Oxomer Formation by the Oxygen Sensors PHD1-3 and FIH

Abstract

The regulation of the response to limited oxygen levels (hypoxia) is crucial for metazoan survival and is primarily governed by the transcription factor hypoxia-inducible factor (HIF). HIF consists of HIF-α and HIF-β subunits. The HIF pathway, regulated by prolyl hydroxylase (PHD) proteins and factor inhibiting HIF (FIH), plays a pivotal role in oxygen-dependent signalling, primarily through hydroxylation of specific proline and asparagine residues on HIF- α, modulating its stability and transcriptional activity. While hypoxia-induced stabilization of HIF-α facilitates the transcriptional activation of genes crucial for adapting to oxygen depletion, the potential broader impact of PHDs beyond the HIF pathway remains uncertain, posing limitations on our understanding of oxygen-dependent physiology. It is critical for the understanding of cellular hypoxia adaptation whether PHD1-3 and FIH exhibit selectivity for HIF-α or have additional non-HIF targets. If PHDs and FIH have target proteins outside the HIF pathway, this would have implications for the interpretation of PHD inhibitor effects in clinical settings and warrants further investigation. We therefore decided to identify and characterise additional putative targets for both FIH and PHDs. We have recently discovered that FIH forms an oxygen-dependent stable protein oligomer (oxomer) with the deubiquitinase OTUB1. Oxomer formation was sensitive to hypoxia and regulated OTUB1 activity, therefore representing a novel mechanism of oxygen-dependent cellular signalling. In addition, it was previously shown that FIH likely forms oxomers also with other proteins, including the inhibitor of NF-κB (IκB) β protein. Here, the FIH-dependent formation of an oxomer with IκBβ was further investigated. It was verified that FIH forms an oxomer with IκBβ, that this novel oxomer formation is highly sensitive to hypoxia and does not occur with other members of the IκB protein family. The FIH-IκBβ oxomer formation occurred selectively through a specific IκBβ amino acid sequence and disrupted the interaction between IκBβ and NF-κB subunits, suggesting a mechanism by which FIH-mediated oxomer formation regulates pro-inflammatory signalling in response to hypoxia. We decided to explore the possible formation of PHD oxomers, prompted by the prior identification of oxomers associated with FIH. We noticed that upon overexpression of PHDs in cells, PHD1-3 form high molecular weight complexes, which are sensitive to oxygen levels and PHD inhibitors. We discovered an oxomer formation between PHD3 and CND2, dependent on PHD3 enzymatic activity and exhibiting a high sensitivity to hypoxia. The PHD3- CND2 oxomer was quickly degraded in hypoxia, quicker than it would be expected based on the CND2 monomer stability. However, in the absence of PHD3 in the neuroblastoma cell line Kelly, the CND2 protein was degraded faster, which could not be explained by oxomer 3 formation. suggesting hydroxylation as a potential mechanism of regulation. This was further investigated with the use of PHD inhibitors which diminish enzymatic activity of the hydroxylases. PHD3 deletion in Kelly cells reduced H3K36me3 histone methylation, indicative of decreased global transcriptional activity, which was further supported by reduced levels of H3K4me1 and decreased incorporation of 5-EU in RNA. Moreover, PHD3 knockout decreased Kelly cell viability and proliferation, potentially via decreasing CND2 stability and subsequently possibly reducing functional condensin I levels, which might lead to a cell cycle block. In summary, this thesis elucidates FIH-mediated oxomer formation with IκBβ, demonstrates high hypoxia sensitivity and specificity, disrupts the interaction between IκBβ and NF-κB subunits and offers insight into FIH's role in regulating pro-inflammatory signalling under hypoxic conditions. Moreover, we show that PHD3 forms an oxomer with CND2, representing the first evidence of PHD oxygen sensors catalysing the formation of a stable protein oligomer. This highlights CND2 as a novel PHD3 target regulated by PHD3 enzymatic activity. PHD3 depletion reduces CND2 stability, leading to impaired cell cycle progression, proliferation, and viability in Kelly neuroblastoma cells. This suggests a potential therapeutic role for prolyl hydroxylase inhibitors in neuroblastoma treatment.