Abstract
Paraquat (PQ) is a highly toxic herbicide to humans. Pulmonary fibrosis is one of the most typical features of PQ poisoning, which develops from several days to weeks after ingestion. However, the mechanism of fibrosis is still unclear. In this study, we aimed to determine expressions of autophagy-related markers Beclin 1, microtubule-associated protein light chain 3 (LC3), and p62 in PQ-poisoned lungs and to explore the role of autophagy in pulmonary fibrosis induced by PQ. We detected markers of lung fibrosis and expressions of autophagy-related protein in the specimens from eight fatal cases of PQ poisoning by hematoxylin and eosin staining, Masson’s trichrome staining, and immunohistochemistry. Based on the staining results of lung fibrosis, these cases were divided into two groups, fibrosis and non-fibrosis groups. The correlation between autophagy protein expressions and pulmonary fibrosis was examined. The results demonstrated that the autophagy-related proteins were significantly expressed in fibrosis group compared with the non-fibrosis group. There was a significantly positive correlation between these protein expressions and severity of lung fibrosis. In conclusion, autophagy dysfunction may be involved in lung fibrogenesis caused by PQ poisoning. This may be a promising clue for understanding the molecular mechanism underlying PQ-induced lung fibrosis and provide evidence for treating fibrosis by regulating the level of autophagy.
Introduction
Paraquat (PQ) is an agricultural herbicide widely used globally, especially in developing countries. 1 In China, it is one of the most toxic poisons involved in suicide attempts, and PQ poisoning has a high mortality rate. 2 Although nearly all PQ poisonings result from ingestion, other routes of poisoning, such as skin contact, intramuscular route, or intraperitoneal route, have been reported. The fatal dose for a 70 kg adult is just a mouthful (∼20 mL) of 20% PQ solution, which corresponds to a dose of 55 mg/kg. 3 Through the polyamine transport system, PQ selectively accumulates in the lungs, leading to alveolitis and irreversible pulmonary fibrosis in a condition known as “paraquat lung”, which also accounts for the main cause of death of PQ poisoning. 4 However, there are no specific clinical preventive methods or treatments available, mainly because of the lack of understanding of the mechanisms underlying PQ-induced acute pulmonary injury or fibrosis. 5
Autophagy is an evolutionarily conserved lysosome-mediated catabolic process that maintains cellular homeostasis through the degradation of damaged organelles and macromolecules. 6 Depending on the way by which intracellular substrates enter lysosomes, autophagy is divided into three modes: macroautophagy, microautophagy, and chaperone-mediated autophagy. The term autophagy typically refers to macroautophagy. 7 The occurrence of autophagy is artificially divided into three stages: the initial autophagosome bilayer formation stage, the bilayer membrane elongation closure and autophagosome formation stage, and the autophagic lysosome formation and degradation stage. 8 The occurrence of autophagy is a complex network system, regulated by more than 30 autophagy-related gene (ATG) proteins. 9 Beclin1, microtubular-associated protein 1 light chain 3 (LC3), and SQSTM1/p62 are the three major proteins involved in the autophagy process, and the most commonly used autophagy-related markers. 8 There is evidence for autophagy dysregulation involved in various disease states, including cancer, neurodegenerative diseases, and infectious diseases, 10 as well as lung diseases. 11
Recently, the role of autophagy in the process of pulmonary fibrosis has become an area of research interest. 12 Increasing studies have indicated that autophagy protects lung tissue from pulmonary fibrosis, and declining or impaired autophagic function contributes to the pathogenesis of fibrosis. It has been demonstrated that insufficient autophagy may contribute to lung fibrogenesis by accelerating epithelial cell senescence and inducing myofibroblast differentiation in idiopathic pulmonary fibrosis (IPF). 13 Other studies on IPF have obtained similar results. 14 –16 Consistently, autophagy model mice (atg4b-/-) treated with bleomycin displayed significantly increased inflammatory responses and fibrosis compared with WT mice. 17 Similarly, mice with mammalian target of rapamycin (mTOR) overactivation in type II alveolar epithelial cells showed mortality and pulmonary fibrosis compared with control mice. 18 On the contrary, some studies have suggested that activated autophagy promotes the development of lung fibrosis. An in vitro study found that silica activates MCPIP1-mediated autophagy in macrophages, and activated autophagy facilitates the secretion of collagen from fibroblasts and induces fibrosis. 19 Another in vitro study on silicosis also yielded the same result. 20 In addition, carbon tetrachloride has been found to activate hepatocyte autophagy, while salidroside reduced liver fibrosis by inhibiting autophagy. 21 Although the role of autophagy in organ fibrogenesis is uncertain, we speculate that autophagy may play a role in pulmonary fibrosis caused by PQ poisoning.
To investigate this, we measured histopathology of lung fibrosis and expression levels of the autophagy-related markers Beclin 1, LC3, and p62 in lung tissues from eight fatal cases of PQ poisoning and analyzed the correlation between them.
Methods
Materials
The specimens used in this experiment were paraffin-embedded lung tissues from victims who died of PQ poisoning that were stored by the Judicial Appraisal Center affiliated with China Medical University. Written informed consent was obtained from the victims’ family, and this investigation was approved by the Ethics Committee of China Medical University (no. [2018] 065). Data on personal information, poisoning, and other parameters were also recorded.
Samples from eight fatal cases of PQ poisoning (20% PQ solution) occurring between July 2008 and March 2018 were gathered from the stored archive. These cases were arranged as 1–8 according to the survival time, from 0.5 day to 20 days (Table 1). Victims who were <18 or >60 years old, pregnant or lactating, treated with immunosuppressive agents other than mycophenolate mofetil, or diagnosed with other respiratory system diseases, including IPF, chronic obstructive pulmonary disease, silicosis, or chronic diseases affecting autophagy such as systemic infection, cancers, diabetes mellitus, and heart diseases, were excluded from the study. One relatively normal lung sample from a victim who died of severe head injury in a traffic accident was used as a negative control. The characteristic features and relevant information of the cases and clinical treatment are presented in Table 1.
Summary of cases, clinical treatment, and period of samples.
aIndicates the date when lung tissues were made into wax blocks, which were collected for study between March 2018 and April 2018.
bOral ingestion.
cSkin contact.
Based on the results of both Masson’s trichrome staining and immunohistochemical staining of alpha-smooth muscle actin (α-SMA), the eight PQ poisoning cases were divided into two groups: (i) the fibrosis group included cases 4, 6–8, which had obvious collagen deposition and significantly elevated expression of α-SMA, and (ii) the non-fibrosis group, which included the other cases. Although a small amount of collagen deposition around a few blood vessels was observed in case 5, there was no obvious expression of α-SMA; therefore, it was classified into the non-fibrosis group.
Hematoxylin and eosin staining
Routine hematoxylin and eosin (H&E) staining was performed. Briefly, the sections were deparaffinized and rehydrated and then stained with hematoxylin solution, followed by 1% acid ethanol. The sections were then stained with eosin solution. After dehydration, the mounted slides were examined and photographed using an Olympus BX41 microscope (Tokyo, Japan).
Masson’s trichrome staining
Lung fibrosis was identified using Masson’s trichrome staining for collagen. After deparaffinization and rehydration, 5-μm-thick sections were stained with Masson’s trichrome. Briefly, the sections were dissolved in Weigert’s iron hematoxylin solution, followed by 1% hydrochloric acid in alcohol for differentiation. After successively staining with acid Ponceau and phosphomolybdic acid aqueous solution, the sections were counterstained with aniline blue, followed by 1% glacial acetic acid. Finally, routine dehydration and mounting were performed. Under a low-power microscope, collagen fibers and nuclei were stained blue, and the cytoplasm, muscle fibers, and red blood cells were stained red.
Immunohistochemical staining
Immunohistochemical staining for α-SMA was performed to detect myofibroblasts, which are important markers for pulmonary fibrosis. Staining was also carried out to determine the expression of Beclin-1, LC3, and p62, all of which are commonly used to reflect the status of autophagy in various studies. Immunohistochemical staining was carried out on lung paraffin sections using an SP kit (ZSGB-BIO, Beijing, People’s Republic of China). Slices were deparaffinized and rehydrated with xylene or gradient ethanol, followed by antigen retrieval in a microwave in 10 mM citrate buffer. To block the activity of endogenous peroxidase, the slices were incubated in 3% hydrogen peroxide for 30 min and nonspecific binding was blocked with goat serum for 2 h, followed by incubation overnight at 4°C with mouse anti-Beclin 1 antibody (1:1000, ProteinTech, Chicago, Illinois, USA), rabbit anti-LC3 antibody (1:500, ProteinTech), rabbit anti-SQSTM1/p62 antibody (1:1000, ProteinTech), and rabbit anti-ACTA2/α-SMA antibody (1:8000, ProteinTech). The following day, the tissues were incubated with goat anti-rabbit/mouse secondary antibody for 30 min, followed by incubation with horseradish enzyme-labeled streptavidin working fluid for 30 min. Hematoxylin was used to stain the nuclei, and an optical microscope was used to observe and obtain images. Cells positive for Beclin-1, LC3, and p62 appeared dark brown under the microscope. Normal lung tissue was used as the negative control. Staining with all antibodies was considered positive if >10% of the cells were stained in a field of view. All the slides were examined and scored independently by two observers (WXL and YH), who were blinded to victim’s case data. Any disagreements in evaluation were resolved through discussion, with or without reevaluating the slides, until a consensus was reached.
Statistical analysis
The correlations between immunohistochemical expression and lung fibrosis were evaluated by Fisher’s exact test. The values of p < 0.05 were considered statistically significant. Two-sided statistical tests were used in the analyses. Statistical analysis software (SPSS 22.0) was used to perform the analyses.
Results
Histopathology of PQ-induced lung fibrosis
HE staining and Masson’s trichrome staining of the control lung tissue showed relatively normal lung morphological structures (Figure 1 (a0)). Pulmonary edema, alveolitis, and alveolar hemorrhage, but no obvious fibrosis, were observed in cases 1–3 (Figure 1 (a1)-(a3)). Cases 4 and 6–8 demonstrated massive fibrosis, leading to alveolar septum widening, alveolar collapse, and indiscernible structures (Figure 1(a) and (b), cases 4 and 6–8). Only a small amount of fibrosis around small blood vessels and bronchi was observed in case 5 (Figure 1(a) and (b), case 5).

Morphopathological changes in paraquat-induced pulmonary fibrosis. H&E staining and Masson’s trichrome staining showed no obvious fibrosis in cases 1–3 (a and b, cases 1–3), a small amount of fibrosis around small blood vessels and bronchi in case 5 (a and b, case 5), and massive fibrosis in cases 4 and 6–8 (a and b, cases 4 and 6–8) compared with the control lung (a and b, Con). Immunohistochemical staining of α-SMA revealed weak expression in the vessel and bronchial walls in the control lung ((c), Con). A slightly enhanced expression of α-SMA in the vessel and bronchial walls in cases 1–3 and 5 ((c), cases 1–3 and 5), and a significant expression of α-SMA in proliferating fibroblasts in fibrosing areas in cases 4 and 6–8 ((c), cases 4 and 6–8) when compared with the control lung (×100 magnification, H&E, Masson, and immunohistochemical staining). Black arrows indicate collagen deposition. Orange arrows indicate significant expression of α-SMA. H&E: hematoxylin and eosin; α-SMA: alpha-smooth muscle actin.
α-SMA was weakly expressed in the vessel and bronchial walls of the control lung (Figure 1(c), Con). Cases 1–3 and 5 showed slightly increased expression of α-SMA in the vessel and bronchial walls when compared with the control lung (Figure 1(c), cases 1–3 and 5). In addition, cases 4 and 6–8 significantly expressed α-SMA in proliferating fibroblasts of fibrosing areas and in the vessel and bronchial walls compared with the control lung (Figure 1(c), cases 4 and 6–8).
Immunohistochemical staining for Beclin 1, LC3, and p62
The control lung showed very slight expression of Beclin 1 in all kinds of parenchymal cells (Figure 2(a), Con), no expression of LC3 (Figure 2(b), Con), and slight expression of p62 in alveolar epithelial cells (Figure 2(c), Con). In the fibrosis group, significantly enhanced expressions of Beclin 1, LC3, and p62 were observed in macrophages and proliferating fibroblasts in fibrosing areas (Figure 2(a) to (c), cases 4 and 6–8). In the non-fibrosis group, only slight expression of these proteins was observed in macrophages, while no obviously enhanced expression was observed in any other sites (Fig 2(a) to (c), cases 1–3 and 5). A significant correlation was obtained between lung fibrosis and immunohistochemical expression of each autophagy-related protein (Table 2), indicating that autophagy dysfunction may mediate PQ-induced fibrogenesis.

Immunohistochemical staining of Beclin 1, LC3, and p62 in the lung tissues from fatal cases of paraquat poisoning. The control lung showed slight expression of Beclin 1 in all kinds of parenchymal cells (a, Con), no expression of LC3 (b, Con), and slight expression of p62 in alveolar epithelial cells (c, Con). In cases 1–3 and 5 (a to c, cases 1–3 and 5), slight expression of all these proteins in macrophages was observed, but no obviously enhanced expression in any other site was observed when compared with the control lung. Significantly enhanced expression of Beclin 1, LC3, and p62 was found in macrophages and proliferating fibroblasts in fibrosing areas in cases 4 and 6–8 (a to c, cases 4 and 6–8) (×400 magnification, immunohistochemical staining). Black arrows indicate significant expression of Beclin 1. Yellow arrows indicate significant expression of LC3. Blue arrows indicate significant expression of p62.
Relationship between fibrosis and immunohistochemical expression.
Discussion
Poisoning by the herbicide PQ commonly led to the development of pulmonary fibrosis and death within a few weeks after exposure. 22 Pulmonary fibrosis induced by PQ poisoning has characteristics of rapidness and irreversibility, different from other fibrotic diseases, such as cystic fibrosis, IPF, and silicosis. There are two distinct phases in the development of pulmonary fibrosis by PQ. The destructive phase is characterized by severe destruction of alveolar epithelial cells within 1–3 days after ingestion. 23 This phase provides a basis for fibrosis. The proliferative phase involves the development of extensive fibrosis associated with proliferation of fibroblasts and myofibroblasts. 24 The exact mechanism of PQ-induced pulmonary fibrosis is still unclear. Altered activation in autophagy has been proved to be involved in a wide range of human diseases, including fibrotic disorders such as IPF, silicosis, kidney fibrosis, and liver fibrosis. 13,19,21,25 It has not yet been elucidated whether autophagy reaction is involved in PQ-induced lung fibrogenesis in humans.
Our immunohistochemical staining results demonstrated that the autophagy-related proteins Beclin 1, LC3, and p62 were consistently and significantly expressed in the lung tissues of PQ-poisoning victims, especially in fibrotic lung tissues. Moreover, we found a significant correlation between pulmonary fibrosis induced by PQ and expression of the three autophagy-related proteins, suggesting that autophagy dysfunction may be involved in PQ-induced pulmonary fibrosis.
Furthermore, we measured Beclin 1, LC3, and p62 expression to determine autophagy status in the lungs. 13,26,27 Beclin 1 is a mammalian homolog of yeast Atg6, which is the first mammalian autophagy protein to be described. 28 Free Beclin 1 is an initiator of autophagy and thus extensively used as a marker for monitoring the onset of autophagy. 29 In addition, Beclin 1–Vps34 complex plays a major role in the formation of autophagosome membrane during autophagy. 30 Therefore, as an important regulator of autophagy, the expression level of Beclin 1 represents autophagy activity to some extent. LC3 is a mammalian homolog of yeast Atg8 and has two subforms—LC3-I and LC3-II. The conversion of LC3-I into LC3-II is a key step in autophagosome formation. 31 Frequently, LC3 is studied as a marker of autophagy. 32,33 p62 has been known as one of the selective substrates for LC3. When autophagy occurs, p62 first binds to the ubiquitinated protein and then combines with LC3-II localized on the inner membrane of the autophagic vacuole to form a complex, which is finally degraded in the autolysosome. 34 In our study, we mainly utilized the recent finding that impairment of autophagy is characterized by accumulation of p62 in the cytoplasm. 35 Our results indicate that, p62 levels increased in the fibrotic area with increase in Beclin 1 and LC3, indicating, at least in part, that impaired autophagy may have an effect on pulmonary fibrogenesis induced by PQ poisoning.
Although autophagy induction or inhibition by PQ is controversial, PQ-induced autophagic dysfunction has been confirmed in previous studies. Intraperitoneal injection of PQ inhibited soluble proteasomal activity and mTOR activation in mice, resulting in decreased level of autophagy, and this result has also been confirmed in human specimens collected from PQ poisoning victims. 36 Another study found that blockage of autophagic flux occurs in P12 cells exposed to low concentrations of PQ, and PQ-induced massive mitochondrial clearance disorders may have contributed to the blockage. 37 Impaired autophagic flux by PQ was also observed in SH-SY5Y cells, and the cytotoxicity caused by PQ poisoning was significantly correlated with the inhibition of lysosomal hydrolase activity. 38 In contrast, some studies reported that PQ activated protective autophagy, and the process may be mediated by endoplasmic reticulum stress. 39,40 The differences in effects of PQ on autophagy may be related to the detection method of autophagy as well as the experimental model and the toxic dose. 41
We propose two possible mechanisms by which impaired autophagy is involved in pulmonary fibrogenesis induced by PQ poisoning. First, impaired macrophage autophagy contributes to pulmonary fibrosis. PQ induces the release of neutrophil chemotactic factor, alveolar macrophage-derived growth factor, transforming growth factors (TGF-β), connective tissue growth factor (CTGF), and fibronectin from alveolar macrophages. 42 –44 The cytokines, in turn, promote lung fibrosis by stimulating multiplication, migration, secretory activity, and collagen production by fibroblasts. 45 Meanwhile, some studies proved that impaired macrophage autophagy may enhance the secretion of inflammatory cytokines, 46,47 which may mediate lung fibrogenesis induced by PQ through autophagy impairment.
Second, impaired autophagy may contribute to PQ-induced lung fibrosis by inducing differentiation of lung fibroblasts into myofibroblasts. Myofibroblasts are a major type of effector cells in organ fibrosis, which drive fibrosis by secreting large amounts of extracellular matrix proteins and signaling molecules. 48 A previous study proved that autophagy inhibition promotes the differentiation of fibroblasts into myofibroblasts in vitro. 13 Our results showed that α-SMA, a marker of myofibroblast, significantly increased in the lungs of death victims of PQ poisoning, especially in those with obvious fibrosis. Based on the above findings, we speculate that impaired autophagy may contribute to pulmonary fibrosis by promoting myofibroblast differentiation in the lungs of PQ poisoning.
There were several limitations to this study. Because this study was performed at a single institution, insufficient sample size made it inappropriate to compare between groups. The samples were confined to paraffin-embedded tissues, so that detection methods that could be used in the study were limited. Because autophagy is a complex, multistep, and highly dynamic process, levels of a single or a small number of autophagy-related markers may not precisely reflect autophagy status. Therefore, in vivo and in vitro studies are required to further verify the research results and explore the mechanism.
In conclusion, autophagy dysfunction may involve in lung fibrogenesis caused by PQ poisoning. This may be a promising clue for understanding the molecular mechanism underlying PQ-induced lung fibrosis and provide evidence for treating fibrosis by regulating the level of autophagy.
Footnotes
Acknowledgements
The samples of this study were provided by Judicial Appraisal Center affiliated with China Medical University, Shenyang, China. We gratefully acknowledge Professor Hu Gengyi of China Medical University for his help in collecting samples in this study. We would like to thank Editage [
] for English language editing.
Author Contributions
Authors GX and XW contributed equally to this work
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
