Understanding signalling pathways mutated in inherited disorders
To ensure we maximise opportunities we frequently collaborate with pharmaceutical companies, drug discovery units and chemical biologists such as Nathanael Gray (University of Harvard) to stimulate the elaboration of chemical inhibitors that specifically target signalling components that we are working with. These tool compounds greatly aid with deciphering the physiological roles that signalling pathways play. They can also be used to determine by what extent inhibiting specific signalling networks suppresses disease. We also aim to collaborate with leading clinicians to ensure that our research is addressing the most important clinical issues of the day. Where possible we will seek access patient derived cells or tissues to learn more about how disruptions of signalling pathways are linked to disease in human patients.
Our laboratory also employs state of the art biochemical and molecular technologies. Our primary focus is to analyse the function of endogenous components. Where possible, we try to avoid having relying on non-robust over-expression or siRNA knock-down methodology that more often than not results in hard to reproduce and physiologically questionable data being obtained. To help with this we make extensive use of genetic knock-in technology including CRISPR/CAS9 gene-editing approaches. We also make extensive use of mass spectrometry analysis of endogenous protein complexes to better understand how pathways are regulated and function.
My aim is to train PhD and Postdoc researchers who have the ambition to go on to become highly successful future research leaders either in academia or pharmaceutical Industry.
Neurodegeneration movement disorders
Our knowledge of the origins of Parkinson’s has been transformed by the identification of genes whose mutation in humans leads to Mendelian inherited disease (Figure 1). Interestingly many of these genes encode for enzymes involved in signalling pathways including Kinases (LRRK2, PINK1, GAK), components of the Ubiquitylation system (Parkin, UCHL1, Fbxo7) and GTPases (Rab7L1 and Rab39B).
Figure 1 Genes Mutated in Parkinson's disease
Our laboratory is currently devoting significant effort to dissecting the regulation and function of these components signal transduction components that have recently been found to be mutated in Parkinson’s disease. The highlight of our recent work undertaken in collaboration with Matthias Mann, identified the first physiological substrate for the LRRK2 protein kinase by showing that it directly phosphorylates a subset of the Rab GTPases on an a residue lying within the middle of the effector interacting-switch II domain (Figure 2)  . The LRRK2 phosphorylation site is conserved throughout evolution in 50 out of the 70 Rab GTPases. Thus far we have strong evidence for five Rab isoforms being phosphorylated by LRRK2 in vivo (Rab 5B, Rab7L1, Rab8A, Rab10 and Rab12)  . We found that pathogenic mutations in LRRK2 including G2019S mutation in the kinase domain as well as R1441G/C and Y1699C mutations in the GTPase regulatory domain markedly increases phosphorylation of these Rab isoforms at the switch II site  . LRRK2 phosphorylation strongly decreases the affinity of Rab isoforms we have tested to regulatory proteins that bind to the Switch II domain including Rab GDP dissociation inhibitors and the GDP/GTP exchange factor for Rab8 termed Rabin-8  . This work suggests that Parkinson’s causing mutations in LRRK2 will inhibit Rab GTPases isoforms that are phosphorylated by LRRK2.
The main focus of future work on LRRK2 will be to investigate the importance that LRRK2 phosphorylation of Rab isoforms plays in biology and to explore links to Parkinson’s disease. We will identify effector proteins specific for the most relevant Rab GTPases we select to study. We will also generate phospho-specific antibody reagents to explore whether Rab phosphorylation in blood cells or CSF can be used as a biomarker for LRRK2 activity in vivo. We would also explore how pathogenic disease-causing mutations of LRRK2 in both mice and humans impact on Rab phosphorylation and whether this can be used as a biomarker for disease progression in Parkinson’s disease.
Figure 2 LRRK2 Phosphorylates and Inhibits Rab GTPases
In collaboration with AstraZeneca we have investigated why a panel of breast cancer cells that had mutations in PTEN or Class 1 PI3Kα were resistant to Akt inhibitors. This resulted in the discovery that a majority of these Akt inhibitor resistant cells expressed high levels of SGK1, a kinase that is closely related to Akt and likely phosphorylate overlapping substrates as well as being activated by the same upstream kinases (PDK1 and mTORC2)  . We elaborated a simple assay based on monitoring the effects that Akt inhibitors have on NDRG1 phosphorylation to determine which cancer cells display high activity of SGK1 and are therefore likely to be insensitive to Akt inhibitors  .
We are also interested in SGK3 which is unique in that it is the only protein kinase known to possesses a PtdIns(3)P-binding PX domain at its N-terminus (Figure 3). We confirmed that SGK3 does indeed bind PtdIns(3)P in vitro and showed that SGK3 localises at endosomes in vivo where PtdIns(3)P is located through its PX domain  . We found that PX domain mutations that ablate binding to PtdIns(3)P, inhibit SGK3 activity by suppressing phosphorylation at the PDK1 and mTORC2 sites  . We have discovered that the lipid kinase which generates the PtdIns(3)P at the endosomal membrane required for triggering the activation of SGK3, is the class 3 PI3K family member termed hVPS34 known to localise at endosome. Deploying structurally diverse highly selective VPS34 inhibitors (VPS34-IN1 and SAR405), we found that SGK3 activity induced by prolonged inhibition of Class 1 PI3K /Akt is indeed controlled through hVps34  .
Figure 3 Can SGK3 activate mTORC1 under conditions where PI3K and Akt are inhibited?
In future work we aim to continue our characterisation of the role that SGK signalling pathways plays in cancer. This would include identifying and characterizing SGK3 substrates employing different approaches. We will explore whether SGK or more likely a combination of SGK/Akt inhibitors have therapeutic potential for treating tumours that display elevated PI3K pathway activity
We have been working a number of years in studying the regulation and function of the WNK isoforms as mutations that increased the expression of WNK1 or WNK4 isoforms caused an inherited hypertension disorder termed Gordon’s syndrome. We have discovered that WNK1 is activated in response to hyperosmotic stress and phosphorylates and activate two other closely related protein kinases termed SPAK and OSR1  . These kinases phosphorylate and regulate the activity of ion co-transporters such as NCC, NKCC1 and NKCC2 that play critical roles in regulating blood pressure and ion homeostasis  . We have now found that SPAK/OSR1 not only activate ion co-transporters (NCC/NKCC1/NKCC2) responsible for influx of Na+ and Cl- ions into cells, but also directly phosphorylate and inhibit potassium co-transporters (KCC1,KCC2,KCC3 and KCC4) that regulate the efflux of K+ and Cl- ions from cells  . This mechanism explains how WNK signalling pathway can cordately regulate net transport of ions into cells.
Figure 4 The WNK1-regulated SPAK/OSR1 kinases stimulate net salt uptake by reciprocally activating sodium co-transporters and inhibiting potassium co-transporters
All of our data points towards SPAK/OSR1 representing good targets for drugs to treat hypertension and/or reduce chloride levels in neurons for the treatment of mood disorders and epilepsy  . There is some debate on whether conventional kinase inhibitors would be sufficiently specific to treat a largely asymptomatic condition such as hypertension, so we have explored other approached which SPAK/OSR1 function could be inhibited. We found that SPAK possesses a conserved carboxy-terminal (CCT) domain, which operates as a docking domain to recognise RFXV/I motifs present in its upstream activator WNKs as well as its substrates (NCC, NKCC1NKCC2)  . To validate the approach of generating CCT domain inhibitors, we generated knock-in mice in which the ability of the CCT domain to bind RFXI/V motif was ablated . The CCT domain defective animals displayed markedly reduced SPAK activity and phosphorylation of NCC and NKCC2 co-transporters at the residues phosphorylated by SPAK. These knock-in mice also displayed markedly reduced blood pressure, suggesting that CCT domain inhibitors would have the potential to effectively inhibit the WNK signalling pathway .
In 2012 it was reported that mutations in one of two genes (KLHL3 and Cul3) encoding a poorly studied E3 ligase complex caused Gordon’s hypertension syndrome the same conditions that was caused by mutations WNK1 and WNK4 genes. In collaboration with Thimo Kurz we discovered that KLHL3 interacted with WNK isoforms and we showed that pathogenic mutations in KLHL3 either abolished binding to CUL3 or WNK isoforms  . We identified the degron motif on WNK4 that bound to KLHL3 that interestingly encompasses the residues on WNK4 that are mutated on Gordon’s syndrome  . In collaboration with Alex Bullock we crystallised the KLHL3 Kelch domain in complex with the WNK4 degron motif which revealed the intricate web of interactions between conserved residues on the surface of the Kelch domain β-propeller and the WNK4 degron motif  . Many of the disease-causing mutations inhibit binding by disrupting critical interface contacts  . These results show that the CUL3-KLHL3 E3 ligase complex regulates blood pressure via its ability to interact with and ubiquitylate WNK isoforms.
Figure 5 Ubiquitylation of WNK isoforms is regulated by CUL3-KLHL3 E3 ligase
We plan to understand in more detail how the WNK signalling pathway senses ionic/osmotic stress. This is likely to be mediated at least in part by the uncharacterised C-terminal domains of these enzymes. We believe this project is important as the ability of mammalian cells to sense and respond to changes in ionic conditions is fundamental for survival and function of biology and this process is very poorly understood.
Professor Dario Alessi
T: 44 1382 385602
F: 44 1382 223778