Hypoxia-Inducible Factor Prolyl-Hydroxylase Inhibitor Roxadustat (FG-4592) alleviates sepsis-induced acute lung injury
Fu Han，Gaofeng Wu, Shichao Han, Zhenzhen Lia, Yanhui Jiaa, Lu Baia, Xiaoqiang Lia, Kejia Wanga, Fangfang Yanga, Jian Zhanga, Xujie Wanga, Hao Guana, Linlin Sua, Juntao Han & Dahai Hu
A Department of Burns and Cutaneous Surgery, Xijing Hospital, Air Force Medical University
Abstract
Acute lung injury (ALI) is one of the most severe outcomes of sepsis which still waiting for effective treatment method. Roxadustat (FG-4592) which is often used for treatment of anemia in patients with chronic kidney disease (CKD), its affection on LPS-induced ALI haven’t been evaluated. MH-S and MLE-12 cell injury and ALI mouse model was induced LPS. Several assays were used to explore the role of FG-4592 in reducing the damage caused by LPS. FG-4592 treatment significantly upregulated HIF-1α and HO-1 and strikingly attenuated inflammation in vivo and in vitro. Furthermore, septic mice overexpressing HIF-1α had high level of survival rate and lower expression of inflammatory factors while down-regulation can enhance the damage of LPS. HIF-1α has a protective effect on acute lung injury in LPS induced septic mice. FG-4592 treatment remarkably ameliorated the LPS-induced lung injury through the stabilization of HIF-1α. Besides the role in treating CKD anemia, the clinical use of FG-4592 also might be extended to ALI.
Introduction
Sepsis is a life-threatening organ dysfunction circumstance resulting from dysregulated host responses to infection, that can lead to multiple organ failure, acidosis, and death (Cecconi et al., 2018) . Although the surgical and pharmacological treatment methods of sepsis are continually improving with the help of increasing studies on its pathogenesis, the mortality of sepsis patients is still very high (Duran-Bedolla et al., 2014). Therefore, it is urgent to develop effective treatments for sepsis.
Sepsis or sepsis-associated multiple organ dysfunction syndrome is the worldwide foremost reason of death in intensive care unit (ICU) by mainly attacking the lung, kidney, and cardiovascular system (Ferreira and Sakr, 2011). Lung is the organ that collapse first when respond to sepsis (Schultz and van der Poll, 2002). Meanwhile, ALI as an inflammation syndrome of the respiratory system, is one of the most severe outcomes of sepsis and responsible for sudden deaths of patients in the ICU (Rubenfeld et al., 2005; Sevransky et al., 2004). The nosogenesis of ALI have been studied previously (Kissoon et al., 2016; Liu et al., 2018). However, there is still a lack of effective drugs for ALI therapy.
Endotoxin is the main component of the outer membrane of gram-negative bacteria and known as a stable complex containing lipopolysaccharide (LPS). The initial pathogenesis of ALI induced by LPS is the production of reactive oxygen species, which makes leukocytes infiltrate into the lung. Plasma proteins leakage with leukocytes infiltration results in hypoventilation and hypoxia. Metabolism pathways related to oxygen homeostasis activate and genes involved up-regulate such as HIFs. HIFs are heterodimeric basic helix-loop-helix transcription factors which were composed of two dimeric subunits (Keith et al., 2011). Among them, HIF-1α is a transcription factor that causes cytoprotection and metabolic changes in response to hypoxia and inflammation (Kaelin and Ratcliffe, 2008). Previous studies have shown that HIF-1α is very important for innate and adaptive immunity (Knowles et al., 2006). In addition, overexpression of HIF-1α have anti-inflammatory effects (Westra et al., 2010). However, HIF prolyl hydroxylase domain proteins (HIF PHDs) can rapidly reduce the HIF activity under normoxic conditions. Thus, the suppression of HIF PHDs activity appears to have considerable clinical perspectives. Researchers investigated the effect of dimethyloxalylglycine (DMOG), a HIF PHDs inhibitor, on ALI while it wasn’t ‘clinical’ PHD inhibitor (Nagamine et al., 2016; Tojo et al., 2018). Therefore, and it is well worthy to give a shot on effect evaluation of the ‘clinical’ HIF PHDs inhibitors for ALI.
Roxadustat (FG-4592) is a transient small molecule HIF PHDs inhibitor (Del Vecchio and Locatelli, 2018). The potency of FG-4592 has been proved by lots of studies. For instance, it can increase HIF activity which is sufficient to stimulate effective erythropoiesis, enhance the endogenous EPO levels near the physiological range, protect the severe premature infant against retinal oxygen toxicity (Hoppe et al., 2016) and protect kidney from the cisplatin induced nephropathy (Yang et al., 2018). Moreover, FG-4592 have the higher potency than DMOG (Singh et al., 2020). However, whether FG-4592 treatment could against acute lung injury is still unknown.
Thus, the role of FG-4592 in LPS induced ALI and its potential mechanisms were examined in this study. We investigated the effects of FG-4592 on mouse alveolar macrophage cell line (MH-S) and mouse lung epithelia type II cell line (MLE-12) in vitro, and an LPS induced sepsis mouse model in vivo. Moreover, the possible mechanisms involved in the protective effects of FG-4592 against LPS-induced inflammation were also been explored.
Material and Methods
1. Cell culture
MH-S and MLE-12 were purchased from BeNa Culture Collection. MH-S cells were grown in 1640 medium and MLE-12 cells were cultured in DMEM F12 medium with 10% FBS supplemented with 100 U/ml penicillin and 100 mg/ml streptomycin at 37℃ in a humidified atmosphere of 5% CO2 (26.4% O2).
2. Cell Counting Kit-8 (CCK-8) assay
Cell viability was determined by CCK-8 assay kit. MH-S and MLE-12 cells were treated with FG-4592 (5μM to 100μM) for 24 hours, then 10μL CCK-8 reagent was added to medium and incubated for 1 hour. OD450 nm was read with a microplate reader.
3. Lentivirus and transfection
Lentivirus overexpressing (HIF-1α and HIF-1α EV) and inhibiting (HIF-1α inhibitor and HIF-1α inhibitor EV) were purchased from Hanbio Biotechnology (China). MH-S and MLE-12 at 50-70% confluence was used for transfection. MH-S and MLE-12 were infected with lentiviral vector at a MOI of 20 in the presence of 5μg/ml polybrene according to the manufacturer’s instruction. After 48h infection, cells were stimulated with LPS (100μg/ml) and/or FG-4592 for 12hours and then cells were collected for analysis.
4. Quantitative real-time PCR (qRT-PCR)
Total RNA from cells and lung tissues were extracted using TRIzol. Expression of different genes was analyzed using SYBR® PremixEx Taq™ II through Bio-Rad CFX System. The reaction procedure is 95°C 10 min, followed by 40 cycles of 95°C 15 s, 60°C 30 s, and 72°C 30 s. Data was normalized to the expression of housekeeping gene using 2-ΔΔCt method. Primers are shown in Table 1.
5. Western blotting analysis
Cells and lung tissues were lysed in RIPA lysis buffer. 30μg total proteins were used for Western blotting analysis. Antibodies specific against P-P65 (1:800), P65 (1:1000), P-IKBα (1:800), IKBα (1:1000), HO-1(1:1000), HIF-1α (1:1000)，FIH(1:1000) and β-actin (1:2000) were obtained from CST.
6. IL-1β and TNF-α level in bronchoalveolar lavage fluid and serum
BALF was collected by irrigating left lungs twice with 2 mL saline at the end of experiments. Blood was collected from the eyes and centrifuged (3000 rpm, 10 min) to get the serum. The expression levels of IL-1β and TNF-α in BALF and serum were determined by ELISA kit according to the manufacturer’s instructions.
7. ROS level in cell culture supernatant
MH-S and MLE-12 were treated with 25μg/ml LPS with or without 5μM FG-4592 for 12hour. Supernatant was collected to detect the ROS level with ELISA kits (R&D Systems, Inc.,Minneapolis, MN) according to the manufacturer’s instructions. OD450nm was measured by ELX808 microplate reader (BioTek, USA).
8. Animals and experimental protocols
Wild-type male 6-8 weeks old BABLC mice were obtained from the Experimental Animal Center of Air Force Medical University and housed under an air-conditioned room (22 ± 2 °C). Animal experiments were approved by Experimental Animal Committee of Air Force Medical University. All procedures were approved by the Institutional Animal Care and Use Committee of the Air Force Medical University in accordance with the Guidelines for the Care and Use of Laboratory Animals. The sepsis mouse model was based on the method reported by previous researchers (Napier et al., 2019). To evaluate the effect of FG-4592 on LPS-induced sepsis lung injury, 32 mice were assigned to 4 groups: control group (control), LPS-induced sepsis lung injury group (LPS), LPS-induced sepsis lung injury plus FG-4592 treatment group (LPS+FG-4592), FG-4592 treatment group (FG-4592). Another 32 mice were assigned to 4 groups: control, LPS, FG-4952 administration after LPS-induced sepsis lung injury one day group (LPS+FG-4592), FG-4592. Mice were intraperitoneally injected with 3 mg/kg LPS (Sigma-Aldrich, #L2880) to establish a sepsis model. The control mice received an i.p. injection of saline. FG-4592 was dissolved in DMSO at 50 mg/mL, then further diluted in sterile PBS to 1 mg/mL, and stored at -80 °C. Mice were treated with FG-4592 for 48 h in LPS+FG-4592 and FG-4592 group at a dose of 10 mg/kg/d via i.p. injection after LPS treatment immediately(Yang et al., 2018). Mice were sacrificed after LPS administration of 48 h. Lung tissues collected for histology were fixed in 4% paraformaldehyde (PFA). Remaining lung tissue was stored at -80°C for protein analysis. Survival rate was monitored every 12h for 3 days. Serum was collected to detect the ROS level with ELISA kits (R&D Systems, Inc.,Minneapolis, MN) according to the manufacturer’s instructions. OD450nm was measured by ELX808 microplate reader (BioTek, USA).
In order to investigate the role of HIF-1α in the protective effects of FG-4592 against LPS-induced inflammation, 96 mice were randomized into four groups which was injected via tail veins with 3×107 TU HIF-1α, HIF-1α EV, HIF-1α inhibitor, HIF-1α inhibitor EV lentiviruses respectively. After three days, mice in each group were randomized into four groups: Control group mice injected with saline; FG-4592 group: mice injected with FG-4592; LPS group: mice injected with LPS; LPS+FG-4592 group: mice injected with LPS and FG-4592. Mice were sacrificed after LPS administration of 48 h in each groups，serum and BALF were collected for cytokine assays and lung tissues were collected for histology or molecular analysis.
9. Haematoxylin and eosin (H&E) staining
Lung tissues were collected and fixed in 4% paraformaldehyde and then embedded in paraffin using an automated processor. Samples were cut into 4 μm thickness sections and stained with H&E. Pathological sections were examined and photographed by Olympus IX71 light microscope (Olympus, Japan).
10. Immunohistochemistry
Lung tissues were embedded in paraffin blocks and cut into 4-μm thick tissue section and then dewaxed. 3 % hydrogen peroxide was used to quench endogenous peroxidase activity for 15 min for immunohistochemistry staining followed by blocking with normal goat serum for 30 min and incubated overnight at 4 °C with primary antibodies against HIF-1α (1:50; CST). Sections were treated with PV6000 Histostain™ kit (ZSGB, Beijing, China) and stained with diaminobenzidine (ZSGB, Beijing, China). Finally, sections were counterstained with hematoxylin. An isotype-matched IgG served as the negative control for each immunostaining procedure. Sections were analyzed and images acquired with FSX100.
11. Statistical analysis
All values were presented as the mean value ± S.D. GraphPad Prism6 software was used for statistical analyses. Differences were considered statistically significant if P < 0.05.
Results
1. FG-4592 alleviates LPS-induced cell inflammation
Firstly, we evaluated the effect of FG-4592 on LPS-induced cell injury in vitro. Cell viability of MH-S and MLE-12 was not affected by FG-4592 treatment at the concentrations of 5 μM, 10 μM, 15 μM, and 25 μM when compared to vehicle control group, while it was depressed around 10% and 50% under 50 μM and 100 μM FG-4592, respectively (Figure 1A). These results indicated that the safe dose of FG-4592 with no noticeable cellular toxicity on MH-S and MLE-12 was within 25 μM. In addition, the effect of FG-4592 on the inflammation of LPS-treated cells was investigated. MH-S and MLE-12 cells were stimulated with LPS at different time points. Results showed that pro-inflammatory cytokines were significantly increased and reached the peak at 12 h after LPS stimulation (Supplementary Figure 1).
Therefore, the cells were finally treated by 100 µg/ml LPS for 12 h with 25 μM FG-4592. Cells only treated by LPS or FG-4592 were set as control groups. LPS-induced pro-inflammatory cytokines of IL-1β, TNF-α and IL-6 were significantly decreased by FG-4592 treatment, while no cell inflammation was presented in FG-4592 treatment control group (Figure 1B).
2. HIF-1α is involved with the FG-4592 effect on LPS-induced cell inflammation
To understand the potential mechanisms involved in FG-4592 against LPS-induced cell inflammation, the effects of FG-4592 on the level of HIF-α in MLE-12 and MH-S cells were evaluated. Results indicated that FG-4592 treatment significantly upregulated the expression of HIF-1α in LPS-treated cells compared with LPS treatment cells, while no changes found in non-LPS group (Figure 2A). To further examine whether FG-4592 reduced LPS-induced cell inflammation through HIF-1α activation, MLE-12 and MH-S cells were transfected with HIF-1α, HIF-1α EV, HIF-1α inhibitor and HIF-1α inhibitor EV lentivirus. Results demonstrated that overexpression of HIF-1α reduced the response of inflammatory factors stimulated by LPS (Figure 2B). Moreover, as shown in Figure 2C, knockout of HIF-1α dramatically enhanced the expression of pro-inflammatory cytokines levels in MH-S and MLE-12 after LPS stimulation. Meanwhile, the depletion of HIF-1α reduced the anti-inflammation of FG-4592 in LPS-induced cell injury. Hence, HIF-1α should be a key gene depressed the LPS-induced inflammatory response which functioned during the treatment of FG-4592.
3. FG-4592 alleviates LPS-induced acute lung injury in mice
LPS-induced mouse model was used to verify the effect of FG-4592 in vivo. The survival rate was higher in the FG-4592 and LPS treatment group than LPS treatment group (Figure 3A). As shown in Figure 3B, inflammatory factors were significantly increased in lung tissues from LPS-induced mice, which were outstandingly obstructed by FG-4592 (Figure 3B). Moreover, no inflammation symptom was presented in mice only oral administrated by FG-4592. Correspondingly, levels of IL-1β and TNF-α in serum and BALF were both significantly decreased in mice from FG-4592 and LPS group than LPS group (Figure 3C & 3D). Meanwhile, no obvious side effects of FG-4592 treatment on inflammatory cytokines were found. H&E staining showed that lung congestion, hemorrhage, and edema lightened significantly declined in FG-4592 and LPS treatment group compared with LPS treatment group (Figure 3E), which indicated that FG-4592 can alleviate the LPS-induced lung injury through reducing inflammatory cell infiltration, relieving edema and hemorrhage and subduing necrosis. We also investigate the effect of FG-4952 administration one day after LPS instillation. Results shown that inflammatory factors were outstandingly obstructed by FG-4592 (Supplementary Figure 2A). Correspondingly, levels of IL-1β and TNF-α in serum and BALF were both significantly decreased in mice from FG-4592 and LPS group. (Supplementary Figure 2B and 2C).
4. HIF-1α, HO-1 and NF-κB is involved in the effect of FG-4592 in vivo and in vitro The protein expression level of HIF-1α from lung tissues were detected. As a result, up-regulation of HIF-α were detected in protein levels after FG-4592 and LPS treatment (Figure 4A and B), while no changed were found in control group only treated by FG-4592. The expression level of HO-1 was measured to investigate whether the protective effect of FG-4592 on LPS-induced inflammation was related to the downstream genes of HIF-1α. We also detected NF-κB inflammation pathway related proteins. LPS treatment significantly increased the expression of p-P65 and p-IKBα compared to control group which was remarkably decreased by FG-4592 (Figure 4C). Additionally, the expression level of HO-1 increased in mice after LPS injection compared to control group mice, while HO-1 expression level increased more in LPS and FG-4592 treatment group (Figure 4D). The effect of LPS and FG-4592 on HO-1 and NF-κB also has been investigated in vitro. LPS treatment significantly increased the expression of p-P65 and p-IKBα in MH-S and MLE-12 cells compared to control group which was notably reduced by FG-4592 (Figure 4E). The level of HO-1 increased after FG-4592 treatment compared to the LPS-treated group both in MH-S and MLE-12 cells (Figure 4F). The expression of HO-1 was increased in the cells treated only by LPS which may due to the host defense mechanism, while the expression remained stable after treatment only by FG-4592. The expression of HIF asparaginyl-hydroxylase wasn’t affected by FG-4592(Supplementary Figure3).
5. HIF-1α is involved with the FG-4592 effect on LPS-induced ALI
To explore whether FG-4592 exerted a protective effect by improving HIF-1α activity, mice was injected via tail veins with 3×108 TU HIF-1α inhibitor, HIF-1α inhibitor-EV. The expression of HIF-1α was down-regulated both in mRNA and protein level (Supplementary Figure 2A). All mice were died within 48 hours after injecting with LPS. However, the survival rate was higher in HIF-1α inhibitor + LPS + FG-4592 group than HIF-1α inhibitor + LPS group while is still lower than LPS + FG-4592 group (Supplementary Figure 4B). The expression level of inflammatory cytokines and levels of TNF-α and IL-1β in BALF in lung tissues from HIF-1α inhibitor + LPS group was higher than HIF-1α inhibitor EV+ LPS group while FG-4592 can reverse both LPS and HIF-1α inhibitor mediated inflammatory response (Supplementary Figure 4C). Moreover, down-regulation of HIF-1α enhanced the levels of IL-1β and TNF-α in serum and BALF in HIF-1α inhibitor + LPS group compared to HIF-1α inhibitor EV+LPS group. FG-4592 can reduce both IL-1β and TNF-α level caused by LPS and HIF-1α down-regulation (Supplementary Figure 4D and 4E). Microscopically, H&E staining revealed that down-regulation of HIF-1α enhanced LPS-induced lung injury while FG-4592 reduced the injury (Supplementary Figure 4F). To confirm the function of HIF-1α in reducing inflammation, mice was injected via tail veins with 3×107 TU HIF-1α, HIF-1α-EV, and the expression of HIF-1α was up-regulated both in mRNA and protein level after injection (Supplementary Figure 5A). The overexpression of HIF-1α could increase the survival rate of mice from 20% in LPS group to 30% in LPS+HIF-1α group. Mice from HIF-1α+ LPS+ FG-4592 group survived more than LPS+FG-4592 group (Supplementary Figure 5B). The expression level of inflammatory cytokines in lung tissues from HIF-1α+LPS group was lower than HIF-1α-EV+LPS group (Supplementary Figure 5C). Moreover, up-regulation of HIF-1α reduced the levels of IL-1β and TNF-α in serum and BALF in HIF-1α inhibitor + LPS group compared to HIF-1α inhibitor EV+LPS group. FG-4592 could reduce both IL-1β and TNF-α level caused by LPS and HIF-1α down-regulation (Supplementary Figure 5D and 5E). H&E staining revealed that up-regulation of HIF-1α reduced LPS-induced lung injury (Supplementary Figure 5F).
6. FG-4592 decreased the ROS levels in vitro and in vivo
The expression level of ROS from serum and cell culture supernatant were detected. The ROS levels in serum and cell culture supernatant were significantly increased after LPS treatment while it was decreased by FG-4592 treatment (Supplementary Figure6).
Discussion
Increasing evidence suggest that lung is the first organ to fail in response to sepsis, and the major cause of death for sepsis patients is respiratory failure. LPS-induced leukocytes infiltration results in generation ROS, hypoventilation and hypoxia, and thus plays important roles in ALI (Rojas et al., 2005). HIF has been identified as an important mechanism of cellular adaptation to low oxygen and ROS condition (Kenneth and Rocha, 2008). However, HIF PHDs can rapidly reduce the HIF activity reduced under normoxic conditions. Hence, it’s important to enhance the expression of HIF-1α under ALI condition. Previous studies have investigated the effect of DMOG on ALI while it wasn’t ‘clinical’ PHD inhibitor. FG-4592 is a small-molecule HIF PHD inhibitor. In the present study, we observed that FG-4592 upregulated HIF-1α expression in lung and improved LPS-induced lung injury and inflammation which provides a theoretical basis for the clinical application of FG-4592 in the treatment of ALI.
FG-4592 has been used for the treatment of CKD-associated anemia. We found that FG-4592 reduced inflammation in a septic mouse model. Previous studies showed that overexpression of HIF-1α had an anti-inflammatory effect on the lung. Our results showed that LPS and FG-4592 could induce HIF-1α expression. We found that LPS alone treatment increased the expression of HIF-1α which was consistent with previous studies (Hu et al., 2015). Furthermore, overexpression of HIF-1α decreased LPS-induced inflammation induced in MH-S and MLE-12 cell model, while down-regulation of HIF-1α increased the inflammation of MH-S and MLE-12 cells. It indicated that the way FG-4592 reduced inflammation of ALI was by increasing HIF-1α expression. The reason that HIF-1α was enhanced by LPS was that ROS were generated after cell or the lung was stimulated with LPS, which increases the expression of HIF-1α dependent on ROS for immune reaction (Koshikawa et al., 2009; Zmijewski et al., 2008). HIF-1α is reported to exert protective roles through regulating downstream genes (Chin et al., 2007). HO-1 is one of downstream genes of HIF-1α which can block the release of TNF-α in LPS-stimulated macrophages (Li et al., 2009). HO-1, as an isoenzyme of heme oxidation rate limiting enzyme, plays an anti-inflammatory role in LPS induced ALI(Athale et al., 2012). Moreover, up-regulation of HO-1 can inhibit the release of LPS-induced pro-inflammatory cytokines in RAW264.7 cells. In our study, we found that FG-4592 enhanced the expression of HO-1. Thus, we could reasonably speculate that the beneficial effect of FG-4592 in treating LPS-induced ALI is through promoting expression of HO-1.
On the other hand, evidence have demonstrated that accumulation and activation of HIF-1α are affected by NF-κB pathway under inflammatory condition (Zepeda et al., 2013). Hence, we speculate that FG-4592 could affect the expression of molecules in NF-κB pathway. Our results showed that the up-regulated phosphorylated P65 and IkBα levels by LPS in MH-S and MLE-12 were decreased by FG-4592. The increase of P65 and p-IKBα in LPS-treated group was prevented by FG-4592 in ALI mouse model. In addition, NF-κB activation is known to play a key role in the production of TNF-α and IL-1β in ALI (Kuo et al., 2011; Li et al., 2012). Our result showed that FG-4592 inhibited LPS-induced production of TNF-α and IL-1β both in vivo and in vitro. These findings indicted that the inflammatory response to LPS-induced ALI were inhibited by FG-4592 via NF-κB pathway. But how FG-4592 affects the expression of NF-κB is still unknown. Previous studies shown that activation of the Nrf-2/HO-1 pathway inactivated NF-κB which reduced airway inflammation(Shin et al., 2013). Moreover, HO-1 can reverse the effect of IL-1β, which can activate the NF-κB, on human chondrocyte metabolism(Zhu et al., 2018). Previous studies also shown that HO-1 suppresses apoptosis through NF-κB pathway(Guillen et al., 2008). All data illustrated that FG-4592 treatment decreased LPS-induced ALI and cell injury by targeting NF-κB, HIF-1α and HO-1.
Conclusions
HIF-1α has a protective effect on acute lung injury in LPS induced septic mice. It’s important to enhance the expression of HIF-1α under ALI condition. FG-4592 had anti-inflammatory effects on the LPS-induced MH-S and MLE-12 cell injury model. FG-4592 has therapeutic effects on LPS-induced ALI mouse model. We firstly found that FG-4592 could reduce cell injury and ALI. We also demonstrated the potent effect of HIF stabilizer on protecting against ALI. Our findings reveal a clinical potential of this anemia-treating drug in treating ALI. Funding