Authors: Thu T. Tran (1Department of Pediatrics, McGovern Medical School, University of Texas Health Science Center at Houston, Houston, TX, USA), Holger K. Eltzschig (2Department of Anesthesiology, Critical Care and Pain Medicine, McGovern Medical School, University of Texas Health Science Center at Houston, Houston, TX, USA), Xiaoyi Yuan (2Department of Anesthesiology, Critical Care and Pain Medicine, McGovern Medical School, University of Texas Health Science Center at Houston, Houston, TX, USA)
Categories: Article, ARDS, HIF, hypoxia signalling, target genes
Source: The Journal of physiology
Doi: 10.1113/JP284599
Authors: Thu T. Tran, Holger K. Eltzschig, Xiaoyi Yuan
Acute respiratory distress syndrome (ARDS) is characterized by bilateral chest infiltration and acute hypoxic respiratory failure. ARDS carries significant morbidity and mortality despite advancements in medical management, calling for the development of novel therapeutic targets. Hypoxia-inducible factor (HIF) is a heterodimeric protein involved in various essential pathways, including metabolic reprogramming, immune modulation, angiogenesis and cell cycle regulation. HIF is routinely degraded in homeostasis conditions via the prolyl hydroxylase domain/von Hippel–Lindau protein pathway. However, HIF is stabilized in ARDS via various mechanisms (oxygen-dependent and independent) as an endogenous protective pathway and plays multifaceted roles in different cell populations. This review focuses on the functional role of HIF and its target genes during ARDS, as well as how HIF has evolved as a therapeutic target in current medical management.
Acute respiratory distress syndrome (ARDS) or acute lung injury (ALI) is composed of multiple aetiologies leading to significant morbidity and mortality even to this day. This syndrome was first described by (Ashbaugh et al. (1967) when he observed that 12 adult patients in the intensive care unit (ICU) were found to have an acute onset of severe hypoxia with minimal response to oxygen and respiratory support, tachypnea, poor lung compliance, and diffuse pulmonary infiltrate on chest radiograph occurring within 1 h to 96 h after toxic stimuli. Of the patients from whom lung histology could be obtained, alveolar atelectasis with intra-alveolar haemorrhage, oedema and gross infiltration of macrophages was observed (Ashbaugh et al., 1967).
Currently, ARDS remains a significant public health and economic burden despite rapid advancement in medical care. A large prospective, multicentre, observational study called LUNG SAFE by Bellani et al. (2016), with 29,144 patients across 459 ICUs from 50 countries across the world, found the prevalence of ARDS admissions to ICUs to be 10.4% with 23.4% of these patients requiring mechanical ventilation. Even with these staggering numbers, ARDS was still considered to be grossly underdiagnosed, with only 60.2% of the patients being identified by clinicians based on the Berlin definition of ARDS, 66% often receiving a delayed diagnosis and 40% of all cases being unrecognized. The mortality risk found in the study was a solemn 40% (Bellani et al., 2016). In 1-year and 5-year longitudinal cohort multicentre studies, 40% of patients reported readmission by 1 year and 83% of patients had another hospitalization by the 5-year follow-up. Median cost per patient at the 1 year follow-up was found to be 58,500; a significant increase considered to account for other long-term services and subspecialty care, including mental health (Ruhl et al., 2017a; Ruhl et al., 2017b). In comparison with the exorbitant healthcare costs that patients with ARDS have to sustain, the average cost of admission per person in the USA was estimated to be $71,004 in 2016 (Siuba et al., 2022).
The definition of ARDS has been critically revised several times over the years (Williams et al., 2021). The most recent update from the ARDS Definition Task Force has defined ARDS as an acute respiratory decline occurring within 7 days of insult, leading to diffuse inflammatory lung injury, pulmonary oedema, atelectasis and increased physiological dead space. Clinical and radiographical hallmarks include bilateral chest opacities, non-cardiogenic hypoxic respiratory failure and decreased lung compliance from volume loss. The severity of ARDS can be further categorized into mild, moderate and severe depending on the positive end-expiratory pressure required, as well as the ratio of PaO2/FIO2 to meet the patient’s oxygenation demands. Hypoxia-inducible transcription factor (HIF) was first discovered by Dr Semenza in (1991) when his group found, under hypoxic stress, ‘binding of nuclear factors to specific DNA sequences at the 3’ of human erythropoietin (EPO) gene results in transcriptional activation (McGettrick & O’Neill, 2020). HIF is stabilized during ARDS via several different mechanisms and provides lung protection (Bowser et al., 2017). This review aims to discuss the functional role of HIF in ARDS, as well as current therapeutic strategies to target hypoxia signalling.
HIF is a heterodimeric protein composed of two polypeptide chains, α and β, where the α subunit is readily degraded in the presence of oxygen, whereas the β subunit remains constitutively expressed under aerobic conditions (Yuan et al., 2018). HIF-1 is highly conserved and widely expressed as a transcription factor essential for the expression regulation of hundreds of target genes, whereas HIF-2 has a more selective expression. During normoxic conditions, HIF-1α and HIF-2α hydroxylation occurs via oxygen-dependent prolyl hydroxylase domain (PHD), allowing HIF to be recognized by von Hippel–Lindau protein and undergo proteasome degradation. Hypoxia, a state of low oxygen availability, is commonly observed during tissue ischaemia and inflammation(Eltzschig, 2022). With hypoxic or inflammatory insults, PHDs cannot hydroxylate HIF and therefore HIFα escapes degradation as it binds to HIF-1β and transcription coactivators CREP-binding protein and 300 kDa coactivator (p300) within the nucleus to increase target gene expression (Fig. 1). Binding of HIF to hypoxia-responsive elements (HREs) at cis-regulatory regions, which act as promoter and enhancer regions, induces transcription of hypoxia-responsive genes (Koeppen et al., 2018; Ruan et al., 2022).
PHDs are not only controlled by oxygen sensitivity, but also by other factors, such as metal chelators to displace iron from directly binding to PHDs, and α-ketoglutarate memetics to act as enzymatic inhibitors (Kaelin & Ratcliffe, 2008). In particular, succinate accumulation through the inactivation of succinate dehydrogenase has been found to inhibit 2-oxoglutarate-dependent PHDs, to stabilize alveolar epithelial HIF-1α (Kaelin & Ratcliffe, 2008). Hypoxia has also been described in the literature to activate nuclear factor-kappa B (NF-κβ), which can directly induce HIF mRNA expression. This phenomenon is seen in hypoxic conditions, polymicrobial infection and immune conditions with the upregulation of inflammatory cytokines and stress response (Rius et al., 2008). For example, Toll-like receptors on immune cells detect pathogen-associated molecular patterns exhibited by microbial pathogens, which trigger the activation of NF-κβ and enhance HIF activity (Fitzgerald & Kagan, 2020). Moreover, reactive oxygen species (ROS) and nitric oxide have also been found to mediate HIF-1 transcription activity and stabilization directly (Kaelin & Ratcliffe, 2008).
HIF affects many target genes to regulate erythropoiesis, neovascularization, antioxidation, wound healing, cellular metabolism and cell cycle arrest (Lee et al., 2020b) (Fig. 2). EPO is one of the most well-known genes targeted by HIF to stimulate red blood production by the bone marrow and it has multiple receptors across different organs (Shu et al., 2019). Vascular endothelial growth factor (VEGF), nitric oxide synthase, heme oxygenase-1, adrenomedullin, fibroblast growth factor 2 (FGF-2), angiopoietin-1 and 2, placental growth factor, stem cell facto, and stromal-derived factor-1 (SDF-1) are also highly upregulated by HIF-1α in response to hypoxia for neovascularization and vascular endothelial cell proliferation (Mammadzada et al., 2020). Transforming growth factor β (TGF-β) consists of a variety of proteins important to cellular proliferation, regulation of immune cell maturation and tumorigenesis (Shi et al., 2022). HIF-1α-dependent TGF-β expression in late cancer stages has been found to promote tumor growth, as well as loss of TGF-β tumor suppression ability, whereas the reverse is true in earlier tumor growth, possibly from its effects on glucose metabolism (Huang et al., 2021). During hypoxia, HIF-1α limits tricarboxylic acid cycle activity by inhibiting pyruvate dehydrogenase phosphorylation and converting from oxidative to glycolytic metabolism by increasing transcription of glucose transporter (GLUT)-1 and 3, responsible for glucose uptake for cellular metabolism, as well as increased expression of other glycolytic enzymes such as aldolase and lactate dehydrogenase (Taylor & Scholz, 2022). In addition to VEGF, EPO, GLUT1 and TGF-β1 being prominent genes in tissue repair, other genes include plasminogen activator-1, connective tissue growth factor and platelet-derived growth factor are all involved in cellular adaptation in hypoxic tissue injury and promote cellular proliferation, migration and repair for increased overall cellular survival (Rankin & Giaccia, 2008). Notably, around two-thirds of HIF-dependent genes are downregulated, rationalizing the possibility of HIF-dependent microRNAs as a mechanism of gene regulation (Ju et al., 2021; Lee et al., 2020a; Neudecker et al., 2016). For example, HIF-1a-dependent repression of equilibrative nucleoside transporter (ENT) is important in maintaining vascular barrier function and the repression is probably mediated by ENT targeting microRNAs.
As alluded to earlier, HIF is stabilized in ARDS through different mechanisms, including inflammation, infection, alveolar and endothelial injury, as well as immune activation. Here, we will further discuss the functional role of HIF in different cellular responses during ARDS.
The alveoli are composed of type I squamous cells, which account for 80% of the alveolar surface areas and are critical in providing an epithelial barrier, and type II cuboidal cells, which are primarily responsible for surfactant production and can proliferate and differentiate into type I cells (Manicone, 2009). Pulmonary oedema, neutrophilic infiltration and epithelial cell apoptosis can impair the alveolar barrier function and alveolar inflammation (Manicone, 2009). Disruption of the alveolar epithelium is an integral complication in ARDS and both in vivo and in vitro models have demonstrated attenuated lung inflammation and improved pulmonary repair governed by HIF stabilization. For example, the stabilization of HIF-1α in type II alveolar epithelial cells after pulmonary injury promotes the proliferation and motility of type II cells during the repair phase through keratinocyte migration and upregulation of VEGF, SDF-1 and chemokine receptor CXCR4 upregulation (McClendon et al., 2017). Stabilization of HIF-1α in type II alveolar epithelial cells during ventilator-induced ALI improves tricarboxylic acid cycle function, enhances mitochondrial stability, increases ATP production, prevents ROS accumulation, and subsequently reduces lung inflammation (Eckle et al., 2013a). Furthermore, HIF-1 binds to HRE sequences located in the 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 (PFKFB3) gene promoter during stretch injury to enhance glycolytic metabolism in the alveolar epithelium (Vohwinkel et al., 2022). Pharmacologic inhibition of PFKFB3 by 3-(3-pyridinyl)−1-(4-pyridinyl)−2-propen-1-one resulted in increased levels of pro-inflammatory cytokines IL-6 and CXCL1 in bronchoalveolar lavage fluid in murine models of ARDS. Mice with ATII cell-specific deletion of PFKFB3 failed to increase carbohydrate glycolysis, and displayed worse alveolar epithelial permeability, alveolar haemorrhage and inflammation compared to controls (Vohwinkel et al., 2022).
HIFs also attenuate alveolar inflammation by promoting extracellular adenosine signalling. Adenosine is a purine nucleoside generated from ATP via CD39 and CD73, accumulating during hypoxic and inflammatory conditions (Koeppen et al., 2011; Le et al., 2019). Adenosine A2B receptor (ADORA2B) has been identified as a HIF-1α target during hypoxia (Aherne et al., 2015; Eltzschig et al., 2003). Induction of ADORA2B leads to activation of adenylyl cyclase with resulting cAMP accumulation. Intracellular cAMP regulates epithelial sodium channels responsible for alveolar fluid clearance through sodium transport across the alveolar epithelium, which is upregulated with increased ADORA2B signalling and attenuates pulmonary oedema (Wang et al., 2020). Deletion of ADORA2B gene in the alveolar epithelial cells was associated with increased pulmonary inflammation, increased albumin leakage in bronchoalveolar lavage fluid, reduced alveolar fluid clearance and diminished gas exchange (Eckle et al., 2008a; Li et al., 2020), whereas myeloid and endothelial specific deletion of ADORA2B did not result in a significant change the mice ALI phenotype (Hoegl et al., 2015).
Endothelial dysfunction is commonly observed in ARDS and HIFs play an essential role in modulating this process. Because HIF-2α is highly expressed in endothelial cells, it is not surprising that it plays a crucial role in endothelial function during ARDS. HIF-2α induces its target gene, vascular endothelial protein tyrosine phosphatase, to promote adherens junction integrity and enhance endothelial barrier function in vitro (Gong et al., 2015). HIF-2α depletion in endothelial cells results in defective lung endothelial junctions by disrupting vascular endothelial cell cadherin stability during endotoxin-induced lung injury in mice (Gong et al., 2015). Along the same line, mice treated with PHD2 inhibitor FG4497 stabilize HIF-2α and maintain vascular endothelial barrier function (Gong et al., 2015). In addition to HIF-2α, HIF-1α also plays a role in endothelial cells during lung injury. For example, a rapid induction of HIF-1α-dependent forkhead box protein M1 expression during polymicrobial-induced endotoxaemia leads to endothelial cell proliferation, endothelial regeneration and vascular repair (Huang et al., 2019). HIF-1α-dependent induction of ADORA2B decreases vascular barrier leak and ischaemic tissue injury in hypoxic settings (Bowser et al., 2017). Moreover, PHD2, in particular, plays an essential role in adherens junction regulation and its deficiency results in the stabilization of both HIF-1α and HIF-2α with concomitant increases of vascular endothelial protein tyrosine phosphatase (Fan et al., 2019).
HIF-1α has been found to be an important regulator of immune cell activity. Targeted deletion of HIF-1α in myeloid cells inhibits ATP production by granulocytes and macrophages through glycolysis, affecting their protective abilities such as adhesion, aggregation, motility and invasion, and further decreasing inflammatory responses (Cramer et al., 2003). Phagocytic abilities were also significantly affected, leading to uncontrolled bacterial overgrowth (Cramer et al., 2003). Mice with high HIF-1α activity had elevated alveolar macrophage inflammatory gene expression as a result of activation of Wnt signalling with formation of β-catenin-HIF-1α complex, whereas HIF-1α deficient alveolar macrophages had diminished inflammatory cytokine expression and suppressed chemotaxis of inflammatory immune cells, suggesting that inflammatory gene expression was dependent on HIF-1α activity (Zhu et al., 2021). Mice deficient in myeloid HIF-1α were found to be more susceptible to Group A Streptococcus (GAS) infection with the decreased intracellular killing of GAS by macrophages and, overtime had a significant increase in viable bacterial load (Peyssonnaux et al., 2005). HIFs also have a critical function in forming and releasing neutrophil extracellular traps to facilitate bacterial killing (McInturff et al., 2012). Furthermore, HIF-1α-dependent netrin-1 production in myeloid cells regulates lung inflammation in endotoxin-induced lung injury through the regulation of leukocyte chemotaxis (Berg et al., 2021). Dendritic cells can also increase expression of HIF-1α during hypoxic insult, which helps to promote dendritic cell maturation, migration, and T cell activation and differentiation (McGettrick & O’Neill, 2020). In addition to innate immunity, HIF-1α plays an important role in adaptive immunity. For example, HIF-1α deletion resulted in age-dependent defects in maturation and differentiation of B cell lymphocytes (Kojima et al., 2002). HIF-1α-dependent induction of ADORA2B has also been found to increase regulatory T cell activation to limit uncontrolled inflammation (Eckle et al., 2013b; Ehrentraut et al., 2012). Furthermore, T cell differentiation has also been found to be impaired with HIF-1α deletion. Th1, Th2 and Th17 require increased glycolysis for production, which can be hindered by a lack of HIF-1α stabilization (Palazon et al., 2014).
Compared to HIF-1α, the functional role of HIF-2α in immune cells has been less well investigated in the setting of ARDS. However, several studies implicated the critical function of HIF-2α in innate immune cells. For example, mice without HIF-2α in myeloid cells were unable to mount a systemic inflammatory response adequately and were more resistant to LPS-induced endotoxaemia because of impaired production of tumor necrosis factor-α, interferon-γ, interleukin-12 and interleukin-1β (Cramer et al., 2003; Imtiyaz et al., 2010; McGettrick & O’Neill, 2020; Palazon et al., 2014). Similarly, HIF-2α deficient mice had reduced tumor-associated macrophage infiltration and recruitment and delayed tumor proliferation and progression in the murine model of hepatocellular carcinoma (Imtiyaz et al., 2010). Finally, overexpression of HIF-2α was also found to decrease neutrophil apoptosis and delay inflammation resolution, but its phagocytic and oxidative burst function remained essentially the same (Thompson et al., 2014). Further studies are needed to elucidate the functional role of HIF-2α in immune cell activation during ARDS.
Besides the protracted pulmonary rehabilitation that burdens ARDS patients, they also suffer from persistent psychological, cognitive and physical impairment (Gorman et al., 2022). Here, we explore the long-term sequelae in ARDS survivors and potential role of HIF in these various co-morbidities.
Post-traumatic stress disorder, anxiety and depression are common long-term sequelae in survivors of ARDS (Gorman et al., 2022). Adrenal steroidogenesis is transcriptionally regulated by HIF-1α during hypoxic stress (Baddela et al., 2020). HIF-1α has been found to upregulate glucocorticoid receptor activity and activate the hypothalamic-pituitary-adrenal axis to enhance downstream transcription of HIF target genes involved in stress regulation (Burtscher et al., 2022). The usage of HIF-1α activators may ameliorate hypoxia-driven mood derangements as a result of the ability of HIF-1α to increase serotonin production under hypoxic conditions (Alam et al., 2016; Burtscher et al., 2022). Cognitive decline and neurological deficits have been documented in long-term follow-up studies for ARDS patients (Gorman et al., 2022; Herridge et al., 2011; Mikkelsen et al., 2012). HIF-1α is generally considered neuroprotective because of its role in optimizing glucose metablisim and improving vascular flow in the brain (Iyalomhe et al., 2017).
Survivors of ARDS were also found to have decreased tolerance for physical activity at 1- and 5-year long-term follow up after ICU discharge even in cases of normal pulmonary function (Herridge et al., 2003; Herridge et al., 2011). Neuromuscular dysfunction, fatigue and tissue scarring were listed as reasons for exercise limitation (Herridge et al., 2003). Muscle ischaemia, hypotonia and skeletal fibrosis are possible etiologies for persistent muscle weakness in ARDS survivors (Nguyen et al., 2021; Valle-Tenney et al., 2020). Stabilization of HIF during muscle ischaemia can lead to upregulation of pro-angiogenic genes such as VEGF and FGF for revascularization of damaged skeletal muscles, as well as myogenic genes such as myf-5 and myog to induce muscle regeneration (Nguyen et al., 2021).
Taken together, HIF can play a critical role in the long-term outcomes of ARDS survivors and pharmacologic enhancement of HIF could be considered as an adjunctive therapy to improve long-term outcomes.
As previously discussed, oxygen availability determines HIF stability and accumulation. Hyperoxia can lead to alveolar injury, interstitial fibrosis, pulmonary oedema, systemic inflammation, damaged end organs, oxygen shunting with vascular maldistribution and increased mortality (Liang et al., 2023). On the other hand, multiple clinical studies have shown that permissive and limited hypoxia has improved stress response, cardiovascular function, aerobic capacity and glucose homeostasis (Liang et al., 2023). Thus, it is perceivable that a conservative oxygenation strategy might convey clinical benefits.
So far, appropriate oxygen administration with targeted SpO2 has been heavily debated, but optimal oxygen therapy for best patient outcomes remains inconclusive. Some studies have shown conservative oxygen therapy with lower SpO2 can reduce atelectasis, result in fewer days on mechanical ventilation and decrease organ failure (Girardis et al., 2016; Helmerhorst et al., 2016; Suzuki et al., 2015). On the other hand, the Liberal Oxygenation Versus Conservative Oxygenation in Acute Respiratory Distress Syndrome (LOCO2) trial, Oxygen-ICU trial and Air Versus Oxygen in Myocardial Infarction (AVOID) trial to investigate conservative vs. liberal oxygen therapy found an increased risk of overall mortality, vascular ischaemia and end organ failure in the conservative oxygen group (Barrot et al., 2020; Girardis et al., 2016; Stub et al., 2015). In the neonatal group, a Cochrane meta-analysis of 10 randomized controlled trials reviewing the difference in outcomes between low vs. high fractions of inspired oxygen delivery found no significant difference in mortality at discharge or neurodevelopmental disability at 2 years (Barbateskovic et al., 2019). The difference between these studies and their reported outcomes can largely be a result of heterogeneity between the different research studies, contrasting approaches to types of respiratory support, no consensus on targeted SpO2 ranges for intervention, and inherent biases such as selection and performance bias. As outlined above, many questions remain on the optimal SpO2 target range and the ideal amount of oxygen supplementation in clinical management.
Stabilization of HIF in the presence of oxygen has been of clinical interest, especially during ischaemic injury. Dimethyloxalylglycine, a non-specific PHD inhibitor, showed evidence of limb ischaemia amelioration after renal ischaemic injury, mitigated renal dysfunction, reduced ROS formation, restored gut barrier function and improved ALI outcomes (Eckle et al., 2013a; Liao & Zhang, 2020; Wang et al., 2014). Studies on myocardial ischaemia–reperfusion injury observed dimethyloxalylglycine to be useful in attenuating proinflammatory cytokine production, increasing EPO production and limiting myocardial infarction area (Eckle et al., 2008b).
Several modern HIF activators have been investigated by clinical studies. For example, the SIERRAS and HIMALAYAS trials administered oral HIF activator/PHD inhibitor roxadustat to adult patients on dialysis and reported that roxadustat significantly lowered low-density lipoprotein cholesterol, reduced frequency of blood transfusions, increased iron stores and was non-inferior to parenteral epoetin alfa with respect to maintaining haemoglobin levels (Haase, 2021). Another PHD inhibitor, vadadustat, has undergone several clinical trials including PRO2TECT and INNO2VATE trials and shown evidence that it is non-inferior to darbepoetin alfa for anaemia maintenance in chronic kidney disease patients that are dialysis-dependent (Haase, 2021). There have also been other clinical trials investigating the use of oral PHD-inhibitors in patients that are not reliant on dialysis but have chronic kidney diseases. Oral medications including vadadustat, molidustat, daprodustat, desidustat, roxadustat and enarodustat have similarly found that these oral HIF-inhibitors are non-inferior to erythropoietin-stimulating agent if not marginally better at improving haemoglobin numerical values (Haase, 2021; Sugahara et al., 2022). Importantly, daprodustat was recently approved by the US Food and Drug Administration as a treatment for renal anaemia in dialysis-dependent chronic kidney disease patients, further validating the feasibility for PHD inhibitors. The application of HIF activator therapy to ARDS is still under evaluation at the same time that vadadustat is undergoing a clinical trial to investigate its therapeutic potential in COVID-19 associated ARDS (NCT04478071).
Since its discovery in the late 1990s, HIF has become an exciting therapeutic target, given its significant function in physiological and pathological conditions. Indeed, both HIF activators and inhibitors have been studied extensively in the clinical setting as treatments for renal anaemia and certain cancers. Therapeutic enhancement of the HIF signalling pathways might also serve as ARDS therapy via optimizing epithelial cell metabolism, promoting endothelial tight junctions and modulating immune cell activation. Its utility, however, depends on various factors, such as HIF-1α or PHD specificity and the timing of treatment. For example, HIF-1α mainly mediated lung protection through enhancing glycolysis and extracellular adenosine signalling, whereas HIF2α might play a contradictory role in immune cell activation. Thus, specific HIF-1α activators might be more advantageous than global HIF activators as a treatment for ARDS. Furthermore, identifying specific targets for HIF transcription factors is essential for understanding the detailed molecular mechanism governed by HIF, facilitating treatments specific to HIF targets. Although questions remain on the functional role of HIFs in different cell types with respect to human physiology and pathology, further research will unlock many biological mechanisms important for its therapeutic targeting in pathological disease.