In this issue of Brain, two articles explore the intriguing relationship between mutations in the lysosomal enzyme glucocerebrosidase (GBA) and Parkinson’s disease. Siebert et al. (2014) present a comprehensive review of the strong evidence for a connection between autophagy-lysosomal dysfunction and neurodegeneration in Parkinson’s disease, while McNeill et al. (2014) provide experimental evidence suggesting that this connection is a plausible target for disease-modifying therapy.
It is a truism that developing effective disease-modifying treatments for neurodegenerative diseases is difficult, with many obstacles to achieving success. Effective treatments must, for example, penetrate the blood–brain barrier with favourable pharmacokinetic properties, and gain access to therapeutic targets that are mostly located intracellularly. Despite the best efforts of both bench and clinical researchers, we seem a considerable distance away from possessing disease models with strong predictive validity. In Parkinson’s disease, model development has been particularly frustrating. Identification of several Mendelian forms of parkinsonism has not led to the development of murine genetic models with strong face validity. Moreover, research over the last couple of decades has revealed features of neurodegeneration that pose additional obstacles, requiring conceptual adjustments in the approach to the development of disease modifying therapies.
Our thinking has been dominated by descendants of Paul Ehrlich’s ‘Magic Bullet’ concept (Yarnell, 2005). Find a key node in a well defined pathological cascade, target it with an effective small molecule, and cure disease. This approach was spectacularly successful for anti-bacterial chemotherapy and has had modest successes for cancer treatment. The great success of anti-bacterial chemotherapy, however, results from marked differences in the biochemistry of prokaryotic and eukaryotic cells. The Magic Bullet model is not likely to apply to neurodegenerative diseases, the majority of which are proteinopathies in which a mutated and/or abnormally conformed native protein species exhibits neurotoxic effects. In the most prevalent neurodegenerative diseases, there are quantitative, as opposed to qualitative, differences between affected and control subjects. In Alzheimer’s disease, Mawuenyega et al. (2010) suggest an ∼30% difference in the clearance of amyloidogenic fragments of amyloid precursor protein is a key factor distinguishing subjects with Alzheimer’s disease from controls. In Parkinson’s disease, the pathogenic consequences of α-synuclein duplications and triplications suggest that changes in the rates of otherwise normal processes are also important. In addition, some of these abnormal protein species likely have multiple pathogenic effects. In the polyglutamine disease Huntington’s disease, credible evidence indicates the presence of multiple downstream pathogenic mechanisms, possibly including both gain-of-function toxicities and haploinsufficiency effects (Zuccato et al., 2010). In the case of the recently described C9orf72 mutations that cause certain forms of frontotemporal dementia and amyotrophic lateral sclerosis, plausible pathogenic mechanisms include haploinsufficiency, RNA toxicity, and toxic peptide species generated from repeat associated non-ATG (RAN) translation (Gendron et al., 2014). In such a scenario, trying to find a Magic Bullet against a single downstream pathogenic cascade may be doomed to failure.
McNeill and colleagues (2014) pursue another approach—taking advantage of cellular defence mechanisms against misfolded proteins to enhance neuronal capacity to deal with proteinopathy. The point of departure for this work is the discovery by Ellen Sidransky and colleagues (Siebert et al., 2014) of a relatively high rate of parkinsonism in individuals with Gaucher disease, an entity caused by GBA mutations. In addition, GBA mutations are a common risk factor for Parkinson’s disease. GBA encodes the lysosomal hydrolase, glucosyl ceramidase. As investigated by Sidransky’s laboratory and other groups, and discussed nicely in the Siebert et al. (2014) review, there seems to be a close inverse relationship between glucosyl ceramidase expression-activity and α-synuclein accumulation, suggesting strongly that autophagy-lysosome dysfunction is a primary feature of Parkinson’s disease pathogenesis, at least for some forms of the disease. In Gaucher disease, primary abnormalities in GBA are suggested to lead to lysosomal dysfunction and α-synuclein accumulation. In sporadic Parkinson’s disease, some combinations of relatively minor changes in α-synuclein accumulation, glucosyl ceramidase expression-activity, and age-related changes in proteostasis could lead to a feed-forward process in which α-synuclein accumulation and further depression of glucosyl ceramidase expression activity perpetuate a vicious cycle. Membrane binding is a key biological feature of α-synuclein, and disruptions in α-synuclein homeostasis may disrupt additional steps in the endolysosomal pathway or membrane trafficking steps in related pathways, a possibility consistent with recent work on genetic risk factors for Parkinson’s disease (Vilariño-Güell et al., 2014).
McNeill et al. (2014) used fibroblast lines from patients with Gaucher disease, and from patients with Parkinson’s disease and GBA mutations, as their model system. Many GBA mutations do not affect the catalytic site of glucosyl ceramidase, but rather result in endoplasmic reticulum retention and destruction of this protein. Ambroxol [trans-4-(2-Amino-3,5-dibromobenzylamino)cyclohexanol hydrochloride] is a clinically used expectorant suggested to act as a ‘chemical’ chaperone facilitating glucosyl ceramidase exit from the endoplasmic reticulum and transport to lysosomes. McNeill et al. (2014) demonstrate that ambroxol treatment improves glucosyl ceramidase activity and reduces production of cytosolic reactive oxygen species in these fibroblast lines. They show also that ambroxol treatment reduces α-synuclein levels in a neuroblastoma line engineered to overexpress α-synuclein. In a particularly intriguing extension of their initial experiments, McNeill et al. (2014) suggest that there is more to ambroxol action than its hypothesized chaperone effect.
GBA is a member of the Coordinated Lysosomal Expression And Regulation (CLEAR) network (Palmieri et al., 2011). The CLEAR network encodes many of the hundreds of genes required for lysosomal biogenesis, regulation, and function. Transcription factor EB (TFEB) seems to act as a master regulator of the CLEAR network, inducing synchronized upregulation of CLEAR network genes. McNeill et al. (2014) present data suggesting that ambroxol treatment induces the CLEAR network, upregulating many genes involved in autophagy-lysosomal function, including TFEB, which can act as a positive feedback inducer of its own transcription. The implication is that ambroxol treatment generally enhances autophagy-lysosomal function. Given the probable connection between autophagy-lysosomal dysfunction and α-synuclein accumulation, boosting CLEAR network activity is a rational therapeutic strategy.
It is important not to exaggerate the effects of ambroxol treatment. As McNeill et al. (2014) are careful to point out, the increases in glucosyl ceramidase activity they document are relatively modest. It is also conceivable that some Parkinson’s disease-related endolysosomal defects could be exacerbated by a general upregulation of the pathway. Ambroxol itself, however, has pragmatic features that enhance its attractiveness as a potential therapy for Parkinson’s disease. It is a clinically used agent, a small pilot trial in type 1 Gaucher disease demonstrated initial tolerability and haematological response in some subjects, and it does cross the blood–brain barrier (Zimran et al., 2013). On the other hand, some data suggest that ambroxol is a sodium channel antagonist (Weiser and Wilson, 2002). Assuming independent replication of the data of McNeill et al. (2014), there is justification for careful tolerability trials of ambroxol in subjects with Parkinson’s disease.
Beyond the merits of this particular agent, the work of McNeill et al. (2014) points to a different approach to disease-modifying treatments for Parkinson’s disease. Protein misfolding is a normal cellular feature. It is estimated that up to 30% of proteins are misfolded after message translation. Cells possess elaborate mechanisms—chaperones, the unfolded protein response, proteasomes, autophagy-lysosomes—to dispose of misfolded proteins. Our very long-lived post-mitotic neurons may be particularly susceptible to minor age-related changes in cellular ability to deal with misfolded proteins, perhaps leading to pathological vicious cycles such as the hypothesized deleterious α-synuclein–glucosyl ceramidase interaction discussed above. Boosting neuronal defences against misfolded proteins by non-specifically enhancing misfolded protein disposal mechanisms might be a way to overcome some of the obstacles we face in the search for effective disease-modifying therapies for Parkinson’s disease and other neurodegenerative disorders. This may mean abandoning the search for ‘Magic Bullets’ in favour of a less targeted approach—a ‘Magic Shotgun’.