Neuroserpin is one of the serpins (serin protease inhibitors), a conserved superfamily of proteins that inhibit serin proteases by a mechanism that requires a high structural flexibility, and which renders serpin proteins very sensitive to point mutations that alter their folding and cellular handling. This molecular mechanism underlies a class of pathologies called the serpinopathies, where point mutations cause serpin polymerisation and retention within the endoplasmic reticulum of the cell of synthesis. Six different polymerisation-causing mutations have been found in neuroserpin, a secreted serpin mainly expressed by neurons, as the cause of a rare but deadly type of dementia called FENIB. Although the pathological manifestations of serpin polymerisation depend on the inhibitory target and place of action of each specific serpin, the molecular mechanism is common and several aspects remain obscure for all serpinopathies. Particularly, little is known about the cell toxicity effects of polymer accumulation inside the endoplasmic reticulum. We have recently published a novel neural expression system, in which RNA sequencing has uncovered the overexpression of several anti-oxidant genes in cells expressing the highly polymerogenic FENIB G392E variant of NS. These cells undergo apoptosis when the anti-oxidant defences are inhibited pharmacologically, and our recent results show alterations in mitochondrial distribution across G392E NS neurons. This phenotype is worsened by a chelator of the antioxidant molecule glutathione, supporting a link between oxidative stress and mitochondrial dysfunction in the neurodegeneration FENIB. This proposal aims to continue our studies on mitochondrial alterations and their relationship with the endoplasmic reticulum where polymers of G392E NS accumulate, inducing a stress response that is still poorly understood.
The molecular bases of the serpinopathies, particularly the dementia FENIB and alpha-1 antitrypsin deficiency due to polymer formation by mutant NS and alpha-1 antitrypsin respectively, are known since more twenty years. Nevertheless, the exact nature of the cellular toxicity exerted by serpin polymers is still incompletely understood, and as yet there is no treatment for the dementia FENIB and only palliative treatments for the pathological manifestations of alpha-1 antitrypsin deficiency. In recent years, we have created a novel cell model overexpressing wild type or the pathological variant G392E of NS in mouse neural progenitor cells, which recapitulates the main features of FENIB and is now being used to study the mechanism behind the toxicity of NS polymers. In this model system, we have recently described for the first time a role for oxidative stress in the neuronal toxicity caused by the FENIB-causing variant G392E NS (Guadagno et al, 2017), thus uncovering part of the mechanism that renders neurons more susceptible to apoptosis in FENIB brains. Moreover, our recent data showing alterations in mitochondrial distribution in G392E NS cells involves these cellular organelles in FENIB for the first time.
Accumulating evidence suggests that protein folding and generation of reactive oxygen species (ROS) as a by-product of protein oxidation in the ER are closely linked events. Conversely, alterations in the redox status and generation of ROS alter ER homeostasis and protein folding (Malhotra et al., 2007). Both ER stress and oxidative stress, through ROS generation, may increase the leak of Ca2+ from the ER, as well as induce protein and lipid oxidation. Moreover, the very close proximity of ER and mitochondria in the mitochondria associated membrane (MAM) areas promotes the accumulation of cytosolic Ca2+ near mitochondria (Jacobson and Duchen, 2002), leading to ROS generation as a consequence of increased mitochondrial Ca2+ loading. High levels of mitochondrial ROS further increase Ca2+ release from the ER, generating a vicious cycle of ROS production that takes place during cellular oxidative stress.
The accumulation of serpin polymers within the ER may upset the redox balance in this organelle. It has been shown that the polymerogenic Z variant of alpha-1 antitrypsin, the archetypical serpin which undergoes polymerisation in the serpinopathy alpha-1 antitrypsin deficiency, leads to increased expression of redox-regulating genes in the liver of a mouse model of the disease, as well as higher levels of ROS and oxidative liver damage in aged mice (Markus et al, 2012). The involvement of oxidative stress in alpha-1 antitrypsin deficiency in humans is supported by the presence of increased oxidative stress markers and reduced antioxidant defences in blood in a cohort of children suffering from this condition (Escribano et al, 2015). Our recent publication has confirmed the activation of an anti-oxidative response in neural cells overexpressing the polymerogenic G392E NS variant that causes severe dementia FENIB (Guadagno et al, 2017), and more recently the alteration of mitochondrial distribution in these cells. In line with these findings, we now propose to look at the effects of antioxidants, the role of Ca2+ and the actin cytoskeleton, and the distribution of MAM sites in response to NS polymer accumulation in the ER. This is supported by previous findings showing the involvement of Ca2+ in the ER stress response to NS polymers (Davies et al, 2009), by our findings of an altered distribution of mitochondria in G392E NS cells differentiated to neurons, and by the identification by RNA sequencing of G392E NS cells of several genes involved with cytoskeleton behaviour that show altered expression. These studies are novel in the field of the serpinopathies, so we expect that our research will shed new light into the intracellular pathways that mediate the toxic effects of NS polymers in the dementia FENIB, and that it will be relevant for other diseases related to serpin polymerisation and other types of neurodegeneration.
Literature cited
Davies MJ et al, 2009. J. Biol. Chem 284:18202-18209
Escribano et al, 2015, Thorax, 70:82-83
Guadagno et al, 2017. Neurob Disease, 103:32-44
Jacobson and Duchen, 2002. J Cell Science, 115:1175-1188
Malhotra and Kaufman, 2007. Antioxid Redox Signal, 9:2277¿2293
Markus et al, 2012. Exp Biol Med, 237:1163-1172