Acrolein inhalation acutely affects the regulation of mitochondrial metabolism in rat lung

Abstract

Exposure of the airways to cigarette smoke (CS) is the primary risk factor for developing several lung diseases such as Chronic Obstructive Pulmonary Disease (COPD). CS consists of a complex mixture of over 6000 chemicals including the highly reactive α,β-unsaturated aldehyde acrolein. Acrolein is thought to be responsible for a large proportion of the non-cancer disease risk associated with smoking. Emerging evidence suggest a key role for CS- induced abnormalities in mitochondrial morphology and function in airway epithelial cells in COPD pathogenesis. Although in vitro studies suggest acrolein-induced mitochondrial dysfunction in airway epithelial cells, it is unknown if in vivo inhalation of acrolein affects mitochondrial content or the pathways controlling this. In this study, rats were acutely exposed to acrolein by inhalation (nose-only; 0− 4 ppm), 4 h/day for 1 or 2 consecutive days (n = 6/group). Subsequently, the activity and abundance of key constituents of mitochondrial metabolic pathways as well as expression of critical proteins and genes controlling mitochondrial biogenesis and mitophagy were investigated in lung homogenates. A transient decreasing response in protein and transcript abundance of subunits of the electron transport chain complexes was observed following acrolein inhalation. Moreover, acrolein inhalation caused a decreased abundance of key regulators associated with mitochondrial biogenesis, respectively a differential response on day 1 versus day 2. Abundance of components of the mitophagy machinery was in general unaltered in response to acrolein exposure in rat lung. Collectively, this study demonstrates that acrolein inhalation acutely and dose-dependently disrupts the molecular regulation of mitochondrial metabolism in rat lung. Hence, understanding the effect of acrolein on mitochondrial function will provide a scientifically supported reasoning to shortlist aldehydes regulation in tobacco smoke.

1. Introduction

Chronic obstructive pulmonary disease (COPD) is a leading cause of morbidity and mortality worldwide (Mathers and Loncar, 2006). Exposure to tobacco smoke, which consists of a complex mixture of over 6000 chemicals, is the main risk factor for developing COPD (Rodgman and Perfetti, 2013; Mannino and Buist, 2007). One particular class of chemicals formed during the pyrolysis and combustion of tobacco is aldehydes, such as acrolein, which, due to its α,β-unsaturated structure, is considered as one of the most reactive (Stevens and Maier, 2008; WHO, 2008). Levels of acrolein in cigarette smoke (CS) vary around 1.3–121.7 μg/cigarette (i.e., 2.1–110.4 ppm) depending on brand and smoking regime (Pauwels et al., 2018).

Acrolein is a primary (respiratory) irritant and considered to represent approximately ninety percent of the smoking-associated known non-cancer disease risk (Haussmann, 2012; Fowles and Dybing, 2003; Yeager et al., 2016). Previous literature suggested a link between acrolein exposure and abnormalities in structure and function of the airways (Costa et al., 1986). Moreover, acrolein is able to induce multiple cellular events considered to be central in COPD pathogenesis (e.g., inflammation, oxidative stress and tissue remodeling) (Yeager et al., 2016). Importantly, recent human studies highlight a link between aldehyde dehydrogenase, an enzyme that detoxifies aldehydes (O’Brien et al., 2005), and lung function. This is illustrated by a study in a Japanese general population which identified an association between the presence of an inactive allele of aldehyde dehydrogenase and incidence of smoking-related chronic airway obstruction (Morita et al., 2015). Although these studies suggest an association between exposure of the airways to acrolein (as a component of CS), the underlying molecular mechanisms that lead to abnormalities in lung structure/function and initiation of COPD-associated cellular processes remain obscure.

In the last few years, mitochondrial dysfunction has been postulated as a key mechanism underlying the pathogenesis of several (smoking-related) lung diseases, including COPD (Cloonan and Choi, 2016; Aghapour et al., 2020).

Indeed, mitochondrial dysfunction is known to trigger multiple cellular processes that are also central to COPD pathogenesis (Cloonan and Choi, 2016; Aghapour et al., 2020) and abnormalities in mitochondrial morphology have been described in airway epithelial cells of COPD patients or in cells of the airways in response to CS extract exposure (Mizumura et al., 2014; Sundar et al., 2019; Hara et al., 2013; Hoffmann et al., 2013; Ahmad et al., 2015; Aravamudan et al., 2014). Moreover, a study by Cloonan et al. supported that mitochondrial dysfunction is causally involved in COPD pathogenesis by demonstrating prevention of CS-induced mitochondrial dysfunction in cells of the airways of mice (e.g., Irp2-deficiency, low-iron diet) protected against the development of COPD-like features (Cloonan et al., 2016).

Mitochondria are dynamic organelles in which homeostasis is controlled by an interplay of processes including mitochondrial biogenesis, mitophagy, and mitochondrial fusion and fission (Cloonan and Choi, 2016). The process of mitochondrial biogenesis is tightly regulated by the Peroxisome Proliferator-Activated Receptor (PPAR) Gamma Coactivator 1 (PPARGC1) signaling cascade, in which Peroxisome Proliferator-Activated Receptor Gamma Coactivator 1 alpha (PPARGC1A) co-activators mediate numerous nuclear and mitochondrial transcription factors responsible for controlling the expression of mitochondrial metabolic constituents (Lin et al., 2005; Scarpulla, 2011). Mitophagy, on the other hand, is the process of selective degradation of damaged or defective mitochondria by autophagy driven by a family of specific proteins. Activation of mitochondrial receptors, such as BCL2 Interacting Protein 3 (BNIP3), BNIP3-Like (BNIP3L), and FUN14 Domain Containing 1 (FUNDC1), triggers receptor-mediated mitophagy, while a loss of mitochondrial membrane potential, accumulation of Phosphatase and tensin homolog (PTEN) Induced Kinase 1 (PINK1) and recruitment/activation of Parkin RBR E3 Ubiquitin Protein Ligase (PRKN) on the outer mitochondrial membrane act as signals to initiate ubiquitin-mediated mitophagy. Subsequently, both pathways require general autophagy proteins to prime proteins for elimination by the autophagic-lysosomal pathways (Fritsch et al., 2020).

As mitochondrial biogenesis and mitophagy require fusion and fission, specific fusion associated proteins Mitofusin 1 and 2 (MFN1 and MFN2), OPA1, Mitochondrial Dynamin Like GTPase (OPA1), and fission related proteins Dynamin-1-Like (DNM1L) and Fission, mitochondrial 1 (FIS1) are crucial to facilitate adequate mitochondrial quality control (Mishra and Chan, 2014).

These processes, (i.e., mitochondrial biogenesis and mitophagy) are critical in maintaining mitochondrial quality control in response to inhaled toxicants in the airways. Although literature regarding the mechanisms of mitochondrial toxicity induced by acrolein is scarce, a few in vitro studies indicate that acrolein is able to impair mitochondrial morphology and mitochondrial function in cells of the airways (Wang et al., 2017a; Agarwal et al., 2013; Luo et al., 2013). However, whether or not in vivo acrolein exposure affects mitochondrial content and the processes regulating this has been discussed to a limited extend in previous literature.

The primary aim of the current study is to assess the impact of acute acrolein inhalation in vivo on mitochondrial metabolic pathways as well as mitochondrial biogenesis and mitophagy in the lung. Therefore, rats were exposed to acrolein by nose-only inhalation (0−4 ppm; 4 h/day) for 1 or 2 consecutive days, after which activity and/or abundance of critical constituents of mitochondrial metabolic pathways as well as expression of key proteins and genes regulating the pathways of mitochondrial biogenesis and mitophagy were investigated in lung tissue.

2. Materials and methods

2.1. Acrolein exposure of rats

The animal exposure experiments were conducted at the United States Environmental Protection Agency (Durham, USA) as previously described by Snow et al. (Snow et al. 2017). All animal experiments were approved by the United States Environmental Protection Agency Institutional Animal Care and Use Committee.

Briefly, male Wistar rats (10-weeks old) were obtained from Charles Rivers Laboratories (Raleigh, North Carolina, USA). The rats were housed in a specific pathogen-free AAALAC-approved animal facility on a 12 h light/12 h dark cycle, in pairs in polycarbonate cages with hardwood chip bedding. Rats were provided ad libitum water and food (Rodent Chow 5001: Ralston Purina Laboratories, St Louis, Missouri, USA). After acclimation to nose-only exposure tubes (Lab Products, Seaford, Delaware, USA) for 2 consecutive days for 1 and 2 h respectively, adult Wistar rats (~12-weeks old) were nose-only exposed to air or acrolein (2 or 4 ppm), 4 h/day for 1 or 2 consecutive days (n = 6/group; in total n = 36) as previously described (Snow et al., 2017). The final desired concentrations of acrolein (CAS No. 107–02-8) were achieved by dilution of acrolein gas (1000 ppm cylinder) with filtered compressed air (supplied by a medical grade air compressor). During the nose-only exposure, chamber pressure, flow rates, temperature and relative humidity were controlled. The acute inhalation dose of 4 ppm acrolein is relevant as it is above the range of the environmental presence of acrolein (3.6–10.7 ppb) (De Woskin et al., 2003) and is in the (lower) range of levels found in one cigarette (i.e., 2.1–110.4 ppm) (Pauwels et al., 2018). Moreover, the concentration of acrolein used in this study is in line with prior in vivo acrolein exposure studies (Yeager et al., 2016).

The rats were euthanized using an overdose of sodium pentobarbital (Fatal-Plus diluted 1:1 with saline; >200 mg/kg; I.P. Vortech Pharmaceuticals, Ltd., Dearborn, Michigan, USA) immediately (within 1 h) following 4 h of acrolein exposure at day 1 or immediately (within 1 h) following 2 days of consecutive acrolein exposure at day 2. The left lung lobe was dissected, snap-frozen in liquid nitrogen and stored at −80 °C for further processing.

2.2. RNA isolation, cDNA synthesis and real time quantitative PCR analysis

Lung tissues were crushed by a mortar while frozen. A handheld PRO Scientific Bio-Gen PRO200 homogenizer was used to homogenize approximately 50−100 mg of powdered lung tissue in 1 mL TRIzol™ Reagent (Invitrogen™, USA) for 2 × 10 s at maximum speed. Next, the homogenates were processed per the manufacturer’s protocol (Catalog number 15596026 and 15596018, Invitrogen™). Quantity and purity of the RNA samples were determined by using the NanoDrop ND 1000 UV–vis spectrophotometer (Isogen Life Sciences, de Meern, the Netherlands).

RNA quality (i.e., degradation) was evaluated by running approximately 400− 1000 ng RNA of each sample dissolved 1:10 in Loading Dye (Invitrogen™) in a 2 % (w/v) agarose gel in 0.5 × Tris-borate EDTA stained with Sybr Safe DNA gel Stain (Invitrogen™) for approximately 60 min and run at 100 V.

Reverse transcription of 400 ng of total RNA was conducted using iScript™ cDNA synthesis kit (Bio-Rad, Veenendaal, the Netherlands) according to the manufacturer’s protocol including a no reverse transcription control and a no template control. The cDNA was diluted in milliQ (1:50) and stored at −20 °C until use.

Real time quantitative PCR amplification of genes of interest was conducted by mixing 4.4 μL of 1:50 diluted cDNA, 5 μL 2 x SensiMix™ SYBR® & Fluorescein Kit (Bioline, Alphen aan de Rijn, the Netherlands) and 0.6 μL target and species-specific primers in white LightCycler480 384 multiwell plates (Roche, Basel, Switzerland). In accordance with the manufacturer’s instructions, the following thermal cycling protocol was run on the Roche LightCycler480 machine (Roche): 10 min at 95 °C, 55 cycles of 10 s at 95 °C, 20 s at 60 °C. Melt curves were analyzed using LightCycler480 software (Roche) and gene expression analysis was conducted using LinRegPCR software 2014.× (the Netherlands). In addition, a correction factor was calculated using GeNorm software 3.4 (Primerdesign, Southampton, USA) based on the expression of a combination of at least four reference genes (B2m, Hprt1, Ppia, Rpl13A, Tuba1a) in order to normalize the expression of mRNA transcripts of interest. A list of primers sequences used is shown in Supplementary Table 1.

2.3. DNA isolation and mitochondrial DNA copy number

Approximately 10−30 mg of powdered left lung lobe was dissolved in 250 μL lysis buffer (0.1 M Tris/HCI pH 8.5, 0.005 M EDTA pH 8.0, 0.2 % (w/v) sodium dodecyl sulphate, 0.2 M NaCl) at room temperature. Thereafter, Proteinase K (Qiagen, USA) (>600 mAU/mL) was added to the lysates (1:50). Following incubation overnight at 55 °C, additional 250 μL lysis buffer was added to the lysates where after they were centrifuged at full speed 25,000 x g for 15 min. The supernatant was taken and DNA was precipitated by adding 500 μL isopropanol and shaking vigorously. The DNA pellets were subsequently washed twice using 70 % ethanol. After evaporation of the ethanol, the DNA pellets were dissolved in 125 μL TE-Buffer (10 mM Tris/HCI pH 8.0, 1 mM EDTA pH 8.0) and incubated at 55 °C for 2 h. Lastly, the samples were incubated overnight at 4 °C and subsequently stored at −20 °C until use.

Quantity and purity of the DNA samples were determined by using the NanoDrop ND 1000 UV–vis spectrophotometer (Isogen Life Sciences). Diluted DNA samples (1:50 in TE-Buffer) were analyzed for expression of mitochondrial DNA (mtDNA), mitochondrially encoded cytochrome c oxidase II (Mt-co2), and genomic DNA, peptidylprolyl isomerase A (Ppia), (Supplementary Table 1) using real time quantitative PCR (see paragraph 2.2.). Assessment of the ratio of mtDNA versus genomic DNA yielded mtDNA copy numbers.

2.4. Western blotting

Whole cell lysates for western blotting were generated by homogenizing approximately 50 mg of powdered left lung lobe in 600 μL Immunoprecipitation lysis buffer (50 mM Tris pH 7.4, 150 mM NaCl, 10 % (v/v) glycerol, 0.5 % Nonidet P40, 1 mM EDTA) including PhosSTOP Phosphatase and cOmplete, Mini, EDTA-free Protease Inhibitor cocktail tablets (both Roche) for 5 s at medium speed using a handheld PRO Scientific Bio-Gen PRO200 homogenizer. Following homogenizing, the whole tissue lysates were rotated for 30 min in the cold room and centrifuged at 20,000 × g for 30 min at 4 °C. The protein content was assessed in the supernatant of the whole tissue lysates with the Pierce™ BCA Protein Assay Kit according to the manufacturer’s protocol (Catalog number 23225 and 23227, Thermo Fisher Scientific, Rockford, USA). The supernatant fraction was diluted (1 μg/μL) in a final concentration of 1 x Laemmli buffer (0.25 M Tris pH 6.8, 8 % (w/v) sodium dodecyl sulphate, 40 % (v/v) glycerol, 0.4 M Dithiothreitol, 0.02 % (w/v) Bromophenol Blue) and boiled for 5 min at 100 °C before storage at −80 °C until further analysis.

The samples were run, 10 μg of protein per lane, through a Criterion XT Precast 4–12 or 12 % Bis-Tris gel (Bio-Rad) in 1 x MES running buffer (Bio-Rad).

Two protein ladders or more were loaded on each gel (Precision Plus Protein™ All Blue Standards #161−0373, Bio-Rad). Electrophoresis (100−130 V for 1 h) was used to separate the proteins, followed by electroblotting (Bio-Rad Criterion Blotter) (100 V for 1 h) in order to transfer the proteins on the gel to a 0.45 μM Nitrocellulose Transfer membrane (Bio-Rad). Nitrocellulose membranes were stained with 0.2 % Ponceau S in 1 % acetic acid (Sigma-Aldrich, Zwijndrecht, the Netherlands) for 5 min, followed by a milliQ wash and imaging using the Amersham™ Imager 600 (GE Healthcare, Eindhoven, the Netherlands) in order to quantify total protein content. After washing off the Ponceau S staining, membranes were blocked for 1 h in 3 % (w/v) non-fat dry milk (Campina, Eindhoven, the Netherlands) in Tween20 Tris-buffered saline (20 mM Tris, 137 mM NaCl, 0.1 % (v/v) Tween20, pH 7.6). Thereafter, membranes were washed with Tween20 Tris-buffered saline and incubated overnight at 4 °C with a target-specific primary antibody (Supplementary Table 2) diluted 1:500−1:2,000 in Tween20 Tris-buffered saline with 3 % (w/v) Bovine Serum Albumin or non-fat dry milk. Membranes were subsequently washed and incubated with a horseradish peroxidase-conjugated secondary antibody (Supplementary Table 2) diluted 1:10,000 in 3 % (w/v) non-fat dry milk in Tween20 Tris- buffered saline for 1 h at room temperature. Next, the membranes were washed and incubated for 3 min with either 0.25 × Supersignal West FEMTO or 0.5 × Supersignal West PICO Chemiluminescent Substrate (Thermo Fisher Scientific, Landsmeer, the Netherlands), and visualized using the Amersham™ Imager 600 (GE Healthcare). Original unaltered images were quantified with Image Quant software (GE Healthcare). Original western blots were corrected for total protein loading content assessed by Ponceau S staining over the entire size range of proteins (250 kDa - 10 kDa). Included representative images in the figures of this manuscript have been adjusted for brightness and contrast equally throughout the picture.

The 75 kDa Ponceau S band in the figures is representative for the whole Ponceau S staining, as well as the selected band of one animal is reflecting the changes observed in the whole group (n = 6/group).

2.5. Metabolic enzyme activity assays

Lysates for assessment of metabolic enzyme activity were generated by homogenizing approximately 10−30 mg of powdered left lung lobe in 350 μL SET buffer (pH 7.4: 250 mM sucrose, 2 mM EDTA, 10 mM Tris) for 5 s at medium speed using a handheld PRO Scientific Bio-Gen PRO200 homogenizer. Following homogenizing, the lysates were incubated on ice for 15 min and centrifuged at 20,000 × g for 10 min at 4 °C. The Pierce™ BCA Protein Assay Kit was used to examine total protein content in the whole cell lysate supernatant according to the manufacturer’s protocol (Catalog number 23225 and 23227, Thermo Fisher Scientific). The supernatant of the homogenates were diluted in 5 % Bovine Serum Albumin in milliQ (1:4) and stored at −80 °C until further analysis of Hydroxyacyl-Coenzyme A Dehydrogenase (HADH) and Phosphofructokinase (PFK).

2.5.1. HADH

Assessment of HADH enzyme activity was conducted as previously described (EC 1.1.1.35) (Bergmeyer et al., 1974). Undiluted samples were mixed with reagent (0.22 mM NADH, 100 mM tetrapotassium pyrophosphate pH 7.3) followed by initiation of the reaction by addition of 2.3 mM acetoacetyl-CoA. Samples were spectrophotometrically analyzed in duplicate at 340 nM at 37 °C using the Multiskan Spectrum plate reader (Thermo Labsystems, Breda, the Netherlands). Total protein content of the samples was used as correction factor in order to calculate HADH enzyme activity by slope determination.

2.5.2. PFK

PFK enzyme activity was assessed as previously described (EC 2.7.1.11) (Ling et al., 1966), by mixing undiluted samples with reagent (48.8 mM Tris, 7.4 mM MgCl₂.6H₂O, 74 mM KCl, 384 μM KCN, 2.8 mM ATP, 1.5 mM DTT, 0.3 mM NADH, 0.375 U/mL aldolase, 0.5625 U/mL glycerol-3-phosphate dehydrogenase and 7.425 U/mL triose phosphate isomerase, pH 8.0). Thereafter, the reaction was initiated by addition of fructose-6-phosphate (30.6 mM) in Tris buffer (49.5 mM, pH 8.0). Analysis of the samples was spectrophotometrically conducted in duplicate at 340 nM at 37 °C using the Multiskan Spectrum plate reader (Thermo Labsystems). PFK enzyme activity was calculated by correction for total protein content of the samples.

2.6. L-lactate assay

Whole tissue lysates for assessment of L-lactate were generated by homogenizing approximately 10−30 mg of powdered left lung lobe in 250 μL lactate assay buffer for 10–15 passes at medium speed using a handheld PRO Scientific Bio-Gen PRO200 homogenizer on ice. Next, the lysates were centrifuged at 20,000 × g for 5 min at 4 °C. The supernatant was collected and the endogenous enzyme lactate dehydrogenase was removed using perchloric acid and potassium hydroxide deproteinization by ensuring a final pH 6.5−8. Subsequently, L-lactate levels were examined in undiluted samples in duplicate according to the manufacturer’s protocol (L-lactate Assay kit colorimetric, ab65331, Abcam, Cambridge, USA). Measurement of L-lactate was conducted at 450 nm.

2.7. Statistical analysis

Data are presented as mean fold change compared to air control ± standard error of the mean (SEM). Extreme outliers were excluded in the mRNA data file based on both boxplot analysis (i.e., values more than three interquartile ranges from the end of a box) in IBM Statistics SPSS 25 and melt curve/peak analysis using LightCycler480 software (Roche). Statistical analyses were conducted in GraphPad Prism 8.0 software (La Jolla, California, USA). Statistical significance was assessed by testing normal distribution of the data with the Shapiro-Wilk test. The difference between three groups (2 or 4 ppm acrolein exposure versus air control) was analyzed in case of normal distributed data by an one-way ANOVA and a Dunnett’s post-hoc test for multiple comparisons or in case of non-normal distribution by a Kruskal-Wallis test followed by a Dunn’s multiple comparisons test. In case of analysis of the difference between two groups (4 ppm acrolein exposure versus air control), a two-tailed unpaired parametric t-test was used. Statistical significance was considered if p-values were below 0.05 (*p < 0.05) or 0.01 (**p < 0.01) and a trend was indicated if ^#p < 0.1. GraphPad Prism 8.0 was used to graph the data.

3. Results

3.1. Altered transcript abundance of inflammation and oxidative stress markers in response to acute acrolein inhalation

In accordance with previous findings, showing the influx of inflammatory cells in response to acrolein exposure in rats (Snow et al., 2017), we observed that exposure to 4 ppm acrolein resulted in increased transcript levels of the inflammatory genes NFKB inhibitor alpha (Nfkbia) (1 Day: FC = 2.34 ± 0.56, p < 0.05) and C-C motif chemokine ligand 2 (Ccl2) (1 Day: FC = 4.25 ± 0.91, p < 0.01 and 2 Day: FC = 13.30 ± 1.70, p < 0.01) (Supplementary Fig. 1A). In contrast, exposure to 2 ppm acrolein resulted in decreased mRNA levels of Nfkbia (2 Day: FC = 0.60 ± 0.03, p < 0.1) (Supplementary Fig. 1A). The mRNA levels of C-X-C motif chemokine ligand 1 (Cxcl1), responsible for attracting neutrophils to the site of inflammation, were decreased after 2 days of 2 ppm (FC = 0.55 ± 0.05, p < 0.01) and following 1 or 2 days of 4 ppm acrolein inhalation (1 Day: FC = 0.71 ± 0.09, p < 0.1; 2 Day: FC = 0.63 ± 0.09, p < 0.05) (Supplementary Fig. 1B). Regarding transcript levels of anti-oxidant genes, although transcript abundance of superoxide dismutase 1 (Sod1) was unaltered in response to acrolein inhalation, mRNA levels of superoxide dismutase 2 (Sod2) were significantly elevated after 2 days of 4 ppm acrolein exposure (FC = 1.56 ± 0.17, p < 0.05) (Supplementary Fig. 1C). No significant differences were found in transcript abundance of aldehyde dehydrogenase 2 (Aldh2) following acrolein exposure (Supplementary Fig. 1D). Moreover, we have analyzed protein levels of cleaved caspase-3 as well as the abundance of the ratio of BCL2 associated X, apoptosis regulator and Bcl2, apoptosis regulator (Bax/Bcl2), all established markers of apoptosis, and observed that levels were unchanged or decreased (Cleaved caspase-3: 2 Day: 4 ppm: FC = 0.54 ± 0.12, p < 0.05) suggesting no activation of apoptosis in response to acrolein in our study (Supplementary Fig. 2).

3.2. Acrolein inhalation acutely affects mitochondrial and non-mitochondrial metabolic pathways in rats

To study the effect of acrolein exposure on indices of mitochondrial content and metabolism, we first assessed mtDNA copy number as a marker of mitochondrial content.

As depicted in Fig. 1A, although mtDNA levels appeared to be higher after 1 day of exposure (4 ppm: p = 0.18), no significant differences in mtDNA copy number were observed in acrolein- versus air-exposed rats.

Fig. 1. Acute acrolein exposure affects the abundance of sub-units of the electron transport chain in rat lung.

Wistar rats were exposed by nose-only inhalation to air (control) or acrolein (2 or 4 ppm) for 4 h/day for 1 or 2 days (n = 5–6/group). (A) Mitochondrial DNA (mtDNA) copy number, (B–D) protein and (F) mRNA levels of nuclear- and mitochondrial-encoded constituents of the electron transport chain (i.e., subunits of oxidative phosphorylation (OXPHOS) complexes I–V), as well as (B, E) protein abundance of TOMM20 in rat lung. Representative western blot images are shown of one animal/group reflective of the changes in the whole group as quantified in the corresponding graph. The dashed line indicates that the samples of day 1 and day 2 were loaded on (and selected from) different blots. Data are presented as mean fold change compared to air-exposed rats ± SEM. Individual animals are represented by open circle (air), triangle (2 ppm acrolein) or square (4 ppm acrolein). Statistical differences between 2 or 4 ppm acrolein versus air were tested using an one-way ANOVA followed by a Dunnett’s post-hoc test for multiple comparisons or in case of non-normal distribution a Kruskal-Wallis test followed by a Dunn’s multiple comparisons test. Statistical significance is indicated as ^#p < 0.1, *p < 0.05 and **p < 0.01 compared to air (control).

Next, the impact of acrolein inhalation on key constituents of mitochondrial and non-mitochondrial metabolic pathways in rat lung was evaluated by investigating the activity and/or abundance of components of the electron transport chain, fatty acid beta-oxidation pathway, tricarboxylic acid cycle, and glycolysis.

Firstly, when investigating protein levels of both nuclear-encoded and mitochondrial-encoded sub-units of electron transport chain complexes, protein levels of mitochondrial-encoded sub-unit of complex IV as well as nuclear-encoded sub-units of complexes I and II were unaltered after acrolein exposure (Fig. 1B–D). In contrast, we observed that the protein abundance of nuclear-encoded sub-units associated with electron transport chain complexes III and V were significantly lower (p < 0.05) following 1 day of exposure to 4 ppm acrolein (Fig. 1B and D). In line, protein levels of Translocase Of Outer Mitochondrial Membrane 20 (TOMM20), a critical outer mitochondrial membrane receptor, were also significantly decreased (p < 0.05) in response to 1 day of acrolein exposure (4 ppm) (Fig. 1B and E). Strikingly, the observed differences in protein abundance of all of these mitochondrial proteins in response to 4 ppm acrolein were no longer observed after 2 days of exposure (Fig. 1B, D, and E). Inhalation of a lower dose of acrolein, 2 ppm, did not alter abundance of these analyzed proteins, with the exception of increased (p < 0.1) protein levels of complex III following exposure for 2 consecutive days (Fig. 1B and D).

In addition, evaluation of mRNA levels of nuclear-encoded genes of the respective electron transport chain sub-units revealed no alterations in transcript levels of those investigated nuclear-encoded electron transport chain constituents, except for succinate dehydrogenase complex iron sulfur subunit B (Sdhb) (subunit of complex II) mRNA levels which were significantly decreased (p < 0.01) in response to acrolein exposure (Fig. 1F). In addition, mRNA levels of cytochrome c oxidase subunit 4i1 (Cox4i1) (sub-unit of complex IV) were significantly elevated (p < 0.01) in rat lung after 2 consecutive days of 4 ppm acrolein exposure (Fig. 1F). In general, similar, but less pronounced, findings were shown in rats exposed to 2 ppm acrolein (Fig. 1F).

Besides assessing protein and transcript abundance of sub-units of the electron transport chain complexes, we also examined activity and abundance of constituents of the fatty acid beta-oxidation and tricarboxylic acid cycle. Exposure to acrolein did significantly alter activity levels of HADH, a rate-limiting enzyme in fatty acid beta-oxidation, in response to 2 ppm acrolein (1 Day: FC = 0.78 ± 0.07, p <0.05; 2 Day: FC = 1.44 ± 0.14, p < 0.05), while acrolein did not affect transcript abundance of hydroxyacyl-CoA dehydrogenase trifunctional multienzyme complex subunit alpha (Hadha) and acyl-CoA dehydrogenase, long chain (Acadl) (Supplementary Fig. 3A–3C). On the other hand, transcript levels of citrate synthase were significantly increased in response to 4 ppm acrolein compared to control both after 1 and 2 days of exposure (1 Day: FC = 1.48 ± 0.14, p < 0.05; 2 Day: FC = 1.79 ± 0.15, p < 0.05) (Supplementary Fig. 3D).

Subsequently, to examine the effect of acrolein exposure on glycolysis, we assessed activity and abundance of key components of the glycolytic pathway. While PFK activity was unaltered, L-lactate levels were significantly elevated (p < 0.05) on day 1 in response to 4 ppm acrolein exposure (Fig. 2A and B). In addition, we evaluated protein and transcript abundance of Hexokinase 2 (HK2), responsible for the first step in glycolysis, and observed increases in HK2 protein as well as transcript levels of Hk2 in response to 4 ppm acrolein exposure for 2 consecutive days (p < 0.05 and p < 0.1, respectively) (Fig. 2C and D).

Fig. 2. Alterations in activity and expression of key constituents of glycolysis in response to acrolein inhalation in rats.

Wistar rats were exposed by nose-only inhalation to air (control) or acrolein (2 or 4 ppm) for 4 h/day for 1 or 2 days (n = 6/group). (A) Relative L-lactate levels are presented. Statistical differences between 4 ppm acrolein versus air were tested using a two-tailed unpaired parametric t- test. (B) PFK activity, (C) protein abundance of HK2, and (D) transcript levels of Hk2 in rat lung homogenates. Representative western blot images are shown of one animal/group reflective of the changes in the whole group as quantified in the corresponding graph. The dashed line indicates that the samples of day 1 and day 2 were loaded on (and selected from) different blots. Statistical differences between 2 or 4 ppm acrolein versus air were tested using an one-way ANOVA followed by a Dunnett’s post-hoc test for multiple comparisons. Data are presented as mean fold change compared to air-exposed rats ± SEM. Individual animals are represented by open circle (air), triangle (2 ppm acrolein) or square (4 ppm acrolein). Statistical significance is indicated as ^#p < 0.1, *p < 0.05 and **p < 0.01 compared to air (control).

Although no alterations were observed in the transcript abundance of lactate dehydrogenase A (Ldha), pyruvate kinase M1/2 (Pkm), pyruvate dehydrogenase kinase 4 (Pdk4) and solute carrier family 2 member 1 (Slc2a1) in acrolein-exposed rats compared to control, mRNA levels of solute carrier family 2 member 4 (Slc2a4) were lower following 1 day of 4 ppm acrolein exposure (FC = 0.46 ± 0.16, p < 0.1) (Supplementary Fig. 3E–H).

Overall, these results suggest alterations in the abundance and activity of constituents of key metabolic processes, especially at the level of the sub-units of electron transport chain complexes, as well as a shift towards a more glycolytic metabolism in acrolein-exposed rat lung.

3.3. The molecular regulation of mitochondrial biogenesis is disrupted in rats acutely exposed to acrolein

In order to examine if the abovementioned alterations in mitochondrial metabolic pathways were associated with changes in regulation of mitochondrial biogenesis, we assessed the abundance of key regulators of mitochondrial biogenesis in response to acute acrolein exposure in rat lung.

Firstly, we studied the abundance of transcriptional co-activators of the PPARGC1 signaling cascade, a central network associated with the regulation of mitochondrial biogenesis and cellular mitochondrial energy metabolism.

As shown in Fig. 3A and B respectively, only 4 ppm acrolein exposure resulted in decreased (p < 0.1) protein as well as transcript levels of PPARGC1A, respectively on day 1 and day 2. In contrast, transcript abundance of Ppargc1b, a homologue of Ppargc1a, was significantly elevated (2ppm: p < 0.05; 4 ppm: p < 0.01) following acrolein inhalation for 2 consecutive days (Fig. 3C), and no differences were observed in transcript levels of PPARG related coactivator 1 (Pprc1) (Supplementary Fig. 4A).

Fig. 3. Abundance of PPARGC1 molecules is altered in rats acutely exposed to acrolein.

Wistar rats were exposed by nose-only inhalation to air (control) or acrolein (2 or 4 ppm) for 4 h/day for 1 or 2 days (n = 6/group). Protein levels of (A) PPARGC1A and gene expression of (B) Ppargc1a and (C) Ppargc1b in rat lung. Representative western blot images are shown of one animal/group reflective of the changes in the whole group as quantified in the corresponding graph. The dashed line indicates that the samples of day 1 and day 2 were loaded on (and selected from) different blots. Data are presented as mean fold change compared to air-exposed rats ± SEM. Individual animals are represented by open circle (air), triangle (2 ppm acrolein) or square (4 ppm acrolein). Statistical differences between 2 or 4 ppm acrolein versus air were tested using an one-way ANOVA followed by a Dunnett’s post-hoc test for multiple comparisons or in case of non-normal distribution a Kruskal-Wallis test followed by a Dunn’s multiple comparisons test. Statistical significance is indicated as ^#p < 0.1, *p < 0.05 and **p < 0.01 compared to air (control).

Next, we assessed whether acrolein exposure affected the abundance of PPARGC1-coactivated transcription factors involved in the regulation of mitochondrial biogenesis and mitochondrial energy metabolism. While protein as well as mRNA levels of some investigated transcription factors of the PPARGC1 network were largely unaltered (Fig. 4A, B, D, E; Supplementary Fig. 4B–E), decreased mRNA levels of nuclear respiratory factor 1 (Nrf1) (1 Day: 2 ppm: p <0.1; 2 Day: 2 ppm: p <0.01 and 4 ppm: p < 0.05) and peroxisome proliferator-activated receptor alpha (Ppara) (1 Day: 2 ppm: p < 0.1 and 4 ppm: p < 0.01) were observed in response to exposure to acrolein as shown in Fig. 4C and F.

Fig. 4. Acute acrolein inhalation affects the abundance of PPARGC1-coactivated transcription factors.

Wistar rats were exposed by nose-only inhalation to air (control) or acrolein (2 or 4 ppm) for 4 h/day for 1 or 2 days (n = 5– 6/group). (A) Protein levels of (B) NRF1 and (D) ESRRA as well as transcript abundance of (C) Nrf1, (E) Esrra, and (F) Ppara in rat lung. Representative western blot images are shown of one animal/group reflective of the changes in the whole group as quantified in the corresponding graph. The dashed line indicates that the samples of day 1 and day 2 were loaded on (and selected from) different blots. Data are presented as mean fold change compared to air-exposed rats ± SEM. Individual animals are represented by open circle (air), triangle (2 ppm acrolein) or square (4 ppm acrolein). Statistical differences between 2 or 4 ppm acrolein versus air were tested using an one-way ANOVA followed by a Dunnett’s post-hoc test for multiple comparisons. Statistical significance is indicated as ^#p < 0.1, *p < 0.05 and **p < 0.01 compared to air (control).

In conclusion, these findings indicate that acute acrolein inhalation disrupts the molecular regulation of mitochondrial biogenesis, in particular at the level of the PPARGC1 signaling cascade.

3.4. The molecular regulation of mitophagy and mitochondrial dynamics following acute acrolein inhalation

Besides mitochondrial biogenesis, mitophagy (i.e., mitochondrial-specific autophagy) plays an essential role in regulating mitochondrial content and function. In order to study the impact of acrolein inhalation on mitophagy, we explored protein and mRNA levels of key regulators of receptor- and ubiquitin-mediated mitophagy.

As depicted in Fig. 5A, B, and D, BNIP3L protein levels were reduced in response to 4 ppm acrolein (1 Day: p < 0.1; 2 Day: p < 0.01), while BNIP3 protein abundance was increased (2 ppm: p < 0.1; 4 ppm: p < 0.05) in response to 2 days of acrolein exposure. In addition, mRNA transcript abundance of these two key receptor proteins involved in receptor-mediated mitophagy, were unaltered in response to acrolein (Fig. 5C and E).

Fig. 5. Altered abundance of key constituents of receptor-mediated mitophagy in response to acrolein exposure in rats.

Wistar rats were exposed by nose-only inhalation to air (control) or acrolein (2 or 4 ppm) for 4 h/day for 1 or 2 days (n = 5–6/group). Receptor-mediated mitophagy-associated (A) protein abundance of (B) BNIP3L, (D) BNIP3, and (F) FUNDC1, as well as transcript abundance of (C) Bnip3l and (E) Bnip3, and (G) Fundc1 in rat lung tissue are presented. Representative western blot images are shown of one animal/group reflective of the changes in the whole group as quantified in the corresponding graph. The dashed line indicates that the samples of day 1 and day 2 were loaded on (and selected from) different blots. Data are presented as mean fold change compared to air-exposed rats ± SEM. Individual animals are represented by open circle (air), triangle (2 ppm acrolein) or square (4 ppm acrolein). Statistical differences between 2 or 4 ppm acrolein versus air were tested using an one-way ANOVA followed by a Dunnett’s post-hoc test for multiple comparisons or in case of non-normal distribution a Kruskal-Wallis test followed by a Dunn’s multiple comparisons test. Statistical significance is indicated as ^#p < 0.1, *p < 0.05 and **p < 0.01 compared to air (control).

Moreover, protein and mRNA levels of FUNDC1, also a key constituent of receptor-mediated mitophagy, were not changed in rat lung after exposure to acrolein (Fig. 5A, F, and G). In addition, no significant changes in protein and mRNA levels of regulators of ubiquitin-mediated mitophagy (PINK1 and PRKN) were observed in rats following acute acrolein inhalation (Supplementary Fig. 5).

As mitophagy requires general autophagy proteins in order to prime proteins for degradation by the autophagic-lysosomal pathways, we studied the abundance of general autophagy proteins. As shown in Supplementary Fig. 6, protein and transcript levels of all investigated general autophagy-related constituents (GABA Type A Receptor Associated Protein Like 1 (GABARAPL1), (ratio) Microtubule-Associated Protein 1 Light Chain 3 Beta (MAP1LC3B), and Beclin 1 (Becn1)) were unchanged in rats acutely exposed to acrolein, however, Sequestosome 1 (SQSTM1) protein levels were increased in response to the highest dose of acrolein for 2 consecutive days (FC = 1.47 ± 0.12, p < 0.01).

In summary, these data suggest minor alterations in the molecular regulation of the mitophagy machinery, in particular in receptor-mediated mitophagy, while in general no changes were identified in the abundance of autophagy-related constituents in response to acute acrolein exposure in rats.

Considering the essential role of mitochondrial dynamics together with mitochondrial biogenesis and mitophagy to maintain mitochondrial homeostasis, we next aimed to explore the impact of acrolein exposure on abundance of key constituents of fusion and fission in rat lung.

Although mRNA levels of regulators of fusion Mfn1 and Mfn2 were not affected by acrolein inhalation, transcript abundance of Opa1 was significantly increased (p < 0.05) in response to 4 ppm acrolein exposure for 1 day (Fig. 6A–C).

Fig. 6. Transcript expression of regulators of mitochondrial dynamics is disrupted in rats exposed to acrolein.

Wistar rats were exposed by nose-only inhalation to air (control) or acrolein (2 or 4 ppm) for 4 h/day for 1 or 2 days (n = 5–6/group). Fusion-associated gene expression of (A) Mfn1, (B) Mfn2, and (C) Opa1, as well as transcript abundance of fission-associated genes (D) Dnm1l and (E) Fis1 in rat lung. Data are presented as mean fold change compared to air-exposed rats ± SEM. Individual animals are represented by open circle (air), triangle (2 ppm acrolein) or square (4 ppm acrolein). Statistical differences between 2 or 4 ppm acrolein versus air were tested using an one-way ANOVA followed by a Dunnett’s post-hoc test for multiple comparisons or in case of non-normal distribution a Kruskal-Wallis test followed by a Dunn’s multiple comparisons test. Statistical significance is indicated as ^#p < 0.1, *p < 0.05 and **p < 0.01 compared to air (control).

In addition, with regard to fission, transcript levels of Dnm1l were significantly elevated after 2 days of acrolein exposure (2 ppm: p < 0.05; 4 ppm: p < 0.01), while Fis1 mRNA levels were decreased (p < 0.01) in response to the highest dose of acrolein after 1 day (Fig. 6D and E). Collectively, these results indicate alterations in the molecular regulation of fusion and fission which are critical processes involved in mitochondrial dynamics.

4. Discussion

In this study, we investigated the acute effect of acrolein inhalation, a component of CS, on key regulators of mitochondrial content and function in rat lung. We observed that acrolein inhalation resulted in a dose-dependent and differential response on the activity and abundance of key constituents of mitochondrial metabolic pathways on day 1 versus day 2 in particular leading to a transient decrease in abundance of several subunits of electron transport chain complexes and to decreased protein levels of critical regulators involved in mitochondrial biogenesis.

Acrolein-induced inflammation and oxidative stress are well described in various in vivo and in vitro airway exposure models (Yeager et al., 2016; Luo et al., 2013; Dwivedi et al., 2018; Cichocki et al., 2014; Xiong et al., 2018; Park et al., 2017; Zhang and Forman, 2008; Moghe et al., 2015). In accordance to previous literature, acrolein exposure in our study induced inflammatory and Sod2 gene expression, as well as a previously reported influx of inflammatory cells in bronchoalveolar and nasal lavage fluid illustrative of effective acrolein administration to the rat lungs (Snow et al., 2017). The observed decrease in expression of Cxcl1 can be clarified by stress hormone-modulated neutrophilic inflammation and cell-specific extravasation of neutrophils as previously suggested by Snow et al. (Snow et al., 2017). Although, previous literature described the impact of acrolein inhalation on cell death and apoptosis in vivo (Kitaguchi et al., 2012; Kim et al., 2018) and it has been shown that acrolein can induce apoptosis in cultured airway epithelial cells (Wang et al., 2017a; Xiong et al., 2018; Roy et al., 2009; Yadav et al., 2013; Zhang et al., 2020), we did not observe acrolein-induced apoptosis in our model.

Next, we focused on the impact of acrolein on the abundance of key constituents of mitochondrial metabolic pathways and observed in particular a decreased abundance of subunits of the electron transport chain. This is in line with previous studies reporting that acrolein exposure resulted in increased intracellular and mitochondrial ROS production, mtDNA damage, and disturbed mitochondrial metabolism in a variety of in vitro airway models and isolated rat liver mitochondria (Wang et al., 2017a; Agarwal et al., 2013; Luo et al., 2013; Sun et al., 2006). Interestingly, comparable findings have been described in multiple in vitro or in vivo (airway) models in response to CS (extract) (Aghapour et al., 2020; Mizumura et al., 2014; Sundar et al., 2019; Cloonan et al., 2016; Wu et al., 2020; Park et al., 2019) or several aldehydes (Clapp et al., 2019; Pei et al., 2014; Farfán Labonne et al., 2009). The fact that mtDNA copy number did not change in our study while we did observe changes in the abundance of proteins involved in mitophagy and mitochondrial biogenesis could be explained by the fact that multiple copies of mtDNA exist in a mitochondrion, lack of evaluation of the activity of mitophagy and mitochondrial biogenesis per se (only the abundance of proteins involved in these processes) and timing of our analyses after exposure. Nevertheless, these results, in concert with data of our study, strongly suggest that aldehydes (i.e., acrolein) may mediate, at least part of, the detrimental effects that CS exerts on the molecular regulation of mitochondrial content and function in airway epithelial cells.

Furthermore, in light of the mentioned elevated ROS levels upon acrolein or CS exposure, interestingly in particular glycolysis is a susceptible process affected by changes in redox balances (Mullarky and Cantley, 2015). This would be in line with our findings of elevated L-lactate levels and increased abundance of HK2. Available literature also supports smoke-induced disruption of glycolytic metabolism, which has been described not only in cultured airway epithelial cells or in vivo models exposed to aldehydes (Agarwal et al., 2013; Sun et al., 2006; Clapp et al., 2019; Fabisiak et al., 2011; Novotny et al., 1994) or CS (Agarwal et al., 2014; Agarwal et al., 2012), but also in COPD (Tu et al., 2014; Kao et al., 2012; Pastor et al., 2013; Agarwal et al., 2019). Our findings of unaltered mtDNA levels, decreased abundance of mitochondrial proteins (e.g., subunits of oxidative phosphorylation (OXPHOS) complexes) and increased lactate levels in response to acrolein suggest that mitochondria become less oxidative, while the number of mitochondria remains the same, which is at least partly in line with the findings from previous studies (Wang et al., 2017a). Taken into account the above-mentioned findings, this can indicate an impaired aerobic respiration resulting in a shift to glycolytic energy metabolism.

To the best of our knowledge, there are only a few in vitro studies examining the effect of aldehydes on the regulation of mitochondrial biogenesis. In line with our data, these studies showed declined PPARGC1A protein levels in acrolein-exposed lung fibroblasts (Luo et al., 2013) or crotonaldehyde-exposed cardiomyocytes (Pei et al., 2014). Accordingly, while increased expression levels of PPARGC1A were observed in lung homogenates of mild COPD patients and in ex-smoking severe COPD patients or short-term CS extract-stimulated bronchial epithelial cells (Hoffmann et al., 2013; Vanella et al., 2017; Li et al., 2010), decreased abundance of regulators of biogenesis were shown in moderate and severe COPD patients (Li et al., 2010) and in in vivo subacute COPD models (Wang et al., 2020; Wang et al., 2017b). However, also opposite findings were reported (Li et al., 2010; Walczak et al., 2020). Differences likely result from variation in duration of CS extract exposure, dose, and models investigated in those studies. Moreover, we observed a significantly increased transcript abundance of Ppargc1b after 2 days of exposure, which may be explained as a compensatory/repair response to mitochondrial damage elicited by acute acrolein exposure. The reductions in abundance of key regulators of mitochondrial biogenesis in response to acrolein that we observed in our study shed some light on the molecular mechanism potentially underlying acrolein-induced impairment of mitochondrial metabolism. Although a causal relationship between these reductions and impairment of mitochondrial metabolism upon exposure to acrolein cannot be concluded from our study, it is feasible that, due to the essential role of these regulators in the molecular control of the mitochondrial biogenesis program that is well described in mammalian cells (Lin et al., 2005; Scarpulla, 2011), these reductions serve as a molecular basis for these metabolic changes. In addition, it is well-known from studies in other cell types than lung cells that activation of the inflammatory NF-κB pathway directly impedes on the activity of the PPARGC1 signaling network that controls mitochondrial biogenesis. For example, P65 (the main transcription factor mediating effects of NF-κB activation) has been shown to be able to physically bind to PPARGC1 and reduces its expression thereby impeding its function (Alvarez-Guardia et al., 2010) and blockade of the NF-κB pathway prevented detrimental effects of inflammatory cytokines on mitochondrial metabolism (Remels et al., 2013). Although, the impact of acrolein on the NF-κB pathway has been described as diverse and conflicting in literature depending on the dose and duration of acrolein exposure (Yeager et al., 2016; Moghe et al., 2015; Yadav and Ramana, 2013), it has been suggested that acrolein can induce an inflammatory response and activate the NF-κB pathway (Moghe et al., 2015), e.g., as demonstrated in an in vivo study (Sun et al., 2014). Collectively, these findings suggest that inflammation-induced impairments in the PPARGC1 signaling network may well serve as the molecular mechanism underlying acrolein-induced abnormalities in mitochondrial metabolism as we observed in our study.

Although no striking differences were shown in the abundance of components of receptor- and ubiquitin-mediated mitophagy, elevated BNIP3 and reduced BNIP3L protein levels were observed in response to acrolein. To the best of our knowledge, only one study showed acute acrolein-induced mitophagy (Wang et al., 2017a) as well as one study showed that CS-induced mitophagy is BNIP3L-dependent in human airway (epithelial) cells (Zhang et al., 2019a). More is known about ubiquitin-mediated mitophagy, as increased PINK1 and decreased PRKN abundance has been demonstrated in various in vitro/vivo (airway) models of CS exposure (Mizumura et al., 2014; Ahmad et al., 2015; Wu et al., 2020; Son et al., 2018; Kyung et al., 2018; Ito et al., 2015) as well as in lung tissue or bronchial epithelial cells of (ex-smoking) COPD patients (Mizumura et al., 2014; Hoffmann et al., 2013; Ahmad et al., 2015; Ito et al., 2015). However, whether or not mitophagy plays a protective role or contributes to CS-induced airway pathology remains to be established as illustrated in previous studies (Mizumura et al., 2014; Sundar et al., 2019; Ahmad et al., 2015; Wu et al., 2020; Son et al., 2018; Kyung et al., 2018; Ito et al., 2015; Araya et al., 2019; Ballweg et al., 2014).

Previous research suggested increased autophagy in response to acrolein or crotonaldehyde in human lung epithelial cells or fibroblasts (Wang et al., 2017a; Wang et al., 2017c; Wang et al., 2019) and after formaldehyde inhalation in rats (Liu et al., 2018). Similar findings were reported in lung homogenates of COPD patients (Chen et al., 2008), in CS (extract)-exposed human cells of the airways (Park et al., 2019; Zhang et al., 2019a; Ito et al., 2015; Chen et al., 2008) and in CS-exposed mice (Mizumura et al., 2014; Chen et al., 2008). However, autophagy-associated indices were mostly reported to be unchanged in our study suggestive of a different impact of acrolein and CS in various exposure models.

Due to the essential role of mitochondrial dynamics in controlling mitochondrial homeostasis, we further studied abundance of fusion and fission regulators. In line with our findings of increased Dnm1l transcript levels in response to acrolein inhalation, Wang et al. also demonstrated increased mitochondrial fission upon acrolein exposure in human lung cells (Wang et al., 2017a). Collectively, previous studies have reported indications for reduced fusion and increased fission in COPD lung homogenates (Mizumura et al., 2014), alveolar epithelial cells of smokers (Kosmider et al., 2019), and in vivo (Mizumura et al., 2014) and in vitro smoke-exposure models (Mizumura et al., 2014; Sundar et al., 2019; Hara et al., 2013; Aravamudan et al., 2014; Son et al., 2018; Kyung et al., 2018; Song et al., 2017). Discrepancies in abundance of regulators involved in mitochondrial dynamics in response to various types of CS in lung epithelial cells (Hoffmann et al., 2013; Ahmad et al., 2015; Park et al., 2019; Ballweg et al., 2014), could be explained by variation in exposure models or duration of CS exposure as illustrated by Walczak et al. (Walczak et al. 2020). Although we showed a disrupting effect of acrolein exposure on the transcriptional regulation of mitochondrial fusion and fission, further studies have to reveal the functional impact of acrolein on mitochondrial dynamics.

As our data, in concert with prior evidence, suggest acrolein-induced disruption of the molecular regulation of mitochondrial content/function and mitochondrial dysfunction is being implicated in the pathogenesis of COPD, acrolein is an interesting potential target to be regulated in CS.

Acrolein, together with acetaldehyde and formaldehyde, are considered to be representative aldehydes for CS in all brands and human smoking regimes due to similar molecular structures, properties, precursors and mechanisms of formation and are extremely toxic due to local peak concentrations (WHO, 2008; Burns et al., 2008; Cheah, 2016; Corley et al., 2015). They are shortlisted for regulation by the World Health Organization Study Group on Tobacco Product Regulation due to their adverse impact on human health for which the current study provides more evidence (WHO, 2008; Burns et al., 2008; WHO, 2012). Potential reduction of aldehyde exposure per puff can be facilitated (WHO, 2008) e.g., by reducing sugar content (Talhout et al., 2006), selection of type of tobacco (Baker et al., 2004), and cigarette design (Pauwels et al., 2018).

Previous studies indicated that the immunologic response and mechanism of cell death (apoptosis, oncosis or necrosis) following acrolein exposure may be dose- and cell type-related (Yeager et al., 2016; Moghe et al., 2015). We found a lower induction of inflammatory gene expression after acute exposure to 2 ppm compared to 4 ppm acrolein. This coincided with the observed less pronounced changes in the expression of regulators of mitochondrial metabolism in response to 2 ppm acrolein, while exposure to 4 ppm resulted in significant deregulation of mitochondrial metabolism indicating that, not only the inflammatory response, but also modulation of other cellular processes depends on the dose and duration of exposure. Previous in vivo studies also reported dose-dependent biological effects ranging from lethal effects after short high dose acrolein exposure in rat up to 40,000 ppm (Ashizawa et al., 2007) to lung inflammation and cell death in animals exposed up to 10 ppm (Park et al., 2017; Kim et al., 2018; Fabisiak et al., 2011). In general, the concentrations of acrolein (2 and 4 ppm) used in this study have been associated with milder effects on lung health in rat (De Woskin et al., 2003; Ashizawa et al., 2007) as for example depression of respiratory rate has been reported in animals exposed to 4.6 ppm (Bergers et al., 1996) or 9.2 ppm (Cassee et al., 1996).

This study, for the first time, provides a comprehensive overview of the impact of acrolein on the regulation of mitochondrial content and function in the lung. Nevertheless, our study has several limitations. Obviously, analyses were conducted in lung tissue homogenates which includes various cell types, so effects of acrolein on specific cell types cannot be dissected from this study. However, previous in vitro studies demonstrated a detrimental impact of acrolein exposure on the regulation of mitochondrial function and quality control (i.e., mitochondrial biogenesis and mitophagy) in various cell types, e.g., lung fibroblasts, alveolar or bronchial epithelial cells, suggesting that these cell types likely are also impacted by acrolein in our study (Wang et al., 2017a; Agarwal et al., 2013; Luo et al., 2013; Roy et al., 2009; Zhang et al., 2020; Moretto et al., 2009; Zhang et al., 2019b). Secondly, reported changes in abundance of key regulators involved in the processes of mitochondrial biogenesis, mitophagy, fusion, and fission are not indicative for actual functional process or mitochondrial content changes, limiting conclusions that can be drawn from our data. Thirdly, different from human smokers exposed to CS, nose-only inhalation of acrolein in our model likely primarily resulted in exposure in the nasal airways as suggested due to alteration of breathing parameters as protecting mechanism in response to acrolein (Snow et al., 2017; Corley et al., 2015; Lee et al., 1992; Perez et al., 2015; Corley et al., 2012). Nevertheless, from the (previous) reported inflammatory profile in the airways of the acrolein-exposed rats, it is still clear that acrolein reached the cells of the airways in our study (Snow et al., 2017). Fourthly, the continuous nose-only exposure model that we deployed does not accurately reflect the puff-topography that smokers usually display (Cheah, 2016; Corley et al., 2015). Fifthly, while no studies are investigating the impact of gender in response to acrolein exposure we cannot exclude that gender affect the observed differences in our study as previous literature reported a sex-dependent ozone-mediated inflammatory and physiological response (Yaeger et al., 2021; Birukova et al., 2019) and inflammation may damage mitochondrial metabolism (Kampf et al., 1999). Besides gender, previous studies indicated age-related deterioration of mitochondrial function in lung disease (Cloonan et al., 2020; Mercado et al., 2015) suggesting an aggravated mitochondrial response following acrolein inhalation. However, by opting for a single gender (male) at a specific (young) age, we prevented confounding by these variables. Furthermore, no pronounced correlations were observed between analyzed parameters related to mitochondrial homeostasis (e.g., mitochondrial biogenesis, mitophagy) correlated with pathological hallmarks associated with inflammation, lung damage or apoptosis (0 versus 4 ppm acrolein; data not shown). Moreover, the differential time course effects observed in our study on 1 day versus 2 day could be explained by the immediate cellular effects analyzed at day 1 mediated by a stress response and cellular metabolic and immune alterations as previously shown in our ozone and acrolein-exposure studies (Snow et al., 2017; Henriquez et al., 2018; Henriquez et al., 2021; Miller et al., 2016), while at day 2 a combination of immediate response including a peak inflammatory response are investigated. Also, conflicting changes were observed at mRNA and protein level in the analyzed markers of mitochondrial metabolism. Potential explanations for this could be that the time point (1 h post exposure) may be sufficient to induce transcriptional changes but may be too short to pick up changes in protein levels at day 1, the low number of animals per group as well as the biological variation (SEM). In addition, it is known, albeit in other cell types, that acrolein is able to induce post-transcriptional (e.g., mRNA splicing (Haberzettl et al., 2009)), and post-translational (e.g., protein-acrolein adducts (LoPachin and Gavin, 2014)) modifications and can impact expression of miRNA’s (O’Toole et al., 2014) potentially clarifying why mRNA changes not correlate with protein changes. Lastly, assessment of the effect of individual aldehydes on mitochondrial metabolism is not representative for exposure of cells of the airways to the complex combination of aldehydes present in CS, indicated to induce synergistic toxicity (Zhang et al., 2019b; LoPachin and Gavin, 2014; Zhang et al., 2018). Therefore, it is relevant to study co-exposure of aldehydes for regulation in the future.

In conclusion, our study shows that acute acrolein exposure has an impact on the molecular regulation of mitochondrial metabolism in rat lung. Further research should focus on assessing the effect of acrolein individually and in combination with acetaldehyde and formaldehyde, on mitochondrial function. Although mitochondrial dysfunction has been implicated in COPD pathogenesis, whether or not this can contribute to COPD development remains to be established. Hence, understanding the effect of acrolein on mitochondrial function will provide a scientifically supported reasoning to shortlist regulation of aldehydes in tobacco smoke.

Data availability

Data will be made available on request.