DYRK protein kinases regulate p62/SQSTM1 to orchestrate cellular responses to oxidative stress, protein misfolding and nutrient starvation

Lead Research Organisation: Babraham Institute
Department Name: Signalling

Abstract

The cells in our body are constantly exposed to chemical and physical 'stress' and accumulate a lifetime of damage to proteins, DNA and various key sub-cellular structures (organelles). It is vital that cells are able to respond appropriately to this damage; failure to do so progressively undermines cellular 'fitness' contributing to age-related declines in cell and tissue function that underpin the normal ageing process and can contribute to age-related diseases such as cancer and dementia.
One way that cells respond to stress is to increase the abundance of enzymes that detoxify the cell by removing potentially harmful chemicals. When a cell receives a stress signal the genes that code for detoxifying enzymes are 'read' by 'transcription factors', discrete proteins that bind to DNA and transcribe the DNA information into RNA molecules, which are in turn 'translated' into the relevant proteins. One such transcription factor, called NRF2, coordinates cellular responses to oxidative stress.
Another way in which cells deal with cellular damage is through a process called 'autophagy' (self eating) in which damaged proteins are targeted to cellular recycling centres (called autophagosomes), where they are broken down to their raw materials, which can then be re-used. This process of autophagy requires 'cargo receptors', which bind to damaged proteins and transport them to the recycling centre for autophagy.
These complex processes are orchestrated by signalling pathways within cells that involve cascades of enzymes called protein kinases. These enzymes 'tag' other proteins with a phosphate group (a process called phosphorylation) and this changes the activity, abundance or localisation of the protein. The tagged protein is referred to as the 'substrate' of the protein kinase enzyme.

This project concerns two protein kinases, called DYRK1B and DYRK2, and a specific 'cargo receptor' called p62 or sequestosome-1; here we'll call it p62. Our new results point to a link between the DYRKs and p62 in coordinating how cells respond to stress and damage:-

1. p62 acts as a scaffold to coordinate the activation of the NRF2 transcription factor. We have now discovered that DYRK1B controls the expression of detoxifying enzymes that are known to be targets of NRF2 suggesting that DYRK1B may control activation of NRF2.

2. p62 acts as a cargo receptor for damaged proteins, transporting them to recycling centres for autophagy. It has also been shown that p62 can shuttle in and out of the cell nucleus to collect damaged nuclear proteins for autophagy. This nuclear shuttling requires phosphorylation of p62 but the kinase responsible for this has remained a mystery. We have now discovered for the first time that p62 is phosphorylated by DYRK1B and DYRK2; we propose that this sends p62 into the nucleus to help collect damaged nuclear proteins.

3. p62 coordinates activation of a protein kinase enzyme called mTOR when cells are starved of nutrients. Indeed, mTOR is a critical regulator of lifelong health and controls the lifespan of organisms such as worms, flies, mice and possibly man. Regulation of mTOR by nutrients takes place at specific organelles called lysosomes. However, the phosphorylation of p62 may take it away from lysosomes to the nucleus. Indeed, we find that DYRK2 and p62 co-locate in cells at aggresomes - sites of damaged proteins, consistent with the cargo function of p62. It is not known what effect this re-location of p62 has on nutrient signalling via mTOR at lysosomes.

In this study we will define how the DYRKs regulate p62 to coordinate cellular responses to stress and damage. This work is critical to understanding how stress contributes to normal ageing but may also have implications for diseases of old age (dementia, cancer); thus, our results may have wider impacts and we will work with scientists in these areas to progress this.

Technical Summary

p62/SQSTM1 is a multi-functional scaffold protein that acts as a signalling hub. Amongst its several functions p62 can:
1. competitively inhibit the KEAP1-NRF2 interaction, thereby promoting NRF2 nuclear entry and ARE-dependent gene expression;
2. bind Raptor, a subunit of the mTORC1 complex; indeed, p62 is required for mTORC1 activity in response to amino acids;
3. bind to aggregated or damaged proteins at aggresomes and deliver them to autophagosomes for autophagic recycling;
4. translocate into the nucleus in a phosphorylation-dependent fashion to recruit and clear nuclear protein aggregates.

We have now found that:
1. The protein kinase DYRK1B can promote expression of canonical NRF2 target genes.
2. DYRK1B and DYRK2 both co-localise with p62 at aggresomes.
3. DYRK2 promotes the assembly p62 foci near the nucleus (typical of aggresomes) and promotes nuclear entry of p62.
4. DYRK1B and DYRK2 can both phosphorylate p62 at sites implicated in its nuclear entry.

Our results define p62 as a substrate of DYRK1B/DYRK2 in vitro and in cells and suggest entirely new functions for these kinases in regulating p62 and its functions, including NRF2 activation, mTORC1 signalling and the trafficking of aggregated proteins. These processes are intimately involved in normal age-related declines in cellular fitness; indeed, p62 KO mice exhibit a premature ageing phenotype. We suggest that DYRK signalling promotes the assembly of p62 foci (aggresomes) and the nuclear entry of p62 as part of the cellular response to misfolded or aggregated proteins. This redistribution of p62 impacts on other p62 functions to regulate NRF2 and mTOR signalling. In this way DYRKs and p62 collaborate to orchestrate cellular responses to oxidative stress, protein misfolding and nutrient starvation.
This is a fundamental biology study but it may also have important implications for diseases where p62 is implicated, including neurodegeneration and cancer.

Planned Impact

The primary impact will come from new knowledge of mechanisms of signal transduction, related to the role of DYRKs and p62 in various biological contexts (see Academic beneficiaries).
Impacts on other stakeholders:

1. Industry: by enhancing the research capacity and knowledge of businesses and organisations. This proposal developed from a PhD studentship between the Cook lab and AstraZeneca, who had developed a 10nM selective DYRK1B inhibitor. All major pharmaceutical companies remain interested in protein kinases as drug targets for a variety of diseases. Several companies have active DYRK inhibitor programmes whilst others have held back due to the lack of clear substrates to provide context for the known biology or to serve as pathway biomarkers. Our research is therefore relevant to a range of BioPharma companies contributing to UK economic competitiveness. We will work with the Babraham Commercialisation Manager to manage any resultant IP and all interactions with industry.

2. BBSRC: meeting national strategic research priorities. We have defined novel regulators of p62, a protein with a central role in stress responses and cell homeostasis, which is clearly a regulator of normal ageing. Within the BBSRC Strategic Plan 2013/14 refresh this research lies at the heart of Strategic Research Priority 3 - Bioscience For Health - and is relevant to the Societal Grand Challenge of 'maintaining health across the whole lifecourse' and the Key Priority 'Generate new knowledge of the biological mechanisms of development and the maintenance of health across the lifecourse'. In addition, this work 'will provide new insights to potential strategies for health monitoring and intervention, including drug targets and pharmaceuticals' consistent with the aspiration that 'basic bioscience funded by BBSRC underpins the pharmaceutical and healthcare industries'.

3. Healthcare and 3rd sector charities: translation of research into the clinic. The DYRKs are implicated processes that underpin disease and infirmity in old age. DYRK1B is implicated in myogenesis and adipogenesis. Age-related loss of muscle mass significantly impairs quality of life in the elderly. Similarly, adipocytes are critical regulators of metabolism and are involved in a variety of metabolic diseases including obesity. DYRKs are directly implicated in Treg biology in the immune system, cardiac hypertrophy and CNS cell death - all of which have ageing as a key risk factor or drive age-related disease. Furthermore, p62, the protein we have identified as a new target of DYRK1B and DYRK2, is intimately linked to age-related disease as a regulator of NRF2, mTOR and autophagy. Thus, our basic biology will be of interest to a variety of disease charities and healthcare professionals.

4. Training: generating a skilled workforce. This project will provide further training for key researchers (Fortian & Cook) in new scientific skills in growth areas (proteomics, genomics, bioinformatics). Through the links the Cook lab has with industry Fortian will also be exposed to drug discovery research providing training for her future contribution to the academic or commercial sector.

5. Science & Society: influencing and informing policy and increasing public understanding of science. We will continue to contribute to public STEM understanding through our public engagement activities. The PDRA, Fortian, will join other members of the Cook lab in PE activities, communicating their knowledge and enthusiasm to policy makers, the next generation of scientists and interested adults through science exhibitions, science festivals and science visits to schools and local community groups. We will work with the Babraham Public Engagement and Science Communications team to incorporate project specific content into resources for our new 'Science of Ageing' theme for PESC activities 2017-22.

Publications

10 25 50
 
Description p62/SQSTM1 is a scaffold protein and signalling hub. The literature tells us that p62: (i) binds KEAP1 to promote NRF2 nuclear entry and ARE-dependent expression of detoxifying genes; (ii) binds RAPTOR to link nutrient abundance to mTORC1 activation (iii) and binds to aggregated or damaged proteins at aggresomes, delivering them to autophagosomes for recycling.
The DYRK family of protein kinases are related to the CDKs and MAPKs. Studies in simple organisms suggest ancient roles in stress and nutrient signaling. In higher eukaryotes there is evidence that some DYRKs are involved in stress responses, but any role for DYRKs in nutrient signalling has not been explored. One recent study has suggested that DYRK3 might be involved in regulation of mTOR but details are scant. Indeed, in contrast to the CDKs and MAPKs, where 100s of substrates are known, relatively few DYRK substrates or gene targets have been defined. Using genomics and phospho-proteomics screens we have now found that: (i) DYRK1B can promote expression of canonical NRF2 target genes; (ii) DYRK1B and DYRK2 can phosphorylate p62 at T269/S272; (iii) DYRK1B and DYRK2 co-localise with p62 at aggresomes; (iv) DYRK2 promotes the assembly p62 foci near the nucleus (typical of aggresomes).

Our high level aims were:
1. To define DYRK:p62 interactions and the effect of DYRK on p62 phosphorylation and location.
2. To define DYRK-dependent gene expression and its relation to NRF2.
3. To investigate how DYRKs regulate mTORC1 signalling.
4. To investigate the regulation of p62 cargo function by DYRKs.

Aim1. To define DYRK:p62 interactions and the effect of DYRK on p62 phosphorylation and location.
We have validated a commercial pT269/pS272 p62 antibody (RNAi and mutagenesis) and confirmed that both DYRK1B and DYRK2 phosphorylate p62 at T269/S272 in cells and in vitro. Indeed, all the DYRKs that we have tested (1A, 1B, 2 & 3) can drive phosphorylation p62 at T269/S272 whereas the closely related kinase HIPK2 does not, indicating that p62 is a relatively selective substrate for DYRKs. In addition, our mass spec data suggested that DYRK2 also increased the phosphorylation of S28 within the PB1 domain of p62. We have now validated a pS28 p62 antibody and confirmed that DYRK does indeed drive phosphorylation of S28; this is observed to a lesser extent for DYRK3 but not the class I DYRKs (1A or 1B). Thus we have defined p62 as a new DYRK substrate and mapped the sites.
The DYRKs must bind to their substrates to phosphorylate them. To define the domains responsible for association we have collaborated with Terje Johansen (University of Tromso, Norway), a world expert on p62. Using recombinant proteins we have shown that DYRK1A, DYRK1B and DYRK2 interact in vitro with purified GST-p62; DYRK1B exhibits the strongest binding so we have focused on this to map the DYRK binding region in p62. Using GST-p62 deletion mutants we find that DYRK1B binds within aa 170-256 of p62, encompassing the previously defined RAPTOR and TRAF6 binding sites. Overlap with the RAPTOR interaction site may be relevant to a role in nutrient signaling (Aim 3 below). Thus we have defined the DYRK binding site within p62.
Consequences of phosphorylation. To assess the consequences of this phosphorylation we are looking at several functions of p62 (see below) and monitoring its subcellular location following DYRK activation. DYRK1B causes a pronounced increase in p62 abundance. We see this by immune fluorescence miscrocopy (IF) and western blot (WB) in cells where we conditionally activate DYRK1B. We also observe higher levels of p62 and pT269/pS272 p62 in cells stably expressing wild type DYRK1B compared to much lower levels in cells stably expressing kinase-dead DYRK1B (DYRK1Bkd); this also suggests that DYRK1Bkd may be a useful interfering mutant. The reduction in p62 in cells expressing DYRK1Bkd can be partially rescued by bafilomycin, which blocks lysosomal degradation. This is consistent with the fact that p62 is degraded during autophagy and suggests that DYRK1B may normally prevent autophagic degradation of p62 and, by extension, its cargo. Finally, we have confirmed that DYRK1B is predominantly nuclear and that it forms insoluble nuclear aggregates in response to proteasome inhibition (MG132); this correlates with loss of pT269/pS272 p62 phosphorylation. DYRK2 is predominantly expressed in the cytoplasm but forms perinuclear foci that co-stain with ubiquitin GFP. Finally, whilst DYRK2 alone cannot drive assembly of aggresomes, it colocalises with p62 in aggresomes in cells subjected to proteasome inhibition (MG132). This work is ongoing but clearly suggests distinct roles for DYRK1B and DYRK2 in proteostasis.
p62 protein-protein interactions. DYRK2, but not DYRK1B, can drive phosphorylation of S28 within the N-terminal PB1 domain of p62. This domain is responsible for the formation of higher order p62 complexes through homotypic and heterotypic interactions. Indeed, we find that DYRK2 promotes the formation of oligomeric pT269/pS272 p62; this is not seen with DYRK1 so we are exploring whether this is dependent on S28 phosphorylation. In addition, we are monitoring heterotypic interactions with other PB1 domain binding partners. For example we have confirmed that PKC-iota binds to p62; this binding is increased upon activation of DYRK2 and this in turn is reduced when we compare WT p62 with a S28A p62 mutant. Thus DYRK2 phosphorylation at S28 my influence PB1 domain-dependent interactions between p62, itself and other partners. This work is ongoing

Aim 2. To define DYRK-dependent gene expression and its relation to NRF2.
We have performed RNAseq to assess both RNA abundance. We have applied this to samples from cells in which we switched on either DYRK1B or DYRK2 expression for 8hrs or 24 hrs. Whilst analysis of these data sets is still ongoing it is apparent that HMOX1 (a classical NRF2 target gene that is involved in stress responses and detoxification) is induced by DYRK1B expression. We have confirmed this by RNAseq, qPCR and WB. Other known NRF2 targets are also induced so we are now proceeding to monitor the effect of DYRK1B on the activity of NRF2 by monitoring its nuclear localization and activation of validated NRF2:Luc reporter genes. The increase in ip62 abundance upon activation of DYRK1B (see Aim 1 above) may be related to regulation of NRF2. KEAP1 and p62 compete for binding to NRF2 so high levels of p62 would sequester KEAP1, causing stabilisation and nuclear entry of NRF2 and ARE-dependent gene expression. This may be an auto-regulatory loop as p62 is itself a NRF2 gene target. This work is currently ongoing.

Aim 3. To investigate how DYRKs regulate mTORC1 signalling.
RAPTOR, a critical regulatory component of the nutrient regulated mTORC1 kinase, interacts with p62; indeed, p62 is proposed to be required for activation of mTORC1 by nutrients, including amino acids. This has prompted us to investigate the role of DYRKs in nutrient responses.
We find that amino acid starvation causes a rapid (1 hour) increase in the abundance of both DYRK1B and DYRK2 and pT269/pS272 p62; in the case of DYRK2 this is accompanied by change in mobility of SDS-PAGE which suggests that DYRK2 undergoes phosphorylation. The increase in DYRK1B and DYRK2 abundance is quickly reversed when amino acids are re-supplemented. These results suggest that the DYRKs are nutrient responsive kinases that are dynamically regulated by amino acid abundance. This is an exciting and highly novel observation and we are pressing ahead to understand the role of DYRKs in nutrient signalling and how they are regulated by amino acid abundance. On the latter point we think the increase in abundance within 1 hour is too fast to be de novo gene expression and we are pursuing the hypothesis that the abundance of these kinases is regulated at the level of protein stability. This would be consistent with the fact that DYRKs autophosphorylate themselves so that they are active once synthesised; thus regulated abundance is likely to be a significant regulatory step.

We are continuing to investigate the role of DYRKs in nutrient signalling and their relation to mTOR signaling.
Further progress on all these aims is now being progressed through the availability of DYRK1B and DYRK2 KO cells that we have generated by CRISPR/Cas9 gene editing

In summary, very good progress is being made towards our aims and objectives.
Exploitation Route Depending on the outcomes, our work may shed new light on how autophagy and the clearance of toxic proteins and chemicals from cells is controlled. This may be relevant to the study of declining health with old age and some diseases of old age including cancer and neurodegeneration
Sectors Education,Healthcare,Pharmaceuticals and Medical Biotechnology

 
Description We are in discussions with a SME Biotech/Drug-Discovery company to progress aspects of this work and our work is also being reviewed by our Entrepreneur in Residence for suggestions on how to progress these impacts
First Year Of Impact 2019
Sector Education,Healthcare,Pharmaceuticals and Medical Biotechnology
Impact Types Economic

 
Title DYRK1B and DYRK2 CRISPR KO cells lines 
Description HEK293 cells with CRISPR/Cas9-mediated deletion of DYRK1B or DYRK2 Fully sequenced and characterisation ongoing 
Type Of Material Cell line 
Year Produced 2019 
Provided To Others? No  
Impact Initial characterisation has revealed some striking phenotypes that are currently being investigated. In the interim these cell lines are available for collaborators 
 
Title Inducible DYRK1B cells 
Description HEK293 cells lines engineered to exhibit Tet-inducible expression of the DYRK1B protein kinase 
Type Of Material Cell line 
Provided To Others? No  
Impact New knowledge of the biological function of the DYRK1B protein kinase, including new substrates. New research papers New collaborations 
 
Title Inducible DYRK2 cell line 
Description HEK293 cells lines engineered to exhibit Tet-inducible expression of the DYRK2 protein kinase 
Type Of Material Cell line 
Provided To Others? No  
Impact New knowledge of DYRK2 Identification of new substrates of DYRK2 
 
Title DYRK1B phosphoproteomics data set 
Description Using HEK293 cells exhibiting inducible expression of the DYRK1B protein kinase (HD1B cells) we have performed Phospho-SILAC mess spectrometry to identify DYRK1B-inducible phosphoproteins. Some of these turn out to be direct DYRK1B substrates 
Type Of Material Database/Collection of data 
Year Produced 2017 
Provided To Others? No  
Impact This data set has allowed us to identify new DYRK1B substrates and so has provided new insights into the function of this protein kinases in controlling gene expression and autophagy. The dataset will ultimately be released and freely available when the first mansucript is published 
 
Title DYRK2 phosphoproteomics dataset 
Description Using HEK293 cells exhibiting inducible expression of the DYRK2 protein kinase (HD2 cells) we have performed Phospho-SILAC mess spectrometry to identify DYRK2-inducible phosphoproteins. Some of these turn out to be direct DYRK2 substrates 
Type Of Material Database/Collection of data 
Year Produced 2017 
Provided To Others? No  
Impact This dataset has allowed us to identify DYRK2-inducible phosphoproteins, including new substrates, providing new insghts into the function of the DYRK2 protein in gene expression, autophagy/proteostasis and cell motility 
 
Description p62 phosphorylation 
Organisation University of Dundee
Country United Kingdom 
Sector Academic/University 
PI Contribution We are examining whether the HIPK family of protein kinases can regulate p62 function by phosphorylation
Collaborator Contribution Our collaborator is studying signalling by the HIPK protein kinasses and their role in oxidaant stress responses. We are studying p62 which coordinates stress respsones to oxidative damage. Our collaboraor has approached us to expand our work to encompass the HIPKs
Impact Too early for any specific outputs or outcomes
Start Year 2018
 
Description p62 phosphorylation 
Organisation University of Tromso
Department Department of Medical Biology
Country Norway 
Sector Academic/University 
PI Contribution Sharing of data; discussion of new experimental directions; experiments We have identified p62/SQSTM1 as a novel substrate of DYRK1B and DYRK2. We are investigating the role of this phosphorylation on p62 functions in proteostasis, stress responses and nutrient signalling
Collaborator Contribution Terje Johansen's lab will share reagents and expertise and perform specific experiments to define the interaction between DYRKs and p62
Impact Too soon for any specific outputs as the collaboration has only just started. However, i have visited the University of Tromso and presented two Lectures as part of a PhD training course
Start Year 2017
 
Description Escape Room Installation 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Public/other audiences
Results and Impact A 'Signalling' Escape Room was designed by students within the Signallign Laboratory, working with the Public Engagement team. This was then presented by studnets and post-docs, including members of the Cook lab at both the Cambridge Science Festival and the Latitude Music Festival.
Year(s) Of Engagement Activity 2019,2020
 
Description Participated in Babraham Institute exhibit at Royal Society Summer Science Exhibition 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Public/other audiences
Results and Impact The Babraham Institute prepared an exhibit - The Ageing Clock - which exemplified aspects of our ageing research portfolio for a public audience. Tjis was selcted tp be part of the prestigious Royal Society Summer Science Exhibition and I was involved in presentign this exhbit to the Public togehter with colleagues.
Year(s) Of Engagement Activity 2018
 
Description School visits 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Schools
Results and Impact The students were enthused about the topic of my presentation and this led to dialogue and discussion about several issues including new cancer therapies, evolution of drug resistance in cancer, the use of animals in research.

Anecdotally, the institute received requests for summer placement students following this visit.
Year(s) Of Engagement Activity 2014,2015,2016,2017,2018,2019
 
Description Science Open Day 
Form Of Engagement Activity Participation in an open day or visit at my research institution
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Schools
Results and Impact Students visited the lab and undertook small lab-based proejcts supervised by students/post-docs and myself. I explained the research that we do and discussed ethical issues such as the use of animals in research.
This precipitated excellent discussion and dialogue.

We received excellent feedback from the schools involved and requests for further outreach activities
Year(s) Of Engagement Activity 2013,2014,2015,2016,2017,2018,2019,2020
 
Description Visits by Teachers 
Form Of Engagement Activity Participation in an open day or visit at my research institution
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact My lab has hosted 6th form Biology teachers who were visiting my Institution during Half Term to update their knowledge as part of their CPD
Year(s) Of Engagement Activity 2016,2017,2018,2019,2020