of inhalable materials and further signaling pathways inducing individual lung diseases since one of the leading and frequent respiratory exposures is PM2.5 inhalation.

Exposure to urban PM2.5 is increasing worldwide. According to various studies, the number of cities having PM2.5 concentrations over international guidelines is increasing. Moreover, the incidence rate of the high level of PM2.5 is increasing in megacities. Also, PM2.5 has various components with confirmed health effects (Hsu et al. 2017; Kim et al. 2019; Lin et al. 2014; Zhang et al. 2018). Therefore, several studies have been tried to evaluate the inflammatory effect of PM2.5 inhalation, and most of them confirmed these effects after exposure to this pollutant. Nevertheless, there are some differences between their experimental approaches and their results. The examples are varied in exposure doses, exposure duration, and inflammatory response levels.

Moreover, these studies measured and reported a wide range of inflammatory mediators, which were involved in different inflammation pathways and could cause different symptoms. Consequently, the possible inflammation pathways after exposure to urban PM2.5 are unclear. However, there are many questions and controversies concerning the relation between PM2.5 exposure and inflammation. The main questions are as follows: Can inhalation and exposure to urban PM2.5 induce inflammation in lung parenchyma? Which inflammatory pathway will be activated after exposure to PM2.5? What are the available procedures to evaluate the inflammatory effect of PM2.5 inhalation? What are the possible immune response pathways that can be activated after exposure to PM2.5?

Therefore, we reviewed available studies that specifically investigated the inflammatory effects of exposure to PM2.5 based on defined search strategy to respond to these questions.

Methodology

Search strategy

This review was conducted in February 2019. Primarily, the main questions were formulated according to the available knowledge, as mentioned in the “Introduction.” The systematic search was carried out according to these questions among four databases: Web of Science, Scopus, PubMed, and Embase in the English language. The search keywords were the following: “lung + inflammation + PM2.5” and “lung + inflammation + particulate matter.” The results were filtered using the inclusion and exclusion criteria. These criteria would be mentioned in the following sections.

Inclusion criteria

Focus on urban air PM2.5 was the main inclusion criteria. Only original articles published in peer-reviewed journals were selected for this study. In vitro, animal, or human studies were all included. All of the selected articles reported an exposure dose, exposure period, and chemical analysis of used particulate matters. We also reviewed references within the returned review articles to find any further appropriate articles.

Exclusion criteria

We excluded all the studies conducted on natural particles such as dust. Moreover, all the studies about occupational situations, accidental exposures, polymorphism, genetic studies, and studies that reported inflammation by serum cytokine measurements were excluded. Besides, we removed studies that did not report the chemical analysis of PM2.5. Finally, some articles were further excluded after careful study of their abstracts or the full article contents.

Data extraction

Extracted data and information from the selected articles included the type of study (in vitro, animal, or human), subjects’ characteristics, PM2.5 source and its chemical composition, exposure procedure (dose, time, groups), type of the measured marker to determine inflammation and response level (cytokine secretion or mRNA expression), results of studies (change in marker level), and outcomes (visible histologic, anatomic, or pathologic changes and conclusion of the study). Then, extracted data/information was classified based on the similarities between studies for better visualization and semi-quantitative comparison.

Results

Included studies

The search on four databases, with inclusion and exclusion filters, returned 555 articles. After the duplication removal (N = 366), 189 articles remained. Finally, 21 articles were selected after screening by title, abstract, and main passage reviewing. Figure 3 is the PRISMA flow diagram presenting the study selection procedure.

Fig. 3
figure 3

PRISMA flow diagram

Full size image

Acquired data

The acquired data from 21 included studies are summarized in Table 1. The included studies could be divided into two main groups. The first group was the animal studies, which included instillation or respiration of ambient PM2.5 by mice or rat for various periods and measurement of inflammatory mediators in serum or bronchoalveolar lavage fluid (BALF). The second group was the in vitro studies using human or animal cell lines. In these studies, PM2.5 was filtered from ambient air, and cell lines were then exposed to them in a controlled manner, and the inflammatory mediators were measured after different exposure periods.

Table 1 Extracted data from included articles. aChemicals reported as μg/mg. HMs heavy metals, WSI water-soluble ions
Full size table

The PM2.5 were collected from urban air in this study. Three articles reported using standard particles in their experiments, including urban and diesel particles. Numerous parameters were measured and reported in the included articles, which were various. Accordingly, we tried to classify particulates’ chemical composition into three categories: heavy metals (HMs); water-soluble ions (WSIs) including SO4−2, NO3−, and NH4+; and PAHs. Besides, we tried to convert and report all concentration values in μg/mg of particle mass. The additional extracted data from articles was exposure dose and procedures.

One of the main problems for comparison between different studies was exposure and the concentration units of inflammatory cytokines. There was no capability of converting all of the reported units to the same unit. Therefore, the exposure units were μg/ml or μg/cm2 of cell culture in in vitro studies and mg/kg or mg/mice for animal studies. Likewise, the concentration of the cytokine was reported as pg/ml or pg/gr of protein for in vitro studies. Various kinds of inflammatory cytokines were measured in the included studies. However, no justification was found for the selected kind of cytokines in each study. The only common mediator in all of the studies was IL-8. The exposure period and dose were different between in vitro and animal studies. Summary of exposure data, characteristics of particulates, and measured mediators are listed in Table 2.

Table 2 Summarized characteristics of included articles
Full size table

Several diversities were observed in extracted information, including the analyzed components of PM2.5, the analysis of more components of PM2.5, and the response to exposure to intact PM2.5 or its components. Therefore, it was impossible to illustrate the data using the plot. Hence, the results were classified based on the observed response, PM2.5 components, and exposure period in Tables 3 and 4. According to these tables, the number of available studies reporting responses to each component is mentioned (N). Besides, the number of direct (D), reverse (R), and non-significant (NS) studies have been mentioned in these tables. The symbols, including +, −, and ±, indicate the effect of exposure period at constant dose.

Table 3 Number of studies dealt with each inflammatory mediator in relation to total PM2.5 and its constituents in in vitro studies
Full size table
Table 4 Number of studies dealt with each inflammatory mediator in relation to total PM2.5 and its constituents in animal studies
Full size table

Discussion

We investigated PM2.5-induced inflammation in lung epithelial cells in this systematic review. The articles included in this review were in vitro and animal studies. There are several human studies on measuring inflammatory cytokines in bronchoalveolar lavage fluid (BALF). However, they cannot be interpreted to determine the inflammatory effects of inhaled PMs because the patients who need bronchoscopy and BALF collection have serious diseases with inflammation as a common complication (El-Bayoumi and Silvestri 2008; Jin and Wang 2019; Sato et al. 2017).

As shown in Table 1, all the included articles have reported the concentration of at least one of the main groups of PM2.5 components. Table 2 shows that the exposure duration in animal studies is significantly longer than the in vitro studies. Reversely, the applied dose of PM2.5 is significantly higher in the in vitro studies.

WSIs are the most frequently measured component. WSIs can dissolve in bronchiolar or alveolar fluids, leave the particulate structure, and change airway surface liquid (ASL) osmolarity and pH (Miller et al. 2009). Dissolved WSIs can cause osmotic stress and induce inflammation. As mentioned in Table 3, the number of in vitro studies that reported a direct relationship between WSIs and cytokine concentration is significantly more than animal studies. The concentration of chemicals is constant in the in vitro studies, while in the animal body, the hemostatic effect moderates the concentration of pollutants.

PM2.5 also contains heavy metals, including Cu, Zn, and Fe, which serve as micronutrients for cells. Nevertheless, other heavy metals such as mercury, lead, cadmium, vanadium, and arsenic do not have any biological function (Gao et al. 2016). All of these toxic metals were detected in the included studies. These metals are water-soluble and can activate the oxidative stress pathway and induce inflammation (Radan et al. 2019). The PAHs are another group of measured components in urban PM2.5. They can be recognized by the aryl hydrocarbon receptor (AHR) in the cytoplasm, transferred to the nucleus, and induce DNA damage (Pieterse et al. 2013). However, their entrance from the surface of the cell into the cytoplasm is unclear. Like WSI, the number of in vitro studies that reported a direct relationship between heavy metals and PAHs and cytokine concentration is clearly more than animal studies.

In addition to chemical features, the activated pathway of inflammation is also essential. There is not robust evidence confirming the PM2.5 transfer to the parenchymal cells using endocytosis. Therefore, the induced inflammation in in vitro studies is probably related to the activation of some signaling pathways from the surface of the cell membrane. It can be true in animal studies, although the source of secreted inflammatory cytokines is not clear and can be from parenchymal cells, connective tissue cells, or immune cells.

Table 3 shows that the most prevalent secreted cytokines in in vitro studies are proinflammatory, including INF-γ, IL-1β, IL-6, IL-8, and TNF-α. Generally, the target cells of these mediators are innate immune cells such as macrophages and neutrophils. Some of them are common between in vitro and animal studies. Moreover, as indicated in Table 4, animal studies show a more complete inflammatory responses to PM2.5 exposure. Some measured mediators in animal studies such as IL-12 and IL-13 induce differentiation of naive T cells into T helper 1 cells and anti-inflammatory effects.

The other parallel mechanism, which can initiate the inflammation, is autophagy. Following the entrance to the cell, PM2.5 can induce autophagy (Deng et al. 2013). In many in vitro studies, autophagy was an uncompleted and unsuccessful mechanism against PM2.5, and it decreased the cell viability and caused cell destruction and release of the particulates into the intercellular space again. In this situation, several damage-associated molecular patterns (DAMPs) are released in the intercellular space (Zhao et al. 2019). These DAMPs are recognized using the pattern recognition receptors (PRRs) in the innate immune cells such as macrophages, dendritic cells (DCs), and monocytes and stimulate the recruitment of additional innate immune cells (Abbas et al. 2019a). After the inflammation induction, the wound healing process or the anti-inflammatory step, including repairing mechanism, would be initiated (Wolters et al. 2014), which is a critical step since the suggested pathophysiology for IPF is an abnormal wound healing process due to the uncontrolled repair. Both types of inflammation (types I and II) can happen in wound healing at normal conditions (Wilson and Wynn 2009). Type I includes secretion of IL-1β, IL-6, IL-8, INF-γ, TNF-α, and IgG2 antibodies and recruitment of innate immune cells, including macrophages and neutrophils. Afterwards, T lymphocytes, especially T helper 1 cells (Th1) and regulatory T cells (Treg), are recruited to balance inflammatory and anti-inflammatory reactions by signaling to fibroblast for collagen secretion and repairing (Abbas et al. 2019a). Type II inflammation includes secretion of IL-4, IL-5, and IL-13 to recruit the eosinophils and basophils and then to signal to Th2 cells to regulate inflammation and then to induce repair mechanisms (Gandhi et al. 2016). In IPF, continuous deposition of collagen and other ECM proteins occurs as an abnormal process, which means that anti-fibrotic mediators are not being secreted sufficiently, or the collagen deposition and fibroblast differentiation activation pathway are activated continuously. In other words, IPF is a completed but unfinished wound healing process. The initiation of this step directly depends on antigen-presenting by innate immune cells (Strieter 2008). Neutrophils do not have the antigen-presenting capability but other cells, including macrophages, natural killer (NK) cells, and DCs, which can make epitope to present antigen to lymphocytes (Abbas et al. 2019a). The PM2.5 can have various components, but, as reported in many studies, there is not any polypeptide structure in their composition (Li et al. 2016). Therefore, the innate immune cells cannot perform the antigen-presenting and lymphocytes, especially Th1, cannot be recruited. Th1 recruitment is necessary for the activation of Treg (Wing and Sakaguchi 2010). Thus, anti-inflammatory responses cannot be activated. Besides, since the lack of antigen-presenting capabilities, the recruitment of Th2 cells is unlikely too. Therefore, the inflammation will continue until full clearance of PM2.5 from the lung. Thus, recruitment and differentiation of myofibroblasts to fibroblast for the repair of the destructed tissue is rare.

Increased bronchiolar level of IL-4, IL-13, and TGF-β is reported in many studies as diagnostic indicators of IPF. However, according to the extracted data, none of the included animal studies reported a significant increase of TGF-β, IL-4, or IL-13 following exposure to PM2.5. Lack of antigen-presenting capability and absence of an increase in the levels of IL-4 and IL-13 show that IPF following the exposure to PM2.5 is rare and maybe non-existent.

The other possible disease following inflammation is emphysema. The first step of phagocytosis by innate immune cells is the secretion of several proteolytic enzymes to destruct antigens and facilitate phagocytosis (Abbas et al. 2019a). The action of proteolytic enzymes can be regulated using anti-elastase enzymes, such as α1 anti-trypsin (α1-AT). Emphysema is possible among individuals with a genetic defect in the production of α1-AT (Taraseviciene-Stewart and Voelkel 2008). In normal individuals, the most approved scenario for the development of emphysema is over-secretion of protease and elastase enzymes such as serine protease, matrix metalloproteinase-12 (MMP-12), and elastase by neutrophils and macrophages following exposure to an antigen and induction of predominant inflammation (Sharafkhaneh et al. 2008). The secretion of these proteolytic enzymes is a robust diagnostic indicator for emphysema (Taraseviciene-Stewart and Voelkel 2008). Besides these enzymes, several studies indicated that MMP-9 would be secreted following the inflammation to destruct the collagen of the basement membrane and facilitate monocyte migration from blood to the alveolar region (McMillan et al. 2004). Several studies reported that the clearance of PM2.5 from the lung takes too long. Moreover, PM2.5 have no polypeptide components. Thus, the secretion of protease enzymes can continue for a long time endlessly. Additionally, several studies reported that PM2.5 phagocytosis could damage the macrophages and destruct them and cause PM2.5 to be released again (Geng et al. 2006). Therefore, the secretion of protease enzymes will continue until the full clearance of PM2.5 from the lung. Because of continuous exposure to PM2.5, these reactions seem to be permanent. There was no data in our included animal studies regarding the measurement of these protease enzymes. However, available results confirm that inflammation in lung follows exposure to PM2.5. Therefore, the destruction of connective tissue proteins seems possible after exposure to PM2.5, while the severity of destruction is very mild and needs a very long time to make a significant tissue damage.

The secretion of INF-γ is another issue concerning the initial response of cells, which are exposed to PM2.5. It has been measured in both types of studies, and almost all of them reported significant increases in INF-γ level following exposure to PM2.5. Generally, after pathogen recognition, especially in viral infections, INF-γ is secreted by infected cells to induce cellular immune response using the stimulation of protein kinase R (PKR) production (Wingate et al. 2009). In the case of PM2.5, the produced PKR has no target antigen due to the absence of any double-stranded RNA in PM2.5 compositions. If PKR deposition continues for a long time, it can induce cell apoptosis or lead the cell to the autophagy pathway (Kim et al. 2007). Both apoptosis and autophagy can signal macrophage recruitment. In normal circumstances such as following the exposure to viruses, the apoptosis has practical function to prevent the virus from spreading, but PM2.5 induce unnecessary apoptosis in pneumocytes (Dagher et al. 2006). This event can reduce the population of parenchymal cells and cause a decrease in lung function, a condition like obstructive diseases.

Finally, the findings of this study showed that, because of the chemical structure of PM2.5, they could induce a permanent inflammation without any antigen-presenting to activate repair mechanisms. Therefore, the tissue destruction and incidence of obstructive diseases are more likely compared to fibrosis. This directly depends on the exposure period to PM2.5 and the dosage. The existence of any mechanism to induce adaptive immunity against PM2.5 is unclear and must be investigated. Also, there is no information about the other components of PM2.5 that may have receptors in pneumocytes. Moreover, it is not clear whether or not there is any pathway for the activation of anti-inflammatory response against the induced inflammation by PM2.5.

Additionally, it is not clear why some of the PM2.5 can enter the cells, and the others cannot. The response to this question could help in distinguishing between initial inflammatory responses of damaged cells and the prediction of the next steps of inflammation. Moreover, because of the absence of any human studies in this field, the pathological effects of long-term exposures to PM2.5 in human lungs are not apparent and must be further investigated to validate these findings (Deng et al. 2013; Moon et al. 2018; Roh and Sohn 2018).

Limitations

Several limitations restricted the conclusions from the included articles. The exposure procedure in several included studies was single-dose exposure, and their results were only compared with their control group. There was no capability to merge their results with other studies to find a correlation between inflammatory response and dose/duration of exposure. Pathological pieces of evidence are significant to illustrate the inflammation in tissues, but just a few articles had pathological investigations. We found several short period animal studies about the inflammatory effects of PM2.5 on lung cells, which included pathological evaluation. However, those articles lacked the chemical analysis of particulate, which is a critical data and had been excluded from this review. Another limitation of this study was related to the selected exposure doses in animal studies. The exposure doses were not selected according to the respiration rates of animals. Therefore, we could not extrapolate the animal results to the human. There was no co-culture study in in vitro studies. Thus, there were no results about innate immune cells that responded to inflammatory cytokines secreted from the lung cells.

Conclusion

According to the in vitro and animal studies, urban PM2.5 activates the inflammatory pathway that involves innate immune responses initiated from exposed/damaged lung parenchymal cells. PM2.5 cannot activate adaptive immune responses because of their chemical nature. However, this condition strengthens the development of obstructive diseases such as emphysema and weakens the probability of restrictive diseases such as IPF because the exposure to PM2.5 and following inflammation is permanent in urban areas. More detailed and robust evidence about the inflammatory effects of long-term exposure to PM2.5 would have to come from human studies such as cohort or cross-sectional studies on highly exposed persons. The adaptation mechanisms to balance inflammatory and anti-inflammatory pathways in humans are unclear and need further investigations. Moreover, the effect of seasonal changes in PM2.5 exposure dose on inflammation is unclear. Also, variations in the composition of particulates need to be investigated further.