Mark R. Cookson, PhD, University of Salford 1995 - US grants

Genetic forms of neurodegenerative disorders can help us to define the pathways associated with cellular dysfunction and damage. We are currently using dominant mutations to model cellular damage in Parkinson's disease (PD), dystonia and amyotrophic lateral sclerosis (ALS). Increasingly, we are also focusing on recessive genes that produce overlapping phenotypes to the dominant mutations. The concept is to understand the molecular links between different genes that cause the same disease, and thus to delineate the pathways that underlie neuronal dysfunction and cell death. In PD, where most of the work has been focused, we have shown that mutant alpha-synuclein causes neuronal death and that there is selectivity in these effects. Neurons that use express a dopaminergic neurotransmitter are selectively vulnerable to mutant alpha-synuclein, at least within the cell population cultured from the midbrain. Furthermore, we found that the recessive gene product Parkin can reverse this damage and we have implicated the ubiquitin-proteasome system in this relationship. Using gene-expression profiling, we have identified neurotransmitter alterations in aalpha-synuclein cell lines. We have also collaborated on a number of studies where new PD mutations have been identified, for example in the alpha-synuclein binding protein synphilin-1. More recently, we have begun working on an additional recently discovered recessive PD gene, DJ-1. We have already identified how a point mutation in DJ-1 causes disease; by altering conformation of an alpha-helix near the C-terminus, the protein is greatly destabilized. The challenge now is to understand the cellular function of the normal DJ-1 protein and develop models for loss of this function within the nervous system. Furthermore, whether DJ-1 has any interactions with parkin or alpha-synuclein is under investigation. Work on other diseases has also continued. Some observations of gene expression changes caused by dominant TorsinA mutations have been published, and more extensive work on these mutations is in preparation. We have shown that a recently reported mutation in TorsinA does not have the same cellular effects as a more common, previously reported deletion and have shown that the heat shock protein system is affected by expression of torsinA in vitro. Work on ALS has continued with a proteomics analysis of SOD1 mutations now published; we are currently applying similar approaches to PD models.

DJ-1 is an evolutionarily conserved protein of unknown function. Mutations in the human DJ-1 gene are associated specifically with a form of recessive parkinsonism, suggesting that despite its widespread expression in the body, the major effect of loss of function is neurodegeneration in the brainstem. The current project attempts to understand why this is the case, and we have been following two major strands. One of the well characterized roles of DJ-1 is in responses to oxidative damage. We have shown previously that DJ-1 has an unusual response to oxidation both in vitro and in vivo in that a specific cysteine residue, C106 in human DJ-1, is modified to form a cysteine-sulfinic acid. Replacement of C106 renders DJ-1 non-functional in a number of assays related to cellular survival in the presence of oxidative stress, which we have interpreted to mean that sulfinic acid formation is critical for DJ-1 function. However, a reasonable alternative interpretation is that the cysteine residue itself is critical and the sulfinic acid is an unimportant modification. To address this, we examined the crystal structure of DJ-1 and noted that a nearby residue that is also evolutionarily conserved, Glu18 in human DJ-1, stabilizes the sulfinic acid sidechain of oxidized C106. By manipulation of Glu18, we were able to produce mutant DJ-1 forms that had decreasing ability to support Cys-106 modification. Although the Cysteine residue is present in all of these variants, ability of DJ-1 to protect cells against oxidative-stress induced mitochondrial dysfunction or cell death correlates with the ability to support the modified sulfinic acid sidechain. This indicates that the modification, not the underlying amino acid is critical for function. The next question is to understand why such an unusual protein modification is important. We have previously proposed that DJ-1 can bind RNA in an oxidation-dependent manner, and have shown recently that this extends to the human brain. However, we have not shown that DJ-1 RNA binding is important for its cellular protective function and we are currently exploring this hypothesis by making directed mutations. We are also attempting to understand why mitochondrial function appears to be a major target of DJ-1 activity in cells.

We have explored the associations between DNA variants in the genome, particularly single nucleotide polymorphisms (SNPs) and expression of nearby mRNA. At a genome wide level, there are many such associations found as expression quantitative trait loci (eQTLs). We have generated, and made publically available, one of the largest eQTL mapping sets in the human brain, with nearly 400 individuals with age range covering most of the human lifespan where genotype and expression data form microarrays is available. Primarily this is focused on two brain regions, the cerebellum and the frontal cerebral cortex, with ancillary datasets in other brain regions and on isolated cell types. Our aim is to provide a comprehensive view of the genomic control of gene expression in this complex organ that can be examined in a number of ways. In our own laboratory, we have begun to examine the relationship of age to epigenetic and gene expression markers. We have found that there are distinct and reproducible gene DNA methylation and expression signatures in the human brain and are following this up by analysis of additional datasets. If this can be confirmed, we will have found novel interesting candidates for neuronal responses to aging. This dataset has been used by several groups to identify the effects of SNPs nominated by genome wide association studies (GWAS) for their effect on lifetime risk of age-related neurodegenerative disorders such as Parkinsons disease or progressive supranuclear palsy. The data has also been used to examine genetic influences on expression of biomarkers related to other diseases. In the future, we will augment these datasets with RNA sequencing level analyses, which is allowing us to examine gene expression with greater precision and better dynamic range than microarrays. This methodology also allows examination of individual exons of a transcript even to the single base level. Current work is aimed at defining bioinformatic tools to query this data and exploit it in the same way that we did for array datasets.

Our main aim in this project is to understand how mutations across many different domains of LRRK2 cause dominantly inherited Parkinsons disease. We have been particularly looking for shared effects of multiple mutations that are found in many different functional domains of the molecule. In the current period, we have followed up on our previous observations that LRRK2 interacts with two proteins at the trans-Golgi network (TGN), Rab7L1 and GAK. Interestingly, these two proteins are in loci for risk of sporadic Parkinson's disease. We have shown that all mutations in LRRK2 promote the retention of LRRK2 at the TGN. Ongoing work is aimed at understanding the mechanistic basis of this observation. We have no completed additional, secondary screens, that further nominate interactions between Golgi-derived vesicles and retromer particles, supporting further links to different familial forms of PD.

Loss of function mutations in either of the genes PINK1 or parkin are associated with autosomal recessive parkinsonism in humans. It is known that these two genes are linked in a single genetic pathway that normally prevents age-dependent mitochondrial dysfunction in Drosophila models. PINK1, which is a mitochondrial kinase has also been shown to support the recruitment of parkin to damaged or depolarized mitochondria. Once recruited to damaged organelles, parkin promotes their turnover via mitophagy, a specialized form of autophagy that acts as a form of mitochondrial quality control. Although several publications have outlined some of the essential molecules involved in PINK1-dependent parkin recruitment and subsequent mitophagy, many details are still unclear. For example, what occurs in depolarized mitochondria to trigger PINK1 activation is poorly understood. The aim of this project is to understand the molecular basis of the relationship between PINK1 and parkin. To this end, we recently have completed a large shRNA screen against depolarization-induced parkin relocalization to find novel participants in this process. We were able to recover PINK1, showing that the screen worked, and are now validating additional genes. To date we have been able to confirm at least one other gene as well as PINK1, which happens to also be a mitochondrial protein. Functional assays in cell lines and in neurons suggest that this protein is upstream of PINK1 genetically although the mechanism by which it promotes PINK1 recruitment is still under investigation.

Our work on a-synuclein is currently focussed on applying large scale screening approaches to understand the pathobiology associated with this protein, which is now known to not only be a marker of disease but also plays an active role in disease progression. In ongoing work, we have been probing the mechanism by which a-synuclein is taken up from one cell to another, a process that has been proposed to be important in the spread of disease between brain regions. We have currently developed reagents to knock down candidate transmembrane receptors in vivo to determine if any contribute to spread of a-synuclein and associated toxicity.

Parkin is a cytoplasmic protein-ubiquitin E3 ligase that is proposed to be involved in regulation of a number of cellular substrates. We have proposed that one potential substrate is phospholipase C and in recent work we have shown that disruption of phospholipase C signaling in cells leads to altered calcium homeostasis. Such phenomena can be triggered by loss of parkin or by high level overexpression of mutant forms of the protein. Given that it is a cytoplasmic enzyme, one of the more surprising results in recent years is that parkin is involved in mitochondrial function. This implies a link between parkin and other genes for recessive parkinsonism, particularly the kinase PINK1. We have also characterized some of the mitochondrial phenotypes in fibroblasts with parkin mutations. We found that these cells have mutliple correlated mitochondrial phenotypes, including altered mitochondrial membrane potential, enzyme activities and morphological changes. How this relates to ubiquitin addition to proposed substrates, the known biochemical activity of parkin, is unclear and future work will be directed at understanding this relationship.

As well as being a kinase associated with age-dependent penetrant forms of Parkinsons disease, Leucine-rich repeat kinase 2 (LRRK2) is also an authentic GTP binding protein. There are mutations in the GTP-binding ROC (Ras of complex proteins) domain and the adjacent COR (C-terminal of ROC) in LRRK2 that cause Parkinsons disease. The aim of this project is to understand why LRRK2 and related homologue LRRK1 bind GTP and what effect this has on the protein. Our ongoing work aims to identify proteins that bind to the ROC:COR domains of LRRK2, ie the GTP-binding region. We predict that understanding what these proteins are will give insight into the function of LRRK2 and may help understand mutations within those domains. We are currently pursuing interactions of LRRK2 with chaperones and co-chaperones, as well as novel interactors, in this region. At this point, we have mapped a series of interactions around the COR domain of LRRK2 that suggest that there is a multivalent protein complex. Ongoing work in the laboratory is aimed at understanding the functional impact of these interactions in cellular models.

We have been interested in the relationship between oxidative stress and DJ-1, a rare cause of recessive Parkinson's disease, for several years and have focussed the relationship between oxidative stress and mitochondrial localization. We have found that in both rat and mouse brains, DJ-1 knockout results in an age-dependent accumulation of hexokinase 1 in the cytosol, away from its usual location at the mitochondria, with subsequent activation of the polyol pathway of glucose metabolism in vivo. Both in the brain and in cultured cells, DJ-1 deficiency is associated with accumulation of the phosphatase PTEN that antagonizes the kinase AKT. In cells, addition of an inhibitor of AKT (MK2206) or addition of a peptide to dissociate association of hexokinases from mitochondria both inhibit the PINK1/parkin pathway, which works to maintain mitochondrial integrity. We have concluded that hexokinases are an important link between three major genetic causes of early onset Parkinson's disease. Because aging is associated with deregulated nutrient sensing, these results help explain why DJ-1 is associated with age-dependent disease.

We have been interested in the relationship between oxidative stress and DJ-1, a rare cause of recessive Parkinson's disease, for several years and have focussed the relationship between oxidative stress and mitochondrial localization. In the current period, we have explored the effects of DJ-1 deficiency in the brains of mice and rats. We find that loss of DJ-1 alters cellular signaling pathways that result, at a cellular level, in the accumulation of changes in metabolic enyzmes. we have focused on crossing DJ-1 and NQO1 animals, both of which have been shown previously to have protective effects against oxidative damage in parkinsonian-related in vitro or in vivo models. To determine whether these targets have a synergistic effect, homozygous DJ-1 (n = 15), NQO1 (n = 16), and DJ-1/NQO1 (n = 18) knockout mice were phenotypically characterized against wildtype (WT; n = 18) mice across a variety of behavioral tasks. Male and female adult mice (12 24 months) were tested for basal differences in the following motor- and cognitive-related tasks: rotarod (motor learning), exploratory motor behavior (open field), grip strength (forelimb strength), cage top wire hang (motor deficit), DigiGait (gait analysis), and Y-maze spontaneous alternation (working memory and spontaneous locomotion). These tasks were chosen to encompass the wide range of motor deficits and cognitive decline seen in PD patients. All behavioral testing has recently been completed and show some mild changes in anxiety-related measures . Very excitingly, assays of fine motor skill appear to show a synergistic effect of the double knockouts, suggesting a PD-like phenotype

Our main aim in this project is to understand how mutations across many different domains of LRRK2 cause dominantly inherited Parkinsons disease. We have been particularly looking for shared effects of multiple mutations that are found in many different functional domains of the molecule. We have now developed methods to measure the endogenous autophosphorylation of LRRK2 in vivo. We have shown that, as expected, mutation in LRRK2 result in enhanced autophosphorylation suggesting a gain of function mechanism. Importantly, we have also been able to block the enhanced activity using specific LRRK2 kinase inhibitors, indicating potential use of this class of compounds in treating LRRK2-related disease. We are currently following up these observations with additional time and concentration-dependency of application of these compounds to understand the long-term consequences of inhibition of kinase activity in vivo.