https://orcid.org/0000-0002-7828-8118
ABSTRACT
Hypoxia is a type of stress caused by an insufficient supply of oxygen. Macroautophagy/autophagy, a well-conserved pathway, is induced during hypoxia; however, the exact mechanism by which autophagy is regulated in a hypoxic environment remains to be elucidated. A recent study by Li et al. shed light on how hypoxia can regulate early steps of autophagy induction. In this study, the authors discovered a novel symmetrical dimethylation of ULK1 at arginine 170 (R170me2s) that accumulates during hypoxia and increases ULK1 kinase activity by promoting autophosphorylation of ULK1 at T180. The authors identified PRMT5 and KDM5C as the primary methyltransferase and demethylase regulating ULK1 R170me2s and show that the lack of oxygen directly leads to reduced activity of KDM5C, which is likely the cause of accumulation of ULK1 R170me2s during hypoxia. Furthermore, the authors showed that ULK1 R170me2s promotes mitochondrial turnover and maintains cell viability in response to hypoxia stress. Together these data provide a new perspective on how the oxygen level regulates autophagy induction and show the physiological role of ULK1 R170me2s.
Maintaining the appropriate oxygen level is critical to the cell as oxygen serves as an electron acceptor in multiple important biochemical reactions, including ATP synthesis and fatty acid desaturation. An insufficient oxygen supply, namely hypoxia, is therefore a stress condition for the cell. In addition, hypoxia also promotes the production of reactive oxygen species (ROS), which, when accumulated, can lead to further harm in the cell. Multiple pathways are implicated in the cellular adaptation to hypoxia, including autophagy [Citation1]. Macroautophagy (hereafter autophagy) has been proposed to replenish intracellular nutrients for catabolic activities when the ATP supply is limited during hypoxia. Autophagy also has the ability to remove oxidized biomolecules (including proteins, DNA and lipids), which can be the downstream effects of ROS. In addition, hypoxia-induced autophagy can promote mitophagy (selective autophagic degradation of mitochondria) to limit cellular oxygen consumption [Citation2]. Although it has been shown that autophagy contributes to cell survival under hypoxia, little is known about how the early steps of autophagy induction are regulated under a hypoxic environment. In their recent paper, Li et al. set out to study the regulation of autophagy induction during hypoxia [Citation3].
The authors first examined the effect of hypoxia on autophagy induction. In LN229 human glioblastoma/GBM cells, Huh7 hepatocellular carcinoma (HCC) cells, and human oral keratinocytes (HOKs) switched from the normoxic environment to the hypoxic environment (1% oxygen) for 12 h, they found induction of autophagy as indicated by increased LC3-II levels based on western blot in all three cell lines and increased GFP-LC3 puncta based on microscopy in the LN229 cells. They then showed that such induction requires ULK1 as mouse embryonic fibroblasts (MEFs) with a ulk1 knockout (KO) have a significantly reduced LC3-II level after a 12-h hypoxia treatment.
How is ULK1 activated during hypoxia? While it is well established that AMPK and MTOR signaling regulate ULK1 activity during energy and nutrient stress, two lines of evidence in this study suggest that the early activation of ULK1 upon hypoxia observed in this paper is independent of these two kinases: 1) Genetic manipulation that leads to constitutively active MTOR or loss of PRKAA1/AMPKα1 and PRKAA2/AMPKα2 do not affect the hypoxia-induced accumulation of LC3-II in MEFs; and 2) the induction of autophagy manifests as early as 12 h after hypoxia treatment, preceding the changes in AMPK- and MTOR-mediated ULK1 modifications. The authors thus investigated other changes in ULK1 at the 12-h time point by subjecting ULK1 immunoprecipitates derived from hypoxia-stimulated LN229 cells to liquid chromatography-tandem mass spectrometry. Through this approach they were able to identify a symmetrical dimethylation at the evolutionarily conserved arginine at position 170 on ULK1 (R170me2s). The level of ULK1 R170me2s is low under the normoxia condition and is enriched as early as 6 h after hypoxia treatment. The authors also showed that the endogenous level of ULK1 R170me2s negatively correlates with oxygen level in cell culture.
Does ULK1 R170me2s participate in autophagy induction during hypoxia? To address this question, the authors used shRNA to knock down endogenous ULK1 and expressed wild-type (WT) ULK1 or a ULK1R170K mutant, which is resistant to R170me2s modification. In all three cell lines they tested (LN229, Huh7, and HOKs), WT ULK1, but not ULK1R170K, is able to sufficiently induce autophagy (indicated by LC3-II level and GFP-LC3 puncta number) after a 12-h hypoxia treatment, despite comparable total exogenous ULK1 protein levels. The same results can be obtained by knockin expression of ULK1R170K in LN229 cells. Thus, ULK1 R170me2s is required for hypoxia-induced autophagy.
The authors next asked what are the methyltransferase(s) and demethylase(s) for hypoxia-induced ULK1 R170 during hypoxia. By knockdown experiments of two arginine methyltransferases, PRMT5 and PRMT7, that catalyze symmetric arginine dimethylation, the authors found that the former, but not the latter, is involved in ULK1 R170me2s, and that PRMT5 can methylate ULK1 in vitro. However, the increase in ULK1 R170me2s during hypoxia seems not to be caused by PRMT5 alteration, as the PRMT5 level is not changed during hypoxia, and in vitro PRMT5 activity is not affected by oxygen level. To investigate whether hypoxia-induced ULK1 R170me2s is governed by a demethylase, the authors tested KDM4E and KDM5C, two demethylases capable of catalyzing de-dimethylation of arginine. KDM5C, but not KDM4E, immunoprecipitates with endogenous ULK1 in LN229, Huh7, and HOK cells. The authors confirmed that KDM5C is responsible for removing R170me2s on ULK1 by showing that knockdown of KDM5C increases ULK1 R170me2s levels under the normoxia condition but does not further increase the ULK1 R170me2s level during hypoxia compared to controls. Importantly, the ability of KDM5C to demethylate ULK1 R170me2s in vitro decreases with oxygen level. Together, these data indicate that PRMT5 and KDM5C jointly regulate the level of ULK1 R170me2s during hypoxia and suggest that hypoxia-induced ULK1 R170me2s is due to repressed KDM5C activity.
What is the effect of ULK1 R170me2s on autophagy? To address this question, the authors performed a kinase activity assay using ULK1 affinity-isolated from cells expressing Flag-tagged ULK1 or ULK1R170K under both normoxia and hypoxia conditions. They showed that although under the normoxia condition both WT ULK1 and ULK1R170K are inactivated, ULK1R170K has significantly reduced kinase activity during hypoxia. As ULK1 can promote autophagy by phosphorylating and activating ATG13 and BECN1, the authors knocked down endogenous ULK1 and introduced WT ULK1 or ULK1R170K. Phosphorylation of ATG13 at serine 335 and BECN1 at serine 15 (phosphorylation sites targeted by ULK1) are significantly reduced in cells expressing ULK1R170K compared to cells expressing a comparable amount of WT ULK1. To further illustrate the role of ULK1 R170me2s, the authors used an in vitro system. Adding PRMT5 to purified WT ULK1, but not ULK1R170K, promotes ATG13 and BECN1 phosphorylation. Similarly, adding WT KDM5C, but not a catalytically dead KDM5C mutant, to WT ULK1 purified from cells under the hypoxia condition significantly reduces the phosphorylation of ATG13 and BECN1 by ULK1. To demonstrate the effect of ULK1 R170me2s on downstream steps of autophagy, the authors showed that knockin expression of ULK1R170K in LN229 cells leads to a significantly reduced hypoxia-induced activity of PIK3C3/VPS34. Together these data indicate that the induction of ULK1 R170me2s promotes autophagy by increasing the kinase activity of ULK1.
How does the R170me2s modification promote ULK1 activity? By examining a structural analysis of ULK1, the authors found that R170 is located within a conserved loop domain and is close to T180. The authors showed that autophosphorylation of ULK1 at T180 is important for its kinase activity and the phosphorylated form accumulates as early as 6 h after hypoxia treatment. Importantly, knockin expression of ULK1R170K significantly reduces autophosphorylation at T180 and in vitro PRMT5 inhibits the autophosphorylation of ULK1 at this site. Together these data suggest that the R170me2s modification increases ULK1 activity by promoting its autophosphorylation at T180.
What are the physiological effects of ULK1 R170me2s? Knockin expression of ULK1R170K is able to suppress the loss of TOMM20-positive mitochondrial area, mitochondria DNA level, and cellular oxygen consumption rate during hypoxia compared to cells expressing WT ULK1. Knockin expression of ULK1R170K also reduces the number of mitochondria-containing autophagosome-like vesicles seen under electron microscopy during hypoxia. These data strongly suggest that ULK1 R170me2s promotes mitochondria turnover during hypoxia. In a trypan blue exclusion assay, a higher death rate is observed in cells with ULK1R170K knockin expression compared to wild-type cells upon hypoxia treatment, suggesting that ULK1 R170me2s promotes cell viability during hypoxia. To investigate the impact of this modification on tumor development, the authors injected either WT LN229 cells or cells with knockin expression of ULK1R170K into athymic nude mice. Tumors expressing ULK1R170K show reduced growth, increased cell death, abolished ULK1 R170me2s modification, and reduced phosphorylation of ULK1 T180, ATG13 S355, and BECN1 S15. These phenotypes are further enhanced when the tumor-bearing mice are exposed to an 11% low-oxygen environment. Clinically, the level of HIF1A, BNIP3, ULK1 R170me2s, ULK1 p-T180, and BECN1 p-S15 are positively correlated in human glioblastoma samples, whereas in glioblastoma patients with a high level of R170me2s, a significantly shorter survival time is found. Together these results suggest that ULK1 R170me2s helps cells cope with hypoxia stress to maintain tumor cell viability and reveals the clinical importance of the ULK1 R170me2s marker.
In multicellular animals, the oxygen level in the microenvironment is a critical factor for cell survival [Citation4]. The research by Li and colleagues provides a new perspective on how autophagy is induced in the early phase of the hypoxia response. Although repressed MTOR and active AMPK were previously detected under hypoxia conditions, the induction of autophagy in the present study was observed to occur prior to the action of MTOR and AMPK. Instead, the accumulation of ULK1 R170me2s precedes the action of these regulatory kinases and activates ULK1 for autophagy induction. More importantly, the oxygen level seems to be directly regulating the activity of KDM5C in the cell, resulting in the accumulation of ULK1 R170me2s during hypoxia. This mechanism presents a more direct way for autophagy to respond to oxidative stress with regard to the well-recognized regulation via signaling [Citation2,Citation5–7]. The fact that R170 of mammalian ULK1 is a conserved residue brings up the possibility that such regulation could have appeared long ago in evolution. Is R170me2s also required for efficient autophagy induction during other stresses? Can ULK1 serve as a platform to allow for more rapid and stress-specific regulation of autophagy in response to different stresses? Further research is needed to answer these questions.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Additional information
Funding
References
- Lee P, Chandel NS, Simon MC. Cellular adaptation to hypoxia through hypoxia inducible factors and beyond. Nat Rev Mol Cell Biol. 2020;21(5):268–283.
- Filomeni G, De Zio D, Cecconi F. Oxidative stress and autophagy: the clash between damage and metabolic needs. Cell Death Differ. 2015;22(3):377–388.
- Li J, Zhang T, Ren T, et al. Oxygen-sensitive methylation of ULK1 is required for hypoxia-induced autophagy. Nat Commun. 2022;13(1):1172.
- Muz B, de la Puente P, Azab F, et al. The role of hypoxia in cancer progression, angiogenesis, metastasis, and resistance to therapy. HP. 2015;83. https://doi.org/10.2147/HP.S93413.
- Bellot G, Garcia-Medina R, Gounon P, et al. Hypoxia-Induced autophagy is mediated through hypoxia-inducible factor induction of BNIP3 and BNIP3L via their BH3 domains. Mol Cell Biol. 2009;29(10):2570–2581.
- Mungai PT, Waypa GB, Jairaman A, et al. Hypoxia triggers AMPK activation through reactive oxygen species-mediated activation of calcium release-activated calcium channels. Mol Cell Biol. 2011;31(17):3531–3545.
- Djavaheri-Mergny M, Amelotti M, Mathieu J, et al. NF-κB activation represses tumor necrosis factor-α-induced autophagy. J Biol Chem. 2006;281(41):30373–30382.