Littermate female C57BL/6 mice aged eight weeks (body weight range: 18.1–20.0 g) were used. The mice were housed with food and water ad libitum at room temperature at a 12:12 light-dark cycle. The investigations conformed to the Guidelines for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH publication, 8th Edition, 2011). Our research protocol was approved by the Fukushima Medical University Animal Research Committee. The study on animals was carried out in compliance with the ARRIVE guidelines (Additional file 1). All efforts were made to minimize the suffering of animals, all of which were sacrificed by cervical dislocation after the experiments.

The experimental protocol is shown in Fig. 1. Experiment 1: The mice were exposed to hypoxic (10% oxygen) or normoxic conditions for four weeks, during which they were given a normal chow. Thereafter, we evaluated the right ventricular systolic pressure (RVSP), right ventricular hypertrophy (RVH), expression of some associated molecules, and histological features. Experiment 2: The mice were exposed to hypoxic condition (10% oxygen) for four weeks, during which they were given a normal chow. Then, normoxia was resumed, and the mice were divided by a person, who was blinded to the investigation, into two groups; the ND group mice were fed with a normal diet (CLEA Rodent Diet CA-1, CLEA Japan Inc., Tokyo, Japan), while the HFD group mice were fed with HFD (High Fat diet 32, CLEA Japan Inc.) for 12 weeks (Table 1). The oxygen concentration of hypoxia was determined based on a previous study [4]. The required sample size was estimated to be 10–20 per group based on the number within the range reported in previous studies [12, 13]. The used animal number was described in each figure legend. The total number of animals used was 62. The mice in all groups were treated at once, and each group was allocated in each cage. After we unified the cage locations, we randomly measured the RVSP to avoid the effect of the order of animals and cages.

Intraperitoneal injection of anesthesia was done using Tribromoethanol (0.25 mg/g of body weight). At 21% O₂ condition, A 1.2-F micromanometer catheter (Transonic Scisense Inc., London, ON, Canada) was inserted to the right jugular vein, and the RVSP was measured without a ventilator and analyzed by LabScribe3 software (IWORX, Dover, NH, USA) [14]. Then, the mice were sacrificed by cervical dislocation; the heart and lungs were removed for RVH evaluation. The right ventricle (RV) was dissected from the left ventricle (LV), including the septum (S), and the RV/LV + S weight ratio (Fulton index) was calculated [14, 15]. The mice with severe bleeding or pneumothorax that affected the hemodynamics during catheterization were excluded from the RVSP analysis, as shown in Fig. 1.

After measuring the RV pressure, the right lung was fixed with 4% paraformaldehyde, embedded in paraffin, and sectioned to 3 μm. After Elastica-Masson (EM) staining or immunostaining of α-smooth muscle actin (α-SMA) (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), the pulmonary arteries (external diameter of 20–50 μm) were randomly selected (60–90 vessels per individual mouse). The medial wall area (the area between the internal and external lamina) was measured using the Image J 1.48 software (National Institutes of Health, Bethesda, MD, USA) and divided by the vessel area (the area surrounded by the external lamina) [14, 15]. Furthermore, each vessel (external diameter of < 20 μm) was classified as non-muscular, partially muscular, or fully muscular, as described in a previous report. The percentage of muscularized pulmonary vessels was determined by dividing the sum of partially and fully muscular vessels by the total number of vessels [14]. Measurements were performed blinded to the mouse information.

The plasma from mice at 24 weeks was isolated by centrifugation (3000 rpm, 15 min), frozen at − 80 °C, and analyzed by Nagahama Life Science Laboratory (Shiga, Japan). The plasma concentrations of high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol (LDL-C) were measured by direct method using Chorestest N HDL® (Sekisui Medical Co., Ltd. Tokyo, Japan) and Chorestest LDL® (Sekisui Medical). The plasma triglyceride (TG) was measured by enzymatic method using L-type Wako TG∙M® (FUJIFILM Wako Pure Chemical Corporation. Osaka, Japan).

To evaluate the mitochondrial function, we measured the mitochondrial adenosine triphosphate (ATP) in the lungs based on a previous study [4]. The removed left lung was homogenized, and the mitochondrial fraction was isolated by centrifugation. Then, the ATP production was measured using a commercially available kit (TOYO B-Net Co., Ltd. Tokyo, Japan), following the manufacturer’s instruction.

Mitochondrial DNA was quantified as described previously [16]. The DNA from the left lung was extracted and purified by QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instruction, with the inclusion of RNAse A digest. Quantitative polymerase chain reaction (PCR) was performed using the power SYBR green PCR master mix with a CFX Connect real-time PCR System (Bio-Rad Laboratories, Inc., Hercules, CA) using mitochondrial DNA-specific primer, cytochrome oxidase subunit 1 (Co1) gene, nuclear DNA-specific primer, and NADH:Ubiquinone oxidoreductase core subunit V1 (NDUFV1). The sequences of primers were as follows: forward 5′-TGCTAGCCGCAGGCATTAC-3′ and reverse 5′-GGGTGCCCAAAGAATCAGAAC-3′ for Co1, and forward 5′-CTTCCCCACTGGCCTCAAG-3′ and reverse 5′-CCAAAACCCAGTGATCCAGC-3′ for NDUFV1. All reactions were run in duplicates. The mitochondrial DNA content was calculated using the delta CT method, and the data were expressed as a fold increase in the ND group.

The removed whole left lungs were used for western blotting, which was performed as described previously [17, 18]. Polyvinylidene difluoride membranes were incubated with mouse monoclonal antibodies to β-actin (Santa Cruz) and heme oxygenase-1 (HO-1) (Novus Biologicals USA, Littleton, CO, USA) diluted to 1:1000 for 1 h at room temperature. Rabbit polyclonal antibodies to active caspase-3 (Cell Signaling Technology, Beverly, MA, USA), PPAR-γ (Cell Signaling Technology), voltage-dependent anion channel (VDAC) (Cell Signaling Technology), hypoxia-inducible factor-1α (HIF-1α) (Cell Signaling Technology), and apelin (Bioss Inc, Boston, MA, USA) were diluted to 1:500. Then, each membrane was incubated with a horseradish peroxidase-conjugated secondary antibody or IRDye® 680RD or IRDye®800CW (LI-COR, Inc., Lincoln, NE, USA) diluted to 1:10,000 for 45 min.

To estimate the apoptosis of pulmonary smooth muscle cells, terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling (TUNEL) assay was performed by CF™640R TUNEL Assay Apoptosis Detection Kit (Biotium, Hayward, CA, USA), following the manufacturer’s instructions. α-SMA was detected by immunofluorescent staining using an antibody against α-SMA (Santa Cruz) and Alexa Fluor 488-conjugated goat anti-mouse secondary IgG (Abcam, Cambridge, United Kingdom). The samples were mounted with ProLong Gold Antifade Reagent with DAPI (Thermo Fisher Scientific, Waltham, MA, USA). Fluorescence was observed with an immunofluorescence microscope (BZ-X700, KEYENCE Co., Osaka, Japan). We randomly selected at least 10 fields in each specimen and counted the nuclei in the medial smooth muscle layer. The results were expressed as percentages of the number of TUNEL-positive nuclei over the total number of nuclei [14].

In experiment 1, we confirmed the effect of hypoxia exposure on RVSP, RVH, and medial thickening at 12 weeks. Both RVSP and Fulton index are elevated in response to hypoxia (Fig. 2A, B). The medial smooth muscle layer of the hypoxia-exposed mice was significantly greater than that of the normoxia-exposed mice (Fig. 2C). Western blot demonstrated that PPAR-γ expression in the normoxia and hypoxia group is comparable, whereas the lung tissue apelin of hypoxia-exposed mice is higher than that of normoxia-exposed mice (Fig. 1D, E). The active caspase-3 in the whole lung tissue of hypoxia-exposed mice is increased. TUNEL staining showed that smooth muscle cell apoptosis is decreased in hypoxia-exposed mice (Fig. 1F, G).

Figure 3A shows that the body weight in the HFD group is significantly heavier than that in the ND group at 24 weeks in Experiment 2. Glucose, LDL-C, and HDL-C, but not TG, in the HFD group are significantly higher than those in the ND group (Fig. 3B). There is a significant difference in the RVSP of the ND and HFD groups at 24 weeks; that in the ND group almost recovered to normal level, whereas that in the HFD group did not (Fig. 3C). The RV weight and Fulton index in the HFD group are larger than those in the ND group, whereas the LV weight is comparable between the groups (Fig. 3D). Figure 3E shows the mitochondrial ATP levels in the lung tissues. The mitochondrial ATP in the HFD group is higher than that in the ND group. However, mitochondrial DNA and amounts of mitochondrial mass reflected by VDAC expression were equivalent between the groups (Fig. 3E).

In the ND group, the hypoxia-induced medial wall thickening of the pulmonary arteries almost normalized at 24 weeks, whereas it remained significantly high in the HFD group (Fig. 4A, left and middle panels). The proportion of fully muscular vessels in the HFD group is significantly higher than that in the ND group. The percentages of partially muscular vessels are nearly equivalent between the two groups (Fig. 4A, right panel). Western blotting demonstrated that caspase-3 activity is decreased in the lung tissue of the HFD group (Fig. 4B). TUNEL staining shows that pulmonary artery smooth muscle cell apoptosis in the HFD group is lower than that in the ND group (Fig. 4C). Moreover, the levels of PPAR-γ and apelin are lower in the lung tissue of the HFD group (Fig. 4D, E). There are no significant differences in the levels of HO-1 between the ND group and HFD group (Fig. 4F). HIF-1α is increased in the HFD group, but it is not statistically significant (Fig. 4G).